The Impact of the International Livestock Research Institute The Impact of the International Livestock Research Institute Edited by: John McIntire and Delia Grace CABI is a trading name of CAB International CABI CABI Nosworthy Way 745 Atlantic Avenue Wallingford 8th Floor Oxfordshire OX10 8DE Boston, MA 02111 UK USA Tel: +44 (0)1491 832111 Tel: +1 (617)682-9015 Fax: +44 (0)1491 833508 E-mail: cabi-nao@cabi.org E-mail: info@cabi.org Website: www.cabi.org © Copyright International Livestock Research Institute and CAB International 2020. Co-published by CABI and the International Livestock Research Institute 2020. This publication is copyrighted by the International Livestock Research Institute (ILRI). It is licensed for use under the Creative Commons Attribution 4.0 Inter- national Licence. To view this licence, visit https://creativecommons.org/ licenses/by/4.0. Unless otherwise noted, you are free to share (copy and redistrib- ute the material in any medium or format), adapt (remix, transform, and build upon the material) for any purpose, even commercially, under the following conditions: ATTRIBUTION. The work must be attributed, but not in any way that suggests endorsement by ILRI or the author(s). NOTICE: For any reuse or distribution, the licence terms of this work must be made clear to others. Any of the above conditions can be waived if permission is obtained from the copyright holder. Nothing in this licence impairs or restricts the author’s moral rights. Fair dealing and other rights are in no way affected by the above. The parts used must not misrepresent the meaning of the publication. ILRI would appreciate being sent a copy of any materials in which text, photos etc. have been used. ISBN-13: 978 1 78924 185 3 (hardback) ILRI ISBN: 92-9146-586-3 (hardback) CABI Commissioning editor: Alexandra Lainsbury CABI Editorial assistant: Lauren Davies CABI Production editor: James Bishop Typeset by SPi, Pondicherry, India Printed and bound in the UK by Bell and Bain Ltd, Glasgow The designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of the International Livestock Research Institute (ILRI) concerning the legal status of any country, territory, city or area or its authorities, or concern- ing the delimitation of its frontiers or boundaries. The mention of specific companies or products of manufacturers, whether or not these have been patented, does not imply that these have been en- dorsed or recommended by ILRI in preference to others of a similar nature that are not mentioned. The views expressed herein are those of the authors and do not necessarily represent those of ILRI. NOTICE: For any reuse or distribution, the license terms of this work must be made clear to others. Any of the above conditions can be waived if permission is obtained from ILRI. Nothing in this license impairs or restricts the author’s moral rights. Fair dealing and other rights are in no way affected by the above. The parts used must not misrepresent the meaning of the publication. ILRI would appreciate being sent a copy of any materials in which text, photos, etc., have been used. Citation: ‘The Impact of the International Livestock Research Institute’. 2020. Edited by John McIntire and Delia Grace. Nairobi, Kenya: ILRI, and Wallingford, UK: CABI. Acknowledgement: ILRI thanks all donors that globally support its work through their contribu- tions to the CGIAR Trust Fund: https://www.cgiar.org/funders/. Contents Foreword: Peter Doherty, ILRI patron xi Preface: A Note from ILRI Director General Jimmy Smith xiii A Note from ILRI Board of Trustees Chair Lindsay Falvey xvi Acknowledgements xv Maps xvii List of Figures xxi List of Tables xxv List of Boxes xxvii List of Abbreviations xxix List of Contributors xxxvii Introduction: The Evolution of IARC Livestock Research, 1975–2018 1 John McIntire and James Smith Part I. anImal GenetIcs, ProductIon and HealtH Preface to Part I: researcH sPendInG and PublIcatIons on anImal GenetIcs, ProductIon, and HealtH 49 1 Livestock Genetics and Breeding 59 J. E. O. Rege, Joel Ochieng and Olivier Hanotte 2 Control of Pathogenesis in Animal African Trypanosomiasis: A Search for Answers at ILRAD, ILCA and ILRI, 1975–2018 103 Samuel J. Black 3 T setse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia: ILRI’s Role in the Field 148 Delia Grace, Ekta Patel and Thomas Fitz Randolph viii Contents 4 Impact Assessment of Immunology and Immunoparasitology Research at ILRAD and ILRI 164 Samuel J. Black and Cynthia L. Baldwin 5 Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 208 Brian Perry, Bernard Bett, Eric Fèvre, Delia Grace and Thomas Fitz Randolph 6 The Management and Economics of East Coast Fever 239 Phil Toye, Henry Kiara, Onesmo ole-MoiYoi, Dolapo Enahoro and Karl M. Rich 7 Transboundary Animal Diseases 274 Delia Grace, Tadelle Dessie, Michel Dione, Henry Kiara, Anne Liljander, Jeff Mariner, Jan Naessens, Edward Okoth, Ekta Patel, Lucilla Steinaa, Phil Toye and Barbara Wieland 8 Zoonoses 302 Delia Grace, Silvia Alonso, Bernard Bett, Elizabeth Cook, Hu Suk Lee, Anne Liljander, Jeff Mariner, Florence Mutua, Hung Nguyen-Viet, Ekta Patel, Thomas Fitz Randolph, Kristina Roesel, Lian Thomas, Phil Toye, Fred Unger and Barbara Wieland 9 Food Safety and Nutrition 338 Delia Grace, Silvia Alonso, Bernard Bett, Johanna Lindahl, Ekta Patel, Hung Nguyen-Viet, Kristina Roesel, Fred Unger and Paula Dominguez-Salas 10 Ticks and Their Control 366 Peter Willadsen Part II. PrImary ProductIon Preface to Part II: researcH sPendInG and PublIcatIons on PrImary ProductIon 387 11 Rangeland Ecology 395 Polly Ericksen, Pierre Hiernaux, Augustine Ayantunde, Philip K. Thornton, Jason Sircely and Lance Robinson 12 Forage Diversity, Conservation and Use 423 Jean Hanson, Rainer Schultze-Kraft, Michael Peters, Peter Wenzl, Ahmed Amri, Ali Shehadeh and Mariana Yazbek 13 T he Impact of CGIAR Centre Research on Use of Planted Forages by Tropical Smallholders 450 Alan J. Duncan, Michael Peters, Rainer Schultze-Kraft, Philip K. Thornton, Nils Teufel, Jean Hanson and John McIntire 14 M ultidimensional Crop Improvement by ILRI and Partners: Drivers, Approaches, Achievements and Impact 480 the late Michael Blümmel, Anandan Samireddypalle, Perez Haider Zaidi, Vincent Vadez, Ramana Reddy and Pasupuleti Janila Part III. troPIcal lIvestock systems Preface to Part III: researcH sPendInG and PublIcatIons on troPIcal lIvestock systems 507 15 African Livestock Systems Research, 1975–2018 515 John McIntire, Tim Robinson and Caroline Bosire Contents ix 16 Ruminant Livestock and Climate Change in the Tropics 601 Polly Ericksen, Philip K. Thornton and Gerald C. Nelson 17 Economics and Policy Research at ILRI, 1975–2018 639 Mohammed Jabbar, Steve Staal, John McIntire and Simeon Ehui 18 The Impact of ILRI Research on Gender 680 Catherine Hill, Nicoline de Haan, Alessandra Galiè and Nelly Njiru Conclusion: The Future of Research at ILRI 699 Jimmy Smith, Iain Wright, Delia Grace, and Brian Perry with Isabelle Baltenweck, Bernard Bett, the late Michael Blümmel, Nicoline de Haan, Alan Duncan, Polly Ericksen, Olivier Hanotte, Jean Hanson, Chris Jones, Steve Kemp, Okeyo Mwai, Gerald C. Nelson, Vish Nene, Lance Robinson, Steve Staal, Philip K. Thornton and Barbara Wieland appendix 1: CIAT and ICARDA Livestock Research 731 Index 735 Foreword It gives me pleasure to provide the foreword to this important scientific record of 45 years of live- stock and livestock-related science conducted by the International Livestock Research Institute (ILRI, 1995–2019) and its two predecessors, the International Laboratory for Research on Ani- mal Diseases, based in Kenya (ILRAD, 1973–1994), and the International Livestock Centre for Africa, based in Ethiopia (ILCA, 1974–1994), as well as selected key partners of these three insti- tutions. ILRI—and ILRAD and ILCA before it—is a member of CGIAR, a global research-for- development partnership of 15 centres working with hundreds of partners around the world for a food-secure future. Co-hosted by Kenya and Ethiopia and working through a network of offices and projects across Asia and Africa, ILRI’s multidisciplinary staff and programs develop and deliver science-based solu- tions, provide evidence for decision-making, and develop capacities of livestock-sector stakeholders for more equitable, broad-based and sustainable livestock development. ILRI’s research for development agenda covers laboratory-based biosciences to field-based re- search to policy support. The main topics include animal productivity (health, genetics and feeds), food safety and zoonoses, livestock systems and the environment, gender and livelihoods, and policy and markets. The institute’s core scientific competencies span the breadth of livestock science, from biosciences to social sciences, agricultural economics to human nutrition, disease epidemiology to environmental sciences. With such diverse disciplinary work conducted under one roof, ILRI works to integrate and share this knowledge and expertise in systems-level solutions that improve global food, nutritional, economic and environmental security. Gender, research methods, communica- tions, knowledge management and business and capacity development are central to ILRI’s mandate and cut across all of its research and development areas. The rich diversity and complexity of livestock livelihoods in Africa and Asia are exceptional. A glance at the chapter titles and contents of this volume will attest to the diversity of research ap- proaches employed over the years to better exploit and sustain that richness for the larger purpose of closing great disparities in global economic and nutritional well-being and human and environmen- tal health. Some big ideas are presented here, along with some approaches that were tested and then let go; there are both success and failure stories, as is only proper in such a record. And there are breakthroughs and scientific and development impacts to laud, including developing-world solu- tions for developing-world problems. xii Foreword This volume can serve as a reference and resource for all interested in the role of livestock in agricultural transformation and sustainable development. It should be useful for distilling, learning from, and building on past work and lessons. We hope it will inform and inspire students, researchers and research managers and their investors. Professor Peter C. Doherty AC, FAA, FRS Animal scientist Nobel Prize Laureate for Physiology or Medicine - 1996 Preface A note from ILRI Director General Jimmy Smith No endeavour can be more important than feeding our world without jeopardizing our future. R epeated, catastrophic hungers have punctuated human history and inadequate nutrition, which still afflicts two billion people, remains a too often invisible health catastrophe. Half a century ago, CGIAR’s research into high-yielding, disease-resistant cereals launched the Green Revolution, sav- ing more than a billion people from starvation. CGIAR, a global network of 15 research centres, was created to perpetuate more socially equitable, and economically and environmentally sustain- able, agricultural revolutions. One of these centres, the International Livestock Research Institute (ILRI), is dedicated to research on livestock in developing countries. Here, as in the following pages, mention of ILRI typically encompasses the work of ILRI’s two predecessors, the International La- boratory for Research on Animal Diseases [ILRAD] and the International Livestock Centre for Africa [ILCA]. As the spectre of global famine began to recede in the middle of the last century, new challenges, as well as new opportunities, emerged, many of which ILRI’s livestock research was uniquely posi- tioned to address. The challenges included increasing disease epidemics, including zoonotic pandemics such as coronavirus disease 2019 (COVID-19); rising demand for meat, milk and eggs, particularly in developing countries; a rise in ‘hidden hunger’ due to micronutrient deficiencies; climate change, poor animal welfare, and widespread ill health caused by unsafe foods. This book describes the evolving and multi-faceted roles of ILRI in addressing these and other global challenges in nearly a half century of research. ILRI researchers and partners took leading roles, for example, in the following. LIVESTOCK REVOLUTION: As more and more people in the global south are moving out of pov- erty and into cities, their demand for livestock products is rising dramatically (especially for the fast-reproducing, easy to intensify pig and poultry systems). Finding ways to meet this rising demand without threatening human, animal and environment health is critically important. Doing so re- quires a deep understanding of the socioeconomics of livestock economies, of the diverse livestock value chains that are the mainstay of smallholder agriculture and an ability to influence livestock policy and practice. Sustainable livestock intensification also requires better livestock breeding, feed- ing and management. LIVESTOCK DISEASE: Tackling neglected livestock diseases of greatest importance to the world’s poor was a major driver for the foundation of ILRI and remains a major constraint and in a globalized xiv Preface world, where diseases can spread fast and severely damage economies. Pastoralist systems are especially vulnerable, due to high contact with animals and often poor access to services. Moreover, most new human diseases originate in animals, livestock as well as wildlife, so controlling these dis- eases in animal populations, before they spread to human populations, is essential. HIDDEN HUNGER: The success of the Green Revolution helped to shift attention from a world shortage of calories to a shortage of essential micronutrients, which are abundant in milk, meat and eggs. This fact is putting these nourishing livestock-derived foods centre stage on the world’s develop- ment agenda. At the same time, obesity and its attendant health issues are growing problems, even in countries were nutrition is insecure. These health problems have been associated with increased consumption of highly processed or unhealthily cooked livestock products. CLIMATE CHANGE: With the threat of climate change looming and the role of livestock produc- tion as an emitter of greenhouses gases taking the limelight, transforming livestock systems into sustainable as well as inclusive systems has become a critical goal of the world’s environmental agenda. At the same time, livestock systems must adapt to changing climates. FOOD SAFETY AND ANIMAL WELFARE: And as more and more people in developing countries are consuming livestock-derived foods that are among the foods most likely to become contaminated and cause human illness, and as more and more people in developed countries are concerned by poor animal welfare, populations are becoming increasingly concerned about the safety and provenance of the food they eat. Over the years, ILRI expanded its remit from better controlling a few deadly tropical animal dis- eases and better understanding livestock production systems in Africa to a broad portfolio of re- search for livestock development globally with a focus on Africa and Asia where poverty rates were highest. This broader agenda covered—in addition to livestock health, feed, genetics, trade, market- ing and consumption—the gender aspects of livestock systems, the impacts of livestock on the nat- ural resource base and planetary boundaries and the roles of livestock in societies. This work produced many outputs. This book documents and quantifies some of these in terms of technologies and innovations discovered and disseminated, papers written and cited, students taught and graduated, policies influenced and written, and global endeavours—such as building sustainable livestock systems and empowering women in livestock agriculture—supported and advanced. More difficult to document, but ultimately more important, are ILRI’s direct contributions to improving lives and livelihoods of the one billion poor people that remain dependent on livestock and the more than six billion poor people (living on less than $10 a day) that use livestock or consume or sell livestock products. Research, of course, is only one part in the complex ‘innovation systems’ that lead to human betterment. This book documents many actual and potential development impacts and many more opportunities for impact as innovation moves from the library and laboratory to the farm and kitchen. This unique book documents 45 years of impacts of livestock research in regions of the devel- oping world, particularly in Africa and Asia. The voices of the authors of these chapters are cultur- ally and scientifically diverse. That was purposeful. The authors of this book are acutely aware that their voices are only representative of the larger work in which they are engaged. Unavoidably, many livestock scientists and important pieces of livestock research are missing or presented only briefly in these pages. It is, of course, the larger canvas of livestock work that matters most in delivering on the global sustainable development agenda. It is our hope that this record—with its many examples of the wealth of scientific and public goods emanating from long-term intellectual and financial livestock research investments and partnerships—helps others to build on this legacy to create more sustainable as well as equitable livestock systems in future years. Preface xv A note from ILRI Board of Trustees Chair Lindsay Falvey As chair of the ILRI Board of Trustees, I am delighted to welcome this important publication, docu- menting the many benefits of the research-for-development work conducted by ILRI and its many partners. This volume captures much of the history of innovation in livestock research in and for low- and middle-income countries, especially in Africa, which could otherwise be lost. By analysing the challenges and failures as well as successes of this mission-oriented livestock research, this book provides an honest and objective guide to managing, planning and implementing agricultural re- search for development. While its focus is retrospective, this publication also sets out a path for ILRI’s future work in these uncertain times. I am confident ILRI is well placed to help meet the world’s 17 Sustainable Development Goals, based on historic successes and on-going efforts and I thank the edi- tors, authors and editorial and research support staff who have devoted so much time to developing and finalizing this important publication. Acknowledgements This book chiefly but not exclusively concerns more than 40 years of work and impacts generated by the International Livestock Research Institute (ILRI, 1995 to date) and ILRI’s two predecessors, the International Laboratory for Research on Animal Diseases (ILRAD, 1973–1994) and the Inter- national Livestock Centre for Africa (ILCA, 1974–1994). We have generally avoided mention of donor organizations in the chapters, preferring to acknowledge their generous and essential finan- cial and intellectual support in one place, in an online appendix (www.ilri.org/dataportal/ impact/investors). As livestock research-for-development work is profoundly collaborative in nature, close partners and collaborators of ILRI, ILCA and ILRAD are gratefully named and cited throughout this work, but no attempt has been made to be exhaustive in these acknowledgements, the comple- tion of which would involve the listing of literally hundreds of organizations worldwide. It is thus possible that we have neglected some significant partner contributions to the work described here, and we ask for your indulgence and understanding that this in no way is meant to undervalue the good work of so many of ILRI’s partners. ILRI thanks the following members of CABI’s book production team: James Bishop, senior production editor; Lauren Davies, editorial assistant; Jane Hoyle, copy-editor; Tabitha Jay, editorial assistant; and Alexandra Lainsbury, commissioning editor. Ekta Patel, ILRI biosciences communications manager, started research assistance for this book back in late 2017 and continued through early 2019, when she helped organize a series of write- shops for the book. Further research assistance was provided by ILRI’s Caroline Bosire, Emily Kilonzi, Emmah Kwoba, Ianetta Mutie, Ann Mureithi, Chi Nguyen, David Oduori, Stephen Oloo, and Lina Wanga. ‘Michael Victor, head of ILRI communications and knowledge management; Ann Mureithi, senior administrative officer in the director general’s office; and Lina Wanga, executive assistant to ILRI’s deputy director general, provided logistical support. ILRI Director General Jimmy Smith and Deputy Director General Iain Wright provided overall institutional support for the production of this book. John McIntire and Delia Grace were scientific and technical editors with dedicated support in all facets of the book from Susan MacMillan, of ILRI’s communications and knowledge management team.’ For comments, advice and additional material, we thank: Nick Abel, Kwaku Agyemang, Jock Anderson, the late Azage Tegegne, Derek Baker, Leyden Baker, Peter Ballantyne, Carolyn Benigno, Berhanu Gebremehdin, the late Hans Binswanger, Derek Byerlee, Peter-Henning Clausen, Cees de Haan, Paula Dominguez-Salas, Steve Franzel, Bruno Gerard, Sita Ghimire, John Gibson, Maggie Gill, xviii Acknowledgements Elaine Grings, Guido Gryseels, Peter Hazell, Tony Irvin, Martyn Jeggo, Steve Kemp, R omano Kiome, Jeffrey Mariner, Peter Matlon, Roger Morris, Ivan Morrison, Jan Naessens, Jerry Nelson, Vish Nene, Hung Nguyen, An Notenbaert, Clare Oxby, Ekta Patel, Bruce Pengelly, Michael Peters, Mark Powell, Ed Rege, Jonathan Rushton, Rainer Schultze-Kraft, Carlos Seré, Emmy Simmons, Werner Stur, Jim Sumberg, Gbassy Tarawali, Shirley Tarawali, Bill Thorpe, Laurian Unnevehr and Trevor Wilson. Nomenclature Trypanosomiasis, also known as trypanosomosis, is the name of several diseases in vertebrates caused by parasitic protozoan trypanosomes of the genus Trypanosoma. This disease in humans in- cludes sleeping sickness (human African trypanosomiasis, or HAT). The disease in livestock is called nagana (African animal trypanosomiasis, or animal African trypanosomiasis, abbreviated to AAT). Because ILRAD used the terms ‘trypanosomiasis’ and ‘African animal trypanosomiasis’, and because ‘trypanosomiasis’ is the term preferred by the Centers for Disease Control and Prevention, the Food and Agriculture Organization of the United Nations and the World Health Organization, that is the term used in this volume. The research centres formerly known as Bioversity Intl. and Centro Internacional de Agricultura Tropical (CIAT) formed the ‘Alliance of Bioversity Intl. and CIAT’ on 1 J anuary 2020. We have noted this in the affiliations of current staff of that Alliance in the ‘List of Contributors’ and at the appropriate places in Chapters 12 and 13. We have not otherwise changed the names of CIAT or of Bioversity Intl. where those names appear throughout the book. Ethiopian names We apologize to the many Ethiopian scientists who contributed to this book for the Western expres- sion of their names. It proved impossible to correct this against the general trend of the scientific literature. Maps 0 2500 5000 10,000 Kilometres Legend N LGY LGH MRY MRH MIY MIH Urban areas LGA LGT MRA MRT MIA MIT Other Map 1. Global livestock production systems. LGY, livestock/grazing/hyperarid; LGA, livestock/grazing/ arid; LGH, livestock/grazing/humid; LGT, livestock/grazing/temperate and tropical; MRY, mixed/rainfed/ hyperarid; MRA, mixed/rainfed/arid and semi-arid; MRH, mixed/rainfed/humid and subhumid; MRT, mixed/rainfed/temperate and tropical; MIY, mixed/irrigated/hyperarid; MIA, mixed/irrigated/arid and semi-arid; MIH, mixed/irrigated/humid and subhumid; MIT, mixed/irrigated/temperate and tropical. xx Preface Inner Delta Western Niono 1976–86 1976–86 Gambia 1985–88 Western Gourma Kaduna 1983–93 1979–86 Simiri–Sadeize Borana Koara 1983–88 1978–91 Fakara 1993–2006 Maasailand 1998–2014 0 1,000 2,000 4,000 Kilometers Legend Study Area LGH MGH MIH Ruminant Production Syetems LGT MRT MIT N LGY MRY MIY Urban areas LGA MRA MIA Other Map 2. African livestock production systems and principal study sites. LGY, livestock/grazing/hyperarid; LGA, livestock/grazing/arid; LGH, livestock/grazing/humid; LGT, livestock/grazing/temperate and tropical; MRY, mixed/rainfed/hyperarid; MRA, mixed/rainfed/arid and semi-arid; MRH, mixed/rainfed/humid and subhumid; MRT, mixed/rainfed/temperate and tropical; MIY, mixed/irrigated/hyperarid; MIA, mixed/irrigated/ arid and semi-arid; MIH, mixed/irrigated/humid and subhumid; MIT, mixed/irrigated/temperate and tropical. Preface xxi N 0 1000 2000 4000 Kilometres Legend Agro-pastoral millet/sorghum Horticulture mixed Rainfed mixed Cereal/root crop mixed Irrigated Rice/tree crop Coastal artisanal fishing Large commercial and smallholder Root crop Dryland mixed Large-scale cereal/vegetable Small-scale cereal/livestock Forest based Maize mixed Sparse (arid) Highland mixed Mixed Tree crop Highland perennial No data Highland temperate mixed Pastoral Map 3. African cropping systems, c.2010. xxii Preface Zones 1,000–1,400m 1,500–1,600m 1,700–1,800m 1,900–2,300m 0 2.5 5 10 Kilometers Map 4. The Ghibe study site, 1990–1993. (Data from Rowlands and Teale, 1994). List of Figures I.1 ( a) Area of sub-Saharan Africa by agroecological zone, mid-1970s. (b) Rural population by agroecological zone, mid-1970s. (c) Zonal rural population shares, mid-1970s. (d) Zonal rural population densities, mid-1970s. 4 I.2 ( a) Numbers of tropical livestock units by agroecological zone, mid-1970s. (b) Tropical livestock unit density by agroecological zone, mid-1970s. (c) Ruminant livestock numbers by agroecological zone, mid-1970. (d) Zonal ruminant populations as percentages of total, mid-1970s. 5 I.3 Agricultural gross domestic product (a) and livestock share of agricultural GDP (b) by agroecological zone, mid-1970s. 6 I.4 (a) Area of tsetse fly infestation by agroecological zone. (b) Zonal area shares infested by tsetse. (c) Ruminant TLU and area by tsetse challenge, mid-1970s. (d) Potential zonal gross benefits from tsetse clearance and trypanosomiasis control, mid-1970s. 7 I.5 ILRAD spending by research domain and period, 1975–1994. 17 I.6 ILRAD publications by research domain and period, 1975–1994. 19 I.7 ILCA spending by research domain and period, 1975–1994. 22 I.8 ILCA publications by research domain and period, 1975–1994. 24 I.9 ILRI spending by source of funds and period (a) and by research domain and period (b) 1975–2018. 28 I.10 ILRI publications by research domain and publication period, 1975–2018. 35 PI.1 T he decreasing importance of ILRI spending on theileriosis and trypanosomiasis, 1975–2018. 50 PI.2 (a) The decreasing ILRI share of global publications on trypanosomiasis, 1977–2018. (b) Frequency of citations of ILRI and global publications on trypanosomiasis, 1977–2018. (c) Domination of the field of trypanotolerance by ILRI papers, 1977–2018. 51 PI.3 (a) ILRI and other publications on theileriosis and related problems, 1977–2018. (b) Frequency of citations of ILRI and other publications on theileriosis and related problems, 1977–2018. 53 PI.4 (a) The domination of ILRI papers in global livestock immunology until recently, 1977–2018. (b) Frequency of citations of ILRI and other publications in immunology, 1977–2018. 55 PI.5 (a) The rapid increase in ILRI papers on animal genetics, food safety, transboundary diseases and zoonoses after the ILCA/ILRAD merger, 1977–2018. (b) Frequency of citations of ILRI publications on animal genetics, food safety, transboundary diseases and zoonoses, 1977–2018. 56 xxiv List of Figures PI.6 Frequency of citations of ILRI and global institutions in animal health research by quantile, 1976–2018. 57 1.1 Evolution of the ILRI genetics portfolio, 1970s to 2020. 65 1.2 Milk production by genotype in dairy production systems of eastern Africa. 83 1.3 Outcomes of the ILRI-SLU capacity development project. 90 2.1 Microhaematocrit and estimation of parasites in the buffy coat. 120 2.2 Linkage mapping to identify genes responsible for trypanotolerance in N’Dama cattle. 133 3.1 An example of extension material to promote rational drug use. 157 4.1 Percentage of manuscripts by citation groups. 172 4.2 Timeline of fundamental developments in T-cell immunology. 179 4.3 Graph showing the survey responses. 194 6.1 Life cycle of T. parva. 243 6.2 Projected ECF vaccine adoption rates by adoption scenario and production system, 2007–2026. 261 6.3 Projected cattle herds by vaccine adoption and mortality avoided, 2007–2026. 263 6.4 Index of agricultural production in Kenya by vaccine adoption and mortality avoided, 2007–2027. 265 6.5 Benefit:cost ratios for ECF vaccination research and development programmes by vaccine adoption levels and mortality avoided. 266 7.1 Space–time cluster analysis of PRRS outbreaks from 2008 to 2016 in Vietnam. 291 7.2 Study area in western Kenya. 292 8.1 Zoonoses and poor livestock keepers. 307 8.2 Participatory epidemiology activities by country. 309 8.3 Network digram of co-publishing patterns among organizations having six or more co-authorship relationships (ILRI in blue). 314 9.1 Impact pathways among livestock keeping and human nutrition and health outcomes among the poor. 358 PII.1 P rimary production research has been a small share of total ILRI spending, 1975–2018. 388 PII.2 ILRI work has been a large share of global publications on rangelands and related problems, 1977–2018. 389 PII.3 I LRI work has been a small share of global publications on planted forages, 1977–2018. 390 PII.4 I LRI has made niche contributions to multidimensional crops research, 1977–2018. 391 PII.5 F requency of citations of ILRI and global institutions in primary production research by quantile, 1977–2018. 392 12.1 R ecipients of samples from the CGIAR forage gene banks. 434 12.2 D istribution of samples from the CGIAR forage gene banks. 434 12.3 C osts by component and type of forage at ILRI in US$. 437 12.4 N umber of associates working on forage genetic resources by nationality at ILRI from 1986 to 2020. 442 13.1 H igh-potential Napier adoption sites in sub-Saharan Africa. 471 14.1 R elationship between cost of sorghum stover and cost of rice straw and their in vivo digestibility. 486 14.2 R elationship between stover N and grain yield (a), between stover IVOMD and grain yield (b), between stover N and stover yield (c) and between stover IVOMD and stover yield (d) in 24 pipeline maize hybrids grown at four locations in India. 491 14.3 B reeding advances in dual-purpose maize stover fodder quality relative to different sorghum stovers traded in rainfed India in the past decade. 493 14.4 R elationships between mean stover crude protein and grain yield (a) and between mean stover crude protein and stover yield (b) in Kharif and Rabi sorghums submitted for cultivar release from 2002 to 2008. 496 List of Figures xxv 14.5 Relationships between N content of pearl millet stover and grain yields (a) and between N content of pearl millet stover and stover yields (b) under high (HF) and low (LF) fertility and high (HP) and low (LP) population density. 497 14.6 Relationship between haulm N content and grain yield (a) and between haulm N content and haulm yield (b) in 280 chickpea cultivars. 497 14.7 (a) Relationships between mean stover in vitro digestibility and grain yield in Kharif and Rabi sorghum cultivars submitted for release from 2002 to 2008. (b) Relationships between mean stover in vitro digestibility and stover yield in Kharif and Rabi sorghum cultivars submitted for release from 2002 to 2008. 498 14.8 (a) Relationships between in vitro digestibility of pearl millet stover and grain yield (a) and between in vitro digestibility of pearl millet stover and stover yield under high (HF) and low (LF) fertility and high (HP) and low (LP) population density. 498 14.9 (a) Relationship between haulm digestibility and grain yield in 280 chickpea cultivars. (b) Relationship between haulm digestibility and haulm yield in 280 chickpea cultivars. 499 PIII.1 ILRI economics and policy research share of total ILRI spending, 1975–2018. 508 PIII.2 I LRI papers in systems research in Africa and globally, 1977–2018. 509 PIII.3 ILRI papers cited in economics and policy research on Africa and global problems related to livestock systems, 1977–2018. 511 PIII.4 ILRI and other publications in climate change research related to livestock systems have become widely cited since 2000, 1977–2018. 512 PIII.5 (a) Frequency of papers of ILRI and other institutions in global systems research, 1977–2018. (b) Frequency of papers of ILRI and other institutions in African systems research, 1977–2018. 513 15.1 Characteristics of global livestock systems by region, 1991–1993. 524 15.2 Characteristics of sub-Saharan Africa livestock systems by agroecology, 1991–1993. 526 15.3 Livestock production index for East and southern Africa (a) and West and Central Africa (b), 1970–2016. 528 15.4 TLU density for East and southern Africa (a) and West and Central Africa (b), 1970–2016. 529 15.5 Sheep and goat numbers relative to cattle numbers for East and southern Africa (a) and West and Central Africa (b), 1970–2016. 530 15.6 A rable land per capita for East and southern Africa (a) and West and Central Africa (b), 1970–2016. 532 15.7 Fertilizer use per hectare for East and southern Africa (a) and West and Central Africa (b), 2002–2016. 533 15.8 Cereal yields per hectare for East and southern Africa (a) and West and Central Africa (b), 1970–2016. 534 15.9 Food production index for East and southern Africa (a) and West and Central Africa (b), 1970–2016. 535 16.1 Livestock coverage in the ‘food security and food production systems’ chapter of AR5, Working Group II. 608 16.2 Global total GHG anthropogenic emissions and livestock’s share. 621 16.3 GHG emissions from global livestock, 1995–2005. 621 16.4 Mitigation potentials of supply-side measures. 624 16.5 Bovine meat (a) and milk (b) production and small-ruminant milk (c) and meat (d) by region. 625 16.6 Regional estimates of feed production for grains (a), grass (b), occasional feeds (c) and stover (d). 626 16.7 G lobal non-CO 2 GHG emissions from ruminant livestock. 627 16.8 Impacts of GHG taxes on food prices, consumption and GHG emissions. 629 xxvi List of Figures F.1 Aggregate meat consumption growth rates by period and region, 1971–2007, 2007–2030 and 2007–2050. 703 F.2 Aggregate milk consumption growth rates by period and region, 1971–2007 to 2050. 704 F.3 Aggregate meat production growth rates by period and product, 1971–2007 to 2050. 705 F.4 Animal numbers by period, region and species, 1961–1963, 2005–2007 and 2050. 706 F.5 G rowth rates of animal numbers by period, region and species, 1961–2007 to 2007–2050. 707 F.6 G rowth rates of animal carcass weights by period, region and species, 1962–2006 to 2006–2050. 708 F.7 Shares of growth rates by period, region and species, 1962–2006 and 2006–2050. 709 F.8 C ontributions of livestock to the 17 Sustainable Development Goals (SDGs). 711 List of Tables I.1 I LCA priorities around the time of its first formal strategy in 1987. 23 I.2 ILRI planned budget by CGIAR allocations, 1999–2003. 29 I.3 P rincipal achievements of ILRI research, 1975–2018. 32 1.1 ILRI-SLU training courses and trainees between 2000 and 2010. 90 2.1 E xperimental questions on AAT vaccine development and control. 112 2.2 T. brucei TL antigens induce a protective innate immune response. 117 2.3 A AT research at ILRAD in the post-trypanosomiasis vaccine era. 118 2.4 Characteristics of experimental cyclic T. congolense infections in N’Dama and Boran cattle. 125 3.1 R esearch on tsetse and trypanosomiasis control in West Africa and East Africa since 1990 by ILRI and partners. 150 4.1 The most highly cited studies on basic bovine immunology. 173 4.2 The most highly cited studies on the bovine immune responses to T. parva. 174 4.3 The most highly cited studies on the immune responses to African trypanosomiasis in the mouse model. 176 4.4 Estimated h-indices of scientists as a measure of research productivity and impact. 176 4.5 Comparisons of diseases for which a vaccine is sought. 177 4.6 Milestones in understanding the bovine immune system. 180 4.7 Immunity to T. parva-infected lymphocytes and identification of candidate vaccine antigens: timeline of substantial findings. 186 4.8 Timeline of substantial findings in immunology of theileriosis: humoral immunity. 187 4.9 Findings about the immunology of AAT in ruminants. 190 4.10 Diseases not targeted at ILRAD/ILRI but benefiting from tools and techniques and approaches developed there. 196 4.11 Answers to questions about impact of immunology research. 197 4.12 Assessment of immunology and immunoparasitology. 197 5.1 ECF risks by region, AEZ and grazing management. 214 6.1 Parameters used for deriving population trajectories in DynMod. 258 6.2 Production and net imports of selected commodities in Kenya, 2005–2026. 264 9.1 Food safety product lines. 347 9.2 ILRI capacity development in food safety, 2012–2015. 348 9.3 Food safety research led by ILRI. 349 9.4 Food safety initiatives to which ILRI contributed. 349 9.5 ILRI food safety interventions in informal markets. 355 xxviii List of Tables 11.1 Contrasts between equilibrium and non-equilibrium systems. 401 12.1 N umber of forage accessions conserved at ILRI, by region. 428 12.2 N umber of forage accessions conserved at CIAT, by region. 429 12.3 N umber of cultivated and wild accessions and taxa of forage germplasm conserved at ICARDA. 429 12.4 M ost distributed forage species from the CGIAR gene banks. 435 12.5 A nnual total cost per accession for four forage types, in current US$ (c.2010). 437 13.1 I ndicative yield and nutritive value of the forage species most commonly requested from the ILRI gene bank, 1984–2016. 455 13.2 D istribution of fodder grasses, legumes and trees in areas with dairy production in Ethiopia and Kenya. 460 13.3 T he most important fodder crops by households in Ethiopia and Kenya. 461 13.4 M ajor countries of recipients of forage seeds from CIAT and ILRI. 465 13.5 G eneral Circulation Model responses for the land areas between latitudes 30°N and 30°S. 469 13.6 H igh-potential sites for Napier grass as a cut-and-carry forage by region, period and climate model (km2). 470 14.1 Prices, relative value and fodder quality traits of cowpea haulm (CPH) and groundnut haulm (GNH) traded in Niger, Mali and Nigeria. 486 14.2 Average N content, IVOMD and prices (CFA franc) of cowpea haulm, groundnut haulm, sorghum stover and pearl millet stover traded over 2 years at four fodder markets in Niger. 486 14.3 Milk potential in Indian dairy buffalo fed two DTM feed blocks based on premium- quality (52% digestibility, 7.39 MJ ME/kg) and low-quality (47% digestibility, 6.52 MJ ME/kg) sorghum stover with total by-product proportion of feed blocks greater than 90%. 488 14.4 E ffect of groundnut and faba bean cultivars on live-weight changes in sheep fed exclusively with haulms offered ad libitum. 488 14.5 M eans of grain yields, CR yields, and CR N content and in vitro digestibility in groundnut and sorghum cultivars grown under water and control condition at Patancheru, India, in 2009 and 2010. 499 PIII.1 Results of the Altmetric search for ILRI livestock systems and related studies. 508 15.1 Selected pastoral systems studies, 1966–2013. 538 15.2 Productivity comparisons of ranching and pastoralism, various years. 544 15.3 Feed recommendations in systems studies, various years. 559 15.4 Mixed crop–livestock systems studies, various years. 568 16.1 Livestock farming system extent and cattle numbers in Africa and Latin America, 2000. 606 16.2 AR5 livestock impacts in the tropics. 609 16.3 Climate change knowledge gaps and research hypotheses. 614 17.1 Livestock policy research impacts by class and problem. 643 17.2 Selected applications of bioeconomic models, various years. 644 List of Boxes I.1 Estimating the economic burden of trypanosomiasis, 1975–2015. 8 I.2 Publications as a measure of scientific impact. 25 1.1 The taurine cattle of West Africa. 68 1.2 Mapping trypanotolerant QTLs in African cattle 74 1.3 Candidate genes involved in the response to T. congolense infection. 74 1.4 Open Nucleus Breeding System for sheep improvement in West Africa. 82 1.5 Horro chicken breed improvement programme in Ethiopia 86 1.6 Community-based sheep and goat breeding programmes. 86 1.7 Red Maasai and Dorper sheep breeding programmes. 87 1.8 Project to identify and deliver adapted chickens for productivity. 87 3.1 Community-based trypanosomiasis control. 156 8.1 A successful intervention to control zoonoses in South-east Asia. 311 9.1 Justification for incorporation of food safety research at ILRI. 343 9.2 Evaluation of a multi-country food safety project. 348 9.3 Smallholder dairy training and certification initiative. 352 16.1 Impact of the MarkSim model. 612 16.2 Impact of ILRI climate research. 616 List of Abbreviations A4NH CGIAR Research Program on Agriculture for Nutrition and Health AAT African animal trypanosomiasis (also called animal African trypanosomiasis) ACGG African Chicken Genetic Gains project ACIAR Australian Centre for International Agricultural Research ADF acid detergent fibre ADGG African Dairy Genetic Gains project ADL acid detergent lignin AEZ agroecological zones AFNETA Alley Farming Network for Tropical Africa AFRNET African Feeds Research Network Ag-ELISA antigen enzyme-linked immunosorbent assay AGTR Animal Genetics Training Resource, of ILRI AICRP All-India Coordinated Research Project on Sorghum AIDS acquired immunodeficiency syndrome ALPAN African Livestock Policy Analysis Network, of ILCA AMMA African Monsoon Multidisciplinary Analysis AnGR animal genetic resources APHISA Animal and Plant Health Information System for Asia AR4 Fourth Assessment Report of the Intergovernmental Panel on Climate Change AR5 Fifth Assessment Report of the Intergovernmental Panel on Climate Change ARC Africa Risk Capacity ASARECA Association for Strengthening Agricultural Research in Eastern and Central Africa ASEAN Association of Southeast Asian Nations ASF African swine fever ASF animal-source food ASFV African swine fever virus ATLN African Trypanotolerant Livestock Network, of ILCA and ILRAD AU-IBAR African Union–Interafrican Bureau for Animal Resources BBM broad-bed maker BBSRC Biotechnology and Biological Sciences Research Council, of the UK BecA Biosciences eastern and central Africa-International Livestock Research Institute Hub BiP binding immunoglobulin protein BMGF Bill & Melinda Gates Foundation xxxii List of Abbreviations BMZ Bundesministerium für wirtschaftliche Zusammenarbeit und Entwicklung (Federal Ministry for Economic Cooperation and Development, of Germany) BoLA bovine lymphocyte antigen BSE bovine spongiform encephalopathy BSL2 Biosafety Level 2 BTA Bos taurus chromosome BTV bluetongue virus CAAM CIMMYT-Asia Association Panel CAAS Chinese Academy of Agricultural Sciences CAFO confined (or concentrated) animal feeding operation CAPES Brazilian Federal Agency for Support and Evaluation of Graduated Education CBPP contagious bovine pleuropneumonia CCAFS CGIAR Research Program on Climate Change, Agriculture and Food Security CCPP contagious caprine pleuropneumonia CD cluster-defined antigen (also called cluster of differentiation) CDC Centers for Disease Control and Prevention, USA cDNA complementary DNA CFA franc Communauté Financière d’Afrique (Financial Community of Africa) CGE computable general equilibrium model CGIAR Consultative Group on International Agricultural Research (former name of CGIAR, which is no longer spelled out) CIAT Centro Internacional de Agricultura Tropical (International Center for Tropical Agriculture), of CGIAR CIDA Canadian International Development Agency CIMMYT Centro Internacional de Mejoramiento de Maíz y Trigo (International Maize and Wheat Improvement Center), of CGIAR CIRAD/EMVT Centre de Cooperation International en Recherche Agronomique pour le Développe- ment/Département d’Élevage et de Médecine Vétérinaire (French Agricultural Research Centre for International Development/Department of Livestock and Veter- inary Medicine) CIRDES Centre International de Recherche-Développement sur l’Elevage en zone Subhumide (International Center for Research and Development on Livestock Production in the Subhumid Zone) CISA Centre for Animal Health Research CK2 casein kinase 2 COCTU Coordinating Office for Control of Trypanosomiasis in Uganda CORPOICA Corporacion Colombiana de Investigación Agropecuaria (Colombian Corporation for Agricultural Research) CP crude protein CP cysteine protease CPH cowpea haulm CRD cross-reacting determinant CRISPR/Cas clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9 CRP CGIAR Research Programme CSF classical swine fever CSIRO Commonwealth Scientific and Industrial Research Organisation, of Australia CSU Colorado State University CTL cytotoxic T-lymphocyte CTTDB Centre for Ticks and Tick-Borne Diseases of Malawi DAGRIS Domestic Animal Genetic Resources Information Systems, of ILRI List of Abbreviations xxxiii DALY disability-adjusted life year DANIDA Danish International Development Agency DARI Dryland Agricultural Research Institute, of Iran DECUMA Decisions under Conditions of Uncertainty by Modelled Agents DEFRA Department for Environment, Food and Rural Affairs DFID Department for International Development, of the UK DGBC dark-ground buffy coat DGEA Dairy Genetics East Africa project, of ILRI DIVA differentiation between infected and vaccinated animals DM dry matter DPRA Direction Provinciale des Ressources Animales (Provincial Directorate of Animal and Fisheries Resources), Burkina Faso DSSAT Decision-Support System for Agro-technology Transfer, of FAO DTMA Drought Tolerant Maize for Africa project, of CIMMYT DTMR densified total mixed ration DVS Department of Veterinary Services, Kenya EADD East Africa Dairy Development programme EARO Ethiopian Agricultural Research Organization EATTRO East African Tsetse and Trypanosomiasis Research and Reclamation Organisation EAVRO East African Veterinary Research Organization ECAPAPA Eastern and Central Africa Programme for Agricultural Policy Analysis ECF East Coast fever EDRSAIA Early Detection, Reporting and Surveillance for Avian Influenza in Africa project EHDV epizootic haemorrhagic disease virus EHNRI Ethiopian Health and Nutrition Research Institute EIAR Ethiopian Institute of Agricultural Research ELISA enzyme-linked immunosorbent assay EMBRAPA Empresa Brasileira de Pesquisa Agropecuária EPMR external programme and management review epoR erythropoietin receptor ESAG expression site associated gene ESI Essential Science Indicators database ESRI Environmental Systems Research Institute ETB Ethiopian birr EU European Union FACS fluorescence-activated cell sorter FAO Food and Agriculture Organization of the United Nations FAOSTAT Food and Agriculture Organization Corporate Statistical Database FAVA Federation of Asian Veterinary Associations FBD food-borne disease FEAST Feed Assessment Tool, of ILRI FERG Foodborne Disease Burden Epidemiology Reference Group, of WHO FLI Friedrich-Loeffler-Institut FMD foot-and-mouth disease FPU food production unit FS full-sibling FTAI fixed-time artificial insemination FU-Berlin Freie Universität Berlin (Free University of Berlin) G×E genotype–environment interaction GALVmed Global Alliance for Livestock Veterinary Medicines GARP Genetic Algorithm for Rule-set Prediction xxxiv List of Abbreviations GASFRA Global African Swine Fever Research Alliance GCM general circulation model GDP gross domestic product GENESYS The global online genetic resources information system that replaced SINGER (Systemwide Information Network for Genetic Resources) GFRA Global Foot and Mouth Research Alliance GFSP Global Food Safety Partnership GHG greenhouse gas GIS geographic information system GLOBIOM Global Biosphere Management Model, of IIASA GPI glycosylphosphatidylinositol GPRA Global PPR Research Alliance GREASE Gestion des Risques Emergents en Asie du Sud-Est (Emerging Risk Management in South-east Asia) GREP Global Rinderpest Eradication Programme GRID Global Resource Information Database, of UNEP GRU Genetic Resources Unit, of CIAT GS genomic selection GWAS genome-wide association study HAT human African trypanosomiasis HDL high-density lipoprotein HIV human immunodeficiency virus HLPE High-Level Panel of Experts HPAI highly pathogenic avian influenza HPS haemophagocytic syndrome Hsp heat-shock protein IAASTD International Assessment of Agricultural Knowledge, Science and Technology for Development IADG Inter-Agency Donor Group IAEA International Atomic Energy Agency IAMA International Food and Agribusiness Management Association IARC international agricultural research centre IBLI Index-based Livestock Insurance project, of ILRI ICAR Indian Council for Agricultural Research ICARDA International Center for Agricultural Research in the Dry Areas, of CGIAR ICIPE International Centre of Insect Physiology and Ecology ICRAF International Centre for Research on Agroforestry, of CGIAR (now known as the World Agroforestry Centre) ICRISAT International Crops Research Institute for the Semi-Arid Tropics, of CGIAR ICT information and communications technology IDL intermediate-density lipoprotein IEMVT L’Institut d’Elevage et de Médecine Vétérinaire des pays Tropicaux (Institute of Livestock and Tropical Veterinary Medicine) IER Institut d’Economie Rurale, Mali IFAD International Fund for Agricultural Development IFDC International Fertilizer Development Center IFN interferon IFPRI International Food Policy Research Institute, of CGIAR IIASA International Institute for Applied Systems Analysis IIMR Indian Institute of Millets Research IITA International Institute of Tropical Agriculture, of CGIAR List of Abbreviations xxxv IL interleukin ILCA International Livestock Centre for Africa, of CGIAR (predecessor of ILRI) ILRAD International Laboratory for Research on Animal Diseases, of CGIAR (predecessor of ILRI) ILRI International Livestock Research Institute, of CGIAR IMPACT International Model for Policy Analysis of Agricultural Commodities and Trade INERA Institut de l’Environnement et Recherches Agricoles, Burkina Faso IPCC Intergovernmental Panel on Climate Change IPMS Improving Productivity and Market Success IRR internal rate of return IRRI International Rice Research Institute, of CGIAR ISVEE International Symposium on Veterinary Epidemiology and Economics ITC International Trypanotolerance Centre ITLN International Trypanotolerance Improvement Network ITM infection-and-treatment method of immunization ITPGRFA International Treaty on Plant Genetic Resources for Food and Agriculture IVEP in vitro embryo production IVOMD in vitro organic matter digestibility KALRO Kenya Agricultural and Livestock Research Organization (formerly KARI) KARI Kenya Agricultural Research Institute (now KALRO) KCC Kenya Co-operative Creameries KDB Kenyan Dairy Board KEMRI Kenya Medical Research Institute KETRI Kenya Trypanosomiasis Research Institute KEVEVAPI Kenya Veterinary Vaccines Production Institute KLIP Kenya Livestock Insurance Project LAMP loop‐mediated isothermal amplification LCIRAH Leverhulme Centre for Integrative Research on Agriculture and Health LCV Laboratoire Central Vétérinaire, Mali LDL low-density lipoprotein LGA livestock/grazing/arid LGACC Local Governance and Adaptation to Climate Change LGH livestock/grazing/humid LGP length of growing period LGT livestock/grazing/temperate and tropical LOD score logarithm of the odds (to the base 10) LRA livestock/rainfed/arid LRH livestock/rainfed/humid LRT livestock/rainfed/temperate and tropical LSR livestock systems research LUCID Land Use Change, Impacts and Dynamics mAb monoclonal antibody MAPK mitogen-activated protein kinase MarkSim rainfall simulation software Mccp Mycoplasma capricolum subsp. capripneumoniae ME metabolizable energy MERS Middle East respiratory syndrome MERS-CoV Middle East respiratory syndrome coronavirus MHC major histocompatibility complex MIA mixed/irrigated/arid and semi-arid MIH mixed/irrigated/humid and subhumid xxxvi List of Abbreviations MIT mixed/irrigated/temperate and tropical Mmc Mycoplasma mycoides subsp. capri Mmm Mycoplasma mycoides subsp. mycoides MRA mixed/rainfed/arid and semi-arid MRH mixed/rainfed/humid and subhumid mRNA messenger RNA MRT mixed/rainfed/temperate and tropical MT metric tonne NAICP Nigerian Avian Influenza Control and Human Pandemic Preparedness and Response Project NARES national agricultural research and extension systems NARS national agricultural research systems NASA National Aeronautics and Space Administration, of the USA ND Newcastle disease NDF neutral detergent fibre NDVI normalized difference vegetation index NIH National Institutes of Health, USA NIRS near-infrared spectroscopy NK natural killer cells NRCS National Research Centre for Sorghum, of India OAU Organisation of African Unity ODA official development assistance ODAP oxalyldiaminopropionic acid ODI Overseas Development Institute OECD Organisation for Economic Co-operation and Development OFAGE orthogonal-field-alternation gel electrophoresis OIE Office International des Epizooties (World Organisation for Animal Health) OPV open-pollinated variety PAAT Programme Against African Trypanosomiasis PANESA Pasture Network for Eastern and Southern Africa PANVAC Pan African Veterinary Vaccine Centre PARC Pan African Rinderpest Campaign, of the African Union—Interafrican Bureau for Animal Resources PBL peripheral blood lymphocyte PBMC peripheral blood mononuclear cell PCR polymerase chain reaction PCV packed cell volume PDSR Participatory Disease Surveillance and Response programme, of FAO and Indonesia PENAPH Participatory Epidemiology Network for Animal and Public Health PHEWS Pastoral Household and Economic Welfare Simulator PIM CGIAR Research Program on Policies, Institutions and Markets PLC phospholipase C PPPS Projet Productivité Primaire au Sahel PPR peste des petits ruminants QQR quinquennial review, by CGIAR QTL quantitative trait locus RABAOC Réseau de Recherche en Alimentation du Bétail en Afrique Occidentale et Centrale RESCAO Réseau d’épidémiosurveillance de la résistance aux trypanocides et aux acaricides en Afrique de l’Ouest (Network of Epidemiomonitoring of Trypanocidal and Mitidicidal Chemoresistance in West Africa) RHoMIS Rural Household Multi-Indicator Survey List of Abbreviations xxxvii RIEPT Red Internacional de Evaluación de Pastos Tropicales (International Tropical Pastures Evaluation Network), of CIAT RVF Rift Valley fever RVFV Rift Valley fever virus SA/IA System Analysis and Impact Assessment group SADC Southern Africa Development Community SAM social accounting matrix SAMPLES Standard Assessment of Agricultural Mitigation Potential and Livelihoods SARS severe acute respiratory syndrome SEACFMD South-East Asia and China Foot and Mouth Disease campaign SEAFMD RCU South East Asia Foot and Mouth Disease Regional Coordination Unit SEAFRAD South East Asian Forage and Feed Resources Network SEARG Southern and Eastern Rabies Group SMTA Standard Material Transfer Agreement SNP single nucleotide polymorphism SPIA Standing Panel on Impact Assessment, of ISPC SPS sanitary and phytosanitary SRES Special Report on Emissions Scenarios SRR-SEA Sub-Regional Representation for South-East Asia STI Swiss Tropical Institute TAC Technical Advisory Committee (superseded by ISPC of CGIAR) TAD transboundary animal disease TB tuberculosis TbTfR Trypanosoma brucei transferrin receptor TCR T-cell antigen-specific receptor TD T-cell-dependent TFP total factor productivity TI-2 T-cell-independent type 2 Tir trypanosome immune response TL Thymus–leukaemia antigen TLU tropical livestock unit TNF tumour necrosis factor Tregs regulatory T cells TRIF TIR-domain-containing adapter-inducing interferon/β UNDP United Nations Development Programme UNEP United Nations Environment Programme UNFCCC United Nations Framework Convention on Climate Change USAID US Agency for International Development USDA US Department of Agriculture VMRD Veterinary Medical Research and Development Centre, of the USA VRL Veterinary Research Laboratory VSF-G Vétérinaires sans Frontières-Germany (Veterinarians without Borders-Germany) VSG variable surface glycoprotein WC1 Workshop Cluster 1 WELI Women’s Empowerment Livestock Index WFP World Food Programme WHO World Health Organization WRL-FMD World Reference Laboratory for Foot-and-Mouth Disease List of Contributors Silvia Alonso, International Livestock Research Institute, Addis Ababa, Ethiopia. Ahmed Amri, International Center for Agricultural Research in the Dry Areas, Rabat, Morocco. Augustine Ayantunde, International Livestock Research Institute, Ouagadougou, Burkina Faso. Cynthia L. Baldwin, University of Massachusetts at Amherst, Massachusetts, USA. Isabelle Baltenweck, International Livestock Research Institute, Nairobi, Kenya. Bernard Bett, International Livestock Research Institute, Nairobi, Kenya. Samuel J. Black, University of Massachusetts at Amherst, Massachusetts, USA. Michael Blümmel, International Livestock Research Institute, Addis Ababa, Ethiopia. Dr Michael Blümmel died in Germany on October 12, 2020. Caroline Bosire, International Livestock Research Institute, Nairobi, Kenya. Elizabeth Cook, International Livestock Research Institute, Nairobi, Kenya. Nicoline de Haan, International Livestock Research Institute, Nairobi, Kenya. Tadelle Dessie, International Livestock Research Institute, Addis Ababa, Ethiopia. Michel Dione, International Livestock Research Institute, Dakar, Senegal. Paula Dominguez-Salas, International Livestock Research Institute Nairobi, Kenya; London School of Hygiene and Tropical Medicine, UK. Alan J. Duncan, International Livestock Research Institute, Addis Ababa, Ethiopia, and Global Academy of Agriculture and Food Security, University of Edinburgh, UK. Simeon Ehui, World Bank, Washington, DC, USA. Dolapo Enahoro, International Livestock Research Institute, Accra, Ghana. Polly Ericksen, International Livestock Research Institute, Nairobi, Kenya. Eric Fèvre, International Livestock Research Institute, Nairobi, Kenya; University of Liverpool, UK. Alessandra Galiè, International Livestock Research Institute, Nairobi, Kenya. Delia Grace, International Livestock Research Institute, Nairobi, Kenya and University of Green- wich, UK. Olivier Hanotte, International Livestock Research Institute, Addis Ababa, Ethiopia and School of Life Science, University of Nottingham, Nottingham, UK. Jean Hanson, International Livestock Research Institute, Addis Ababa, Ethiopia. Pierre Hiernaux, Caylus, France. Catherine Hill, Vancouver, Canada. Mohammad Jabbar, Dhaka, Bangladesh. Pasupuleti Janila, International Crops Research Institute for the Semi-Arid Tropics, Patancheru, India. Chris Jones, International Livestock Research Institute, Nairobi, Kenya. Steve Kemp, International Livestock Research Institute, Nairobi, Kenya. xl List of Contributors Henry Kiara, International Livestock Research Institute, Nairobi, Kenya. Hu Suk Lee, International Livestock Research Institute, Hanoi, Vietnam. Anne Liljander, EUROIMMUN Medical Laboratory Diagnostics AG, Lübeck, Germany. Johanna Lindahl, International Livestock Research Institute, Hanoi, Vietnam, and Uppsala Uni- versity, Sweden. Susan MacMillan, International Livestock Research Institute, Nairobi, Kenya. Jeff Mariner, Cummings School of Veterinary Medicine at Tufts University, Grafton, Massachusetts, USA. John McIntire, Santa Barbara, California, USA. Florence Mutua, International Livestock Research Institute, Nairobi, Kenya. Okeyo Mwai, International Livestock Research Institute, Nairobi, Kenya. Jan Naessens, De Haan, Belgium. Gerald C. Nelson, University of Illinois, Urbana-Champaign, Illinois, USA. Vish Nene, International Livestock Research Institute, Nairobi, Kenya. Hung Nguyen-Viet, International Livestock Research Institute, Hanoi, Vietnam. Nelly Njiru, International Livestock Research Institute, Nairobi, Kenya. Joel Ochieng, College of Agriculture & Veterinary Sciences, University of Nairobi, Kenya. Edward Okoth, International Livestock Research Institute, Nairobi, Kenya. Onesmo Ole-MoiYoi, International Centre of Insect Physiology and Ecology, Nairobi, Kenya. Ekta Patel, International Livestock Research Institute, Nairobi, Kenya. Brian Perry, University of Oxford, Oxford, UK; University of Edinburgh, Edinburgh, UK. Michael Peters, Alliance of Bioversity and CIAT, Nairobi, Kenya. Thomas Fitz Randolph, International Livestock Research Institute, Nairobi, Kenya. Ramana Reddy, International Livestock Research Institute, Patancheru, India. J.E.O. Rege, Emerge Centre for Innovations-Africa, Nairobi, Kenya. Karl M. Rich, International Livestock Research Institute, Dakar, Senegal. Lance Robinson, Ruwaza Sustainable Development, Stratford, Ontario, Canada. Tim Robinson, FAO, Rome, Italy. Kristina Roesel, International Livestock Research Institute, Frankenberg, Saxony, Germany. Anandan Samireddypalle, Indian Council of Agricultural Research, Bangalore, India. Rainer Schultze-Kraft, Alliance of Bioversity and CIAT, Cali, Colombia. Ali Shehadeh, International Center for Agricultural Research in the Dry Areas, Terbol, Lebanon. Jason Sircely, International Livestock Research Institute, Addis Ababa, Ethiopia James Smith, University of Edinburgh, Edinburgh, UK. Jimmy Smith, International Livestock Research Institute, Nairobi, Kenya. Steve Staal, International Livestock Research Institute, Nairobi, Kenya. Lucilla Steinaa, International Livestock Research Institute, Nairobi, Kenya. Nils Teufel, International Livestock Research Institute, Nairobi, Kenya. Lian Thomas, International Livestock Research Institute, Nairobi, Kenya. Philip K. Thornton, CGIAR Research Programme on Climate Change, Agriculture and Food Security (CCAFS), International Livestock Research Institute, Nairobi, Kenya; University of Edin- burgh, Edinburgh, UK. Philip Toye, International Livestock Research Institute, Nairobi, Kenya. Fred Unger, International Livestock Research Institute, Hanoi, Vietnam. Vincent Vadez, International Crops Research Institute for the Semi-Arid Tropics, Patancheru, India. Barbara Wieland, International Livestock Research Institute, Addis Ababa, Ethiopia. Peter Willadsen, Brisbane, Australia. Peter Wenzl, Alliance of Bioversity and CIAT, Cali, Colombia. Iain Wright, International Livestock Research Institute, Nairobi, Kenya. Mariana Yazbek, International Center for Agricultural Research in the Dry Areas, Terbol, Lebanon. Perez Haider Zaidi, Centro Internacional de Mejoramiento de Maíz y Trigo, Patancheru, India. Email addresses and other information about contributors may be found at http://www.ilri.org/ dataportal/impact/contributors Introduction: The Evolution of IARC Livestock Research, 1975–2018 John McIntire1 and James Smith2 1Santa Barbara, California, USA; 2University of Edinburgh, UK Contents Introduction 2 The Research Challenge of African Livestock Systems in the mid-1970s 3 Land use and rural population 3 The growth potential of livestock 3 The problem of trypanosomiasis 5 Areas infested 6 Numbers of animals affected 8 Productivity effects 8 Economic benefits of control 9 Origins of the International Livestock Research Institutions 10 The influence of the Green Revolution 10 The role of the Rockefeller Foundation 11 The Beck and Tribe reports 11 The Beck report 12 The Tribe report 12 Establishment of ILRAD 13 Establishment of ILCA 13 The International Laboratory for Research on Animal Diseases 13 Priorities 13 Approach to trypanosomiasis control at ILRAD’s founding 15 Approach to East Coast fever control at ILRAD’s founding 16 Resources 16 Institutional evolution 17 Achievements 18 The International Livestock Centre for Africa 20 Priorities 20 Resources 21 Institutional evolution 22 Achievements 23 Merger of ILRAD and ILCA into ILRI, 1995 27 New priorities at the merger 28 Species mandate 29 Scientific priorities 29 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 1 2 J. McIntire and J.Smith Resources 29 Reorientation of ILRI after 2000 30 The reform of CGIAR 30 Achievements, 1975–2018 31 Animal genetics, production and health 31 Trypanosomiasis and trypanotolerance 31 Theileriosis 36 Achievements in other fields of animal genetics and health 36 Primary production 37 Livestock systems, the global environment and gender 38 Livestock and the global environment 38 Economics and policy 38 Impact Analysis 39 Limitations of Citation Indices 39 Construction of Search Expressions 40 References 41 Introduction (ICARDA) and other international institutions2. In describing that history, it discusses their re- The international community invested more search priorities, budgets, institutional evolu- than US$1.8 billion in global livestock research tion and achievements from 1975 to 2015. It from 1975 to 20181. Most of this investment then frames the thematic parts of the book, has been publicly financed in one institution – which evaluate ILRI’s scientific and develop- what is now the International Livestock Research ment impacts. Institute (ILRI) – and most of that investment The remaining parts of the book cover the has been in sub-Saharan Africa. The impact of major themes of ILRI’s work: ILRI research is therefore an important subject, Part I: Animal Genetics, Production and Health: given the size of the investments, the effects of livestock production and consumption on income, • Preface to Part I: Research Spending and wealth, the environment and health, both Publications on Animal Genetics, Produc- human and animal, and the potential benefits of tion and Health research for production costs, consumer welfare • Chapter 1: Livestock genetics and breeding and the environment. • Chapter 2: Control of pathogenesis in African This introduction traces the creation, evo- animal trypanosomiasis: a search for answers lution and achievements of ILRI and its prede- at ILRAD, ILCA and ILRI, 1975–2018 cessors as background to the thematic chapters • Chapter 3: Tsetse and trypanosomiasis con- that explore impacts in specific scientific fields. trol in West Africa, Uganda and Ethiopia: The chapter begins by introducing the scale of ILRI’s role in the field the livestock research problem in sub-Saharan • Chapter 4: Impact assessment of immun- Africa in the mid-1970s at the time of the creation ology and immunoparasitology research at of ILRI’s predecessors, the International Livestock ILRAD and ILRI Centre for Africa (ILCA) and the International • Chapter 5: Veterinary epidemiology at Laboratory for Research on Animal Diseases ILRAD and ILRI, 1987–2018 (ILRAD) and the subsequent changes in demog- • Chapter 6: The management and economics raphy, land use and input use as they affected ru- of East Coast fever minant livestock production and productivity. • Chapter 7: Transboundary animal diseases The chapter then describes the history of the • Chapter 8: Zoonoses international livestock research institutions in sub- • Chapter 9: Food safety and nutrition Saharan Africa, focusing on ILRI (1995–present), • Chapter 10: Ticks and their control ILCA (1974–1994) and ILRAD (1973–1994), Part II: Primary Production: with some reference to the Centro Internacional de Agricultura Tropical (CIAT), International • Preface to Part II: Research Spending and Center for Agriculture Research in the Dry Areas Publications on Primary Production The Evolution of IARC Livestock Research: 1975–2018 3 • Chapter 11: Rangeland ecology There were some 22.4 million km2 of land in • Chapter 12: Forage diversity, conservation sub-Saharan Africa in the mid-1970s, of which and use 37% was classed as ‘arid’, 18% as semi-arid, 22% • Chapter 13: The impact of CGIAR centre as subhumid, 19% as humid and the balance of research on use of planted forages by trop- 4% as highland. The classification variable is the ical smallholders growing period for the zones, as indicated by • Chapter 14: Multidimensional crop improve- aridity, except for the highlands. Arid zones ment by ILRI and partners: drivers, ap- have rainfall of less than 200 mm annually (with proaches, achievements and impact less than 90 days suitable for crop production); semi- arid, 200–600 mm (90–179 growing days); Part III: Tropical Livestock Systems: subhumid, 600–1200  mm (180–269 growing • Preface to Part III: Research Spending and days); and humid, 1000–2000 mm (at least 270 Publications on Tropical Livestock Systems growing days). The highlands are considered as • Chapter 15: African livestock systems being above 1500 m above sea level; their rain- research, 1975–2018 fall would be in the same broad range as that of the • Chapter 16: Ruminant livestock and climate subhumid zone, but colder temperatures would re- change in the tropics duce growing periods at some points during the • Chapter 17: Economics and policy research year and Jahnke (p. 152) defines them as having at ILRI, 1975–2018 ‘a mean daily temperature of less than 20°C dur- • Chapter 18: The impact of ILRI research on ing the growing period’. The arid and highlands gender zones are alike in that their tsetse challenges are • Conclusion: The Future of Research at ILRI usually ‘low’ or ‘very low’ (see Fig. I.4b,c). More detailed estimates of livestock systems Map 1 (p. XVII) shows the principal African scale were incomplete until the path-breaking research sites of ILCA, ILRAD and ILRI from work of Seré et al. (1996), which became the 1975 to the present. base of later classifications. Subsequent work in- cluded that of Delgado et al. (1999) and the land The Research Challenge of African cover maps by Otte and Chilonda (2002) and Livestock Systems in the mid-1970s Kruska et al. (2003), while Thornton (2010) and Robinson et al. (2011, pp. 8, 11, 38–39 and This part of the introduction reviews the status of 145–152) refined the Seré and Steinfeld typ- the main African ruminant livestock systems dur- ology using remote sensing data, national sur- ing the decade of the founding of ILRAD and ILCA veys and an expert consultation (Map 1). While – land use, human population, ruminant animal the Seré et al. (1996) classification included numbers, animal productivity, and the threats of non-ruminants, we concentrate only on rumin- trypanosomiasis and theileriosis. Jahnke’s land- ants here. mark book (Jahnke, 1982) summarized the basic characteristics of African ruminant livestock pro- The growth potential of livestock duction at the founding of ILCA and ILRAD. From that work and other published material from that era, Figure I.2 shows tropical livestock units (TLUs) and we can summarize the scale of the African livestock ruminant numbers by agroecological zone. The research challenge as it existed in the mid-1970s. arid and semi-arid zones, dominant in Western and Eastern Africa, had the highest numbers of ru- Land use and rural population minants and TLU while having the lowest stocking rates, the smallest rural populations and the lowest Figure I.1 shows, as of the mid-1970s, the basic rural population density. Their historical potential relationships among land use, human population depended largely on growth of TLU numbers, an and population density, using Jahnke’s system idea that has since been confirmed by the long-term classification which was based on that of the Food experience with animal numbers and productivity and Agriculture Organization of the United per animal (see this volume, Chapter 15). Nations (FAO) (Higgins and Kassam, 1981), and The growth potential of the highlands – which used data compiled from the mid-1970s. found almost entirely in eastern Africa and having 4 J. McIntire and J.Smith (a) (b) 8 60 6 40 4 20 2 0 0 Arid Semi-arid Subhumid Humid Highlands Arid Semi-arid Subhumid Humid Highlands Agroecological zone Agroecological zone (c) (d) 50 40 20 30 20 10 10 0 0 Arid Semi-arid Subhumid Humid Highlands Arid Semi-arid Subhumid Humid Highlands Agroecological zone Agroecological zone Fig. I.1. (a) Area of sub-Saharan Africa by agroecological zone, mid-1970s. (b) Rural population by agroecological zone, mid-1970s. (c) Zonal rural population shares, mid-1970s. (d) Zonal rural popula- tion densities, mid-1970s. Arid zones have rainfall of less than 200 mm annually, with (< 90 days suitable for crop production); semi-arid, 200-600 mm (90-179 growing days); sub-humid, 600-1200 mm (180-269 growing days); and humid, 1000-2000 mm (>= 270 growing days). Jahnke (1982, p. 152) defines ‘tropical highlands’ as having ‘a mean daily temperature of less than 20 C during the growing period.’ (Data from Jahnke, 1982: Annex tables 1-11). the highest TLU and human population densities the share of livestock in agricultural GDP. Given – would depend almost entirely on a shift from the abundance of land in most of sub-S aharan meat animals to dairying and on rising productiv- Africa, it was thought in the mid-1970s that the ity per animal in meat and in dairy given the exist- livestock growth potential of the humid and ing land constraints at the time. Trypanosomiasis subhumid zones would follow an extensive path was a production risk on less than 5% of the given that those zones had lower population dens- highlands area and theileriosis was unknown in ity and lower TLU density. Exploiting that growth Ethiopia, which was the most populous highland potential therefore depended on the control or even nation. elimination of trypanosomiasis, which affected Figure I.3 shows Jahnke’s mid-1970s estimates one-third or more of the area of each. At the time of agricultural GDP in sub-Saharan Africa and of of the founding of ILCA and of ILRAD, and of the Percentages of total national population Area in millions of km2 Rural population density in km2 Rural population in millions The Evolution of IARC Livestock Research: 1975–2018 5 (a) (b) 25 8 20 6 15 4 10 2 5 0 0 Arid Semi-arid Subhumid Humid Highlands Arid Semi-arid Subhumid Humid Highlands Agroecological zone Agroecological zone (c) (d) 120 30 90 20 60 10 30 0 0 Arid Semi-arid Subhumid Humid Highlands Arid Semi-arid Subhumid Humid Highlands Agroecological zone Agroecological zone Fig. I.2. (a) Numbers of tropical livestock units by agroecological zone, mid-1970s. (b) Tropical livestock unit density by agroecological zone, mid-1970s. (c) Ruminant livestock numbers by agroecological zone, mid-1970s. (d) Zonal ruminant populations as percentages of total, mid-1970s. (Data from Jahnke, 1982: Annex tables 1-11.) initial entry of the International Crops Research was achievable under certain assumptions about Institute for the Semi-A rid Tropics (ICRISAT) into management and its effects on calving rate, calf African research, the climate costs of land clearing survival and adult mortality (Dahl and Hjort, for agriculture were not well understood. 1976, and the discussion therein on pp. 66–75). Jahnke’s summary (Jahnke, 1982, p. 35) of Significant increases in herd growth, from exist- the (extremely limited) data on growth in livestock ing breeds under African conditions, depended numbers in the 1960s and 1970s indicated that on levels of management, market access and the principal ruminants, plus camels, were growing sequences of good rainfall years that would have at about 1–2% annually. The weak growth in Af- been uncommon at the time. rica was attributed entirely to an increase in TLUs, A summary of historical growth (Anteneh, not to higher yield of meat or milk per animal 1984) after almost a decade of ILCA’s work (Jahnke, 1982, pp. 42–43). An early simulation of concluded that growth rates of the numbers of East African cattle herds argued that an an- livestock units were about 1.2% annually in nual growth rate of female cows of about 4% sub-Saharan Africa during the 1970s, a rate Head in millions TLU in millions Perccentages of total ruminant population TLU density in number km2 6 J. McIntire and J.Smith (a) 8,000 6,000 4,000 2,000 0 Arid Semi-arid Subhumid Humid Highlands (b) Agroecological zone 50 40 30 20 10 0 Arid Semi-arid Subhumid Humid Highlands Agroecological zone Fig. I.3. Agricultural gross domestic product (a) and livestock share of agricultural GDP (b) by agroecologi- cal zone, mid-1970s. (Data from Jahnke, 1982: Annex tables 1-11.) lower than that of population growth rates; only trypanosomiasis control and admit (pp. 810– in beef was there any sign of higher productivity 811) that ‘…the tsetse problem is too general per animal. Evidence over the period 1970– or our knowledge of the ecology of the flies is 2016 (see Chapter 15, this volume) confirmed still too limited, to enable an economic form of the finding that livestock production growth was control or extermination to be devised which a function of herd size and that productivity could be justified by the productive capacity of growth per animal was weak. the reclaimed area’. The potential benefits to tsetse and trypanosomiasis control had been coarsely estimated by Jahnke (1982), using The problem of trypanosomiasis data from the mid-1970s (Fig. I.4). Beyond the evident facts that trypanosomiasis threatened The biology of trypanosomiasis and the tsetse a much smaller share of the arid, semi-arid vector had been well-studied (Mulligan and and highlands zones compared with the wetter Potts, 1970) before the founding of ILRAD. ones, there was a lack of information on losses Mulligan and Potts (1970), however, say prac- to this disease and on potential benefits from tically nothing about economic aspects of various control methods. Livestock GDP as percentage of agricultural GDP Millions of current US$ The Evolution of IARC Livestock Research: 1975–2018 7 (a) (b) 3,000 75 2,000 50 1,000 25 0 0 Arid Semi-arid Subhumid Humid Highlands Arid Semi-arid Subhumid Humid Highlands Agroecological zone Agroecological zone (c) (d) Area by challenge level in km2 6,000 5.0 4,000 2,000 4.0 0 3.0 TLU by challenge level in numbers 25,000 2.0 20,000 15,000 10,000 1.0 5,000 0 0.0 High Medium Low Very Low Arid Semi-arid Subhumid Humid Highlands Tsetse challenge level Agroecological zone Fig. I.4. (a) Area of tsetse fly infestation by agroecological zone. (b) Zonal area shares infested by tsetse. (c) Ruminant TLU and area by tsetse challenge, mid-1970s. (d) Potential zonal gross benefits from tsetse clearance and trypanosomiasis control, mid-1970s. (Data from Jahnke, 1982: Annex tables 1-11.) Much has been achieved in understanding Areas infested the control of trypanosomiasis in sub-S aharan Africa since 1970. By the turn of the century it The work of Ford and Katondo (1977) is the ref- was understood, partly through the efforts of erence point for modern estimates of infested ILCA/ILRAD/ILRI, that the economic benefits of areas. Jahnke (1982) declared, based on the control would be large. Box I.1 highlights as- work of Ford and Katondo (1977), the affected pects of the economic burden of trypanosomia- areas to be roughly 10.3 million km2 with data sis from the mid-1970s to the vaccine modelling from the mid-1970s, of which 7.0 million km2 done around the year 2000 to the most recent were said to be in the more seriously infested estimates of economic benefits from control. humid and subhumid zones. An approximation Rural area in 000s km2 Rural area in 000s km2 Potential benefits as percentages of zonal livestock GDP Percentages of rural area Infested 8 J. McIntire and J.Smith to 10 million km2 became a standard figure in humid zones. Kristjanson et al. (1999, p. 88), the past century. A study by Katondo (1984) using data up to the mid-1990s, assumed tsetse- gave revised data from ten countries but did not infested areas of 8.7 million km2 and approxi- report a total infested area and concluded that mately 48 million cattle in those areas, or about ‘a programme of intensive revision is needed 32% of the cattle in sub-Saharan Africa, in a before it can be claimed that tsetse situation in model of the ex ante return to a (hypothetical) vac- Africa is well understood’. An estimate included cine against trypanosomiasis. Reid et al. (2000, in the first published collection of the ILCA/ p.  231) did not predict animal numbers in their ILRAD African Trypanotolerant Livestock work, but they did argue that ‘5–17 million people’ Network (ATLN) gave the infested areas as 9.5 would still live ‘in fly-infested areas’ by the year million km2, based on updated information since 2040. Gradual cross-breeding is occurring be- the 1970s (Jahnke et al., 1988). tween Bos indicus (Zebu) cattle types and Bos taurus Wint and Rogers (2000), using some of the (trypanotolerant cattle, such as the N’Dama) in original Ford and Katondo (1977) data, devel- West and Central Africa and this may reduce over- oped a prediction function and prepared detailed all susceptibility to trypanosomiasis, although the maps of infestation by species but did not state productivity effects of such introgression are not different total areas. Rogers and Robinson (2004) well quantified (Traoré et al., 2017). used satellite imagery to arrive at a continental estimate of 8.7 million km2. Shaw (2004) sum- marized turn-of-the-century estimates as being in Productivity effects the range of ‘8–10 million km2’. Reid et al. (2000) used a sophisticated model The productivity effects of trypanosomiasis were: of human population density, land use and tsetse (i) the inability to raise livestock in heavily infested species to project the impact of human popula- areas; (ii) the substitution of less productive tion growth on the density and species compos- trypanotolerant animals for more productive ition of tsetse infestation to the year 2040. They trypanosusceptible stock in infested areas; (iii) the estimated the land areas with ‘high populations inability of trypanotolerant animals to reach of savannah and forest flies’ to be between 6 mil- their genetic potential under tsetse challenge; lion and 7 million km2 in 2000. Reid et al. (2000) (iv) losses owing to increased mortality and mor- forecast the area ‘infested with riverine flies’ to bidity in infected animals; and (v) the costs of remain ‘nearly 7 million’ km2 in 2040. They sprays, drugs, traps and other resources needed projected wide changes in species composition, to maintain livestock under challenge. Older es- with ‘[s]avanna and forest flies’ to be in general timates of these effects were even less reliable decline, with no riverine types in southern Africa, than estimates of infested areas because the data and with the potential for human population requirements of productivity studies were stricter; growth to have eradicated the fly in some places. it remains impossible, even using modern survey Reid et al. (2000, p. 232) predicted an infested tools, to calculate a reliable total effect and to area of the order of 5–5.5 million km2 by the partition it into the five components. A coarse year 2020, but there is no current appraisal of estimate from Jahnke’s mid-1970s data (Fig. I.4d) the accuracy of that prediction. was made assuming that tsetse and trypanosom- iasis control in the arid, semi-arid, subhumid and humid zones allowed TLU density gains per zone of 10%, 10%, 25% and 25%, respectively, with Numbers of animals affected gains in productivity per animal of 5% in each zone. These modest gains in stocking rates and Jahnke (1982) estimated numbers of ruminants in yield per animal would, hypothetically, in- affected by trypanosomiasis by mapping animal crease livestock GDP by zone in the range of inventories to the level of tsetse challenge in the 1% annually (semi-arid and highland zones) to five agroecological zones of his study, as shown 4-5 % in the subhumid and humid zones. in Fig. I.4a–d. He reported approximately 148 One appraisal at the outset of the ATLN was million cattle of which perhaps 40 million were from Trail et al. (1979a, p. 91)3, who reviewed 30 in the more severely infested subhumid and studies of the performance of trypanotolerant The Evolution of IARC Livestock Research: 1975–2018 9 stock at varying levels of management under ex post calculations of benefit:cost ratios from v illage (n  =  5) and ranch/experiment station spraying from Nigeria and Cameroon to be well (n  =  25) conditions in West and Central Africa. above 1. Putt et al. (1980) reviewed the costs and Trail and colleagues found productivity effects per benefits of tsetse control in Nigeria and found herd to range from –20% to –50% at ‘low’, ‘me- benefit:cost ratios ranging from 2.7 to 8.0 for dium’ and ‘high’ trypanosomiasis risk compared selected ground and helicopter spraying oper- with situations of ‘zero’ risk. They noted that the ations. The dissertation of Itty (1992, p. 270) ‘zero risk’ levels were confounded with ‘very high analysed the biological results from the ATLN in levels of feeding and management’ and con- an economic model of herd productivity at seven cluded that the direct trypanosomiasis effect on humid and subhumid sites in sub-Saharan Africa. productivity per animal would have been over- He found economic rates of return varying from estimated. 15% (Muhaka, Kenya) to 53% (Ghibe, Ethiopia) The review by Swallow (2000, p. 7–12) for treatments involving sprays, traps, trypano- of 13 studies of trypanotolerant and mixed cides and trypanotolerant stock. Using data (trypanotolerant and trypanosusceptible) cattle from a cross-section of ten humid African ‘coun- concluded that: ‘The general implication is that tries completely infested by tsetse’, Swallow the incidence of trypanosomiasis: (i) reduces (2000, p. 35) constructed figures purporting to calving rates by 1–12% in tolerant breeds of cat- show that the impact of trypanosomiasis would tle and by 11–20% in susceptible breeds; and be to lower agricultural GDP by 8–16%. (ii) increases calf mortality by 0–10% in tolerant Shaw (2003) compared tsetse eradication breeds of cattle and by 10–20% in susceptible with trypanocides in a hypothetical subhumid breeds of cattle.’ situation in West Africa at a range of human Swallow further noted three herd studies population densities. She found that a population (in southern Burkina Faso; Ghibe, Ethiopia; and density above 50 or so persons/km2 reduced the northern Côte d’Ivoire) that have shown annual incidence of the fly to a low, manageable level. herd growth effects of control in a range of 1–3%. Population densities below 10–20 persons/km2 Kamuanga et al. (2001) reported a herd size gain made tsetse control impractical. In the intermedi- of some 25% in a study of control with trypa- ate zone of 20–50 persons/km2, a combination of notolerant stock, traps and pour-on treatments; trypanocides and tsetse elimination gave profit- the herd size effect was attributable to decreased able benefit:cost ratios. Box 3.1 in Chapter 3 high- mortality. Swallow’s broad conclusion was that lights successes and failures in community-based ‘trypanosomiasis reduces cattle population by control of trypanosomiasis using several methods 30–50%’ and productivity from affected cattle by and notes the sustainability problems of both vec- 50%. The model of Kristjanson et al. (1999, tor control and drug treatment. p. 84) of a hypothetical vaccine against trypano- somiasis achieved productivity gains through Box I.1. Aspects of the economic burden of steep declines in cattle mortality – and therefore, trypanosomiasis, 1975–2015 a commensurate rise in stocking rates and in commercial offtake – and with only modest gains At the founding of ILRAD and ILCA, it was gen- in live weight, calving rates and lactation lengths. erally understood that trypanosomiasis was a stout barrier to expanding agricultural productiv- ity in much of sub-Saharan Africa. However, beyond that general understanding, calculations Economic benefits of control of the costs of trypanosomiasis were sparse and unreliable. Variance in such calculations arose from unsolved problems in: (i) estimating the Jahnke’s book contained little that was specific areas infested by the tsetse fly; (ii) calculating on the economics of control except to say that numbers of animals affected; (iii) productivity ef- long-term control with drugs had been success- fects of trypanosomiasis; and (iv) benefits and ful on some African ranches (Jahnke,1982, p. 198). costs of control. Nearly half a century has Tacher et al. (1988, pp. 331–335) reviewed the passed since the founding of ILRAD and ILCA costs of traps, sprays and other methods in rep- and the same economic problems are now only partly resolved. resentative challenge situations and noted a few 10 J. McIntire and J.Smith Shaw’s compilation (2004) found benefit: a proportionate change in agricultural GDP, cost ratios as follows: (i) Ethiopia in 1999, ex post and in the regional models of Shaw et al. (2015, evaluation of a pour-on trial, 4.3–8.0; (ii) Central p. 207 for 0.7 m km2 in five countries of East African Republic in 1995, ex post evaluation of Africa), which allows product and input prices to trapping, 3.1–5.9; (iii) Côte d’Ivoire in 1994, mixed remain unaffected by large changes in livestock ex post and ex ante evaluation of trapping, 2.3–4.0; output over many years. The major microeco- (iv) Burkina Faso in 1988, various sterile insect nomic analysis of trypanotolerance (Itty, 1992) techniques over 10 years, 0.49–1.56; (v) Nigeria advocates a joint strategy of trypanotolerant in 1980, two districts, spraying with other local cattle with chemotherapy without considering control practices, 2.7–8.0; and (vi) Kenya, prophy- the land use effects of higher stocking density in laxis with isometamidium chloride, 6.0–52.3. areas of moderate trypanosome prevalence or Kristjanson et al. (1999) analysed the net po- without adequately discussing the evolution of tential benefits of a (hypothetical) vaccine against resistance in the vector. trypanosomiasis. That work integrated field data Except for Shaw’s work on East Africa, there from experimental sites of the ATLN, GIS data on is little in the literature of the decade after 2010 the continental distribution of tsetse challenge on the economic problems of controlling tsetse and affected animals, and an economic surplus and animal trypanosomiasis in Africa. The costs model to calculate the aggregate welfare effects of gathering and analysing time-series data on of a (hypothetical) vaccine. The study estimated the animal populations, disease incidence and the annual production and control costs of tryp- the effects of control methods have stopped nearly anosomiasis in sub-Saharan Africa to be of the all new research in this area. order of US$1.3 billion (in 2000 prices). Elimin- Box I.1 describes some problems in estimat- ating those costs with a vaccine adopted over 12 ing the economic burden of trypanosomiasis. years by a maximum of 30% of producers would Jahnke (1982) approximated the potential gross give a benefit:cost ratio of 34:1. A fatal defect of benefits to research from reducing trypano- this model is that it projected a vaccine adoption somiasis in cattle by: (i) projecting additional period of 12 years when experiments to develop numbers where tsetse challenge had been reduced a vaccine had already failed for more than modestly in each zone except the highlands, 25 years and hence it is highly unrealistic to im- where challenge was already low; and (ii) project- pute benefits, and the temporal pattern of bene- ing productivity gains per animal at low levels of fits, to such a hypothetical vaccine. management where tsetse challenge had been Recent work by Shaw et al. (2014, 2015) pre- reduced. As shown in Figure I.4(d), the largest sented benefit–cost models of trypanosomiasis gains relative to zonal GDP from livestock would control in eastern Africa – Uganda, Kenya, be achieved in the subhumid and humid zones Ethiopia, Somalia, Sudan and South Sudan – dif- even in the absence of a vaccine. This early calcu- ferentiated by farming system, population dens- lation did not consider the costs of control from ity, and control strategy (‘continuous control, and drugs, traps, sprays or other methods. The cost elimination’). Although there was no single con- of maintaining vector control operations over trol strategy applicable to all farming systems in time in less densely populated areas to prevent the sample, high benefit:cost ratios were more recolonization by the fly has meant, in practice, likely in densely cultivated mixed farming areas that the attractive benefit:cost ratios of vector in which oxen were used for power. Control was control have not been sustainable over long typically unprofitable in areas of low cattle dens- periods and this fact has long been used by advo- ity. Problems of programme management imply cates of vaccine development, despite the failure to that elimination of tseste through large-scale develop an effective vaccine. campaigns would be infeasible. A problem with some aggregate projections of trypanosomiasis control, by any method or Origins of the International Livestock combination of methods, is omission of general Research Institutions equilibrium effects. This is notable in the review of Swallow (2000), who postulates that a doub- The creation of the international livestock ling of trypanosomiasis control inputs gives research centres was due in part to the Green The Evolution of IARC Livestock Research: 1975–2018 11 Revolution in crops. It was believed that new scientific solutions to those problems. Green animal research could produce gains like those Revolution packages of improved seeds, fertilizer achieved in rice and wheat, and that combating and irrigation had a dramatic impact on agricul- diseases in African ruminant livestock offered ture. Between 1961 and 1985, cereal production the greatest immediate potential for such gains. in the Asian developing countries grew rapidly, The mandate of the new livestock institutions with the increase attributed to seeds, fertilizer did not include swine or poultry, which were and irrigation, each factor having a research then of less importance in most African systems component. or whose research needs could be met by tech- The Asian successes in crops had bypassed nology transfer4. sub-Saharan Africa up to about 1990 because inadequate infrastructure, expensive market access, unfavourable agro-climates, lack of irri- The influence of the Green Revolution gation and weak, unstable governments had blocked the spread of new technologies. Beyond The roots of the Green Revolution were in tech- the general diagnosis of why the Green Revolu- nical assistance programmes introduced into tion had not occurred in sub-Saharan Africa Latin America in the 1940s. In 1960, the Inter- was the belief that livestock were understudied, national Rice Research Institute (IRRI) was and that African animal disease and rangeland opened in the Philippines with a global mandate management problems were unusually difficult. to increase rice production to ‘solve this problem These considerations confirmed the view that a of rice’ (Anderson et al., 1991). The opening of new research approach to livestock and livestock IRRI together with that of the Mexico-based systems was needed in sub-Saharan Africa. CIMMYT (Centro Internacional de Mejoramiento de Maíz y Trigo) produced the improved cultivars and management innovations of the Green The role of the Rockefeller Foundation Revolution (Ruttan, 1977). The initial successes of rice and wheat led to The Rockefeller Foundation traced its global the creation of a network of international agri- agricultural influence to Dr Norman Borlaug’s cultural research centres (IARCs), under the um- wheat programme in Mexico in the 1940s. The brella of CGIAR.5 This network became the dom- Foundation believed that ‘international cooper- inant model of publicly funded and conducted ation in any field of science … would contribute agricultural research in the tropical climates. towards a common pattern of global living’ The rationale of CGIAR was that increased (Chandler, 1992, pp. 2–3). In addition to its role and stable funding of research with specific in CIMMYT, the Foundation had funded the commodity agendas could supplement the limited establishment of IRRI, and was a logical organ- capacity of developing-country national agricul- ization to advance agricultural research institutes tural research systems and produce public goods in sub-Saharan Africa. that might otherwise never have been produced. The Rockefeller Foundation was active in The IARCs were conceived as engines to drive the creation of ILRAD and ILCA, commissioning agricultural modernization and rural develop- a series of meetings and reports, and later sign- ment, using science: ing the initial memorandum of agreement with … two types of activity … make sense: first, the Government of Kenya for ILRAD, on behalf activities which explicitly face up to the complex of CGIAR. The Foundation further convened a and interrelated problems of ignorance and series of conferences between 1968 and 1970 tradition, and to seek to attack those problems; that led to the task force that set the basis for and second, isolable technical problems which are so important that their solution would find what became ILCA. acceptance and application even under present circumstances. Harrar et al. (1952, pp. 25–26). The Beck and Tribe reports Harrar’s6 perspective was influential both in defining agricultural problems in poor countries – Discussions of an African livestock institution ‘ignorance and tradition’ – and in proposing had begun as early as 19637. At that time, IRRI 12 J. McIntire and J.Smith had been established, CIMMYT was nascent and research stations’. The new centre would have complementary needs for livestock research two ‘thrusts’ – a ‘livestock production re- were recognized. In 1966, a paper in the Pro- search laboratory’ in West Africa, working at ceedings of the National Academy of Sciences USA sites in Niger and in Nigeria to cover a range noted the ‘need for the establishment at an early of agroecological conditions; and a proposed date of a major research centre located in the ‘International Laboratory for Animal Disease tropics devoted to tropical diseases of animals’ R esearch in Africa’ in East Africa with a ‘program (Pritchard et al., 1972, p. 368). A US Presiden- in ECF and trypanosomiasis’ (Beck, 1971, pp. 28–29). tial Panel on the World Food Supply recommended The new institution would have one board and the establishment of international centres for one Director General, with associate directors at research on animal production and diseases in the East Africa and West Africa sites. the humid and arid tropics, with ‘at least one The new centre would have the goals of centre devoted primarily to research on epizootic ‘improving the social and economic welfare of diseases’ (President’s Science Advisory Commit- pastoralists’ and of increasing the contribu- tee, 1967: Vol. 1, p. 93). Donors were reluctant tion of livestock to African GDP by inducing because of the potential high costs. technical change to increase meat production In 1968, a Rockefeller Foundation confer- per animal, lift offtake rates, optimize produc- ence in Bellagio on East African rangelands recom- tion systems and reduce losses to animal disease mended the establishment of an ‘International (Beck, 1971, p. 34). Range Land Institute’ (Longhurst and Heady, In light of these goals, an initial task of the 1968). In 1970, the Ford Foundation proposed new centre was declared as (Beck, 1971, p. 35): an ‘International Centre for Rangelands Research A study of livestock production systems of Africa and Development in Africa’, with the goal of is an area of high priority. New research is likely increasing ‘the yield of animal protein (meat to have most impact if planned within the context and milk) and other animal products from such of these production systems. An Analysis and rangelands to its maximum economic level, Planning Unit would be required from the while carefully monitoring all such programmes inception of the Centre for this purpose. This of management and development’ (Robin and would involve surveys and data analysis covering Brown, 1970). The ambition of this Ford Foun- sociology, range ecology, water resources, animal dation proposal appeared smaller than that from production, economics and marketing. the Rockefeller Foundation; the proposal re- The Beck report (Beck, 1971, p. 27) further assured readers that ‘with no capital spending defined the role of what became ILCA: and no capital assets beyond its office furniture and equipment, the Centre can be terminable at The most important function of the Centre would be to assemble a multidisciplinary team any time’ (Robin and Brown, 1970, p. 3). The of scientists to develop research programmes specific title of the Ford Foundation proposal – an designed to solve the basic production and ‘International Centre for Rangelands’ – revealed socio-economic problems that are serving as the bias of the international partners in the dec- constraints to livestock development. ade before the founding of ILCA and ILRAD towards animal production on extensive grasslands and Beck proposed specific research domains in against smallholder mixed farming. ‘range and pasture studies’, ‘animal production studies’ (including such problems as ‘causes of low reproductive performance and high calf The Beck report mortality’) and animal disease studies cover- In 1971, a Rockefeller Foundation-sponsored ing ‘parasitism, streptothricosis, and fascioliasis’ mission, led by Professor Glenn Beck8 and includ- (Beck, 1971, p. 38). ing scientists with wide African experience, wrote what became known as the Beck report. The Beck report explored the possibility of an The Tribe report ‘African Livestock Center’ and suggested one institution with a central headquarters in Kenya, More detail about a potential second institution working through a ‘network of associated African was required following the Beck report, and this The Evolution of IARC Livestock Research: 1975–2018 13 was provided by a mission led by Professor Derek p roduction systems. It was generally believed Tribe. The Tribe report (Tribe et al., 1973, p. 1) that success was more likely if ILRAD were to focused on systemic and multidisciplinary focus on one or two diseases, and on one prob- approaches in making its recommendations: lem, that of immunization. If ILCA were to come In particular, there is a need for more detailed into being as a centre to study livestock systems, study of animal production systems of tropical then ILRAD should logically focus on livestock Africa before existing knowledge can be fully disease. The evident synergy between animal utilized or future research priorities defined. This production and animal disease led the enabling work must give full consideration to those committee for the establishment of ILCA to con- aspects of biology, economics and social clude that ILCA and ILRAD should be integrated anthropology that relate to animal production. from the start, although this ultimately did not The Tribe mission advocated several activities – happen for 20 years. research, capacity building, advice at different That the centres were established at the levels and development of a knowledge base of same time, one seeking the laboratory solutions ‘all relevant information on animal production for two animal diseases and the other analys- in tropical Africa’ – in close ‘… cooperation with ing production systems, underlined two issues. existing institutions in Africa … Its essential role First was the prevailing optimism that science should be complementary, cooperative and cata- could make a fundamental contribution to lytic’ (Tribe et al., 1973, p. 28). l ivestock production and related agricultural The Tribe report gave a comprehensive problems in Africa. Second was the view that and ambitious overview of what ILCA would African livestock research structures were inad- become and recommended that ‘… an Inter- equate to solve problems of livestock disease national Centre for the development of a nimal and management. production in Africa should be established i mmediately’. Establishment of ILRAD The limited history of international agri- cultural research institutes made it difficult for The Rockefeller Foundation, on behalf of CGIAR, donors to reach a consensus on the decentral- began work with the Government of Kenya to ized, network model recommended by the Tribe establish ILRAD. Bruce MacKenzie, the then report. Discussions about a second livestock Minister of Agriculture of Kenya, co-drafted a institute in Africa therefore continued, partly letter with John F. McKelvey of the Rockefeller under the influence of an earlier view of the Foundation for President Jomo Kenyatta to send need for both an ‘ILRAD’ and a second insti- to Robert McNamara, the then President of tute, working on production (Pino, 1970). the World Bank. The letter recommended that By the beginning of 1973, there was a con- ILRAD be established in Kenya; this ultimately sensus among potential donors that Africa would allowed CGIAR to appoint an enabling commit- host two international centres of livestock re- tee, financed by the Rockefeller Foundation, for search. Initial prospects for merging them on the establishment of ILRAD in Kenya. An agree- one campus had not been realized – not neces- ment was signed in 1973 between the Govern- sarily for scientific reasons but rather due to pol- ment of Kenya and the Rockefeller Foundation itics and timing – and within a few years two (Gray, 1984). new campuses would exist, one in Nairobi and the other in Addis Ababa. Establishment of ILCA Despite the forcefulness of the Tribe report, the Rockefeller Foundation chose not to invest in ILCA was established by the signing of a memo- ILCA, considering it too risky, and ultimately in- randum of understanding between the Ethiopian vested in what became ILRAD. The Rockefeller Minister of Agriculture and CGIAR in February Foundation’s early experience was influential in 1973. The first Director General of ILCA, Jean creating two institutions because of its focus on Pagot, arrived in Ethiopia in September 1974 ‘isolable problems’ such as raising rice and wheat and construction began on the new centre yields rather than focusing on agricultural soon after. 14 J. McIntire and J.Smith The International Laboratory for effective vaccine against trypanosomiasis was Research on Animal Diseases ‘conservative’ (Anderson et al., 1991). The choice of two mandate diseases did not ILRAD was established in 1973 and inaugur- close the question of how to attack them. There ated in 1978 as a new centre with independent had been extensive discussion in the Bellagio facilities and staff. meetings about ILRAD’s initial priorities. Ultim- ately, the decision was made that its initial emphasis would be on haemoprotozoan parasites and immunological aspects of African animal Priorities diseases9. A third disease, cysticercosis, was considered for ILRAD’s initial programme and One initial vision of ILRAD was to focus only on rejected 10. Underlying the choice of the two dis- a vaccine for theileriosis, commonly known as eases, the first paragraph of ILRAD’s founding East Coast fever (ECF), partly because it was memorandum of understanding with the Gov- believed that no existing facility could create ernment of Kenya began with a reference to such a vaccine (Pritchard 1991, interview)7. ‘basic research’. The proposed single disease focus raised ques- A second issue at ILRAD’s founding was tions from the Rockefeller Foundation and translation of research results into commercial the  United Nations Development Programme applications. There was discussion about what (UNDP), given the potential competition with an ECF vaccine would cost, who would pay, existing efforts. Research at the East African Vet- whether it would be more cost-effective than erinary Research Organization (EAVRO), funded dips and whether it would be commercially by the UNDP, had concentrated on ECF, and v iable. scientists there had been the first to grow Thei- Both ECF and trypanosomiasis were diseases leria parva, the causative agent, in tissue cul- that could possibly be controlled through new ture. The choice was ultimately made to focus molecular techniques involving similar scientific ILRAD’s mission on two problems – trypano- skills, even if ECF was the ‘short-term problem’ somiasis and ECF. and trypanosomiasis the ‘long-term problem’. The limited mandate of two diseases at IL- Earlier research on the causative agents of the RAD simplified the definition of research. Al- two diseases, tick-transmitted apicomplexans of though the questions about ILRAD’s mandate the T. parva spp., and tsetse-transmitted kine- had been resolved, other problems remained. toplastids of the species Trypanosoma brucei, One was that the global outlook of CGIAR de- Trypanosoma congolense and Trypanosoma vivax, rived from a US or European worldview and respectively, had shown that, while it was pos- hence was at times detached from the countries sible to immunize livestock against reinfection where the centres worked (Fitzgerald, 1986; with that specific immunizing strain, immunity Anderson et al., 1991). A second was relation- was not conferred to other strains of trypano- ships with other research institutions in East somes or Theileria spp. This inevitably meant that Africa. ILRAD would have a different mandate from the Documentation from a meeting in Rome in other early centres, which were primarily man- 1971 suggests other considerations in expanding dated to conduct applied and adaptive research. ILRAD’s initial mandate. One was the view that ILRAD, too, had an adaptive and applied man- an ECF vaccine would soon be available com- date, but it also had basic research to do, notably mercially: ‘One half or perhaps three fourths of establishing immunity against parasites and the the research towards vaccine production has mechanisms causing its failure. Disease control been accomplished but to complete the final would be – indeed, still is – a long-term and am- stages of this research will probably require five bitious goal, and would leave ILRAD and ultim- to ten years’ (McKelvey, 1971, pp. 285–286). ECF ately ILRI vulnerable to new priorities in how vaccine development was clearly seen as a ‘short- ‘science for development’ should be conceived. 11 term programme’ and trypanosomiasis research McKelvey believed that: as a ‘long-term problem’. Indeed, some felt that … to focus sharply on one, possibly two, a 10-year time frame for the development of an diseases East Coast fever and African animal The Evolution of IARC Livestock Research: 1975–2018 15 trypanosomiasis, and on one problem, the field impact of those advances was mixed in immunization techniques, to combat the the two ILRAD mandate diseases. diseases would afford greater chance of success Protozoan and helminthic diseases affect than to range widely over many problems of millions of people and animals in the developing cattle production in Africa. The Rockefeller world, implying that a targeted investment in Foundation successes in the medical sciences, their control could have a significant impact on combating yellow fever, for example, and in the human and animal health with potential spillover agricultural sciences, maize and wheat effects from scientific advances into control of improvement, reinforced this belief. malaria. The challenge, however, was formidable. Research capacity was strong in East Unlike bacteria and viruses, unicellular proto- Africa. African institutes, such as EAVRO and zoa and the cells of worms resemble the cells of the University of Nairobi, plus foreign univer- their human and animal hosts; these pathogens sities and colonial programmes, had made sig- have accordingly developed sophisticated mech- nificant research investments. Donor agencies, in anisms to evade attack by host immune systems. particular the British Overseas Development Parasitology emerged relatively late as a sci- A dministration and the UNDP, were keen on a entific discipline and, by the early 1970s, was r egional East African institution, and assisted in about to undergo profound change. There was a forming a research network serving Kenya, Tanza- recognition that modern biology, especially nia and Uganda. This communitarian ideal soon molecular biology, could make an important faded with the differing political systems of the contribution to parasitology. This interest in three nations; matters reached the point where parasitology was to some extent prompted by colo- it was expedient to communicate bilaterally with nialism, which had seen livestock diseases as Kenya over ILRAD. This simplified the initial threats to the profitability of colonial invest- decision to locate ILRAD at Muguga under the ments (Lyons, 1992). auspices of EAVRO (McKelvey, 1991). East African trypanosomiasis research and control had started around the turn of the Approach to trypanosomiasis control century and had been strengthened in 1946 at ILRAD’s founding with the formation of what would become the East African Tsetse and Trypanosomiasis In 1895, David Bruce identified T. brucei as a R esearch and Reclamation Organisation (E ATTRO), causative agent of animal trypanosomiasis. In at Tororo in Uganda, and EAVRO, based at Mu- 1903, Bruce identified tsetse-transmitted Trypa- guga, Kenya. Research was redirected equally nosoma gambiense as a causative agent of human at ‘ entomology, protozoology, biochemistry trypanosomiasis. In 1906, Robert Koch showed and medical and veterinary studies’ (Onyango, that aminophenyl arsonic acid (atoxyl) could 1971). While EATTRO’s work had faltered for cure trypanosomiasis in people, albeit at the lack of resources, new external funds had r evived risk of blindness. These findings, as well as sub- it to some extent (Clarke, 2007) and I LRAD’s sequent advances in understanding of tsetse work would clearly impinge upon the mandate and trypanosome taxonomy, trypanosome host of EATTRO. range and virulence, and identification of less The timing of ILRAD’s foundation was also toxic trypanocidal drugs, underpinned the strat- fortunate in terms of advances in immunology egies used in colonial Africa to control epidemic and molecular biology, which promised a revolu- human trypanosomiasis – diagnosis and treat- tion in vaccine development. Advances in recom- ment, vector control, and spatial separation of binant DNA technology and DNA sequencing, host and vector. alongside other technologies, allowed closer study Imposition of these historical strategies on of parasites and host immune systems, and of human populations throughout humid and semi- the genomes of parasites and hosts, and the humid Africa, while dehumanizing (Headrick, invention of new tests for characterizing para- 2014), had led by 1940 to a decline in human sites and diagnosing infections. As we shall see in trypanosomiasis and improved human health. the thematic chapters, especially those in Part I, Despite a resurgence of human trypanosomiasis 16 J. McIntire and J.Smith during the post-colonial period because of the might be found and were influential in ILRAD’s breakdown in control, the human disease is now early priorities12. contained in most regions and the World Health The state of knowledge of the animal tryp- Organization (WHO) projects that it is possible, anosomiases at the time of ILRAD’s founding with appropriate control measures, to eliminate was good. The comprehensive review of Mulli- the human disease as ‘a public health problem’ by gan and Potts (1970) had described vector biol- 2020 (www.who.int/trypanosomiasis_african/ ogy and behaviour, pathogenesis of the disease en/; accessed 21 January 2020). in animals and clinical aspects of disease man- While human African trypanosomiasis is agement, in addition to measures for control of now a less serious health risk in Africa, African the disease in livestock, and there was exten- animal trypanosomiasis (AAT) is still quite im- sive field knowledge of the disease across sub- portant. AAT is endemic throughout the tse- Saharan Africa. One outstanding gap of the tse habitat, which encompasses the humid and Mulligan and Potts book was the field of trypa- semi-humid regions of Africa, a landmass cover- notolerance, which become a major ILCA/ ing about one-third of the continent. AAT ILRAD theme in the late 1970s. causes high mortality in cattle and other domes- tic livestock and excludes cattle-based agricul- ture from all but the fringes of the tsetse habitat where tsetse and trypanosome challenge are Approach to ECF control at ILRAD’s relatively low; even there, animal agriculture re- founding13 quires tsetse control and the application of tryp- anocidal drugs and can still incur substantial ECF is a fatal bovine disease caused by the proto- production losses. Exotic breeds of livestock that zoan parasite T. parva. The disease occurs in 12 had been selected for growth rate and milk and countries in eastern, central and southern Af- meat production tended to develop more acute rica where the vector, the brown ear tick (Rhipi- forms of AAT than indigenous African breeds, cephalus appendiculatus), is found. ECF causes and this usually limited the use of such exotic major economic losses by affecting both dairy stock where tsetse was endemic unless trypa- cows and young Zebu cattle among pastoralists notolerant breeds could be introduced. and on ranches. It is among the most serious Much of ILRAD’s mandate was to discover constraints to cattle productivity in the coun- new ways of controlling AAT, with a focus on tries in which it is found. vaccination. It was clear from the inception of At the time of ILRAD’s founding, ECF had ILRAD that the development of an efficacious been managed with acaricides, but this treat- AAT vaccine would be difficult because blood- ment was expensive and not always successful. stream-stage African trypanosomes were known An alternative was for farmers to keep local to have remarkably high antigenic variation. cattle breeds, which tended to be more disease However, there was evidence to suggest that ani- resistant but were less productive than exotic mal infective ‘metacyclic’ trypanosomes derived breeds, especially for dairying. It was widely ac- from tsetse, and the first bloodstream-stage cepted that vaccination against ECF would be parasites that they generated, might express a the most attractive control option, and develop- common antigen that could be targeted by a vac- ment of a vaccine was a founding aim of ILRAD. cine (summarized by Gray, 1970, pp. 113–116). Work to develop an ECF vaccine had begun Thus, molecular characterization of trypano- in the 1960s at EAVRO, located in Muguga, some surface coat antigens, and analysis of their Kenya, under the auspices of the East African diversity within and between strains, species and Community. At about the time of ILRAD’s estab- differentiation stages of trypanosomes, was an lishment, a vaccination procedure was being de- immediate focus at ILRAD, and was paralleled by veloped at EAVRO. The infection-and-treatment deeper analysis of the bovine immune system, method (ITM) was an immunization procedure and by the investigation of immune responses to against ECF involving inoculation of live sporo- putative AAT vaccine antigens and to trypano- zoite forms of T. parva, usually in the form of a some infections. These scientific developments semi-purified homogenate of T. parva-infected renewed hope that a trypanosomiasis vaccine ticks, combined with simultaneous treatment The Evolution of IARC Livestock Research: 1975–2018 17 60 40 20 10 1975–80 1981–85 1986–90 1991–94 Period totals Domain Management and other Theileriosis Trypanosomiasis Total ILRAD spending Fig. I.5. ILRAD spending by research domain and period, 1975–1994. Total ILRAD spending, 1975–94, of $262 million in 2015 US$. (Data from ILRAD Annual Reports, various years). with a dose of a long-acting formulation of oxy- tetracycline. Production and delivery of this live Resources ECF vaccine was complicated, expensive and time consuming, and at the end of the 1970s, there Figure I.5 summarizes ILRAD’s spending over its were doubts as to whether the ITM process would life as an independent centre. ILRAD began with a be commercially viable. complement of perhaps 50 international scientists Despite uncertainties about the commercial in the 1970s (CGIAR/TAC, 1981, Table 1, p. 23)14. possibilities of ITM, it was recognized that Its annual budget (in 2015 US$) averaged acaricide treatments had their own shortcom- US$11.5 million between 1975 and 1981 (the ings, including cost and sustainability given the year of ILRAD’s initial external review). In the need for repeated applications. ILRAD therefore early years of ILRAD, the numbers of scientists focused on the immunopathology of theileriosis by discipline (CGIAR/TAC, 1981, Table 2, p. 26) and began the years of work in its laboratories, were (of a total of 38): parasitology, seven; cell and with East African partners, that eventually biology, six; immunology, nine; molecular biol- led to the commercial introduction of the ITM ogy, two; biochemistry, seven; and pathology, ‘Muguga cocktail’ vaccine in several East Afri- seven. We estimate that ILRAD spent nearly 60% can countries between 1998 (Tanzania pastoral of its research commitments on trypanosomiasis sector) and 2012 (Kenya dairy sector). and related immunological work and the balance Spending in 2015US$ millions 18 J. McIntire and J.Smith on ECF; and, further, that ILRAD spent about one- programme review (EPMR) noted a lack of field sixth of its real budget on management and ad- impact d uring 1986, it expected that such an ministration over its lifetime. ILRAD investment in impact would occur within the next 5 years systems, economics and policy was limited to based on what had been achieved scientifically trypanotolerance work before 1987, when its in the first decade. veterinary epidemiology unit commenced. ILRAD The second external programme review of spending on systems, economics and policy was ILRAD did not recommend major changes in pri- less than 3% of its lifetime total. orities. It rejected major new work on heartwa- ter, a tick-borne disease caused by Ehrlichia rumi- nantium. It referred to collaboration with ILCA in Institutional evolution the context of the Joint ILRAD/ILCA Trypa- notolerance Programme but did not recommend The first external evaluation of ILRAD (known changes in institutional collaboration or struc- at the time as a quinquennial review (QQR) ture. While the second EPMR stated that new (CGIAR/TAC, 1981) praised the centre’s science15. priorities on trypanosomiasis or ECF should be It stated: funded from existing resources, ILRAD manage- ment sought additional funds for the study of … the Panel is unanimously satisfied with the chemotherapy against trypanosomiasis and for very rapid progress achieved by ILRAD in building up a first-class research tool, with the research on trypanotolerance (CGIAR/TAC remarkable advances made in the scientific 1986a, p. 4 of letter from ILRAD Board Chair to knowledge of trypanosomiasis and theileriosis the Technical Advisory Committee (TAC) Chair). and with the high quality of the scientific Given the limits of funding and the changes leadership. in CGIAR priorities, developing ILRAD’s capacity to assess the research impact was important. While proposing no change in ILRAD’s disease Accordingly, in 1987, ILRAD established a small mandates, the panel recommended changes in socio-economic and epidemiology programme, management and priorities, including organiz- whose goal was to estimate the economic impact ing research themes into projects, doing more on of trypanosomiasis and ECF and to evaluate the trypanotolerance, building a greater presence in potential impact of their control. This programme West Africa, and establishing five new subprojects soon became a leader of ILRAD’s research, and on trypanosomiasis and another five on ECF. its successors have been a highly productive part Throughout the 1980s, research on ECF of ILRAD/ILRI over the past 30 years (see Chap- moved at a faster pace than trypanosomiasis ter 3, Chapter 5 and Chapter 6, this volume). research, as had been expected at ILRAD’s foun- dation. Already, however, ECF vaccine develop- ment was behind earlier ambitious timescales. The third external review of ILRAD Nevertheless, by 1989, a review group recom- The third external review of ILRAD (CGIAR/ mended the establishment of a Project Area on TAC, 1993a) further acknowledged the quality of vaccine formulation. A vaccine against trypano- the institution’s work. It urged widening ILRAD’s somiasis proved more intractable. mandate to apply its findings from molecular biology and immunology to other diseases and to The second external programme do more capacity development. It set a deadline review of ILRAD of five years for a theileriosis vaccine. It recom- mended that immunology research on trypano- The second external programme review of IL- somiasis be restricted to four candidate antigens RAD (CGIAR/TAC, 1986a) again praised the before ‘any new ones are isolated’ (CGIAR/TAC, laboratory’s work. It found scientific advances16 1993a, p. ii). It suggested that ILRAD do ‘a modest in studying the two diseases. It expressed con- expansion’ of research on other tick-borne dis- tinued hope for ‘immunological solutions’ to eases. While recommending no major change in trypanosomiasis and ECF, although it recog- ILRAD’s mandate, it recognized the need for new nized that prospects for a trypanosomiasis vac- investment in a vaccine development facility cine were fainter. While the second external (CGIAR/TAC, 1993a, p. iii). This third external The Evolution of IARC Livestock Research: 1975–2018 19 Papers 60 40 20 0 Citations 2000 1500 1000 500 0 1975–80 1981–85 1986–90 1991–94 Publication period Domain Theileriosis Trypanosomiasis Total ILRAD Fig. I.6. ILRAD publications by research domain and period, 1975–1994. Sample size of 632 papers. (Data from www.scopus.com and www.scholar.google.com). review of ILRAD did not endorse a merger of scientists and partners and as confirmed by ILCA and ILRAD, but it did encourage better co- three positive external reviews of the laborato- ordination between ILCA and ILRAD across ry’s work. sub-Saharan Africa. Figure I.6 shows the numbers of the princi- pal ILRAD papers in two fields – trypanosomiasis Achievements and ECF – plus the total. ILRAD’s chief achieve- ments in its early period were in diagnostic and other methods, as summarized in Fig. I.6, and The start-up costs of building facilities, ap- in the fields of genetics, immunology and pointing scientists and developing a programme pathology. The thematic chapters in Part I of meant that initial progress at ILRAD was slow. this volume review the scientific and develop- These start-up costs soon made the projected ment impact of that work. 5–10-year period to develop a vaccine against Achievements for the two major ILRAD AAT unrealistic, a finding that was confirmed by diseases (see Chapters 1–6, this volume) were: the first ILRAD external review (CGIAR/TAC, 1981). Scientific achievements during ILRAD’s • The successful continuous cultivation of initial 15 years were none the less significant, as T. brucei in vitro (Hirumi and Hirumi, 1989; shown by the many papers published by ILRAD Hirumi et al., 1977). Counts by publication period 20 J. McIntire and J.Smith • Advances in the diagnosis of trypanosomia- • Novel field and participatory epidemiology. sis (Paris et al., 1982; see see Chapters 2 The focus of ILRAD’s work was almost exclu- and 3, this volume). sively laboratory based from 1975 until the • Advances in understanding the genetics merger with ILCA in 1995. Two significant of T. brucei and T. vivax in (see Chapters 2, exceptions were the ATLN on trypanotoler- this volume). ant animals, which started in 1977, and field • Improvements in the management of AAT epidemiology on ECF, which started in 1986 in the absence of a vaccine (see Chapters 2 (see Chapters 2–6, this volume, for discus- and 3, this volume). sions of this field work). • Parasitology and immunology of trypano- • Development of partnerships with national somiasis and theileriosis (Baldwin et al., programmes in Kenya and Ethiopia. 1986); despite the eventual failure to pro- duce a successful vaccine against AAT, major global advances in bovine immun- Capacity development ology and immunopathology were made at ILRAD (see Chapter 4, this volume). Training of scientists and technicians became • The successful example of the long trad- an immediate ILRAD contribution, given the ition of research on the economic aspects time needed for research results. Moreover, cre- of controlling tsetse and trypanosomiasis, ating a well-trained alumnus was important to which ILCA/ILRAD/ILRI work has con- if ILRAD was to extend its research findings tributed in part over many years (see Chap- across Africa. Accordingly, technicians re- ter 3, this volume). This began with Hans ceived bespoke training in partnership with the Jahnke’s path-breaking work (1974, 1976, Kenya Agricultural Research Institute and the 1982), continued through the African University of Nairobi. Technicians from East Trypanotolerance Livestock Network (ILCA/ Africa were trained, with ambitions to extend ILRAD, 1988; Itty, 1992), the summary of opportunities to francophone and anglophone Swallow (2000), and the detailed investiga- West Africa. Courses on theileriosis, trypano- tions by Shaw (2004, and 2015, for example). somiasis and immunology were offered to asso- This work has allowed the definition of ciates and veterinary graduates. ILRAD alumni control models and has provided valuable ultimately made major contributions to investi- advice to extension services on the applica- gations of trypanosomiasis and bovine immun- tion of those models. ology (see Chapters 2–4, this volume). • Advances in cultivation of T. parva (1970s and 1980s), continuing the work of EAVRO. • Understanding of the biological mechan- isms of trypanotolerance (Murray et al., The International Livestock 1982; ILCA/ILRAD, 1988; Rowlands and Centre for Africa Teale, 1994). • Understanding of the behaviour and con- The CGIAR Task Force’s site recommendation trol of the ECF vector, the brown ear tick was for Addis Ababa, Ethiopia, or Yaoundé, (R. appendiculatus) (see Chapter 10, this vol- Cameroon, with the former being the apparent ume). preference because Ethiopia had the largest • Understanding of tsetse fly (Glossina spp.) cattle population in Africa and a ‘wide range of behaviour and control (see Chapters 2 and ecological conditions’ according to the Tribe re- 3, this volume). port (Tribe et al., 1973, p. 45). Further advan- • Definition of the epidemiology of tryp- tages of Ethiopia were its emergence as a hub anosomiasis and theileriosis (Perry and for international institutions and the fact that Hansen, 1994, and Norval et al., 1992, hosting both ILRAD and ILCA in anglophone on theileriosis; see Chapters 5 and 6, this countries would have been politically unaccept- volume). able. ILCA was established in 1974 in Addis The Evolution of IARC Livestock Research: 1975–2018 21 Ababa, Ethiopia, with the first phase of its cam- attention in future surveys and research, and pus inaugurated in 1980. suggest new or amended systems of animal production... Such techniques and systems will almost certainly require validation and further Priorities investigation, either at the Centre itself or within a cooperative programme at national stations, so that a constant interplay can be expected ILCA had a similar genesis to ILRAD – the idea of between research and development planning. applying science to agriculture, in a public insti- tution, with wide international collaboration, The initial work was diagnostic, developing and with the expectation of stable core funding. a ‘problem analysis’ as a basis to formulate im- ILCA’s founding premise was that solutions to provement packages at the farm level, undertake Africa’s livestock problems existed, but product- more intensive studies, experiment on components ivity had not grown because of a ‘failure to inte- and assess alternative systems. Accordingly, sub- grate the biological, economic and sociological Saharan Africa was divided into three broad components of research and development agroclimatic zones, first by Beck and later by ILCA programmes’ (Tribe et al., 1973, p. 1). The Tribe scientists – arid, humid and highland – which report stated (p. 31): were the sites of ILCA’s initial systems research. Multidisciplinary teams, based in Nigeria, Technical answers are available to many of the Ethiopia, Kenya and Mali, subsequently pre- specific problems facing livestock development in Africa. The major constraint lies rather in the pared systems surveys in four agro- climates 17: difficulty of introducing change into existing • The middle highlands (near Debre Zeit) and socio-economic systems, combined with upper highlands (near Debre Berhan) of inexperience in adapting technologies to suit Ethiopia, where mixed crop–livestock farm- local conditions. ing was dominant. The initial job of ILCA became one of as- • The subhumid zone in north-central Ni- sembling multidisciplinary teams to study geria, around Kaduna, to cover mixed physical and economic constraints to livestock crop-and-livestock farming in the humid and development. A basic choice of ILCA’s mandate subhumid regions. was to work only on ruminants and only in • The semi-arid zones in Botswana, north- sub-Saharan Africa. Some work on camels central Mali (near Niono) and later in began in the early 1980s and had significant south- western Niger, to cover agropastoral capacity development effects, but the original systems in the arid and semi-arid range- ruminant mandate of ILCA and ILRAD changed lands; and the semi-arid pastoral zone in little until the 1995 merger into ILRI. Kajiado County, Kenya, known as the Maa- ILCA was quite different from ILRAD. It sailand study area. worked on many problems where ILRAD worked • The humid zone of south-western Nigeria, on only two. It focused on creating new field near Ibadan. knowledge, and on mapping existing informa- tion, instead of creating new laboratory know- 18 ledge, by reviewing literature, by evaluating Resources development projects and by conducting new surveys (Tribe et al., 1973, p. 192). ILCA grew ILCA spent about US$374 million (in constant out of systems thinking and out of work recog- 2015 US dollars) from 1975 to 1994 (Fig.  I.7). nizing the importance of indigenous know- ILCA spent about US$16.9 million annually ledge, as opposed to the laboratory approach from 1975 to 1987 (the latter is the date of the of ILRAD. second EPMR) and another US$22.0 million an- The Tribe report proposed (p. 31): nually from 1988 to 1994. ILCA’s real spending Having established broad frameworks for over its lifetime was about 43% more than that systems studies … a more analytical phase will of ILRAD in its lifetime. One reason was that soon follow, which will both identify areas of ILCA had a broader scientific mandate than specific ignorance which deserve priority ILRAD. A second was that the Tribe report (Tribe 22 J. McIntire and J.Smith 125 100 75 50 25 0 1975–80 1981–85 1986–90 1991–94 Period totals Primary production Livestock systems Management and other Domain Economics and policy Animal production health and genetics Total ILCA spending Fig. I.7. ILCA spending by research domain and period, 1975–1994. Total ILCA spending of US$374 million in 2015 US$. (Data from ILCA Annual Reports, various years.) et al., 1973, pp. 1 and 102) had proposed a de- 23%. There were no important differences in centralized structure at ILCA; ILCA therefore those shares over subperiods of ILCA’s 20 year created subcentres at Kaduna, Ibadan, Bamako existence. A crude attribution of ILCA’s spending and Nairobi, in addition to the three Ethiopian by country is: Ethiopia, 70% of the total from sites at Addis Ababa, Debre Zeit and Debre Berhan, 1980–94; Ibadan, Nigeria, 3%; Kaduna, Nigeria, thus increasing its total costs compared with 5%; Niono and Bamako, Mali, 10%; Nairobi, those of ILRAD with its one Nairobi site. Kenya, 10%; and the sum of other Africa coun- Scientific spending was about 77% of IL- tries plus occasional work in Asia, Latin America CA’s total from 1975–1994, of which 12% was and the Caribbean, 1%19. devoted to capacity development; the balance was spent on management, administrative and support services. ILCA lifetime spending shares Institutional evolution by domain were: (i) livestock systems, 19% (much of which was at sites outside Ethiopia); The findings of ILCA’s QQR covering the period (ii) economics and policy, 13%; (iii) primary pro- 1975–1981 and issued in 1982 (CGIAR/TAC, duction, 7%; (iv) animal genetics, health and 1982) differed sharply from those of ILRAD. The production, 27%; (v) capacity development, panel concluded that ILCA’s research was of 12%; and (vi) management and administration, poor quality and its management was worse. Spending in 2015 US$ millions The Evolution of IARC Livestock Research: 1975–2018 23 The QQR recommended that ILCA move away manures with work on both large and small from ‘systems description’ and towards compo- ruminants. nent analysis. The panel argued that the identi- • Semi-arid (Niono, Mali; Niger; Kenya; Ethi- fication of constraints had been done and that opia): selective harvesting and handling of the logical next step was to develop technical crop residues to improve livestock nutrition options to overcome these constraints. These and soil management, water and grazing were especially important as early research had management, and on-farm fattening of showed that, contrary to the erroneous as- ruminants. sumptions of the founding Tribe report, intro- • Humid (Ibadan, Nigeria): alley cropping with duced technologies offered no significant leguminous fodder trees and trypanotoler- a dvantages over t raditional methods given the ant stock, which focused mainly on small economic and ecological constraints facing ruminant production (ILCA, 1979a,b). producers (Gryseels and Anderson, 1983, for highland Ethiopia; Cossins and Upton, 1987, The second phase of ILCA – roughly 1981– and Cossins and Upton, 1988a,b, for the Borana 1987 – saw a strengthening of the central re- rangelands in southern Ethiopia; Wilson et al., search units, including the Livestock Economics 1983, Wilson, 1986, and Wagenaar et al., 1986, Unit, the Small Ruminant and Camel Group, and for central Mali). the Forage Agronomy Section. There was an ex- ILCA’s research gradually evolved to focus pansion of the Livestock Productivity and Tryp- on component technical changes to release sys- anotolerance Group, which formed part of the tem constraints. These included the following: ATLN jointly coordinated by ILCA and ILRAD. The network established study sites under differ- • Highlands (Debre Zeit and Debre Berhan, ent levels of tsetse challenge and trypanosomia- Ethiopia): cross-bred cows for dairying, single- sis risk in West Africa (Côte d’Ivoire, the Gambia, oxen traction, simple mechanization to im- Togo and Zaire) and East Africa (Ethiopia and prove the drainage of waterlogged soils, and Kenya). establishment of a forage gene bank in 1983. The second EPMR of ILCA (CGIAR/TAC, • Subhumid (Kaduna, Nigeria): making bet- 1987) was again quite critical of ILCA’s perform- ter use of indigenous feeds, establishment ance. In fact, the second EPMR was drafted but of ‘fodder banks’ of leguminous pasture for withheld from publication by TAC until the new dry-season grazing, crop rotations with ILCA management could prepare a ‘long-term cereals and legumes, and nutrient cycling strategy as a basis for setting of short and alternatives with crop residues and animal medium-term priorities’ (CGIAR/TAC, 1987, Table I.1. ILCA priorities around the time of its first formal strategy in 1987. Budget shares, Publication shares, Citation shares, Field 1988–1993 1975–1994 1975–1994 Cattle 27 40 44 Small ruminants 18 21 32 Animal traction 17 4 10 Feed and animal nutrition 14 37 5 Trypanotolerance 13 19 15 Economics and policy 11 13 32 Publication shares are numbers of ILCA publications by field, divided by the total number of all ILCA publications in all fields; citation shares are numbers of citations of ILCA publications by field, divided by the total of citations of ILCA publications in all fields. Publication and citation shares add up to more than 100 because multiple citation classes per publication occur often. Sources: Budget shares for 1988–1993 were calculated from data in the 1993 EPMR (CGIAR/TAC, 1993b, p. 68). Publication and citation shares were calculated from the Scopus and Google Scholar databases for the 406 published papers (publications database updated through March 2020) in which at least one member of ILCA staff participated as an author of any rank (e.g. first author, third author, etc.). 24 J. McIntire and J.Smith Papers 50 40 30 20 10 0 Citations 1,500 1,000 500 0 1975–80 1981–85 1986–90 1991–94 Publication period Animal science Feed and forage Rangelands Domain Economics and policy Livestock systems Total ILCA Fig. I.8. ILCA publications by research domain and period, 1975–1994. Sample size of 406 papers. (Data from www.scopus.com and www.scholar.google.com.) p. 10). The second EPMR did not recommend that Achievements the centre change its mandate, which remained limited to sub-Saharan Africa and to ruminants. Figure I.8 summarizes the chief achievements The review did insist that ILCA ‘should focus its of ILCA in its 20 years. A bibliometric analysis, research activities … and avoid spreading its re- using the Scopus publications and Google Scholar sources too thinly’ (CGIAR/TAC, 1987, p. 10). databases, was done as described in Box I.2. A fourth senior management team arrived These achievements were as follows20. at ILCA in November 1986. It prepared a scien- Animal genetics and health: tific strategy in 1987 (Table I.1). The strategy organized its work into six ‘thrusts’: three ‘com- • Contributions with ILRAD and national modity thrusts’ – cattle milk and meat, small partners in elucidating the genetic and ruminant meat and milk, and animal traction – physiological basis of trypanotolerance in and three ‘strategic thrusts’ – animal feed, tryp- African livestock and identifying condi- anotolerance, and livestock policy and resource tions in which those stock could be more use. This structure and strategy remained in productive (ILCA/ILRAD, 1988; Rowlands place until the merger with ILRAD in 1995. and Teale, 1994). Counts by publication period The Evolution of IARC Livestock Research: 1975–2018 25 Box I.2. Publications as a measure of scientific • Complete novel feed quality and animal impact nutrition research (Reed et al., 1988, 1990), which later led to a classification of nutritional Publications as a measure of scientific impact toxicology in forage legumes (Reed, 1995). apply to many research projects because the lat- • Building the forage gene bank and herbage ter produces knowledge that may never have seed unit (see Chapters 12 and 13, this been translated into economic output. Examples of such knowledge are understanding soil nutrient volume). management on small farms or mapping vector Livestock systems: behaviour across farming environments. While it is possible to attribute costs to research pro- • Characterization of the principal animal grammes, the problem of estimating benefits to production systems of sub-Saharan Africa outputs from those programmes has typically (Jahnke, 1982). proven intractable in most of the work done at ILRI • Complete innovative systems studies of graz- and its international partners in livestock research. ing (Mali, Kenya, lowland Ethiopia) and One solution to the problem of assigning mixed systems (Kaduna, Ibadan, highland benefits to knowledge outputs is to estimate Ethiopia). the numbers of papers published and citations per paper using ‘bibliometrics’ tools (Aria and • Explaining the successes and failures in Cuccurullo, 2017). A related tool is Altmetric pastoral development in poor countries, with (www.altmetric.com/; accessed 13 March 2020), a focus on sub-Saharan Africa (Sandford, which uses a briefer period of analysis but takes 1983a) and many papers in the African a broader view of the form of scientific impact Livestock Policy Analysis Network (ALPAN) as it includes social media, which bibliometrics and Overseas Development Institute (ODI) generally does not. Papers and citations are a Pastoral Networks. measure of scientific productivity and an – ad- • Analysis of the resources, constraints and mittedly imperfect and partial – indicator of the potential for livestock water in sub-Saharan potential economic gains from research. In the Africa (Classen et al., 1983; King, 1983; Prefaces to each part of this volume, we pre- sent analyses of published papers and citations Sandford, 1983b). with the purposes of estimating the relationship • Explanation of the evolution of African between scientific output and research cost and mixed farming systems in terms of land use of situating ILRI’s published outputs in the Afri- and technology choice (McCown et al., can and global contexts. 1979; McIntire et al., 1992). Related scientific impacts were: predictions • An analytic review of the determinants of of system evolution from the climate change lit- the reproductive performance of African erature, particularly warnings about the more Zebus (Mukasa-Mugerwa, 1989). rapid conversion of grazing systems into mixed farms in East and West Africa (see Chapter 15, Primary production: this volume); and on the effect of population growth on tsetse infestation (see Box I.1). • Identification of potential technical changes Much of ILCA’s scientific legacy was in in grazing and water management (Borana characterizing African livestock systems and in and Maasailand), in alley farming with making predictions of system evolution, par- leguminous trees (humid west-central ticularly about the more rapid conversion of Africa), in sown forages (subhumid west- grazing systems into mixed farms in East and central Africa) and in herd and flock man- West Africa (see Chapter 15, this volume). agement (semi-arid Mali), the integration of livestock into mixed farms through semi- Systems evolution arid and subhumid Africa, and research on the role of women in pastoral and mixed Beginning with the work of Hans Jahnke on Tse- agropastoral farming systems. tse Flies and Livestock Development in East Africa • Extended knowledge of the range ecology (Jahnke, 1976) and continuing with Jahnke’s in East Africa and across the continent milestone book (Jahnke, 1982) on Livestock (ILCA, 1975). Production Systems and Livestock Development in 26 J. McIntire and J.Smith Tropical Africa, ILCA scientists and partners de- • Advances in forage agronomy in Ethiopia, scribed the evolving role of animals in African Nigeria and Mali, related to the systems grazing and mixed farming systems. Prominent studies, based on a combination of on- station works in the early history of ILCA included stud- and on-farm trials (Kang et al., 1990; Powell ies on pastoralism and on the role of water in et al., 1995). livestock development (Sandford, 1983b). Work • Defining the research inputs needed to raise on farming systems across sub-Saharan Africa productivity of animal traction among by ILCA, or in collaboration with ILCA, included smallholders in Ethiopia and Mali, includ- Pingali et al., (1987) on agricultural mechaniza- ing the adaptation of a broad-bed maker for tion and McIntire et al. (1992) on crop–livestock vertisols in central Ethiopia. integration. A major piece in the area of land • Estimating the trade-off between meat and management was the work by Powell et al. milk production in grazing systems in Kenya, (1995) on livestock and nutrient cycling in Ethiopia, the Gambia (Agyemang et al., mixed farming areas. The impact of these books 1993, 1997), Mali (Wagenaar et al., 1986) was to describe the constraints to technical and Botswana (Cartwright et al., 1982). change and to set the physical, demographic, • Elucidating the interactions between small economic and agronomic boundaries with which ruminant parasitology and leguminous trees higher productivity could be achieved in the in alley farming systems for humid West principal mixed crop–livestock systems of Africa. Africa (Kang et al., 1990). • Creating the first ley farming model in the Systems characterization lowland tropics of sub-Saharan Africa (von Kaufmann et al., 1986), which integrated As discussed in Chapter 15 (this volume), initial Stylosanthes hamata cultivars into fodder systems studies consisting of field surveys and banks for cattle. data analysis led by ILCA, with varying collabor- ation with national and international institu- tions, defined the principal constraints to animal Trypanotolerance production. These included low dry- season feed quantity and quality, inadequate water supplies, The joint ILRAD/ILCA African Trypanotolerant and competition between people and calves for Livestock Network (ATLN) began in 1979 after limited milk supplies in arid pastoral systems. background studies in sub-S aharan Africa. By Animal diseases, poor feed quality and low soil 1986, under the leadership of John Trail at fertility were barriers to higher productivity in ILCA and Max Murray at ILRAD, it had ana- humid and subhumid zones. Availability of ani- lysed data on tsetse group and challenge level, mal draught power, high mortality of young livestock species, and type and management re- stock, liver fluke in sheep, inefficient water con- gime from 11 sites in seven countries of humid servation and utilization, and inadequate sup- and subhumid West and Central Africa (ILCA, plies of protein supplements were noted in high- 1994, pp. 69–75). Important results were found land zones. on the genetics, diagnosis, pathology and man- The main scientific accomplishments were agement of trypanotolerant livestock under as follows: varying conditions of trypanosomiasis chal- • lenge, as summarized in ILCA/ILRAD (1988).Characterization of mixed farming systems The principal landmarks in trypanotolerance in semi-arid central Mali (Wilson, 1986), research, done jointly by ILCA and ILRAD in col- the Inner Delta of the Niger River in semi- laboration with national and international pro- arid central Mali (Wagenaar et al., 1986), in grammes in sub-Saharan Africa, were as follows: subhumid central Nigeria (von Kaufmann et al., 1986), and two sites in the highlands of • Understanding the genetics and pathology Ethiopia (Gryseels and Anderson, 1983). of trypanotolerance (ILCA/ILRAD, 1988; • Characterization of grazing systems in Kenya Rowlands and Teale, 1994; also Chapters 2, (‘Maasailand’, Kajiado County, Kenya; 3 and 4, this volume). Bekure et al., 1991) and in southern Ethi- • Developing a comprehensive model – genetics, opia (‘Borana’; Coppock, 1994). vector control and disease management – The Evolution of IARC Livestock Research: 1975–2018 27 for using trypanotolerant stock (Rowlands conditions, notably by showing the effects of and Teale, 1994). phenolics in the digestibility of roughages, in • Devising the first empirical benefit–cost ana- quantifying the value of crop residues in cereal lyses of the combined use of trypanotolerant and legume improvement programmes (Reed animals and chemotherapy based on reli- et al., 1988), and in later work on the toxicology able productivity in several situations in of phenolics in tropical legumes (Reed, 1995). sub-Saharan Africa (Itty, 1992). Forage gene bank and herbage Merger of ILRAD and ILCA seed unit into ILRI, 1995 ILCA created the forage gene bank in Ethiopia in 1983. From an early stock of some 10,000 Discussions in the early 1970s considered and a ccessions in 1986, the collection had grown to rejected the option of one institution, although it over 18,600 in 2017. An herbage seed unit was was widely believed that the two centres would established in 1989 to enhance forage use in eventually merge. The QQR of ILCA (CGIAR/ sub-Saharan Africa, given the evident fact that TAC, 1982, p. 5) considered the relevance of a seed was a constraint to adoption of new forages. ‘change in the relationship of ILCA and ILRAD’ The scientific impact of the forage gene but saw no need for a ‘change in the structural banks, including those at ICARDA and CIAT, has relationship between the two centres’ (CGIAR/ been strong (Amri et al., 2018, for ICARDA; TAC, 1982, p. 69). The second EPMRs, respect- Schultze-Kraft and Peters, 2017, for CIAT; ively, of ILRAD (CGIAR/TAC, 1993a) and ILCA Labarta et al., 2017 for CIAT; Lynam and Byerlee, (CGIAR/TAC, 1993b) agreed that the two should 2017 for CIAT; see Chapter 12, this volume). Sci- not be joined, although this view soon changed. entific impacts included: (i) an understanding of The second EPMR of ILRAD (CGIAR/TAC, the distribution and biology of tropical forages; (ii) 1993a) was generally positive about ILRAD’s germplasm management in relation to production work in its first two decades. While praising the constraints, such as establishment; and (iii) use of research of ILRAD, it took a cautious approach modern genetic tools to elucidate phylogenetics. about the probability of success with a vaccine against trypanosomiasis; it did advocate, how- ever, a production facility for an ECF vaccine. Feed quality and nutrition research It recommended that the lessons from research on theileriosis and trypanosomiasis be applied to ILCA reported: new diseases such as heartwater and anaplas- Studies on the nutritive value of feeds covered not mosis. It said that ILRAD should gradually only traditional forages but also crop residues and extend its work outside sub-Saharan Africa. The foliage and seeds of browse trees. A major report did not recommend a merger between research area was in identifying polyphenolic ILRAD and ILCA. compounds in crop residues and browse and The second EPMR of ILCA (CGIAR/TAC, determining their effects on feed intake and 1993a) was again critical of the quality of ILCA’s utilisation. … Assessments of the nutritive value research and management. The review, unlike of browses was largely directed at determining the the second EPMR of ILRAD, argued that ILCA’s content and effects of polyphenolic compounds in mandate should be narrowed, given the diffi- different browse species and accessions. culty of managing such a complex institution ILCA (1994, pp. 146 and 149). across several subregions of sub-Saharan Africa. Related work on forage agronomy, includ- The second EPMR of ILCA again rejected a mer- ing grasses and legumes, was done at several ger, while agreeing to ‘appropriate collaborative sites in Nigeria (von Kaufmann et al., 1986; projects’ between ILCA and ILRAD (CGIAR/ ILCA, 1994, pp. 121–124), Ethiopia (Ka- TAC, 1993b, pp. ii and x). hurananga, 1987) and Mali (ILCA, 1994, pp. Following the 1993 EPMRs of ILRAD and 56–58). ILCA also made significant advances in ILCA, the TAC undertook another global knowledge of ruminant nutrition under tropical priority setting exercise in 1992–1993 (CGIAR/ 28 J. McIntire and J.Smith TAC, 1994). It compared the CGIAR’s research International, 1992) argued for tighter research investment to the value of production of the focus in ILCA with the implication that this could major food commodities. This analysis suggested be achieved by incorporating the animal health more resources for resource management, pol- work of ILRAD. icy research and germplasm conservation. Fol- A CGIAR Steering Committee on Livestock lowing the TAC priority exercise and taking note Research was formed in 1993 to ‘identify prior- of CGIAR investment in livestock research and ity activities for international livestock re- of the second EMPRs of ILCA and ILRAD, TAC search, which would be managed through a agreed that the centres should remain separate single institution and be constrained by the while expanding ‘ecoregional programmes’ to join current proportion of CGIAR resources allo- crop-and-livestock work in a systematic way. cated to livestock’ (ILCA, 1994, pp. 2–3). The Despite TAC’s views against a merger, as steering committee subsequently recommended expressed in the second EPMRs and in the priority the creation of a new, consolidated centre. setting exercise, external opinion was building in CGIAR approved the steering committee recom- favour of a merger for programmatic and man- mendation at its meeting in October 1993. agerial reasons. An external assessment of ani- CGIAR later requested a Rockefeller-appointed mal agriculture in sub-Saharan Africa (Winrock task force (an ‘implementing agency’) to develop (a) Before merger into ILRI After merger into ILRI 200 150 400 100 200 50 0 0 1975–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Period totals Funds Window 1 and Window 2 Restricted funds Unrestricted funds All sources (b) Before merger into ILRI After merger into ILRI 200 150 400 100 200 50 0 0 1975–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Period totals Primary production Economics and policy Management and other Domain Total spending Capacity development Livestock systems Animal sciences Fig. I.9. ILRI spending by source of funds and period (a) and by research domain and period (b) 1975–2018. Lifetime (1975–2018) spending of US$1753.2 million, in 2015 US$ millions. (Data from ILRI Annual Reports and Financial Reports, various years.) Spending in 2015 US$ millions Spending in 2015 US$ millions The Evolution of IARC Livestock Research: 1975–2018 29 a unified strategy for global livestock research and Asia 28%, in a projected 2010 budget of to be conducted within CGIAR by the merged US$47 million. The budget targets for the first ILRAD and ILCA. This implementing agency in decade of the 21st century were met, but pro- 1994 developed a medium-term strategy and a jected regional shares for 2010 were not strategic plan for the new CGIAR global live- achieved; the great majority of ILRI’s budget stock research institution – ILRI, which was in the years 2000–2010 was spent in legally established in September 1994 and sub-Saharan Africa, as had always been the began work in January 1995. case for ILCA and ILRAD, and the Asia staff never exceeded 5% of the total. New priorities at the merger Species mandate The nascent ILRI shifted from an exclusive focus TAC’s commentary on the first EPMR of ILRI on sub-Saharan Africa to a global one under a (CGIAR/TAC, 2000) only advised ILRI to con- rapid projected growth in budget. The indicative sider work on non-ruminants, seeking impact on medium-term plan for ILRI (CGIAR/TAC, small producers in Asia while using results, where 1994a, p. 30) projected that ILRI staff would ex- relevant, from its ruminant research. The 2000– pand from 88 scientists (100% in sub-Saharan 2010 Livestock Strategy (ILRI, 2000, p. 68) fore- Africa) in 1994 to 107 scientists (81% in cast a non-ruminant share of 10% for the 2010 sub-Saharan Africa, 11% in Asia, 5% in Latin budget (in a projected total of US$47 million), America and the Caribbean (LAC), and 3% in but this was not achieved. The share of swine West Asia and North Africa (WANA) by 1998. and poultry research in ILRI papers grew from Starting from a combined 1994 (ILCA and IL- roughly zero in the mid-1990s to a cumulative RAD) nominal budget of US$24.8 million, share of about 3% by 2016. Research on non- CGIAR projected that ILRI’s budget would grow ruminants consisted mainly of poultry pathology by 25% to 1998 (CGIAR/TAC, 1994a: Table 5). in the first decade after the merger with later Despite optimistic budget projections at the work on poultry genetics, the immunology and launch of ILRI, the merger coincided with a fall in pathology of African swine fever (ASF), food budget from an annual average of US$36.4 million safety risks related to non-ruminants and value (constant US$2015) in 1988–1994 to an average chain analytics. of US$31.5 million in 1995–2001. The 1995– 2001 core budget fell from 89% of the total from Scientific priorities 1995 to mid-year 1998 to 69% from mid-year 1998 to 2001. The falling trend in unrestricted re- The merger and the more restrictive budget sources continued until 2011 when core budget meant a change in priorities. Rough estimates of was eliminated across all centres (Fig. I.9a). the priorities of ILCA and ILRAD before the mer- The ILRI 2000–2010 Strategy (ILRI, ger can be created, from budget data and from 2000, p. 68) forecast regional shares to be: publications and citations data (Table I.1)21. sub-Saharan Africa 64%, LAC 6%, WANA 2% Subsequent ILRI priorities (Table I.2) are some- Table I.2. ILRI planned budget by CGIAR allocations, 1999–2003. (From ILRI, 2000, Table 2.) Field Shares, % of 1999–2003 planned budget Improving productivity (total) 58 Germplasm 9 Production systems 49 Protecting the environment 12 Saving biodiversity 7 Improving policies 9 Strengthening national agricultural research systems 15 30 J. McIntire and J.Smith what difficult to compare on spending because of Reorientation of ILRI after 2000 changes in institutional structure and spending reporting but can be compared more easily on In the early part of this century, ILRI underwent output metrics, such as papers and citations. four transformations. First, it began to explicitly Figure I.9(b) shows ILRI spending in the princi- address poverty, which had not been a major pal fields of ‘animal sciences’, ‘economics and r esearch theme before 2000, largely in response policy’, ‘systems’, ‘feed and the environment’, to donor demands for more immediate impact ‘capacity development’ and ‘other’ over its entire and to join the spirit of the recently an- history. nounced Millennium Development Goals. The goal of research grew beyond higher productiv- ity – it became the broader development goal of Resources ‘better lives through livestock’. Landmark books by Perry et al. (2002) and Thornton et al. (2002) The ILRI Medium-term Plan 1998–2000 proposed established empirical relationships among agri- budgets in terms of CGIAR resources allocation cultural environments, livestock production, categories at the time. It declared post-merger animal health and poverty. priorities in terms of ‘products’, including the Second, there was a shift from laboratory- following: based animal science working on genetics and immunopathology to field-based animal science • Ruminant genetics (mapping the bovine focusing on veterinary epidemiology with new genome; identifying areas of the mammalian programmes in food safety, zoonoses and emer- genome that control trypanosomiasis). ging infectious diseases. At the same time, the • Ruminant health (laboratory culture for work of the institute expanded to include research growing trypanosomes in vitro; immun- on swine, poultry and previously understudied ity work on ECF; genetic map of T. parva; diseases such as ASF, contagious bovine pleuro- community trypanosomiasis control in pneumonia (CBPP) and contagious caprine Ethiopia; better estimates of the economic pleuropneumonia (CCPP). Work in this century impacts of trypanosomiasis and tick-borne on zoonoses, food safety, and transboundary dis- diseases). eases became prominent and replaced much of • Ruminant feed resources (distributing the earlier laboratory work on a trypanosomia- materials to the national programmes; sis vaccine and on trypanotolerance. identifying more productive forages through Third, was the emergence after 2000 of a selection; analysing the use and quality of new field – the interactions between livestock crop residues). and climate change. This field had been a small • Crop–livestock systems (characterizing the share of the ILCA/ILRAD portfolio and of the ILRI major African systems; quantifying nutrient portfolio before the turn of the century. Since exchanges among crops, animals and soils; about 2000, research on climate change has be- reducing range degradation). come a fairly large share of the institute’s work, • Strengthening collaboration with national and papers in this field have been among the programmes (training; bibliometrics; col- most cited of ILRI’s outputs. laborative research). A fourth shift was towards becoming a ser- • Regions outside Africa projected to grow vice provider. In 2002, ILRI began to host the from about 20% of the research portfolio in Biosciences eastern and central Africa (BecA)- 1996 to about 30% in 2000. ILRI Hub whose purpose was to become a bio- An estimate of priorities after the merger sciences centre of excellence for East and Central can be derived from the ILRI Medium-Term Africa. As such, the BecA-ILRI hub would provide Plans (e.g. ILRI, 2000), the first EPMR of ILRI capacity development services for regional scien- (CGIAR/TAC, 2000), the second EPMR of ILRI tists in the biosciences related chiefly to agricul- (CGIAR/TAC, 2008) and from publications after ture and the environment. 1994. Table I.2 shows ILRI’s projected budget Part of the shift towards service provi- after the merger. sion was a new orientation to work directly in The Evolution of IARC Livestock Research: 1975–2018 31 development projects. Budget shifts, begin- partners). The former core funding was now ning in the mid-1990s and accelerating until channelled through the CRPs as programmatic about 2010, were associated with a change funding. Although the CRPs had been designed from core (untied) resources to special pro- on the assumption that 50–60% of the funding jects (tied) resources. These changes in total would be provided as programmatic funding funding and in its composition induced a with the balance being from bilateral projects, move away from longer-term research, for ex- this never happened, and the programmatic ample on vaccines, parasitology and genet- funding declined from about 40% of ILRI’s ics, to more short-term research and to more budget in 2012 to about 20% in 2018. Bilaterally engagement in development projects in the funded projects now make up most of the port- hope of short-term benefits. Part III of this folio of the CRPs. Seven programmes were com- book, with chapters on systems research and modity based, three focused on environment policy, evaluates the effects of the shift in and natural resource management, three dealt orientation from research to development. with agricultural systems in different agroeco- A millennial change related to shifts in port- logical zones, one dealt with policies, institutions folio was in information and communication and markets, one with agriculture, nutrition technology (ICT). ICT has also, of course, been and health and one supported the 11 gene banks transformational with regard to science and in CGIAR. allows communications among scientists and ILRI was engaged in seven CRPs. It led a research that would have been impossible recently. CRP on Livestock and Fish, with WorldFish, CIAT Examples are lowered costs of collaboration with and ICARDA as partners, and was a partner in partners in Africa and on other continents and Climate Change, Agriculture and Food Security lowered costs of scientific analysis, such as gene (CCAFS), Agriculture for Nutrition and Health sequencing of plants, animals and pathogens. (A4NH), Policies, Institutions and Markets (PIM), These lowered costs have allowed more efficient Dryland Systems, Humidtropics and Genebanks. collaboration and increased the scientific prod- While the CRPs were effective at fostering collab- uctivity of research spending. The scientific im- oration, the transaction costs of being engaged pact of modern ICT has been high at ILRI and its in many CRPs were high for ILRI and for the partners by allowing more rapid analysis of data other centres. and communication of results. The direct devel- A second phase of the CRPs started in 2017, opment impact related to ILRI’s use of ICT is with a reduced number (12 plus three platforms – unknown, but broader development impacts are Big Data, Excellence in Breeding and Gene- obviously large. banks). The need to reduce transaction costs, coupled with the observation from an independ- ent review (CGIAR/ISPC, 2014) that CGIAR The reform of CGIAR livestock research was dispersed across many CRPs, led to the consolidation of livestock re- The gradual reduction of core funding (Fig. I.9a) search. Livestock work was concentrated in one that had begun around the time of the merger, Livestock CRP, led by ILRI, from 2017. ILRI was continued in the first decade of the new century, still engaged in CCAFS and A4NH and has se- and culminated in 2011 with the elimination of lective involvement in PIM and the three plat- core funding. The reform of CGIAR, beginning forms. As the CRPs were designed as long-term in about 2011, reversed the trend of declining programmes, it is too soon to judge their devel- overall funding but at the cost of a complex and opment impact, although the first evaluation burdensome systemic change that changed the of the Livestock and Fish CRP was generally allocation and composition of funding and the positive. structure and type of research. ILRI’s second-generation research extended A second major change in 2011 was the its work on animal disease, especially in veterin- advent of the CGIAR research programmes ary epidemiology, although it abandoned field (CRPs). These organized much of CGIAR research work on trypanotolerance in the later 1990s into 16 programmes, each being a multicentre and did less on identifying constraints through consortium (in most cases with non-CGIAR the earlier systems studies. At the same time, 32 J. McIntire and J.Smith Table I.3. Principal achievements of ILRI research, 1975–2018. Domain/sub-domain Field/system Era Achievements Animal health and genetics Cattle genetics Bovine genome 1990s–present Mapping of the bovine genome (Barendse et al., 1997; see Chapter 1, this volume) Genetics of African pastoralism (Hanotte et al., 2002) Genetic structure of cattle breeds (Bovine HapMap Consortium, 2009). Genetics of Subhumid and humid Africa 1990s–present Quantitative trait loci controlling trypanotolerance in N’Dama and trypanotolerance and Borana (Hanotte et al., 2003; see Chapters 1–3, this volume) trypanosomiasis Genetics of the infection response to Trypanosoma congolense (Noyes et al., 2011) Genes controlling resistance to trypanosomiasis in mice (Kemp et al., 1997) Cloning Boran cattle (Yu et al., 2016) Diagnostic methods Diagnostics; in situ phenotyping tools 2000s Use of SNP tools for in situ characterization of smallholder livestock Description of African bovine genome Sheep genetics Sub-Saharan Africa; highland Ethiopia Recent Genetic history of African sheep and full genetic characterization of Ethiopian sheep Inter-breed differences in resistance to endoparasites Yak genetics Mongolia and Russia rangelands 2000s Molecular characterization of yak genomes Genetics of ASF Sub-Saharan Africa 2000s Genome sequencing of ASF (Bishop et al., 2015) Swine pathology including ASF Has very high development potential Control of GIS methods for mapping control methods: trypanocides, tsetse Trypanocide resistance in West Africa (Affognon et al., 2009; trypanosomiasis control, and trypanotolerant animals Clausen et al., 2010) Sources of trypanocide resistance; control of resistance Vector biology Tick dynamics 1980s–present Less emphasis on tick and tsetse biology after merger Acaricide resistance, (see Chapter 10, this volume) Possible field impact through tsetse control (Leak, 1999) Ticks (see Chapter 10, this volume) Population growth and tsetse dynamics (Reid et al., 2000) Veterinary LGA, LGH, LGT, MRA, MRH, MRT 1995–present Problems of scale, policy changes and technical changes (see epidemiology Chapters 5–7 and 9, this volume) The Evolution of IARC Livestock Research: 1975–2018 33 Domain/sub-domain Field/system Era Achievements ECF ITM 1970s–present Identification of the polymorphic immunodominant molecule of See Chapter 6 (this volume) Theileria parva Potential vaccine development CBPP and CCPP Potential vaccine development Peste des petits High development potential of a thermostable vaccine (Mariner ruminants et al., 2017) Heartwater control LGH, MRH 1990s See Chapter 5 (this volume) Highly pathogenic Diagnostic and surveillance tools 2000s Mariner et al. (2014) avian influenza Zoonoses and Global analysis of 13 zoonotic 2010s Policy, capacity, priority setting and control methods (Grace et al., emerging infectious diseases 2012; see Chapter 8, this volume) diseases Food safety Surveys of main risks; identification of risk profiles and prevention Policy, capacity and priority setting (Grace et al., 2012; see methods Chapter 9, this volume) Rift Valley fever Economic analysis of control options 2010s See Chapters 5 and 7 (this volume) Mapping of disease 2010s Rich and Wanyoike (2010) Economics of animal Modelling economics of parasitic 1990–present Perry and Randolph (1999) health and disease diseases Primary production Forage gene bank Global tropics (LGA, LGT, MRA, MRH, 1995–present Materials collected, characterizations partly completed and wide MRT, MIA, MIH, MIT) distribution Sub-Saharan Africa, WANA, LAC More than 70,000 accessions in ILRI, CIAT and ICARDA forage gene banks More than 60% of collections characterized on morphology New information generated on conservation and production characteristics Herbage seed unit Global tropics (LGA, LGT, MRA, MRH, MRT, MIA, MIH, MIT) More than 138,000 samples distributed Multidimensional crop India and Ethiopia, MRA, MRH, MRT, 1999–present Quality factors identified in pearl millet, sorghum and maize (see research MIA, MIH, MIT Chapter 14, this volume) Used in Indian national breeding programmes Importance of new multidimensional model for cereals selection and breeding. Finding variation in crop residue quality; near-infrared spectroscopy methods Planted forages Sub-Saharan Africa; South Asia; Forage development strategies produced (see Chapter 13, this tropical LAC volume) Wide uptake of forages by commercial farmers in LAC Continued 34 J. McIntire and J.Smith Table I.3. Continued. Domain/sub-domain Field/system Era Achievements Livestock systems, LGA, LGT, MRA, MRH, MRT, MIA, 1990s–present Baltenweck et al. (2003) economics and policy MIH, MIT Nutrient cycling and allocation (Powell et al., 1995) East Africa Understanding complementarities between wildlife and domestic stock Arid/semi-arid Niger, LGA, MRA 1990s and 2000s Hiernaux and Ayantunde (2004); Hiernaux et al. (2009) Grazing management Geography of crop–livestock interactions. Humid zones Alley farming Abandoned in 2000s Plant yields, animal nutrition and nutrient cycling Subhumid zones Borana and Kaduna Abandoned in 1990s Kenya coast Mombasa dairying (MRT) 1990s and 2000s Nicholson et al. (1999) Highlands Kenyan dairy policy (MRT, MRH) See Chapter 17 (this volume) East Africa poverty mapping See Chapter 17 (this volume) (MRT, MRH) Ethiopia dairy experiments (MRH) See Ethiopia dairy impact studies discussed in Chapter 17 (this volume) Kenya index-based Arid, semi-arid Kenya (LGA) 2009–present Created analytic, empirical and operational basis of an index- livestock insurance based livestock insurance instrument (Chantarat et al., 2013) Global environment Crop yield and Maize yield modelling 2000–present Climate change impacts on maize in Africa and LAC (Jones and environmental Thornton, 2003) indices MarkSim model 2000–present Software to generate daily climate data in sub-Saharan Africa and LAC, Climate change Livestock technology and policy 2000–present Characterization, policy guidance and estimates of mitigation and adaptation effects; see Chapter 16 (this volume) Livelihood transitions between cropping and herding. Identification of climate and livestock productivity relationships (see Chapter 16, this volume). Identification of mitigation and adaptation options. ASF, African swine fever; CBPP, contagious bovine pleuropneumonia; CCPP, contagious caprine pleuropneumonia; ECF, East Coast fever; GIS, geographic information system; ITM, infection-and-treatment method; LGA, livestock/grazing/arid; LGH, livestock/grazing/humid; LGT, livestock/grazing/tropical highlands; LAC, Latin America and the Caribbean; MIA, mixed, irrigated, arid/semi-arid; MIH, mixed, irrigated, humid/subhumid; MIT, mixed, irrigated, temperate; MRA, mixed, rainfed, arid/semi-arid; MRH, mixed, rainfed, humid/subhumid; MRT, mixed, rainfed, temperate; WANA, West Asia and North Africa; SNP, single-nucleotide polymorphism. The Evolution of IARC Livestock Research: 1975–2018 35 Papers Papers Before merger into ILRI After merger into ILRI 300 90 200 60 100 30 0 0 1975–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Citations Citations Before merger into ILRI After merger into ILRI 6,000 3,000 4,000 2,000 1,000 2,000 0 0 1975–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Rangelands Economics and policy Feed and forage Domain Livestock systems Animal production, health and genetics Total ILRI Fig. I.10. ILRI publications by research domain and publication period, 1975–2018. Sample size of 4,367 papers. (Data from www.scopus.com and www.scholar.google.com). ILRI became a global centre and posted scientists by the three parts of this volume: (i) animal to Latin America, East Asia and India. The Latin genetics, production and health; (ii) primary American programmes and several Asian pro- production; and (iii) livestock systems, includ- grammes were short lived. The principal sus- ing economics and policy, climate change and tained effort outside sub-Saharan Africa was the g ender (Fig. I.10). programme on multipurpose crops at ICRISAT in ILRI’s scientific spending was about 77% of Hyderabad22. ILRI created a model of ‘livestock as its total from 1995 to 2018, of which about 11% a pathway out of poverty’, doing less on production was devoted to capacity development. Spending by constraints and more on large- scale systems ana- scientific domain over ILRI’s lifetime has been: lysis and on the animal link to climate change. (i) livestock systems, 10%; (ii) economics and policy, 11%; (iii) primary production and the environment, including climate change, 6.4%; Achievements, 1975–2018 (iv) animal genetics, health and production, about 39%; (v) capacity development, 11%; and We organize the discussion of ILRI’s principal (vi) management, administration and technical scientific achievements as outlined in Table I.3, support, 23%. Counts by publication period 36 J. McIntire and J.Smith Animal genetics, production and health parasitaemic episodes in cattle and in Cape buffalo. Trypanosomiasis and trypanotolerance • Trypanocides and vector resistance to tryp- anocides; field studies in Ethiopia, other Our best estimate of trypanosomiasis spending East African nations, and in West Africa, is US$234 million (US$5.3 million annually), which developed cost-efficient models of which is approximately 13% of the ILCA/ trypanocide use. ILRAD/ILRI total from 1975 to 2018 (see Pref- • Definition of the epidemiology of trypano- ace to Part I, this volume). That investment led to somiasis (see Chapters 5 and 6, this many notable achievements: volume). • No realistic overall economic analysis is Characterization of the pathology of tryp- possible of trypanosomiasis and trypanotoler- anosomiasis, trypanotolerance and Thei- ance research at ILCA/ILRAD/ILRI. None the leria and description of related immuno- less, this work has produced economic benefits logical problems (see Chapters 2–4, this in several areas: volume). • Definition of molecular markers for trypa- • Generating information about the profitabil- notolerance (see Chapter 2, this volume). ity of trypanotolerant stock – economic ana- • Identifying genes controlling resistance lysis done in the ATLN (a joint effort of ILCA to trypanosomiasis in mice (Kemp et al., and ILRAD) showed internal rates of return 1997). ranging from 15 to 53% (see Chapter 3, this • Describing the genetics of infection response volume; Itty, 1992) at seven subhumid and to T. congolense (Noyes et al., 2011). humid sites of sub-Saharan Africa. • Showing the steps in cloning Boran cattle • Producing information about drug resist- (Yu et al., 2016). ance and optimal drug treatment regimens • Establishment of protocols for cultivation under various levels of tsetse and trypano- in  vitro of bloodstream stages of T. brucei somiasis challenge. and subsequently T. congolense. • Work led by partners, gathering information • Recognition that trypanosome strain com- valuable for targeting tsetse and trypanosom- plexity and surface coat antigenic variation iasis control by disease pressure and type of precludes development of an effective con- treatment. ventional vaccine that targets the immuno- dominant coat antigens. Theileriosis (see Chapters 5, 6 and • Developed a comprehensive suite of mono- 10, this volume) clonal antibodies that identified the major bovine T-cell subpopulations, immunoglobu- The closest estimate we can make of theileriosis lin isotypes, phagocytic cell populations, spending is US$181 million (US$4.1 million several bovine lymphocyte antigens (BoLAs; annually), which is approximately 10% of the also called major histocompatibility com- 1975–2018 total for ILCA/ILRAD/ILRI (see plex (MHC) antigens, and the circum- Preface to Part I, this volume). This investment sporozoite coat of T. parva sporozoites resulted in: used in host-cell invasion. • Genetic characterization of T. parva strains. • Development and maintenance in vitro of • Publication of a landmark book (Norval et al., clones of bovine CD4+ and CD8+ T-cells that 1992), which has become the global refer- facilitated analysis of monoclonal antibody ence on the epidemiology of theileriosis. specificity, investigation of T. parva antigenic • The sequencing of the T. parva genome diversity, and T. parva peptide–BoLA com- (Gardner et al., 2005). This was the second plexes that induce T. parva strain- and host apicomplexan to be sequenced and MHC-specific protective immunity. was  essential in screening for cytotoxic • Elucidation of host immune responses that T-lymphocyte antigens. control the frequency, magnitude and dur- • An understanding of tick distributions and ation of African trypanosome (and T. parva) their determinants, allowing cheaper and The Evolution of IARC Livestock Research: 1975–2018 37 more effective vector control through acari- • Identifying inter-breed differences in resist- cides and grazing management. ance to endoparasites in small ruminants. • Models of vector distribution that allowed • Creating and developing the Domestic Ani- better targeting of vector controls (Ducha- mal Genetic Resources Information System teau et al., 1997). (DAGRIS), a web tool for indigenous farm • Application of molecular technologies to animal genetic resources (Ayalew et al., improve the scientific understanding of ticks 2003; Hanotte et al., 2010). and tick control, including capacity devel- • Using single-nucleotide polymorphism (SNP) opment for a new generation of scientists. tools for in situ characterization of small- • Further development of the ITM method of holder livestock. vaccinating cattle against ECF, as discussed • Estimating the relationships between poverty in Chapter 6 (this volume). Chapter 6 of this and livestock in a spatial framework (Barendse volume presents a model of ITM use in Ken- et al., 1997; Perry et al., 2002). yan cattle, stratified by production system. • Modelling the economics of parasitic dis- Results from that model show benefit:cost eases (Perry and Randolph, 1999). ratios to ILRAD/ILRI research in the range • Long-term support to participatory epidemi- of 1.6–16.5 under justifiable assumptions ology and participatory disease surveillance about vaccine adoption and mortality in Kenya, including its use in control of avian avoided and with use of actual data on out- influenza and Rift Valley fever. Later efforts put prices and quantities. developed global capabilities in the manage- • There are two other results of ILRI research ment of zoonoses, through participatory epi- on ECF for which economic analysis is demiology techniques in Egypt, Indonesia, impossible but which probably produced Nigeria and Pakistan. ILRI has published a development benefits: (i) influencing the landmark review (Grace et al., 2012), which government of Ethiopia about the risk of R. is now a global reference on the subject. appendiculatus, which led to a change in • Innovative work on food safety in sub- Ethiopian policy on live cattle imports from Saharan Africa and South-east Asia, not- Kenya; and (ii) development of sustainable ably on aflatoxins, including publication of tick control strategies for field application in a major book on Africa (Roesel and Grace, most of the East African countries where 2014; Chapter 9, this volume). ECF is present. • Achieving thermostability in the existing peste des petits ruminants virus vaccine (Jones et al., 2016; see Chapter 7, this v olume). Achievements in other fields of animal Defining a model of CBPP (Mariner et al., genetics and health (see Chapters, 5–9, • 2006; Jores et al., 2013; see Chapter 7, this this volume) volume). The closest estimate we can make of spending • Analysing acaricide resistance in ticks (see on animal genetics, production and health other Chapter 10, this volume) and trypanocide than projects clearly labelled as ‘trypanosomiasis’ resistance in trypanosomes (see Chapter 3, or ‘theileriosis’ is US$261 million (US$6 million this volume). annually), which is approximately 15% of the • Advances in the genetics and epidemiology 1975–2018 total. of several important transboundary diseases (see Chapter 7, this volume), notably for • Creating a genetic linkage map of the bovine ASF, including: genome (Barendse et al., 1997). • Analysing the genetics of African pastoral • Isolation and genetic characterization cattle (Hanotte et al., 2002). of CBPP and ASF with better under- • Elucidating the genetic structure of cattle standing of related disease dynamics breeds (Bovine HapMap Consortium, 2009). and of immunity in African breeds and • Presenting the genetic history of African European breeds. sheep and full genetic characterization of • Developing a new vaccine challenge Ethiopian sheep. model for CCPP, improved diagnostics for 38 J. McIntire and J.Smith CBPP and ASF, and using synthetic biol- million, or 10% of the 1975–2018 total of ogy to identify vaccine candidates. US$1.75 billion. This spending double counts • Modelling of the dynamics of ASF, some work done in the ‘primary production’ do- CBPP and rinderpest, and investigating main, specifically on rangelands systems, but reservoirs of ASF and rinderpest and this has been unavoidable. Achievements from the role of carriers in ASF. this spending include: • Characterizing the principal grazing and Primary production mixed livestock systems in sub-Saharan Africa (Wilson et al., 1983; von Kaufmann The best estimate we can make of spending on et al., 1986; Wilson, 1986; Bekure et al., ‘primary production’ – broadly defined as work on 1991; McIntire et al., 1992; Coppock, 1994; primary production of four broad types – range- Hiernaux and Ayantunde, 2004). land production systems, forage diversity, plant- • Defining nutrient cycling functions in ed forages, multidimensional crops – is US$111 mixed and grazing systems (Powell et al., million (US$2.5 million annually), which is ap- 1995; Buerkert and Hiernaux, 1998; Giller proximately 6.4% of the 1975–2018 total. et al., 2011). Achievements in the field of primary produc- Estimating crop–livestock interactions tion are covered in Part II (see Chapters 11–14, • across continents (Baltenweck et al., this volume): 2003). • Completing landmark studies of grazing • Delineating the relationships between live- systems in Mali, Niger, Ethiopia, and Kenya stock and poverty (Thornton et al., 2002; (see Chapters 11, 15 and 16, this volume). Randolph et al., 2007). • Elucidating the competition between do- • Completing characterization of a mixed mestic livestock and wildlife in East Africa farming and grazing system in semi-arid (see Chapter 11, this volume). western Niger where rural poverty and land • Expanding the forage gene bank, involv- degradation are acute (Hiernaux and Ayan- ing the collection and characterization of tunde, 2004; Hiernaux et al., 2009). nearly 20,000 accessions, and distribut- ing approximately 18,600 accessions and Livestock and the global distributing some 138,000 samples to more environment than 180 countries (see Chapter 12, this volume). ILRI’s limited investment in global environmental • Contributing modestly to the development research has been very productive in this cen- of improved forage cultivars for tropical tury and the institute is now a world leader in smallholders, while adding much more to this area. Achievements include: the characterization of forage options for Mapping livestock systems – scope, product- diverse tropical farming systems. • • ivity, potential for growth and relationship Completing work on multidimensional crops to income poverty – refinement of green- ( cereals, legumes and oilseeds), with important house gas emission estimates by systems (Seré global results in research methods and in dis- et al., 1996; Kruska et al., 2003; Robinson tribution of materials to national and regional et al., 2011). programmes (see Chapter 14, this volume). • • Analysing the interaction between live-Estimating the development impact of stock and the global environment (Reid planted forages for smallholders, notably et al., 2000; Thornton et al., 2009; Herrero fodder banks in the West African subhumid et al., 2013). systems (see Chapter 13, this volume). • Quantifying the greenhouse gas emission mitigation potential of tropical livestock Livestock systems, the global systems (see Chapter 16, this volume). environment and gender • Significant contributions to understanding policy measures to mitigate climate change An upper bound estimate of ILRI lifetime spend- or to adapt to its effects in livestock systems ing on livestock systems has been about US$177 (see Chapter 16, this volume). The Evolution of IARC Livestock Research: 1975–2018 39 • Under the leadership of CIAT, identifying Impact Analysis sources of biological nitrification inhibition in Brachiaria spp. (Subbarao et al., 2006, The thematic chapters in Parts I–III of this 2009). volume analyse the impact of international live- stock research in the areas led by ILRI and its partners, including other centres working on Economics and policy livestock. Each chapter first presents an executive ILRI lifetime spending on economics and policy summary, which is a concise non-technical research has been about US$198 million since statement of the principal problems, scientific 1975, or 11.3% of the 1975–2018 total of US$ impacts and development impacts and, where 1.75 billion. ILCA/ILRAD/ILRI research has con- information was available, an estimate of ex- tributed to the understanding of policy determin- penditures at ILRI and its predecessors in the ants and applications (see Chapter 17, this vol- given field. The chapters continue with a re- ume), especially on grazing systems, water view of the history of research at ILRI and its development and management, and public fi- collaborators, of the main achievements, and nance. Achievements include: an analysis of how those achievements were • Contributing to the global movement to translated into scientific and development reduce policy barriers to growth of African impact. agriculture (see Chapter 17, this volume). Each chapter answers some common questions: • Providing the analytic base and advice (Kaiti- • What were the research problems? bie et al., 2010) for the reform of Kenyan • How relevant were those problems? For dairy marketing, producing national benefits example, did ILRI and its predecessors work in net present value terms of US$230 million on the scientific problems with the highest (see Chapters 9 and 17, this volume). potential scientific and development returns? • Description of policy responses to the rapid • What were the principal scientific achieve- expansion of global livestock trade begin- ments? ning in the 1990s (Delgado et al., 1999). • What were the principal development • Developing the analytic and operational achievements? basis of an index-based livestock insurance • What were the principal capacity develop- policy, which is now in commercial use in ment achievements? eight counties of Kenya (Chantarat et al., 2013; and, most recently, Jensen et al., In answering these questions, each chapter 2019); a major achievement of the IBLI presents a history of the research and its princi- work is the collection and publication of pal findings. Each history examines the results unusually rich household data and many of ILRI and its predecessors, with this concen- high-quality papers, over several years, tration being obviously more thorough in the without which policy making is impossible. chapters on trypanosomiasis, ECF, ticks, zoo- • Estimating the rate of return to research noses and other themes that were rarely or never and development costs of the broad-bed studied by other IARCs. There is also an analysis maker tool (Rutherford et al., 2001; Ruther- of the problems studied – for example, animal ford, 2008; see Chapter 15, this volume) in health, farming systems characterization, pri- Ethiopia. mary production, livestock and climate change, • Advancing the introduction of a gender animal genetics, bovine immunology, the con- perspective into programmes across CGIAR, trol of trypanosomiasis, the behaviour and con- which forced changes in the design and trol of ticks, forage crop improvement and others conduct of research programmes that less- – as constructed from the publications of the in- ened the bias against the economic poten- stitutions and of collaborators. Each section pre- tial of women, which is particularly acute sents briefly the resources invested – money, sci- in the livestock sector given the importance entists and sites – where it has been possible to of animal production to women’s wealth reconstruct the financial and staff data in ad- and income. equate detail. 40 J. McIntire and J.Smith The book broadly defines two types of to date and Farrell’s (1957) paper on the meas- impact: urement of productive efficiency has more than 21,000 citations; (iv) the frequency of some key- • Scientific impact, as indicated by papers, words, which vary over time with new research citations and specific achievements as technologies, e.g. earlier papers would have shown in Box I.2. fewer hits on ‘DNA sequence’ and related terms; • Development impact, as indicated by net ex less obvious is the fact that expressions related to post economic benefits of given technologies income and wealth (‘poverty’, ‘poor’) rarely ap- (e.g. the ITM vaccine against ECF and the peared in ILRI published work before 2000; and broad-bed maker tool in highland Ethiopia), (v) double counting of papers among research of policy reforms promoted by ILRI (e.g. the domains – for example, many papers will hit on Kenya dairy market), by poverty effects (the ‘trypanosomiasis’ and on ‘bovine immunology’. numbers of people lifted out of poverty) and capacity development effects. Construction of Search Expressions Limitations of Citation Indices A database of some 4367 ILCA/ILRAD/ILRI pa- Publications databases, such as Scopus, Google pers was extracted from Scopus, Google Scholar Scholar, and Web of Science, have common and CGSpace for the period 1975–201823. Cit- limitations. The metadata submitted to journals ation counts from Scopus were supplemented are sometimes incomplete or in various formats. with counts from Google Scholar for papers that The algorithms used to generate the databases did not appear in Scopus or were obviously are naturally imperfect, and their outputs are not undercounted there. The bibliometrix software always completely comprehensive or accurate. of Aria and Cuccurullo (2017) was used to The databases do not always cover all journals. generate master lists of search terms from the Some papers, even from high-impact journals, database. A clustering technique was then ap- escape the search algorithms. An important plied to the master lists to identify correlations shortcoming of the Scopus databases used in among search terms and to create aggregate this volume is that they give a biased underesti- expressions from the clusters. The code and data mate of citations of many older ILCA and ILRAD can be found at www.ilri.org/dataportal/ books (e.g. Wilson, 1986, on livestock in central impact/bibliometrix. Mali; von Kaufmann et al., 1986, on mixed graz- The analysis of impact uses four models: ing and livestock systems in central Nigeria; ILCA/ILRAD, 1988, on livestock in tsetse-affected • Literature reviews by field (e.g. bovine areas) and even some of the most important immunology, forage crops, zoonoses, eco- ILRI works (Norval et al., 1992; Perry et al., nomic and policy research), including 2002; Thornton et al., 2002) These forms of analyses of citations. ‘Literature’ is usually sampling bias affect the evaluation of impact of limited to reviewed papers but includes ILRI work over time and with respect to other at times other formats, such as newspaper scientific centres. articles, conference papers and social Citation scores can also be influenced by: (i) media. the age of the paper: older papers obviously tend • Meta-analyses of research and pro- to accumulate more citations; (ii) field effects: ject data; these are limited because the some domains have higher citation scores than data are sometimes incomplete or poorly others, e.g. mathematics and computer science reported, or the experiments were them- have fewer citations than molecular biology be- selves not properly designed for impact cause the latter tends to have more references analysis. per paper; (iii) the type of paper: methodological • Cost–benefit analyses; these are few and papers tend to have the highest citations, e.g. most are ex ante, not ex post; even in the Or- Lowry et al., (1951), describing a method to ganisation for Economic Co-operation and measure protein has more than 250,000 citations Development (OECD) countries, with much The Evolution of IARC Livestock Research: 1975–2018 41 more comprehensive information on the spending infeasible. Second, even where it conduct and effects of research, individual is possible to attribute results to spending cost–benefit analyses of livestock commodity from a given project, results frameworks or system research are much less common have rarely been done in a standardized than those for crops despite the economic manner, as they are in the develop- importance of animal products. There are few ment banks. Even when project evalua- African examples of individual commodity tions are done to scientific standards, the or livestock systems research that lend criteria cannot be summed to estimate an themselves to cost–benefit analysis. aggregate return to a research problem, sci- • Project evaluations of the effective- entific field or institution. ness/efficiency/sustainability types, such as are done in the development banks and in FAO. Such evaluations could be used to measure both scientific and development Acknowledgements impact. Project evaluations were, until the past decade, of limited use in the CGIAR The authors thank Jock Anderson, Derek context for several reasons. First, most of Byerlee, Cees de Haan, Guido Gryseels, Gbassi the centres’ budgets were unrestricted until Tarawali, Shirley Tarawali, Trevor Wilson and about 2005, making attribution of specific Iain Wright for valuable comments. Notes 1 Spending data for ILCA, ILRAD and ILRI are shown at www.ilri.org/dataportal/impact/finance 2 A Standing Panel on Impact Assessment (SPIA) 2016 review calculated the following shares of Inter- national Agricultural Research Centre (IARC) livestock research spending from 1990 to 2015: ILRI, 91.7%; CIAT; 4.2%; ICARDA, 1.1%; CIP, 0.1% and other (not attributed to a single centre) of 2.8% (Jutzi and Rich, 2016: p. 14, Table 2) in a total of US$948.5 million (current dollars). 3 Trail et al. (1979a,b) were the first major reviews of trypanotolerance and the foundations of the innovative field work of the ATLN. These reviews were a collaborative effort of ILCA, ILRAD, the United Nations Environment Programme (UNEP) and the FAO. 4 A short summary of CIAT and ICARDA livestock research, based on background papers from CIAT and ICARDA scientists at www.cgspace.com. 5 To be followed in 1967 by the International Centre for Tropical Agriculture (CIAT, 1973) in Colombia and the International Institute for Tropical Agriculture (IITA) in Nigeria. 6 J. George Harrar was an agricultural scientist who worked on the original Rockefeller Foundation pro- gramme in Mexico and subsequently became president of the foundation. 7 Pritchard, in an interview (24 February 1991), traced discussions about a forerunner to ILRAD to 1963, following his attendance at a forum that considered a livestock centre for research. 8 The formal title of the Beck report is ‘An International African Livestock Centre: Task Force R eport’, 15 October 1971. 9 Influenced by Elvio Sadun, who was designated as founding Director General but died before he assumed the post. 10 ILRAD’s Board of Trustees considered work on African swine fever in 1979 but rejected the idea. 11 Personal communication to ILRAD’s Peter Gardiner in 1991. 12 It is important to recognize the legacy of EAVRO in developing the infect-and-treat ‘Muguga cocktail’ ECF vaccine, later taken up by ILRI. This is discussed in Chapter 6 (this volume). 13 Material in this section is taken from Chapter 6 (this volume) on ‘The Management and Economics of East Coast Fever.’ 14 All categories – core scientists, visiting scientists, post-doctorates and research associates – are in- cluded. The 1981 quinquennial review of ILRAD noted that the ‘number of core scientists is unusually low’ (CGIAR/TAC, 1981: p. 26). 15 The names of the external evaluations of ILRAD and ILCA changed from time to time. The first external evaluation of ILRAD (and likewise of ILCA) was called a ‘quinquennial review’ (QQR). The second external 42 J. McIntire and J.Smith evaluation of ILRAD (CGIAR/TAC, 1986a) was called the second ‘External Programme Review’ and was con- ducted in parallel to the ‘External Management Review’ (CGIAR/TAC, 1986b) of ILRAD. The second external evaluation of of ILCA (CGIAR/TAC, 1987) was called the first ‘External Programme and Management Review’ (EPMR). The third external evaluation of ILRAD (TAC/CGIAR, 1993a) was called the third ‘External Pro- gramme and Management Review,’ as was the third external evaluation of ILCA (CGIAR/TAC, 1993b). 16 These advances are discussed in Chapters 1 and 2 (this volume). 17 See Chapter 15 (this volume) on ‘African Livestock Systems Research.’ 18 Spending is in 2015 US$, unless otherwise stated; any small amounts of ILCA or ILRAD spending before 1975 were added to the 1975 spending. 19 Data on scientific fields and staff locations are shown in www.ilri.org\dataportal\impact\finance. 20 ILCA (1994) is a history of ILCA, as written by Paul J.H. Neate; Thornton and Odero (1998) is a compen- dium of ILCA/ILRAD/ILRI impact studies in various stages of completion to 1998. 21 ILCA had no formal written strategy for more than a decade after its founding. 22 ILRI-led research at ICRISAT’s Hyderabad, India, headquarters has made significant contributions to estimating the relationships among greenhouse gas emissions and biomass (see Chapter 16, this volume) and to quantifying the feed value of stover in multidimensional crops (see Chapter 14, this volume). 23 Another 340 papers were published in 2019, but they are excluded from this database because citations data would be incomplete. References Affognon, H., Coulibaly, M., Diall, O., Grace, D., Randolph, T. and Waibel, H. (2009) Étude des politiques relatives aux stratégies de gestion de la chimiorésistance dans le cadre de la lutte contre la trypano- somose en Afrique de l’Ouest: cas du Mali. ILRI Research Report 17. ILRI, Nairobi. Agyemang, K., Clifford, K. and Little, D.A. (1993) An assessment of the biological and economic efficiency in conversion of milk to growth in N’Dama calves. Animal Science 56, 165–170. Agyemang, K., Dwinger, R.H., Little, D.A. and Rowlands, G.J. (1997) Village N’Dama Cattle Production in West Africa: Six Years of Research in The Gambia. ILRI, Nairobi. Amri, A., Ali, S. and Yazbek, M. (2018) Conservation and use of temperate Mediterranean forage genetic resources at ICARDA. School of Biological Sciences, University of Birmingham, Birmingham, UK. Anderson, R.S., Levy, E. and Morrison, B. (1991) Rice Science and Development Politics: Research Strat- egies and IRRI’s Technologies Meet Asian Diversity. Oxford University Press, Oxford. Anteneh, A. (1984) Trends in sub-Saharan Africa's livestock industries. ILCA Bulletin 18, 7–15. Aria, M. and Cuccurullo, C. (2017) bibliometrix: an R-tool for comprehensive science mapping analysis. Journal of Informetrics 11, 959–975. Ayalew, W., Rege, J.E.O., Getahun, E., Tibbo, M. and Mamo, Y. (2003) Delivering systematic information on indigenous animal genetic resources – the development and prospects of DAGRIS. In: Wollny, C., Deininger, A., Bhandari, N., Maas, B., Manig, W. and Muuss, U. (eds) Proceedings of the Deutscher Tropentag 2003: Technical and Institutional Innovations for Sustainable Rural Development, 8–10 Oc- tober, University of Göttingen, Göttingen, Germany. Available at: www.tropentag.de/2003/abstracts/ full/28.pdf (accessed 20 January 2020). Baldwin, C.L., Teale, A.J., Naessens, J.G., Goddeeris, B.M., MacHugh, N.D. and Morrison W.I. (1986) Char- acterization of a subset of bovine T lymphocytes that express BoT4 by monoclonal antibodies and function: similarity to lymphocytes defined by human T4 and murine L3T4. Journal of Immunology 136, 4385–4391. Baltenweck, I., Staal, S.J., Ibrahim, M.N.M., Herrero, M., Holmann, F. et al. (2003) SLP Project on transre- gional analysis of crop-livestock systems: crop-livestock intensification and interactions across three continents: main report. ILRI, Nairobi. Barendse, W., Vaiman, D., Kemp, S.J., Sugimoto, Y., Armitage, S.M. et al. (1997) A medium-density genetic linkage map of the bovine genome. Mammalian Genome 8, 21–28. Beck, G.H. (1971) An International African Livestock Center: Task Force Report. Rockefeller Foundation, New York. Bekure S., de Leeuw, P.N., Grandin, B.E. and Neate, P.J.H. (1991) Maasai herding: an analysis of the live- stock production system of Maasai pastoralists in eastern Kajiado district, Kenya. ILCA Systems Study 4. ILCA, Addis Ababa. The Evolution of IARC Livestock Research: 1975–2018 43 Bishop, R.P., Fleischauer, C., de Villiers, E.P., Okoth, E.A., Arias, M. et al. (2015) Comparative analysis of the complete genome sequences of Kenyan African swine fever virus isolates within p72 genotypes IX and X. Virus Genes 50, 303–309. Bovine HapMap Consortium (2009) Genome-wide survey of SNP variation uncovers the genetic structure of cattle breeds. Science 324, 528–532. Buerkert, A. and Hiernaux, P. (1998) Nutrients in the West African Sudano-Sahelian zone: losses, trans- fers and role of external inputs. Journal of Plant Nutrition and Soil Science 161, 365–383. Cartwright, T.C., Anderson, F.M., Buck, N.G., Nelsen, T.C., Trail, J.C.M. et al. (1982) Systems modelling in cattle production – an application in Botswana. FAO World Animal Review 41, 40–45. CGIAR/ISPC (2014) Strategic review of livestock in the CGIAR (including ISPC synthesis and commen- tary): ISPC White Paper. TAC Secretariat, FAO, Rome. CGIAR/TAC (1981) Report of the TAC quinquennial review of the International Laboratory for Research on Animal Diseases. TAC Secretariat, FAO, Rome. CGIAR/TAC (1982) Report of the TAC quinquennial review of the International Livestock Centre for Africa (ILCA). TAC Secretariat, FAO, Rome. CGIAR/TAC (1986a) Report of the external program review of the International Laboratory for Research on Animal Diseases. TAC Secretariat, FAO, Rome. CGIAR/TAC (1986b) Report of the external management review of the International Laboratory for Research on Animal Diseases. TAC Secretariat, FAO, Rome. CGIAR/TAC (1987) Report of the second external program review of the International Livestock Centre for Africa. TAC Secretariat, FAO, Rome. CGIAR/TAC (1993a) Report of the second external program review of the International Laboratory for Research on Animal Diseases. TAC Secretariat, FAO, Rome. CGIAR/TAC (1993b) Report of the third external programme and management review of the International Livestock Centre for Africa (ILCA). CGIAR/TAC. Rome. CGIAR/TAC (1994a) An Indicative Medium-term Plan for the International Livestock Research Institute. CGIAR, Washington, DC. CGIAR/TAC (1994b) Review of CGIAR priorities and strategies. CGIAR and TAC, Rome. CGIAR/TAC (2000) Report of the first external programme and management review of the International Livestock Research Institute (ILRI). TAC Secretariat, FAO, Rome. CGIAR/TAC (2008) Report of the second external programme and management review of International Livestock Research Institute (ILRI). TAC Secretariat, FAO, Rome. Chandler, R. (1992) An Adventure in Applied Science: A History of the International Rice Research Insti- tute. IRRI, Los Baños, Philippines. Chantarat, S., Mude, A.G., Barrett, C.B. and Carter, M.R. (2013) Designing index-based livestock insur- ance for managing asset risk in northern Kenya. Journal of Risk and Insurance 80, 205–237. CIAT (1973) External review team: beef production systems program. CIAT, Cali, Columbia. Clarke, S. (2007) A technocratic imperial state? The colonial office and scientific research, 1940–1960. Journal of American History 18, 453–480. Classen, G.A., Edwards, K.A. and Schroten, E.H.J. (1983) The water resource in tropical Africa and its exploitation. ILCA Research Report No. 6. ILCA. Addis Ababa. Clausen, P.H., Bauer, B., Zessin, K.H., Diall, O., Bocoum, Z. et al. (2010) Preventing and containing trypanocide resistance in the cotton zone of West Africa. Transboundary and Emerging Diseases 57, 28–32. Coppock, D.L. (1994) The Borana plateau of southern Ethiopia: synthesis of pastoral research, develop- ment, and change, 1980–91. Systems Study No. 5. ILCA, Addis Ababa. Cossins, N.J. and Upton, M. (1987) The Borana pastoral system of Southern Ethiopia. Agricultural Systems 25, 199–218. Cossins, N.J. and Upton, M. (1988a) The impact of climatic variation on the Borana pastoral system. Agri- cultural Systems 27, 117–135. Cossins, N.J. and Upton, M. (1988b). Options for improvement of the Borana pastoral system. Agricultural Systems 27, 251–278. Dahl, G. and Hjort, A. (1976) Having Herds: Pastoral Herd Growth and Household Economy. Stockholm Studies in Social Anthropology. University of Stockholm, Stockholm. Delgado, C.L., Rosegrant, M., Steinfeld, H., Ehui, S. and Courbois, B. (1999) Livestock to 2020: the next food revolution. Food, Agriculture and the Environment Discussion Paper No, 28. IFPRI, Washington, DC, FAO, Rome, and ILRI, Nairobi. 44 J. McIntire and J.Smith Duchateau, L., Kruska, R. and Perry, B.D. (1997) Reducing a spatial database to its effective dimension- ality for logistic-regression analysis of incidence of livestock disease. Preventive Veterinary Medicine 32, 207–218. Farrell, M.J. (1957). The measurement of productive efficiency. Journal of the Royal Statistical Society A 120, 253–281. Fitzgerald, D. (1986) Exporting American agriculture: the Rockefeller Foundation in Mexico, 1943–53. So- cial Studies of Science 16, 457–483. Ford, J. and Katondo, K.M. (1977) Maps of tsetse fly (Glossina) distribution in Africa, 1973 according to sub-generic groups on scale of 1:5,000,000. Bulletin of Animal Health and Production in Africa 25, 199–194. Gardner, M.J., Bishop, R., Shah, T., de Villiers, E.P., Carlton, J.M. et al. (2005) Genome sequence of Theileria parva, a bovine pathogen that transforms lymphocytes. Science 309, 134–137. Giller, K.E., Tittonell, P., Rufino, M.C., van Wijk, M.T., Zingore, S. et al. (2011) Communicating complexity: integrated assessment of trade-offs concerning soil fertility management within African farming sys- tems to support innovation and development. Agricultural Systems 104, 191–203. Grace, D., Mutua, F., Ochungo, P., Kruska, R., Jones, K. et al. (2012) Mapping of poverty and likely zoo- noses hotspots. Zoonoses Project 4. Report to the UK Department for International Development. ILRI, Nairobi. Gray, A. (1970). Antigenic variation. In: Mulligan, H.W. (ed.) The African Trypanosomiases. Wiley, London, pp. 113–116. Gray, A.R. (1984) Activities of the International Laboratory for Research on Animal Diseases. Preventive Veterinary Medicine 2, 13–21. Gryseels, G. and Anderson, F.M. (1983) Research on farm and livestock productivity in the central Ethiopian highlands: initial results, 1977–1980. ILCA Research Report No. 4. ILCA, Addis Ababa. Hanotte, O., Bradley, D.G., Ochieng, J.W., Verjee, Y., Hill, E.W. and Rege, J.E.O. (2002) African pastoralism: genetic imprints of origins and migrations. Science 296, 336–339. Hanotte, O., Ronin, Y., Agaba, M., Nilsson, P., Gelhaus, A., et al. (2003) Mapping of quantitative trait loci controlling trypanotolerance in a cross of tolerant West African N'Dama and susceptible East African Boran cattle. Proceedings of the National Academy of Sciences USA 100, 7443–7448. Hanotte, O., Dessie, T. and Kemp, S. (2010) Time to tap Africa’s livestock genomes. Science 328, 1640–1641. Harrar, J.G., Mangelsdorf, P.C. and Weaver, W. (1952) Notes on Indian agriculture. Rockefeller Archive Center, Rockefeller Foundation, New York. Headrick, D.R. (2014) Sleeping sickness epidemics and colonial responses in East and Central Africa, 1900–1940. PLoS Neglected Tropical Diseases 8, e2772. Herrero, M., Havlik, P., Valin, H., Notenbaert, A., Rufino, M.C., et al. (2013) Biomass use, production, feed efficiencies, and greenhouse gas emissions from global livestock systems. Proceedings of the Na- tional Academy of Sciences USA 110, 20888–20893. Hiernaux, P. and Ayantunde, A. (2004) The Fakara: a semi-arid agroecosystem under stress. Report of re- search activities. First phase (July 2002–June 2004) of the DMP-GEF Program (GEF/2711-02-4516). ILRI, Nairobi. Hiernaux, P., Ayantunde, A., Kalilou, A., Mougin, E., Gérard, B., et al. (2009) Trends in productivity of crops, fallow and rangelands in Southwest Niger: impact of land use, management and variable rainfall. Jour- nal of Hydrology 375, 65–77. Higgins, G. and Kassam, A. (1981) The FAO agroecological zone approach to determination of land po- tential. Pedologie 31, 147–168. Hirumi, H. and Hirumi, K. (1989) Continuous cultivation of Trypanosoma brucei blood stream forms in a medium containing a low concentration of serum protein without feeder cell layers. Journal of Parasit- ology 75, 985–989. Hirumi, H., Doyle, J.J. and Hirumi, K. (1977) African trypanosomes: cultivation of animal-infective Trypano- soma brucei in vitro. Science 196, 992–994. ILCA (1975) Evaluation and Mapping of Tropical African Rangelands. Proceedings of the Seminar: Bam- ako, Mali, 3–8 March. ILCA, Addis Ababa. ILCA (1979a) Livestock production in the subhumid zone of West Africa: A Regional Review. ILCA Sys- tems Study 2. Addis Ababa. ILCA (1979b) Small ruminant production in the humid tropics. ILCA Systems Study No. 3. ILCA, Addis Ababa. The Evolution of IARC Livestock Research: 1975–2018 45 ILCA (1994) Improving Livestock Production in Africa: Evolution of ILCA’s Programme 1974–94. ILCA, Addis Ababa. ILCA/ILRAD (1988) Livestock production in tsetse affected areas of Africa: proceedings of a meeting held in Nairobi, Kenya from the 23rd to 27th November 1987. ILRAD, Nairobi, and ILCA, Addis Ababa. ILRI (2000) ILRI medium-term plan for 2001–2003. Making the livestock revolution work for the poor. ILRI. Nairobi. ILRI (2013) ILRI forage gene bank: 30 years 1983–2013. ILRI Project Brochure. ILRI, Nairobi. Itty, P. (1992) Economics of Village Cattle Production in Tsetse Affected Areas of Africa: A Study of Tryp- anosomiasis Controlling Trypanotolerant Cattle and Chemotherapy in Ethiopia, Kenya, Côte d’Ivoire, the Gambia, Zaire and Togo. Hartung-Gorre Verlag, Konstanz, Germany. Jahnke, H.E. (1974) The Economics of Controlling Tsetse Flies and Cattle Trypanosomiasis in Africa: Examined for the Case of Uganda. No. 48. Weltforum Verlag, Munich, Germany. Jahnke, H.E. (1976) Tsetse Flies and Livestock Development in East Africa: A Study in Environmental Economics. Weltforum Verlag. Munich, Germany. Jahnke, H.E. (1982) Livestock Production Systems and Livestock Development in Tropical Africa. Kieler Wissenschaftsverlag Vauk, Kiel, Germany. Jahnke, H.E., Tacher, G., Kiel, P. and Rojat, D. (1988) Livestock production in tropical Africa, with special reference to the tsetse-affected zone. In: Livestock Production in Tsetse-affected Areas of Africa. ILCA/ILRAD Nairobi, pp. 3–21. Jensen, N., Stoeffler, Q., Fava, F., Vrieling, A., Atzberger, C., et al. (2019) Does the design matter? Comparing satellite-based indices for insuring pastoralists against drought. Ecological Economics 162, 59–73. Jones, B.A., Rich, K.M., Mariner, J.C., Anderson, J., Jeggo, M., et al. (2016) The economic impact of eradi- cating peste des petits ruminants: a benefit–cost analysis. PLoS One 11, e0149982. Jones, P.G. and Thornton, P.K. (2003) The potential impacts of climate change on maize production in Africa and Latin America in 2055. Global Environmental Change 13, 51–59. Jores, J., Mariner, J.C. and Naessens, J. (2013) Development of an improved vaccine for contagious bo- vine pleuropneumonia: an African perspective on challenges and proposed actions. Veterinary Re- search 44, 122. Jutzi, S.C. and Rich, K.M. (2016) An evaluation of CGIAR Centers' impact assessment work on live- stock-related research (1990–2014). Standing Panel on Impact Assessment (SPIA), CGIAR Inde- pendent Science and Partnership Council (ISPC), Rome. Kahurananga, J. (1987) End of year review of forage research for the past five years. ILCA, Addis Ababa. Kaitibie, S., Omore, A., Rich, K. and Kristjanson, P. (2010) Kenya dairy policy change and economic import- ance. World Development 38, 1494–1505. Kamuanga, M., Sigué, H., Swallow, B., Bauer, B. and d’Ieteren, G. (2001) Farmers’ perceptions of the im- pacts of tsetse and trypanosomosis control on livestock production: evidence from southern Burkina Faso. Tropical Animal Health and Production 33, 141–153. Kang, B., Reynolds, L. and Atta-Krah, A.N. (1990) Alley farming. Advanced Agronomy 43, 315–359. Katondo, K. (1984) Revision of second edition of tsetse distribution maps. International Journal of Tropical Insect Science 5, 381–388. Kemp, S.J., Iraqi, F., Darvasi, A., Soller, M. and Teale, A.J. (1997) Localization of genes controlling resist- ance to trypanosomiasis in mice. Nature Genetics 16, 194–196. King, J.M. (1983) Livestock water needs in pastoral Africa in relation to climate and forage. ILCA Research Report No 7. ILCA, Addis Ababa. Kristjanson, P.M., Swallow, B.M., Rowlands, G.J., Kruska, R.L. and de Leeuw, P. (1999) Measuring the costs of African animal trypanosomosis, the potential benefits of control and returns to research. Agri- cultural Systems 59, 79–98. Kruska, R.L., Reid, R.S., Thornton, P.K., Henninger, N. and Kristjanson, P. (2003) Mapping livestock-orient- ed agricultural production systems for the developing world. Agricultural Systems 77, 39–63. Labarta, R., Andrade, R., Marin Salazar, D., Rivera, T., Orrego, M. and Pinillos, J. (2017) The Impacts of CIAT’s Collaborative Research. CIAT, London. Available at: https://cgspace.cgiar.org/bitstream/han- dle/10568/89160/CIAT50_The_Impacts_of_CIAT%27s_Collaborative_Research.pdf?se- quence=1&isAllowed=y (accessed 20 January 2020). Leak, S.G.A. (1999) Tsetse Biology and Ecology: Their Role in the Epidemiology and Control of Trypano- somosis. CAB International, Wallingford, UK. Longhurst, W. and Heady, H. (1968) Report of a Symposium on East African Range Problems. Held at Villa Serbelloni, Lake Como, Italy, 24–28 June. Rockefeller Foundation, New York. 46 J. McIntire and J.Smith Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry 193, 265–275. Lynam, J. and Byerlee, D. (2017) Forever pioneers: CIAT: 50 years contributing to a sustainable food fu- ture… and counting. CIAT Publication No. 444. CIAT, Cali, Colombia. Lyons, M. (1992) The Colonial Disease: A Social History of Sleeping Sickness in Northern Zaire, 1900– 1940. Cambridge University Press, Cambridge. Mariner, J.C., McDermott, J., Heesterbeek, J.A.P., Thomson, G. and Martin, S.W. (2006) A model of con- tagious bovine pleuropneumonia transmission dynamics in East Africa. Preventive Veterinary Medi- cine 73, 55–74. Mariner, J.C., Jones, B.A., Hendrickx, S., El Masry, I., Jobre, Y. and Jost, C.C. (2014) Experiences in par- ticipatory surveillance and community-based reporting systems for H5N1 highly pathogenic influenza: a case study approach. Journal of EcoHealth 11, 22–35. Mariner, J.C., Gachanja, J., Tindih, S.H. and Toye, P. (2017) A thermostable presentation of the live, attenu- ated peste des petits ruminants vaccine in use in Africa and Asia. Vaccine 35, 3773–3779. McCown, R., Haaland, G. and de Haan, C. (1979) The interaction between cultivation and livestock production in semi-arid Africa. In: Hall, A.E., Cannell, G.H. and Lawton, H.W. (eds) Agriculture in Semi-arid Environ- ments. Ecological Studies (Analysis and Synthesis), Vol. 34. Springer, Berlin/Heidelberg, pp. 297–332. McIntire, J., Bourzat, D. and Pingali, P. (1992) Crop–Livestock Interaction in Sub-Saharan Africa. World Bank. Washington, DC. McKelvey, J. Jr (1971) East Coast fever and related diseases: a technical conference. Available at: http:// library.cgiar.org/bitstream/handle/10947/486/tc7107e.pdf?sequence=1 (accessed 20 January 2020). Mukasa-Mugerwa, E. (1989) A review of the reproductive performance of female Bos indicus (Zebu) cattle. ILCA Monograph No 6. ILCA, Addis Ababa. Mulligan, H.W. and Potts, W.H. (1970) The African Trypanosomiases. Wiley, London. Murray, M., Morrison, W.I. and Whitelaw, D.D. (1982) Host susceptibility to African trypanosomiasis: tryp- anotolerance. Advances in Parasitology 21, 1–68. Nicholson, C.F., Thornton, P.K., Mohammed,L., Muinga, R.W., Mwamachi, D.M., et al. (1999) Smallholder dairy technology in Coastal Kenya: an adoption and impact study. ILRI Impact Assessment Series. no. 5. ILRI, Nairobi. Norval, R.A.I., Perry, B.D. and Young, A.S. (1992) The Epidemiology of Theileriosis in Africa. Academic Press, London. Noyes, H., Brass, A., Obara, I., Anderson, S., Archibald, A.L., et al. (2011) Genetic and expression analysis of cattle identifies candidate genes in pathways responding to Trypanosoma congolense infection. Proceedings of the National Academy of Sciences USA 108, 9304–9309. Onyango, R.J. (1971) Brief history of trypanosomiasis research in East Africa and the development of E.A.T.R.O. Annales des Sociétés Belges de Médecine Tropicale, de Parasitologie, et de Mycologie 51, 559–567. Otte, M.J. and Chilonda, P. 2002. Cattle and Small Ruminant Production Systems in Sub-Saharan Africa: A Systemic Review. FAO, Rome. Paris, J., Murray, M. and McOdimba, F. (1982) A comparative evaluation of the parasitological techniques currently available for the diagnosis of African trypanosomiasis in cattle. Acta Tropica 39, 307–316. Perry, B.D. and Hansen, J.W. (eds) (1994) Modelling Vector-borne and Other Parasitic Diseases. Proceed- ings of a workshop organised Jointly by the International Laboratory for Research on Animal Diseases and the Food and Agriculture Organization of the United Nations, held at ILRAD Nairobi, Kenya, 27 November 1992. ILRAD, Nairobi. Perry, B.D. and Randolph, T.F. (1999) Improving the assessment of the economic impact of parasitic dis- eases and of their control in production animals. Veterinary Parasitology 84, 145–168. Perry, B.D., Randolph, T.F., McDermott, J.J., Sones, K.R. and Thornton, P.K. (2002) Investing in Animal Health Research to Alleviate Poverty. ILRI, Nairobi. Pingali, P., Bigot, Y. and Binswanger, H.P. (1987) Agricultural Mechanization and the Evolution of Farming Systems in sub-Saharan Africa. Johns Hopkins University Press, Baltimore/London. Pino, J.A. (1970) Discussion paper on livestock production in tropical Africa. In: Bellagio IV Conference, 3–4 December, New York. Powell, J.M., Fernández-Rivera, S., Williams, T.O. and Renard, C. (eds) (1995) Livestock and Sustainable Nutrient Cycling in Mixed Farming Systems of sub-Saharan Africa, Vol. II: Technical Papers. Pro- ceedings of an International Conference held 22–26 November 1993 in Addis Ababa, Ethiopia. ILCA, Addis Ababa. The Evolution of IARC Livestock Research: 1975–2018 47 President’s Science Advisory Committee (1967) The World Food Problem. Report of the Panel on the World Food Supply. The White House, Washington, DC. Pritchard, W., Robertson, A. and Sachs, R. (1972) Proposal for an International Laboratory for Research on Animal Diseases. Report Commissioned by the Rockefeller Foundation and Consultative Group on International Agricultural Research. Putt, S.N.H., Shaw, A.P.M., Matthewman, R.W., Bourn. D.M., Underwood, M., et al. (1980) The social and economic implications of trypanosomiasis control – a study of its impact on livestock production and rural development in northern Nigeria. Study no. 25. VEERU, Reading University, Reading, UK. Randolph, T.F., Schelling, E., Grace, D., Nicholson, C.F., Leroy, J.L., et al. (2007) Invited review: Role of livestock in human nutrition and health for poverty reduction in developing countries. Journal of Animal Science 85, 2788–2800. Reed, J.D. (1995) Nutritional toxicology of tannins and related polyphenols in forage legumes. Journal of Animal Science 73, 1516–1528. Reed, J.D., Capper, B. and Neate, P.J.H. (1988) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a workshop held 7–10 December 1987 at ILCA, Addis Ababa. ILCA, Addis Ababa. Reed, J.D., Soller, H. and Woodward, A. (1990) Fodder tree and straw diets for sheep: intake, growth, di- gestibility and the effects of phenolics on nitrogen utilisation. Animal Feed Science and Technology 30, 39–50. Reid, R.S., Kruska, R.L., Deichmann, U., Thornton, P.K. and Leak, S.G. (2000) Human population growth and the extinction of the tsetse fly. Agriculture, Ecosystems & Environment 77, 227–236. Rich, K.M. and Wanyoike, F. (2010) An assessment of the regional and national socio-economic impacts of the 2007 Rift Valley fever outbreak in Kenya. American Journal of Tropical Medicine and Hygiene 83 (Suppl.), 52–57. Robin, J.P. and Brown, L.H. (1970) An international center for rangeland research and development in Africa south of the Sahara. A proposal for cooperative international action put forward for discussion by the Nairobi Office, Ford Foundation. Ford Foundation, Nairobi. Robinson, T., Thornton, P., Franceschini, G., Kruska, R., Chiozza, F., et al. (2011) Global Livestock Pro- duction Systems. FAO, Rome, and ILRI, Nairobi. Roesel, K. and Grace, D. (2014) Food Safety and Informal Markets: Animal Products in Sub-Saharan Africa. Routledge, London. Rogers, D.J. and Robinson, T.P. (2004) Tsetse distribution. In: Maudlin, I., Holmes, P.H. and Miles, M.A. (eds) The Trypanosomiases. CAB International, Wallingford, UK. Rowlands, G.J. and Teale, A. (eds) (1994) Towards increased use of trypanotolerance: current research and future directions. Proceedings of a Workshop organized by ILRAD and ILCA held 26–29 April 1993 at ILRAD, Nairobi, Kenya. ILRAD, Nairobi, and ILCA, Addis Ababa. Rutherford, A.S. (2008) Broad bed maker technology package innovations in Ethiopian farming systems: an ex post impact assessment. ILRI Research Report 20. ILRI, Nairobi. Rutherford, A.S., Odero, A.N. and Kruska, R.L. (2001) The role of the broadbed maker plow in Ethiopian farming systems: an ex post impact assessment study. ILRI Impact Assessment Series 7. ILRI, Nairobi. Ruttan, V.W. (1977) The Green Revolution: seven generalizations. International Development Review 19, 16–22. Sandford, S. (1983a) Management of Pastoral Development in the Third World. Wiley, Chichester. Sandford, S. (1983b) Organization and management of water supplies in tropical Africa. ILCA Research Report No. 8. ILCA, Addis Ababa. Schultze-Kraft, R. and Peters, M. (2017) CIAT forages – background paper. ILRI Impact Book. ILRI, Nairobi. Seré, C., Steinfeld, H. and Groenewald, J. (1996) World livestock production systems: current status, issues and trends. FAO Animal Production and Health Paper No. 127. FAO, Rome. Shaw, A.P. (2003) Economic Guidelines for Strategic Planning of Tsetse and Trypanosomiasis Control in West Africa. FAO, Rome. Shaw, A.P. (2004). The economics of African trypanosomiasis. In: Maudlin, I., Holmes, P.H. and Miles, M.A. (eds) The Trypanosomiases. CAB International, Wallingford, UK, pp. 369–402. Shaw, A.P.M., Cecchi, G., Wint, G.R.W., Mattioli, R.C. and Robinson, T.P. (2014) Mapping the economic benefits to livestock keepers from intervening against bovine trypanosomosis in Eastern Africa. Preventive Veterinary Medicine 113, 197–210. Shaw, A.P.M., Wint, G.R., Cecchi, G., Torr, S.J., Mattioli, R.C. and Robinson, T.P. (2015) Mapping the benefit– cost ratios of interventions against bovine trypanosomosis in Eastern Africa. Preventive Veterinary Medicine 122, 406–416. 48 J. McIntire and J.Smith Subbarao, G., Ito, O., Sahrawat, K., Berry, W., Nakahara, K., et al. (2006) Scope and strategies for regu- lation of nitrification in agricultural systems – challenges and opportunities. Critical Reviews in Plant Sciences 25, 303–335. Subbarao, G.V., Nakahara, K., Hurtado, M.P., Ono, H., Moreta, D.E., et al. (2009) Evidence for biological nitrification inhibition in Brachiaria pastures. Proceedings of the National Academy of Sciences USA 106, 17302–17307. Swallow, B.M. (2000) Impacts of trypanosomiasis on African agriculture. PAAT Technical Scientific Series. FAO, Rome. Tacher, G.J., Rojat, H.E. and Keil, D.P. (1988) Livestock development and economic productivity in tset- se-infested Africa. In: Proceedings of a meeting held in Nairobi, Kenya from the 23rd to the 27th No- vember 1987 organized by the International Livestock Centre for Africa and the International Labora- tory for Research on Animal Diseases. ILCA/ILRAD, Nairobi, pp. 329–349. Thornton, P.K. (2010) Livestock production: recent trends, future prospects. Philosophical Transcripts of the Royal Society Biological Sciences 365, 2853–2867. Thornton, P.K. and Odero, A.N. (eds) (1998) Compendium of ILRI research impacts and adoption, 1975– 1998. ILRI Impact Assessment Series. No. 1. ILRI, Nairobi. Thornton, P.K., Kruska, R.L., Henninger, N., Kristjanson, P.M., Reid, R.S., et al. (2002) Mapping poverty and livestock in the developing world. ILRI. Nairobi. Thornton P.K., van de Steeg J., Notenbaert A. and Herrero M. (2009) The impacts of climate change on livestock and livestock systems in developing countries: a review of what we know and what we need to know. Agricultural Systems 101, 113–127. Trail, J.C.M., Hoste, C.H., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979a) Trypanotolerant Livestock in West and Central Africa. Vol. 1. General Study. ILCA Monograph No. 2. ILCA, Addis Ababa. Trail, J.C.M., Hoste, C.H., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979b) Trypanotolerant Livestock in West and Central Africa. Vol. 2. Country Studies. ILCA Monograph No. 2. ILCA, Addis Ababa. Traoré, S.A., Markemann, A., Reiber, C., Piepho, H.P. and Zárate, A.V. (2017) Production objectives, trait and breed preferences of farmers keeping N’Dama, Fulani Zebu and crossbred cattle and implications for breeding programs. Animal 11(4), pp. 687–695. Tribe, D., Nestel, B., Pratt, D.J. and Thomé, M. (1973) Animal production and research in tropical Africa: report of the Task Force commissioned by the African Livestock Sub-Committee of the Consultative Group on International Agricultural Research. Report submitted by Task Force Team Leader, Profes- sor D.E. Tribe. CGIAR. Washington, DC. von Kaufmann, R., Chater, S. and Blench, R. (1986) Livestock Systems Research in Nigeria’s Subhumid Zone. Proceedings of the second ILCA/NAPRI Symposium held in Kaduna, Nigeria, 29 October–2 November 1984. ILCA, Addis Ababa. Wagenaar, K.T., Diallo, A.K. and Sayers, A.R. (1986) Productivity of transhumant Fulani cattle in the inner Niger delta of Mali. ILCA Research Report No. 13. ILCA, Addis Ababa. Wilson, R.T. (1986) Livestock production in central Mali: long-term studies on cattle and small ruminants in the agropastoral system. ILCA Research Report No. 14. ILCA, Addis Ababa. Wilson, R.T., de Leeuw, P.N. and de Haan, C. (1983) Recherches sur les systèmes arides: résultats préliminaires. Rapport de Recherche No. 5. ILCA, Addis Ababa. Winrock International (1992) Assessment of Animal Agriculture in Sub-Saharan Africa. Winrock Inter- national Institute for Agricultural Development, Morrilton, Arkansas. Wint, W. and Rogers, D. (2000) Consultants report on predicted distributions of tsetse in Africa. FAO, Rome. Yu, M., Muteti, C., Ogugo, M., Ritchie, W., Raper, J. and Kemp, S. (2016) Cloning of the African indigenous cattle breed Kenyan Boran. Animal Genetics 47, 510–511. Preface to Part I: Research Spending and Publications on Animal Genetics, Production and Health This preface first reviews the estimated spending trypanotolerance, theileriosis, most commonly in the principal fields corresponding to the ten known as East Coast fever (ECF), intestinal para- chapters in the domains of animal genetics, sites and, recently, transboundary diseases – production, and health. It then sketches ‘scien- committed real spending on animal genetics, tific impact’ as a function of publications and production and health of some US$334 million citations of the International Livestock Centre (in 2015 US$) in the 20  years before ILRI, or for Africa (ILCA), International Laboratory for 50% of the total of the two predecessors to ILRI. Research on Animal Diseases (ILRAD) and Inter- Lifetime spending (1975–2018) on animal health national Livestock Research Institute (ILRI) ex- and genetics was some US$676 million, or roughly tracted from the Scopus and Google Scholar 39% of the total of US$1.75 billion. The mean of databases using search keywords relevant to the real annual spending over the lifetime of ILCA/ four fields (Aria and Cuccurullo, 2017). ILRAD/ILRI was roughly US$15 million. The shares of spending on animal genetics, production and health fell after the mid-1990s; during the Research Spending period 2011–2018, this share was 42% of the 2011–2018 total of US$529 million as ILRI di- versified into other research domains. Data from the financial and annual reports of ILCA, ILRAD and ILRI were used to compile a spending database for 1975–2018. Current spending for each year and institution was assigned Trypanosomiasis to scientific domains using spending detail by project, by scientists’ fields of expertise and, The closest estimate we can make of trypano- sporadically, from cost accounting by the institu- somiasis spending is US$234 million (US$5.3 tions. Current annual spending in US$ was million annually), which is approximately 13% converted to constant annual spending in 2015 of the total from 1975 to 2018. A paper in the US$ using the global manufacturers’ unit value trypanosomiasis field cost about US$334,000 (MUV) index. and generated about 77 citations per US$ million. ILRI and its predecessors, ILCA and ILRAD, The mean number of citations per trypanosom- made major investments in animal health and iasis paper was 26, the median was 15, and the genetics (Fig. PI.1)1. ILCA and ILRAD scientists top ten individual papers generated 19% of all cit- working on animal genetics, production and ations of ILRI papers in that field (Fig. PI.2a, b). health of all types – notably trypanosomiasis, Most ILRI papers (1975–2018) had few 49 50 Research Spending and Publications on Animal Genetics, Production and Health 400 200 0 1975–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Period Domain Theileriosis Trypanosomiasis Animal production health and genetics Total all domains Fig. PI.1. The decreasing importance of ILRI spending on theileriosis and trypanosomiasis, 1975–2018. Data from ILRI Annual Reports and Financial Reports, various years. Total, all domains spending (1975–2018) of US$1753.3 million, in 2015 US$ millions. citations; for example, 85% had fewer than 41 work of Hirumi and Hirumi (1989) on cultiva- citations, a share similar to that of global papers tion in vitro of Trypanosoma brucei, of Murray on the subject (87%). Trypanosomiasis papers et al. (1977) and Paris et al. (1982) on diagnostic were 16% of all ILRI papers and 17% of all techniques, of Murray et al. (1982) on host sus- c itations. A coarse estimate of ILRI’s lifetime ceptibility and trypanotolerance, and of Bald- contributions to research on animal trypano- win et al. (1986) and Clevers et al. (1990) on somiasis, including work on the tsetse fly vector, bovine immunology. Subsequent work focused is roughly 7% of global papers and 8% of global on estimating the economic burden of trypano- citations. It is notable that ILRAD and the Inter- somiasis, on control methods, on understanding national Trypanotolerance Centre (ITC) created and managing trypanotolerance, on project- the modern field of trypanotolerance research, ing the future of the disease. Swallow’s (2000) which had been neglected before the creation of review of trypanosomiasis and African agri- ILRAD (Fig. PI.2c); ILRI and its two predecessors culture illustrated the difficulties of estimat- contributed 68% of global papers and 91% of ing the impacts of this disease by summarizing global citations in this field (see Chapters 2–5, the variability in such basic parameters as this volume). morbidity, mortality and calving rates associ- The major papers on trypanosomiasis are ated with trypanosomiasis. The model of Reid older and tend to report advances in methods et al. (2000) projected that the tsetse fly vector and in understanding of the basic mechanisms would continue to cover wide areas of sub- of trypanosomiasis. Highlights include the Saharan Africa at least until 2040 despite Spending in millions of 2015 US$ Research Spending and Publications on Animal Genetics, Production and Health 51 (a) Citations 40,000 30,000 20,000 10,000 0 Papers 3,000 2,000 1,000 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Global (b) 6,000 4,000 2,000 0 1–20 21–40 41–80 81–200 >200 Ranges of citations per paper Institution ILRI Global Fig. PI.2. (a) The decreasing ILRI share of global publications on trypanosomiasis, 1977–2018. ILRI sample = 691 papers; global sample = 9,681 papers. (b) Frequency of citations of ILRI and global publications on trypanosomiasis, 1977–2018. Merged ILRI and global sample size = 10,372 papers. (c) Domination of the field of trypanotolerance by ILRI papers, 1977–2018. ILRI sample = 127 papers; global sample = 185 papers. (Data from www.scopus.com/.) Counts of papers Counts 52 Research Spending and Publications on Animal Genetics, Production and Health (c) Citations 800 600 400 200 0 Papers 40 30 20 10 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Global Fig. PI.2. Continued. dramatically higher rural population density ECF papers were 11% of all ILRI papers and and cropping intensity. The field work reported 11% of all citations of ILRI papers. The top ten in Chapters 2 and 3 showed the difficulty of papers generated 16% of all ILRI papers in the managing trypanosomiasis by controlling vec- field of ECF. A coarse estimate of ILRI’s lifetime tors with insecticides and by treating infected contributions to theileriosis research, includ- animals with drugs in the absence of a vaccine. ing work on the tick vector, is roughly 13% of global theileriosis papers and 19% of global theileriosis citations. Theileriosis Most of the major epidemiology (e.g. Norval et al., 1992, and the work cited in Chapters 5 and The closest estimate we can make of theilerio- 6, this volume) involved the study of Theileria sis spending in the total is US$181 million parva. Notable advances were made in eluci- (US$4.1 million annually), which is approxi- dating the genetics of Theileria spp., leading mately 10% of the 1975–2018 total. A paper eventually to the sequencing of T. parva (Gard- in the ECF field cost about US$384,000 and ner et al., 2005) and of T. annulata (Pain et al., generated about 64 citations per US$ million. 2005). A particular innovation in ILRAD/ILRI The mean number of citations per ECF paper epidemiology on theileriosis was the initial was 25, and the median was 14 (Fig. PI.3a, b). development of bioeconomic models of ECF Counts Research Spending and Publications on Animal Genetics, Production and Health 53 (a) Citations 12,500 10,000 7,500 5,000 2,500 0 Papers 1,000 500 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Global (b) 2,000 1,000 0 1–20 21–40 41–80 81–200 >200 Ranges of citations per paper Institution ILRI Global Fig. PI.3. (a) ILRI and other publications on theileriosis and related problems, 1977–2018. ILRI sample = 471 papers; global sample = 3510 papers. (b) Frequency of citations of ILRI and other publications on theileriosis and related problems, 1977–2018. Merged ILRI and global sample size = 3,981 papers. (Data from www.scopus.com/.) Counts of papers Counts 54 Research Spending and Publications on Animal Health and Genetics (Mukhebi et al., 1992), integrating field data fever’, ‘African swine fever’, ‘peste des petits ru- collection into geographic information systems minants’ and for several mycoplasmas. The (GIS) data derived from remote sensing and mean of citations per ‘other problems’ paper other sources. was 25 and the median was ten. ‘Other prob- lems’ papers were 29% of all ILRI papers and 26% of all ILRI citations (Fig. PI.5, b). Immunology The keywords for the genetics field were limited to recent terminology because of the Spending on immunology cannot be clearly obvious possibility of including practically all distinguished from other fields given that it was papers based on laboratory work. For example, always integrated into trypanosomiasis and ECF the keyword ‘sequencing’ in the context of DNA/ investigations and given that ILRAD/ILRI scien- RNA analysis appears in fewer than 20 papers tists worked across disciplines. ILRI produced before the year 2000 but appears frequently after 894 papers on some aspect of immunology, with that year. Given the explosive growth in modern a mean number of citations per paper of 23 and genetics, we decided not to estimate a share of a median of 14 (Fig. PI.4a, b). Immunology ILRI papers in global publications, but it would papers were 20% of lifetime ILRI papers and evidently be quite small given the modest research 20% of citations. The top ten ILRI immunology investment made by ILRI over its lifetime in the papers produced 10% of ILRI lifetime citations genetics subfield. Despite contributing a small in immunology. A rough estimate of ILRI’s life- share of global publications in animal genetics, time contributions to research on immunology ILRI has several landmark papers in that domain, involving human and animal trypanosomiasis as shown in Chapter 1 (this volume), notably on and theileriosis is roughly 20% of global papers the history and structure of the bovine genome and 24% of global citations. and on the genetics of trypanotolerance. Swine The failure to develop an effective vaccine and poultry genetics have only become more against trypanosomiasis should not prevent important at ILRI over the past decade, following recognition of the wide scientific impact of the with a long lag the addition of pigs and poultry immunology and immunoparasitology work to ILRI’s mandate in 1995. of ILRAD in its 20 years of life (see Chapter 4, The remarkable expansion of ILRI’s work this volume). Although the contribution of on food safety, zoonoses and transboundary published ILRI work to the global effort on ani- diseases in this century, as shown by the spend- mal immunology has weakened noticeably ing and bibliometrics data, has not produced a after the mid-1990s, as immunology became a measurable economic rate of return, but it is lower priority in the views of ILRI manage- likely that it will do so soon (see Chapters 7–9, ment, those earlier contributions remain im- this volume). portant to this day. One measure of major ILRI scientific contri- butions is the share of ILRI papers in the top 5% of citations in a given field, compared with the Genetics, food safety, transboundary share of all papers – ILRI plus global – in the top diseases and zoonoses 5% (Fig. PI.6). This subsample was limited to pa- pers having at least ten citations. ILRI’s global Spending on the wide domain including animal leadership in studies of Theileria spp. and of bo- genetics, food safety, transboundary diseases vine immunology are notable using this metric. and zoonoses cannot be quantified accurately. The ILRI share of all trypanosomiasis papers was The closest estimate we can make of spending about 9%; the ILRI share of papers in the top 5% on animal genetics, production and health of citations was about 6%. The corresponding other than projects clearly labelled as ‘trypano- shares of ILRI papers in the Theileria field were somiasis’ or ‘theileriosis’ is US$261 million 16% and 12%, respectively; the shares were 21% (US$6 million annually), which is approxi- and 13%, respectively, for immunology papers. mately 15% of the 1975–2018 total. We cre- The heterogeneity of the ‘other problems’ field ated an ‘other problems’ category with key- made it unrealistic to calculate the place of ILRI words for ‘zoonoses’, ‘food safety’, ‘Rift Valley papers in the distribution of global publications. Research Spending and Publications on Animal Genetics, Production and Health 55 (a) Citations 20,000 15,000 10,000 5,000 0 Papers 2,000 1,000 1,000 500 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Global (b) 3,000 2,000 1,000 0 1–20 21–40 41–80 81–200 >200 Ranges of citations per paper Institution ILRI Global Fig. PI.4. (a) The domination of ILRI papers in global livestock immunology until recently, 1977–2018. ILRI sample = 894 papers; global sample = 4,428 papers. (b) Frequency of citations of ILRI and other publications in immunology, 1977–2018. Merged ILRI and global sample size = 5,322 papers. (Data from www.scopus.com/.) Counts of papers Counts 56 Research Spending and Publications on Animal Health and Genetics (a) Citations 10,000 5,000 0 Papers 600 400 200 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Domain Genetics Food safety, transboundary, zoonoses (b) 1,000 750 500 250 0 1–20 21–40 41–80 81–200 >200 Ranges of citations per paper by publication period Domain Genetics Food safety, transboundary, zoonoses Fig. PI.5. (a) The rapid increase in ILRI papers on animal genetics, food safety, transboundary diseases, zoonoses after the ILCA/ILRAD merger, 1977–2018. Genetics sample = 1,544 papers; food safety, transboundary, zoonoses = 1,108 papers. (b) Frequency of citations of ILRI publication on animal genetics, food safety, transboundary diseases and zoonoses, 1977–2018. Merged sample size = 2,652 papers. (Data from www.scopus.com/.) Counts of papers Counts Research Spending and Publications on Animal Genetics, Production and Health 57 Theileriosis Trypanosomiasis 400 150 300 100 200 50 100 0 0 Genetics, other Immunology 300 200 150 200 100 100 50 0 0 First Second Third Fourth Top 5% First Second Third Fourth Top 5% Quantiles of merged ILRI and global institutions sample Institution ILRI Global Fig. PI.6. Frequency of citations of ILRI and global institutions in animal health research by quantile, 1976–2018. Sample size = 11,430 papers having at least ten citations. (Data from www.scopus.com.) Note 1 Spending data from the International Center for Tropical Agricultural (CIAT) and the International Center for Agricultural Research in the Dry Areas (ICARDA) were not available and are not included in this discussion. References Aria, M. and Cuccurullo, C. (2017) bibliometrix: an R-tool for comprehensive science mapping analysis. Journal of Informetrics 11, 959–975. Baldwin, C.L., Teale, A.J., Naessens, J.G., Goddeeris, B.M., MacHugh, N.D. and Morrison W.I. (1986) Characterization of a subset of bovine T lymphocytes that express BoT4 by monoclonal antibodies and function: similarity to lymphocytes defined by human T4 and murine L3T4. Journal of Immun- ology 136, 4385–4391. Clevers, H., MacHugh, N.D., Bensaid, A., Dunlap, S., Baldwin, C.L., et al. (1990) Identification of a bovine surface antigen uniquely expressed on CD4–CD8– T cell receptor γ/δ+ T lymphocytes. European Jour- nal of Immunology 20, 809–817. Counts of papers > 9 citations 58 Research Spending and Publications on Animal Health and Genetics Gardner, M.J., Bishop, R., Shah, T., de Villiers, E.P., Carlton, J.M., et al. (2005) Genome sequence of Thei- leria parva, a bovine pathogen that transforms lymphocytes. Science 309, 134–137. Hirumi, H. and Hirumi, K. (1989) Continuous cultivation of Trypanosoma brucei blood stream forms in a medium containing a low concentration of serum protein without feeder cell layers. Journal of Parasitology 75, 985–989. Mukhebi, A.W., Perry, B.D. and Kruska, R. (1992) Estimated economics of theileriosis control in Africa. Preventive Veterinary Medicine 12, 73–78. Murray, M., Murray, P.K. and McIntyre, W.I. (1977) An improved parasitological technique for the diagnosis of African trypanosomiasis. Transactions of the Royal Society of Tropical Medicine and Hygiene 71, 325–326. Murray, M., Morrison, W.I. and Whitelaw, D.D. (1982) Host susceptibility to African trypanosomiasis: trypanotolerance. Advances in Parasitology 21, 1–68. Norval, R.A.I., Perry, B.D. and Young, A.S. (1992) The Epidemiology of Theileriosis in Africa. Academic Press, London. Pain, A., Renauld, H., Berriman, M., Murphy, L., Yeats, C.A., et al. (2005) Genome of the host-cell transforming parasite Theileria annulata compared with T. parva. Science 309, 131–133. Paris, J., Murray, M. and McOdimba, F. (1982) A comparative evaluation of the parasitological techniques currently available for the diagnosis of African trypanosomiasis in cattle. Acta Tropica 39, 307–316. Reid, R.S., Kruska, R.L., Deichmann, U., Thornton, P.K. and Leak, S.G., (2000). Human population growth and the extinction of the tsetse fly. Agriculture, Ecosystems & Environment 77, 227–236. Swallow, B.M. (2000) Impacts of trypanosomiasis on African agriculture. PAAT Technical Scientific Series. FAO, Rome. 1 Livestock Genetics and Breeding J.E.O. Rege1, Joel Ochieng2 and Olivier Hanotte3 1Emerge Centre for Innovations-Africa, Nairobi, Kenya; 2College of Agriculture & Veterinary Sciences, University of Nairobi, Kenya; 3International Livestock Research Institute, Addis Ababa, Ethiopia and School of Life Science, University of Nottingham, Nottingham, UK Contents Executive Summary 60 ILRI’s contribution in the global context 60 Scientific impacts 61 Livestock genetics and breeding 61 Genetic characterization and history of livestock 61 Breeding technologies 62 Development impacts 62 Capacity development 62 Partnerships 62 Introduction 63 Genetic improvement of indigenous livestock 64 Characterization of Animal Genetic Resources 66 Phenotypic characterization and breed surveys 66 Tools and methods for conducting breed surveys 66 Classification of African livestock populations 67 Molecular genetic characterization 68 First genetic history of cattle in Africa, linking livestock to human history 69 Genetic history and geography of African sheep 69 Genetic history and geography of African chickens 70 Genetic history and geography of the African dromedary 70 Establishment of a joint laboratory with CAAS in Beijing and expansion into Asia 70 ILRI’s genetic characterization as a catalyst for international interest 71 Genetics of Disease Resistance 72 Genetics of trypanotolerance 72 Genetics of resistance to gastrointestinal parasites 74 Resistance to gastrointestinal nematodes in Red Maasai sheep 75 Menz versus Horro sheep of Ethiopia 76 Small East African versus Galla goats 76 ILRI studies inspire and provide benchmarks for national experiments 76 Link between resistance and productivity 77 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 59 60 J.E.O. Rege, J. Ochieng and O. Hanotte QTLs for resistance to endoparasites in small ruminants 77 DAGRIS 78 Biorepositories/biobank facilities 79 Economic valuation 79 Tools for conservation decisions 80 Genetic Improvement of Livestock 81 Analysis of long-term breeding programmes 82 Towards the optimization of breeding programmes in Africa 83 Breeding programme design 84 Appropriate germplasm for smallholder dairy farmers 84 Optimum breed combinations 84 Sustainable germplasm supply 84 Supporting on-farm selective breeding by smallholders 84 Demonstrating potential for within-breed selection 85 Towards adapted and productive chickens for African smallholders 85 Reproductive technologies for smallholders 88 Strengthening NARS Capacity in Livestock Genetics 88 Capacity development 89 Group training 89 ILRI-SLU project 89 Conclusions and the Future 91 References 92 Executive Summary ILRI’s contribution in the global context Five challenges informed the historical choices of The initial work of the International Laboratory for research priorities in livestock genetics and breed- Research on Animal Diseases (ILRAD) concen- ing at the International Livestock Research Insti- trated on the genetic aspects of pathology and im- tute (ILRI). The first challenge was to develop munopathology of two major diseases of livestock: vaccines against two protozoan parasites causing African animal trypanosomiasis and theileriosis. African animal trypanosomiasis and theileriosis ILRI’s characterization of animal genetic resources (East Coast fever), which led to novel work on cat- (AnGR), including work on diversity studies, trypa- tle genetics. The second challenge, related to the notolerance in cattle and endoparasite resistance in first, was the need to understand trypanotoler- small ruminants, has since created unprecedented ance, the genetic ability of some animals to toler- levels of awareness in national programmes of the ate infection with the trypanosome parasite that uniqueness and potential of these genetic resources. causes trypanosomiasis in cattle. The third chal- The greater attention paid to indigenous livestock lenge, which came much later, was understand- genetic resources led to more i nformed interven- ing the genetics of environmental adaptive tions to conserve and use these resources. The in- attributes (especially diseases and climate chal- creased visibility resulted largely from research lenges) in A frican livestock and their potential results, tools and methods by ILRI and its partners. contribution to improving livestock productivity. The ‘awakening’ that ensued has enabled the na- The fourth challenge, on which ILRI’s predeces- tional agricultural research systems (NARS) of Af- sor, the International Livestock Centre for Africa rica and Asia to play significant roles in expanding (ILCA), began work in 1992, was exploiting the the characterization of their native livestock, and potential of the genetic diversity of indigenous Af- this has accelerated the coverage of many livestock rican livestock (ILCA, 1992). The fifth challenge populations across these continents. was amassing research data on the comparative While ILRI’s substantive entry into design- performance of breeds and genotypes (purebreds ing on-farm breeding strategies is more recent, and cross-breds) in different production environ- the progress already achieved has demonstrated ments to inform cross-breeding programmes. promising possibilities for setting up working Livestock Genetics and Breeding 61 models that can be used for data and informed molecular mechanisms and markers for defining breeding programmes leading to long-term gen- trypanotolerant cattle types – polymorphisms etic improvements within breeds and in cross-bred that differentiate breeds such as the trypanotol- livestock populations. erant N’Dama from non-trypanotolerant African According to altmetrics, by 2019, ILRI had Zebu types. They identified candidate chromo- contributed to 27% of the global research out- somal regions by mapping second-filial gener- puts on African livestock genetics and is a sub- ation (F2) N’Dama × Boran resource families, stantial contributor to the Scopus database on and subsequently candidate genes in pathways African animal genetics in general and on endo- responding to infection with the causative para- parasite resistance in particular. ILRI’s classifi- site, Trypanosoma congolense, through genetic cation of African cattle and small ruminant and expression analysis. This work examined breeds remains a widely used standard refer- variation in products of the major histocompati- ence. Some of the top-cited papers include: Clev- bility complex (MHC) loci, biochemical variants ers et al. (1990) on the bovine antigen structure of a variety of gene products, and both mito- (97th percentile in Scopus); Barendse et  al. chondrial and nuclear DNA polymorphisms. (1997) on the genetic linkage map of the bovine This work described a strategy for the discovery genome (97th percentile); Hanotte et al. (2002) of candidate genes responsible for phenotypic on the genetic history and structure of African variation in trypanotolerance, paving the way livestock (98th percentile); Hanotte et al. (2003) for subsequent expression analysis, in vitro test- on mapping the quantitative trait loci (QTLs) ing of candidate pathways, genome mapping controlling trypanotolerance in an N’Dama × and population genetic work. Anticipating that Boran cross; Pain et al. (2005) on the compara- identification of causative trypanotolerant gen- tive genetics of T. annulata and T. parva (97th per- ome mutations in cattle would come to light centile); and the Bovine HapMap Consortium shortly, as such mutations had already been (2009) on a genome-wide survey of variation in identified in wild species, ILRI led the cloning of single- nucleotide polymorphisms (SNPs) and the first African indigenous transgenic livestock the genetic structure of cattle breeds. by somatic cell nuclear transfer using primary embryonic fibroblasts. Tumaini (Kiswahili for ‘hope’) is the result – a Kenyan Boran bull born Scientific impacts on ILRI’s research facility in Nairobi. This cloning success has opened up the possibility of genetic- ILRI’s research findings on the genetics of dis- ally engineering African livestock using foreign ease resistance and genome mapping are widely genes for desired traits (such as disease resist- recognized. The institute has made significant ance) through genome editing at the fibroblast contributions to the discovery of candidate level followed by somatic cell nuclear transfer. genes responsible for the phenotypic variation in livestock diseases and for phenotypic and genetic Genetic characterization and history characterization of indigenous livestock using of livestock neutral markers. Together, these contributions ILRI delivered the first comprehensive classifica- have helped unravel the genetic history and tion of African cattle breed and strain groups, geography of African livestock (cattle, sheep, including the status of their risk of extinction, goats, camelids and chickens), and have made and the first and most compelling reconstruc- significant contributions to our understanding tion of the genetic history of indigenous African of Asian livestock (yak, camelids, banteng, cattle. One of the papers (Hanotte et al., 2002), chickens, cattle, sheep and goats). Many publi- published in Science, won a CGIAR Science cations of this research are highly cited globally, Award as ‘Outstanding Scientific Article’ in 2003 and the tools and methods are now used across and remains a key reference on the genetic his- Africa, Asia and other parts of the world. tory of African cattle, ranking in the 98th per- centile in Scopus citations in its field to March Livestock genetics and breeding 2020. Similarly, ILRI described the first complete ILRI delivered ‘firsts’ in many areas of livestock molecular characterization of Mongolian and genetics and breeding. ILRI scientists identified Russian yak populations, supporting their 62 J.E.O. Rege, J. Ochieng and O. Hanotte distinct genetic entities for conservation and innovation platform of the Dairy Genetics East breeding programmes, including the first mo- Africa (DGEA) project helped identify business lecular validation of cattle introgression into the opportunities in the dairy value chain (heifer yak genome. Using these novel approaches, ILRI production, semen production and field delivery), scientists and collaborators made the first report which are now operational. The DGEA project of multiple introduction routes of chicken on the provided compelling evidence that innovative African continent (terrestrial and maritime) and application of genomic and associated technolo- documentation of a directional gene flow from gies can be transformative. domestic chickens into free-ranging wild red ILRI has made global contributions to the junglefowl. Similarly, ILRI and its collaborators development of economic valuation tools to in- developed a comprehensive genetic history of Af- form the prioritization of breeds and resource rican sheep and classified Ethiopian sheep breeds. allocations for conservation programmes and ILRI also validated the first set of molecular the Global Strategy on Animal Genetic Re- markers for the genetic characterization of cam- sources for Food and Agriculture led by the elids and undertook the first molecular charac- Food and Agriculture Organization of the United terization and classification of Kenyan dromedary Nations (FAO). ILRI also established and con- breeds. These are now papers of reference for the tinues to maintain the Domestic Animal Genetic history of the dromedary, revealing the dynamics R esources Information System (DAGRIS) as a re- of their domestication and cross-c ontinental dis- search and development tool providing access persal. Similarly pioneering was the finding of to much-needed information and data for re- Ethiopia as one of the centres of genetic diversity searchers and development practitioners. (and potentially of domestication) for domestic donkeys in the Horn of Africa. Capacity development Breeding technologies ILRI has built significant capacity in livestock genetics and breeding through its master’s and Of direct relevance to millions of farmers across doctoral trainings as well as through its mentor- Africa (and Asia), ILRI made the first application ship of practising scientists in collaborative pro- of SNP technology to identify genotypes in situ jects and group training courses. While livestock in smallholder systems (disclosing what geno- genetics and breeding research remain com- types farmers currently have) and to estimate paratively weak throughout Africa, ILRI’s cap- genotype differences among a diverse range of acity strengthening efforts, made through genotypes kept by these farmers. ILRI has also graduate training, secondments, group training demonstrated the feasibility of making genetic and advisory services, have created a cadre of improvements through within-breed selection high-level experts in African countries who are in village poultry systems. generating new outputs in livestock genetics and breeding. ILRI has supported at least 200 Bachelor of Development impacts Science students, 70 Master of Science students, 65 doctoral students, 35 visiting scientists and Much of ILRI’s work has had policy and develop- postdoctoral fellows, and 150 short-term train- ment impacts in animal genetics and breeding. ees in genetics, breeding and related fields. Many Elucidation of the genetic uniqueness of indi- of the beneficiaries of these capacity strengthen- genous livestock and the special adaptive attri- ing efforts are driving the livestock genetics butes of certain breeds have provided critical research agenda across the sub-Saharan contin- evidence of the potential of indigenous livestock, ent, with some now internationally recognized which has catalysed actions for their conserva- experts in their own right. tion and use in Africa and Asia. In the absence of breed-level livestock census data, breed survey Partnerships tools developed by ILRI are being applied by countries to estimate populations and trends in Partnerships with national programmes have individual indigenous breeds. More recently, an been the basis of nearly all research at ILRI. Livestock Genetics and Breeding 63 Other than the resource populations used in than the exotics, and many efforts were made to genetics of disease resistance – the N’Dama cat- ‘upgrade’ them, the vision then being that they tle herd, the Red Maasai and Dorper sheep flocks, would gradually transition into ‘grade livestock’ the Galla and Small East African goat flocks, and and eventually into pure exotic types. This was the mouse lines in ILRI’s laboratories – all of widely considered to be the main strategy for im- ILRI’s genetics and breeding research was based proving livestock productivity in Africa and Asia. on animals owned by NARS institutions or farm- From the 1950s and 1960s through to the late er-owned animals accessed through NARS col- 1970s, the word shenzi, which is K iswahili for laboration. ILCA and ILRAD worked closely with uncivilized, uncultured, uncouth or even filthy, the International Trypanotolerance Centre (ITC) and its equivalents in other parts of Africa were in The Gambia and the Centre International de commonly used in reference to indigenous live- Recherche-Développement sur l’Elevage en zone stock. Sub-humide (CIRDES/ILRI/ITC, 2000) in Bur- At the species level, national livestock stat- kina Faso. ILRI’s work has also benefitted from istics continued to be presented at the level of collaboration with universities and national native or indigenous versus exotics and crosses. p rogrammes in the Global North (Europe, North While some results across the continent showed America and Australia), in Asia (China and that the productivity of local types could be im- Korea), and in Latin America (Brazil). FAO has proved under better management, the baseline been a critical partner of ILRI, especially was still so limited that the importation of exotic through engagement of ILRI experts in formu- breeds was seen as the means to breed more pro- lating the Global Plan of Action for Animal ductive livestock (Blench, 1997). The few suc- Genetic Resources, for example, by contributing cessful European settler farms using European to the development of tools, protocols and guide- breeds, although usually based on economically lines for characterizations across the globe. unrealistic management practices (Dunbar, 1970), seemed to validate such a strategy. Endowed with more resources than was Introduction previously available for livestock research and development on the continent and guided by Scientific understanding of African and Asian more forward-looking research agendas in- livestock has improved remarkably since ILRAD formed by stakeholder consultations, ILCA and and ILCA were established in 1973–1974. At ILRAD were able to put together well-trained that time, countries classified livestock by species teams and agricultural research and develop- with limited reference to breeds, strains or types ment programmes with the ability to mobilize (Payne, 1970). The only widely applied within- and apply available and emerging scientific tools. species distinction was that made among indi- ILCA and ILRAD programme priorities in their genous types, exotics and crosses between the early days‚ from the 1970s and to the early two. Exotic animals were uniquely and consist- 1980s, had one thing in common: a focus on ently identified as distinct breeds and further understanding context. In relation to genetics classified into output- or commodity-specific and livestock improvement, ILCA’s entry point types – dairy, meat or wool – reflecting the fact was understanding the performance of different that long-term within-breed genetic improve- livestock genotypes (indigenous, exotics and ment had successfully produced these specialized crosses) in different production systems. ILRAD’s European breeds. focus in this domain was more specific: confirming While the works of Epstein (1971) and Mason indigenous breeds that were reported to be able (1988) on classification of African livestock to survive and produce under trypanosomiasis were cited, indigenous livestock continued to be challenge (i.e. possessing the trypanotolerance lumped together as though they were a uniform trait). These initial priorities developed over time genetic group. Impressionistic accounts of indi- into major programme initiatives over three genous ‘breeds’ suggested – indeed, emphasized – d ecades – the AnGR programme, which in- that their productivity was low compared with cluded molecular genetic characterization, and European livestock. The indigenous animals were the genetics of disease resistance programme, generally considered less worthy of attention focusing on trypanosomiasis and endoparasites. 64 J.E.O. Rege, J. Ochieng and O. Hanotte While not ‘complete’ by any measure, much have tended to be short-term atomized projects more is known today about indigenous livestock with no opportunity to demonstrate tangible in developing countries; many breeds and strains genetic progress. The end result is that local have now been identified with specific attributes breeds have remained uncompetitive, especially and some anecdotal claims are now supported because the metrics used for comparisons have with data. focused on productivity as measured in terms of ILRI has contributed significantly to docu- quantities of milk, meat, eggs, growth rate, etc., menting diversity among populations within with little reference to the economics of produc- species, providing evidence to match genotypes tion under different production constraints, to environments, and unravelling the genetics such as disease, poor feeds and excessive heat. underpinning observed differences, including in The lack of performance and pedigree recording adaptive attributes anecdotally referenced in in developing countries, which has been a major earlier reports. In major ways, ILRI research has driver of the success of breeding programmes in provided the evidence now used widely in ‘calls the Global North, has compounded the problem. for action’ for the conservation of the unique In the absence of robust recording systems, the genetic diversity found in indigenous livestock, ‘eye-balling’ and memory approach used in and these results are now informing conserva- traditional livestock systems does not have the tion strategies. same discriminatory power to deliver the re- quired selection differentials and accuracies, es- pecially in view of the large environmental noise Genetic improvement of indigenous that is typical in smallholder systems. Continued livestock inferior economic performance of local breeds reduces farmers’ interest in those breeds, and The rate of progress made in understanding indi- they become threatened with extinction (Rege, genous livestock – and the role that ILRI has 2008). played in this – is much greater than that in the At the time of ILRI’s founding, two mutually genetic improvement of livestock in Africa (and reinforcing actions were needed in developing- Asia). There are at least two major and inter- country situations: (i) to increase the economic related reasons why genetic improvement of performance of indigenous animals, with more local African and Asian breeds has received output and/or less cost; and (ii) to conserve live- comparatively little attention: (i) the decades- stock breeds not yet supported by market forces. long perception that these breeds are inferior to While the early work of ILRI provided useful in- exotic breeds; and (ii) the original (and persist- formation on genotype comparisons under dif- ing) belief that cross-breeding is an effective and ferent production environments, ILRI’s direct sustainable shortcut to productivity improve- and explicit entry into the area of livestock gen- ment, with inadequate considerations given to etic improvement is more recent (starting in the important role of indigenous livestock adap- 2000). This work is taking advantage of recent tations to their environments. This continues to developments in the areas of genomics and in- be the case, even after studies have shown that formation and communication technologies to time and capital invested in selection within speed up the pace of genetic improvements in, local breeds can, and does, eventually lead to and for, smallholder livestock systems. adapted and profitable populations (FAO, 2013). This chapter summarizes ILRI’s work in Even where attempts have been made in recent livestock genetics and breeding, with a focus on decades to pay attention to local breeds, in- scientific and development impacts since the es- formed by the results of genotype characteriza- tablishment of ILCA and ILRAD. The chapter is tion, the primary focus of such initiatives has organized in the following sections: (i) charac- been to conserve and use local animals as they terization of livestock genetic resources; (ii) gen- are, with very little investment in selective etics of disease resistance/tolerance, focusing on breeding. trypanotolerance in cattle and endoparasites The few attempts made to genetically im- in small ruminants; (iii) genetic improvement prove indigenous breeds through various nu- strategies; and (iv) strengthening the capacity of cleus breeding schemes (and other approaches) NARS to conduct and use this work. Figure 1.1 Livestock Genetics and Breeding 65 History and geography Genetic gain through Genomic selection in African of African cattle diversity CBBP in small livestock (dairy cow, mapped; Mapping of ruminants. Pilot on-farm chicken). Environmental trypanotolerant QTLs in breeding platform suitability maps for cattle. A mouse model established in EA; indigenous and improved established for successful cloning of genotypes. Cattle and sheep trypanosomiasis Boran calf. Full First and second generation resource populations resistance/susceptibility genomes of African SNPs for chicken and dairy cow. Genes and for QTL search with advanced livestock sequenced polymorphisms for adaptive established. intercross lines. First (chicken, cattle, small traits indentified. Genome Comprehensive set of results on Asian ruminants). Signature of editing for types and classification of African livestock diversity. selection for adaptive ECF tolerance. Microbiome cattle breeds on Gene expression of and productivity traits for analysis of cattle and chicken Trypanotolerance phenotypes, including trypanotolerance African cattle, sheep of Djallonke sheep, understood at genome and chicken identified breeds at risk. WAD goat, and wide level in cattle Endoparasite 0s some WA 202 resistance of Red Massai shorthorn cattle sheep and Trypanotolerance confirmed; SEA Goat confirmed s of N’Dama performance of 0102 confirmed; some African livestock breeds and 00s production systems cross-breds 20 characterized and documented Genome editing, documented 90s improved and new19 breeds Breeding 980 s strategies; 1 Expansion of cloning/gene genetic editing s characterization 197 0 On-farm breed surveys; phenotypic to non- and genetic ruminants; deeper Comparative characterization of understanding of perfomance of ruminants – incl. livestock expansion to Asia; history and geography of Understanding genotypes; genetics genetics of disease genetic diversity in production systems, of trypanosomes; resistance; gene discovery Africa and Asiaand trypanotolerance host genetics cattle breeds and traits Fig. 1.1. Evolution of the ILRI genetics portfolio, 1970s to 2020. EA, East Africa; ECF, East Coast fever; CBBP, community-based breeding programme; SEA, Small East African; WA, West African; WAD, West African Dwarf. O U T C O M E S FOC US AMM E PROG R 66 J.E.O. Rege, J. Ochieng and O. Hanotte summarizes the past and present outcomes and Phenotypic characterization impacts of ILRI’s genetics programme portfolio and breed surveys since the mid-1970s and ILRI’s projected future focus in this area. Early work in phenotypic characterization of indigenous breeds and their crosses with exotic types formed the foundation of what later evolved to become animal genetics and breeding Characterization of Animal Genetic at ILRI. This work initially focused predomin- Resources antly on production traits – growth, milk yield and reproductive traits. The only substantive The developing regions of Africa and Asia have work on adaptive traits (in the broad sense) in rich reservoirs of AnGR. Until the 1990s, infor- the 1980s was the work on trypanotolerance – mation on the genetic diversity of this reservoir the ability of some livestock breeds and strains to was lacking, limiting development of livestock survive and produce under trypanosomiasis conservation and use strategies. Characteriza- challenge (Trail et  al., 1979b; ILCA, 1979; see tion of AnGR – the distillation of all available Chapters 2 and 3, this volume). Trypanotoler- knowledge, both published and unpublished, ance was important due to the high livestock that contributes to the reliable prediction of gen- productivity losses associated with trypanosom- etic performance in defined environments – in- iasis. These huge losses effectively curtail the cludes measurable/observable performance as use of livestock across vast regions of Africa that well as the underlying genetic diversity, whether are infested by the tsetse fly, which transmits related to the observable performance or not. the disease-c ausing trypanosome parasites to Work to characterize AnGR includes defining the animals when the flies take a blood meal. both the genetic attributes of a population pos- The first continent-wide effort to characterize sessing a unique genetic identity and the envir- African AnGR was spearheaded by ILCA start- onment to which it is adapted or known to be ing in 1992 (Rege and Lipner, 1992), and this partially adapted (Rege, 1992). While breed set the stage for continental AnGR characteriza- evaluation and comparison (based on pheno- tion efforts, eventually covering all the major typic information) is an important aspect of livestock species of Africa and later expanding characterization that provides information on into Asia. The initial phase of this research was performance (including adaptation), genetic aimed at developing, in collaboration with characterization refers to the description of at- NARS and FAO, baseline information on Africa’s tributes that involve specific DNA sequences. l ivestock – the number of breeds, sizes of popula- The conservation of AnGR entails actions de- tions, their indicative performance and their typ- signed to prevent loss of the genetic distinctive- ical traditional production environments. ness of different species, subspecies, breeds or populations due to concerns that existing pheno- Tools and methods for conducting types, genetic integrity and local adaptations breed surveys need to be maintained. It is, however, increas- ingly recognized that in the face of environmen- Previously, a major limitation to the phenotypic tal change the conservation of adaptability and characterization mission was the lack of the processes sustaining and providing diversity methods for conducting breed surveys and might be of greater general importance than a phenotypic characterization. In collaboration focus only on conserving specific local adapta- with FAO and other partners, ILRI scientists tions. Knowledge, including local indigenous made a major contribution to this initiative by knowledge, of population differentiations and developing new instruments for breed surveys. the temporal and spatial extent of gene flows is They developed random cluster and transect therefore essential to management decision survey methods for estimating herd/flock struc- making. This section highlights both direct and tures and partitioning species-level population indirect ILRI contributions to the characteriza- statistics into breed-level census data for use in tion of African and Asian AnGR over the past determining the risk status of ruminant breeds. four decades. They also published breed survey guidelines Livestock Genetics and Breeding 67 (Rowlands et al., 2003; Ayalew et al., 2004) for cattle. Subsequent molecular work demon- use throughout Africa and Asia. strated that African Zebu cattle are ancient hy- brids of African Bos taurus with Bos indicus of Classification of African livestock p opulations historical Asian origin. Contemporary African cattle populations were assigned to four broad Another notable contribution by ILRI and part- categories: the humpless B. taurus; the humped ners has been the comprehensive classification African Zebu, distributed widely in Africa; B. of African livestock populations through breed taurus × B. indicus derivatives (sanga), found surveys and phenotypic characterization. ILRI mainly in eastern and southern Africa; and spearheaded the first continent-wide initiative sanga × Zebu types (termed the ‘zenga’), such as to characterize African AnGR starting in 1992 the Fogera and Horro of Ethiopia and the Ngan- (Rege and Lipner, 1992). ILCA’s systematic breed da of Uganda. The taurine (humpless) type has surveys led to the first comprehensive classifica- two subgroups – longhorns (B. taurus longifrons) tion of African cattle (Rege et  al., 1994a,b,c; and shorthorns (B. taurus brachyceros) – both of Aboagye et al., 1994; Rege, 1999a). This set the which were (and still are) restricted to West and stage for what were to become major continental Central Africa. While the longhorns are repre- AnGR characterization efforts. These efforts sented by two breeds only – the N’Dama and the i nvolved many international and national Kuri – the shorthorn subgroup has numerous stakeholders and linked with a global plan under representatives (Box 1.1). The survey revealed development by FAO ‘to preserve the ancestral that sub-Saharan Africa is home to a total of at gene pool of domestic animals in the developing least 145 cattle breeds/strains comprising two world’. This work built on breed evaluation taurine longhorns, 15 taurine shorthorns, 75 a ctivities that ILCA was already undertaking in Zebu (B. indicus), 30 sanga, eight zenga, nine ‘re- collaboration with NARS under a Cattle and cent’ breeds derived from interbreeding of indi- Small Ruminant Thrusts initiative and expanded genous breeds/strains located in close proximity the scope to encompass breed surveys focusing to each other and six systematically created on identification, quantitative and qualitative composite breeds. Of the 145 breeds identified description of breeds and the natural habitats from these studies, using the FAO’s classifica- and production systems to which they were tion, 47 (about 32%) were considered to be at adapted. risk of extinction. Of the breeds identified as at The first efforts focused on cattle. The first risk of extinction, six were in the ‘rare’ category, major description of African domestic cattle had ten were ‘vulnerable’, another ten were ‘endan- been prepared by Doutressoulle (1947), while gered’ and 15 were ‘critical’. A total of 22 breeds Epstein (1971) carried out the most comprehen- previously recognized on the continent were de- sive review of the era before ILRAD. Attempts termined to have become extinct in the past cen- had previously been made to construct outlines tury. This number excluded some populations with the aid of archaeological (Epstein and that were considered to have lost individual Mason, 1984; Muzzolini, 1983; Clutton-Brock, identities due to admixtures involving two or 1989) and other (Blench, 1993; Doutressoulle, more originally distinct breeds. 1947; Epstein, 1971) evidence. The European (commercial) taurine breeds The first papers (Trail et al., 1979a,b; Shaw and their cross-breds were also found in many and Hoste, 1987) surveyed the distribution of parts of the continent, but their populations trypanotolerant cattle of West and Central Africa. then, as today, were relatively low compared Subsequent efforts by Rege et  al. (1994a,b,c), with the indigenous breeds. Taking a cue from Aboagye et  al. (1994), Rege (1999b) and Rege these surveys, NARS scientists in many coun- and Tawah (1999) represented the most system- tries, working independently or in collaboration atic classifications of African cattle and cattle with ILRI, started to undertake similar on-farm populations, and included production systems, surveys and detailed studies of individual breeds. characteristics and extinction risk. This contin- Examples include Mekonnen et al. (2012) work- ental survey confirmed that phenotypically ing on Horro cattle; Tawah and Rege (1996a,b) humped cattle – the ‘non-native’ Zebu cattle of on Gudali and white Fulani cattle; Tawah et al. Africa – constituted the majority of African (1997) on Kuri cattle; and Rege et  al. (2001), 68 J.E.O. Rege, J. Ochieng and O. Hanotte Box 1.1. The taurine cattle of West Africa. The review of West African taurine breeds sought to determine the extent to which those breeds are endangered and to inform conservation decisions. The reviews addressed origin, distribution, popula- tion statistics, habitat, management and production systems, description, adaptability, disease resist- ance and performance characteristics. West African taurine cattle breeds include the hamitic longhorns (Bos taurus longifrons) and the shorthorns (Bos taurus brachyceros). Only two longhorn breeds are found in West Africa – the N’Dama and the Kuri. The shorthorn population is much more varied and their breeds/strains are known by a variety of names, depending largely on their location, including Muturu (Liberia and Nigeria), Lagune (Benin, Côte d’Ivoire and Togo), Ghana shorthorn, Baoulé (Burkina Faso and Côte d’Ivoire), Somba (Benin and Togo), Bakosi, Bakweri, Doayo and Kapsiki (Cameroon), Lobi (Burkina Faso and Côte d’Ivoire), Manjaca (Guinea Bissau) and Logone (Chad). However, it is not clear whether all of these populations are different enough to deserve different names (this classification challenge pointed to the need for genetic diversity studies, which were to follow). With the possible exception of Kuri cattle, which live in a special environment where tsetse flies have not been recorded, these breeds have developed in tsetse-infested areas and are thought to have developed various degrees of trypanotolerance. The N’Dama is the best known, most numerous and most widely spread of the trypanotolerant breeds, with a total of nearly 5 million head spread throughout West and Central Africa. However, the population of all West African shorthorns is only around 2 million head, and several breeds, including the Muturu and Lagune, are at risk of extinction, mainly due to cross-breeding (especially with Zebus), to neglect and to reduction of pastoral lands as human population pressure increases. Mwacharo and Rege (2002) and Mwacharo African cattle. A concerted effort was initiated in et  al. (2006a) on East African Zebu. A clear 1995 to collect biological samples for DNA extrac- p attern of the different phenotypic characteris- tion for this purpose. Coinciding with the emer- tics and distribution of different types of African gence and increasing application of a wide range cattle populations emerged from these studies. of genetic markers, these studies attracted many Breed characterization soon thereafter scientists to the power of ‘neutral’ molecular became a major area of livestock research. Most markers in unravelling genetic structure and esti- of these initiatives were conducted by ILRI in mating diversity in livestock populations. collaboration with NARS (Rege and Lebbie, By 1993, scientists had developed and tested 2000), or were inspired or directly influenced by a panel of short sequence repeat markers and the work of ILRI. The work expanded to sheep methods for global use in genetic diversity studies and goats (Farm Africa/ILRI, 1996; Warui, of different livestock species (Brezinsky et  al., 2008; Abegaz et al., 2013; Taye et al., 2016) and 1993a,b,c; Kemp et  al., 1992, 1995). As such, thereafter to chickens (Halima et al., 2007; Dana ILRI was a major contributor from the start to the et  al., 2010; Dessie et  al., 2012) and camels FAO-coordinated ‘Molecular Domestic Animal (Ishag et al., 2010; Tandoh et al., 2018). Diversity’ – which identified and published panels of markers to be used globally for genetic diver- Molecular genetic characterization sity studies of different livestock species (FAO, 1993) – and to the secondary guidelines devel- When ILRAD and ILCA merged in 1995, the ILCA oped (ISAG/FAO, 2004); the FAO marker panel Animal Genetic Resources programme and the included many of those developed and tested by ILRAD Genetics of Disease Resistance programme ILRI. These markers and methodologies cata- were already organizing a global programme on lysed significant investments in AnGR character- molecular genetic diversity of livestock. The large ization in Africa using a variety of markers. The phenotypic variation observed among African co-development of the 30-microsatellite marker populations of cattle, sheep and goats was used to panel, related capacity development efforts, formulate the hypotheses that formed the basis of and the wider domestication and use of these the genetic diversity studies. As was the case with methods has led to accelerated and high-quality phenotypic characterization, the first major set of characterization of a large number of indigenous ILRI-led molecular diversity studies started with livestock breeds in Africa and Asia. This has Livestock Genetics and Breeding 69 consequently led to increased understanding of African cattle inhabit more than five dis- genetic diversity and thus informed conservation tinct major agroecological zones. Overall, Zebu efforts, including regional gene banks, Sheko cattle are common in the arid and semi-arid nor- semen banks in Ethiopia, Ankole semen banks in thern Sahelo-Sudanian zone, as well as in the Uganda and local/indigenous chicken germ-cell eastern part of the continent, including the banks in ILRI Kenya, among other efforts. highlands, whereas taurine cattle today form the majority of the herds in the subhumid and First genetic history of cattle in Africa, humid regions of West Africa, which are heavily linking livestock to human history infested with tsetse fly. Sanga are predominant in the western region of Central Africa, around Utilizing the tools and methods developed, ILRI the Great Lakes region, and in the southern part scientists established the genetic relationship of of the continent. The highest genetic diversity is more than 31 breeds of African and Asian cattle among African Zebu and sanga cattle (Kim et al., and reconstructed the first genetic history of 2017). No pure indigenous B. indicus cattle cattle in Africa (Hanotte et  al., 2002), linking occur on the African continent. These findings livestock to human history and providing a have informed strategies for genetic character- glimpse into the distant past of the peoples and ization and have direct and important implica- civilizations of Africa and Asia. The major cu- tions for AnGR conservation and use, as they mulative result of this work has been the know- point to population location phenotypes and ledge base of the history and geography of the uniqueness. genetic diversity of domestic livestock of Africa and Asia. Genetic history and geography Taken together, these studies led to the con- of African sheep clusion that the earliest cattle of Africa were of taurine B. taurus type. Subsequent waves of mi- Indigenous African sheep genetic resources grations of humped Zebu (B. indicus) animals have been classified into two main groups, fat- then reshaped the genomic landscape of African tailed and thin-tailed sheep. The fat-tailed cattle (Bradley et al., 1994; Hanotte et al., 2002; sheep are the most widely distributed, being Freeman et  al., 2004). African and European found in a large part of North Africa (from cattle seem to be more closely related and quite Egypt to Algeria) and in eastern and southern distinct from Indian cattle, the relatively large Africa (from Eritrea to South Africa). The thin- divergence providing evidence for two separate tailed sheep are present mainly in Morocco, d omestication events, presumably of different Sudan and West Africa. Historically, African subspecies of the aurochs, Bos primigenius sheep were domesticated outside of Africa. (Loftus et al., 1994). Today, the African continent They share a common ancestry with European is uniquely rich in cattle diversity, with an esti- and Asian sheep. Archaeological information mated 145–150 African cattle breeds or popula- supports separate introductions and dispersion tions recognized (Rege, 1999b; Mwai et  al., histories for the African thin-tailed and fat- 2015). Importantly, it is now well established tailed sheep. The first sheep entered Africa via that African cattle carry a taurine maternal an- the Isthmus of Suez and/or the southern Sinai cestry originating from the Near East taurine do- Peninsula between 7500 and 7000  BP. They mestication centre(s), while the possible genetic were likely of the thin-tailed type. Fat-tailed contribution of the now-extinct African auroch sheep entered Africa through its north-eastern (B. primigenius opisthonomous) remains unclear part and the Horn of Africa. Mitochondrial (Epstein, 1971; Bonfiglio et  al., 2012; Decker DNA analysis supports a common maternal et al., 2014). The pattern of introgression of the a ncestral origin for all African sheep, while Zebu genome across the southern, eastern and autosomal and Y-chromosome DNA analysis north-western parts of sub-Saharan Africa has indicates a distinct genetic history for African been well documented using autosomal and thin-tailed and fat-tailed sheep, and the main Y-chromosome-specific microsatellite loci (Brad- ancestral population of southern African fat- ley et al., 1994; Hanotte et al., 2000a; Freeman tailed sheep probably originated in East Africa et al., 2004). (Muigai and Hanotte, 2013). 70 J.E.O. Rege, J. Ochieng and O. Hanotte Genetic history and geography which could support genetic improvement pro- of African chickens grammes. It is anticipated that further conti- nent-wide studies combining archaeological, ILRI has made a major contribution to the body ancient and/or modern genetic information are of knowledge of African chicken origin and likely to shed new insights on the history of present-day diversity (e.g. Lyimo et  al., 2014). chickens in Africa, as well as globally. Sociocultural, linguistic, archaeological and his- torical data together suggest a complex history Genetic history and geography of present-day African chicken populations. of the African dromedary Introductions evidently occurred via land and sea routes followed by several subsequent disper- The dromedary has a higher initial diversity sal routes across the continent (Mwacharo et al., relative to the native distribution of its wild an- 2013b). Molecular genetic studies support the cestor on the Arabian Peninsula and to the brief origin and migration models for African domes- coexistence of early-domesticated and wild indi- tic chickens and suggest centres of origin in viduals compared with other livestock, which Asia, including South Asia and South-east Asia. show a long history of gene flow with their wild Studies have consistently found that indigenous ancestors. Mburu et al. (2003) reported results chickens are uniquely adapted to their local not supporting the classification of the indigen- agroecologies and are distinct from commercial ous Kenyan dromedary into four distinct breeds broiler and egg layer lines (Bettridge et al., 2018). based on socio‐geographical criteria; instead, More localized studies have been used to test spe- their results pointed to only two separate genetic cific hypothesis on origins. For example, results entities, the Somali population on the one hand of a study reported evidence for a dual geo- and a group including the Gabbra, Rendille and graphical and genetic origin for the indigenous Turkana populations on the other. Extending Malagasy chickens, specifically, their relation- this work to Africa and Asia, phylogenetic ana- ship to Asian breeds and, more particularly, lyses of ancient and modern dromedaries sug- to Indonesian chickens (Razafindraibe et  al., gest a history of restocking from wild relatives 2008). However, a subsequent in-depth study from the south-east coast of the Arabian Penin- (Herrera et al., 2017) has reported findings sug- sula following domestication (Almathen et  al., gesting a much stronger relationship between 2016). The results suggested that the classic Malagasy and East African chicken populations, models of domestication from multiple centres putting an interesting spin on the earlier link of and from wild conspecific individuals in isola- the genetic origin due to an important func- tion may not be applicable to the dromedary. tional genetic trait – susceptibility/resistance to Molecular genetic analyses have revealed a rela- viral (avian influenza) infection (Razafindraibe tive lack of genetic differentiation among drom- et  al., 2008). Using autosomal microsatellite edaries living in different areas; this is a possible markers, Mwacharo et  al. (2013b) identified legacy of ancient trading routes converging at three distinct autosomal gene pools (I–III) in Mediterranean ports where goods, including the eastern African chicken populations, possibly packing animals carrying them, were exchanged representing genetic signatures of separate (Almathen et al., 2016). events in the history of the continent that relate to the arrival and dispersal of village chickens Establishment of a joint laboratory with and humans across the region. Earlier, analysis CAAS in Beijing and expansion into Asia of mitochondrial DNA indicated a probable Indian subcontinent origin for the commonest Understanding the genetic diversity of domestic haplogroup and a maritime introduction for the stock requires attention to Asia both because of next commonest one from South-east and/or the importance of Asia in the history of many East Asia (Mwacharo et  al., 2011). These find- African livestock populations and because of ings not only support ancient historical mari- Asia’s endowment with livestock genetic diver- time and terrestrial contacts between Asia and sity. Thus, ILRI’s early sampling strategies for East Africa but also indicate the presence of genetic diversity studies included coverage of large maternal genetic diversity in the region, strategic areas in Asia and included more than Livestock Genetics and Breeding 71 just reference populations for African livestock et  al., 2007; Muigai et  al., 2009; Muigai and species; the coverage included species not farmed Hanotte, 2013), goats (Chenyambuga et  al., in Africa, such as water buffalo, yak, banteng 2004; Tarekegn et  al., 2018), chickens (Mwa- and Bactrian camels. Work in Asia was signifi- charo et  al., 2007; Razafindraibe et  al., 2008; cantly strengthened when, starting in 2005, Mwacharo et  al., 2011; Dessie et  al., 2012; ILRI established a joint laboratory in Beijing Wragg et al., 2012; Desta et al., 2013; Mwacha- with the Chinese Academy of Agricultural Sci- ro et  al., 2013a,b; Bettridge et  al., 2018; Park ences (CAAS) to support livestock and forage et al., 2018) and camelids (Jianlin et al., 2000; genetic resources work. Subsequent funding en- Mburu et  al., 2003; Jianlin et  al., 2004; hanced the quality of science and capacity in Almathen et  al., 2016), applying a range of AnGR for Bangladesh, Sri Lanka and Vietnam, markers, starting with the first-generation DNA among other countries, allowing the wider ap- markers (restriction fragment length polymor- plication of FAO’s panel of genetic markers. The phisms, minisatellites, microsatellites and mito- ILRI-CAAS Joint Laboratory significantly in- chondrial DNA). creased ILRI’s presence in, and impact on, the Demonstrating the unique attributes of Asian continent. Although quantitative esti- i ndigenous livestock has provided a strong ra- mates were variable, the results across all species tionale for conservation and use of indigenous began to draw attention to the uniqueness of animals. Other scientific outcomes included im- various local populations or ‘breeds’ not just in proving the body of knowledge of the origin and Africa but also across Asia, such as yak (Jianlin diversity of African cattle (Hanotte et al., 2002), et al., 2002; Xuebin et al., 2005; Qi et al., 2010), sheep (Muigai and Hanotte, 2013) and chicken camelids (Jianlin et  al., 2000, 2004); banteng (Mwacharo et  al., 2006b; Mwacharo et  al., (Nijman et  al., 2003) and cattle (Dorji et  al., 2013a), grossly expanding the knowledge base 2003). Going beyond its intended role, the joint of the history and geography of the genetic laboratory and its associated capacity working d iversity of the domestic livestock of Africa and with other laboratories enabled China, in 2018 Asia. These ‘pathfinder’ studies triggered subse- during the African swine fever outbreak, to put quent studies (Decker et al., 2014), which ana- in place polymerase chain reaction (PCR)-based lysed an expanded cattle data set worldwide diagnostics to support disease surveillance. Spe- (134 breeds), showing ancient African cattle to cifically, sequencing analysis of a 417 bp region have been first domesticated in the Fertile Cres- of the B646L/p72 gene of the virus helped trace cent of the Middle East and then brought to the outbreak to the Georgian strain of the virus A frica by migrating peoples thousands of years in Russia and Eastern Europe (Wang et  al., ago. Today, the study of the genetic diversity of 2018). livestock at the molecular level has developed into an active area of research, with African ILRI’s genetic characterization results receiving considerable new attention by as a catalyst for international interest the scientific community. ILRI’s genetics and breeding work on Afri- ILRI’s efforts in molecular characterization cata- can cattle included many ‘firsts’, such as com- lysed significant investments in this area, while prehensive classification of breed/strain groups, beginning to provide genetic links to populations including their status of risk of extinction (Rege of other regions using a variety of markers to et  al., 1994a,b,c) and reconstruction of their study subsets of populations, ranging from sin- genetic history (Hanotte et  al., 2002). Other gle countries to groups of countries in a sub- ‘firsts’ included: the complete molecular charac- region of the continent to the whole continent. terization of Mongolian and Russian yak popu- Indeed, as part of ILRI’s entry into Asia, its gen- lations, supporting their distinct genetic entities etics programme activities were among the first for conservation and breeding programmes to move ILRI’s work into Asia and to strengthen (Xuebin et  al., 2005); molecular validation of its work on cattle (Okomo et al., 1998; Hanotte cattle introgression into the yak genome (Jianlin et al., 2000b; Hanotte et al., 2002; Hassen et al., et al., 2002; Qi et al., 2010); reports of multiple 2007; Ndumu et al., 2008; Kugonza et al., 2011; introduction routes of chicken on to the African Kim et al., 2017; Taye et al., 2018), sheep (Gizaw continent (terrestrial and maritime) (Mwacharo 72 J.E.O. Rege, J. Ochieng and O. Hanotte et al., 2013a,b); a comprehensive genetic history low milk yield. This helped to draw attention to of African sheep (Muigai and Hanotte, 2013) the trypanotolerance promise (based on early and a classification of Ethiopian sheep breeds field results) of a particular Zebu cattle breed, (Gizaw et al., 2007); documentation of a direc- the Orma Boran of East Africa. Farmer inter- tional gene flow from domestic chickens into views at the time already suggested that intro- free-ranging wild red junglefowl, calling atten- ducing tolerance into larger breeds of cattle, tion to interpretation of genetic data applied to a including improved exotic (European) breeds, genetic approach to the conservation of the red would be acceptable, and this held promise as a junglefowl (Thakur et al., 2018); the molecular way to use trypanotolerance traits (see Chapters 2 characterization of the dromedary, which re- and Chapter 3, this volume). classified the four traditional Kenyan dromedary In a search of genetic markers for trypa- breeds (Somali, Turkana, Rendille and Gabbra) notolerance, two polymorphic systems of bovine into two genetic entities, the Somali and a se- lymphocyte antigens were studied in 1988 by cond group comprising the Gabbra, Rendille and ILRAD in collaboration with ILCA. These sys- Turkana populations (Mburu et al., 2003); val- tems were the MHC and a more limited poly- idation of the first set of molecular markers for morphic system of common leucocyte antigens, the genetic characterization of camelids (Jianlin detected in cattle only. The first objective of the et al., 2000); the paper of reference for the his- study was to survey the MHC and common tory of the dromedary, revealing the dynamics leucocyte antigen phenotypes of populations of of domestication and the cross-continental dis- N’Dama cattle in Zaire and The Gambia and to persal of the dromedary (Almathen et al., 2016); compare these phenotypes with corresponding and the finding of Ethiopia as one of the centres profiles of trypanosensitive Boran cattle in of genetic diversity (and potentially domestica- Kenya. The second objective was to look for tion) for domestic donkeys in the Horn of Africa a ssociations between these MHC and common (Kefena et al., 2014). leucocyte antigen phenotypes, trypanotoler- ance and the productivity of N’Dama cattle. Significant correlations were found between the two classes of lymphocyte markers and the Genetics of Disease Resistance d egree of resistance shown by trypanotolerant cattle exposed to trypanosomiasis by natural Genetics of trypanotolerance challenge. These results, which suggested the existence of genetically selectable markers for The earliest major collaboration between ILRAD the trypanotolerant trait, were investigated fur- and ILCA was field characterization of trypa- ther using larger numbers as well as family notolerant livestock – breeds reported to be able groups of cattle. The results indicated a central to survive and produce under trypanosomiasis role for immunity in the manifestation of the challenge – in Africa. The African Trypanotoler- trypanotolerant trait. ant Livestock Network (ATLN) of NARS in West, From 1987, N’Dama heifers produced at Central and eastern Africa was an important ILRAD from frozen N’Dama embryos brought tool for this work. Within the network, in-depth in 1983 from The Gambia were regularly in- investigations were undertaken at many sites duced to superovulate using Folltropin, a covering a range of trypanotolerant and trypa- follicle-s timulating hormone (Jordt and Loren- nosusceptible livestock breeds under different zini, 1990). By implanting the best of the levels of tsetse/trypanosomiasis risk and man- N’Dama embryos in Boran foster mothers, agement. ILRAD scientists recognized the tryp- ILRAD produced 24 N’Dama calves, which anotolerant attributes of certain breeds of were used in studies of trypanotolerance and B. taurus (humpless cattle), specifically the N’Da- bovine genetics. ma. Despite evidence from the ATLN that N’Da- In the 1990s, ILRAD scientists undertook ma could be productive under trypanosome research to identify molecular mechanisms or challenge without treatment, there remained markers for defining trypanotolerant cattle reluctance among farmers in East Africa to types – polymorphisms that differentiate breeds adopt the N’Dama because of its small size and such as the N’Dama and Baoulé (both B. taurus) Livestock Genetics and Breeding 73 from non-trypanotolerant African Zebu types. body weight and parasitaemia. Overall, the pro- This work, among other things, examined portion of the phenotypic variance of the trypa- variation in products of the MHC loci, biochem- notolerance traits explained by these QTLs ical variants of a variety of gene products, and remains relatively low. Moreover, the estimated both mitochondrial and nuclear DNA polymor- effects of QTLs that are detected in experiments phisms. Following an idea inspired by Morris of only moderate power tended to be overesti- Soller, of the Hebrew University of Jerusalem, mated. This suggested the presence of other to use molecular tools to map traits in cattle, QTLs affecting trypanotolerance that are segre- ILRAD created a resource F 2 population segre- gating in the N’Dama, as well as other QTLs with gating for trypanotolerance even before the effects too small to be detected. To determine technology was available for actual exploitation whether these QTLs could be used in marker- of such a resource population. assisted selection, relevant genes or markers Among the early results were identifica- tightly linked to these would need to be resolved. tion of informative polymorphisms, especially Thus, as was the case with endoparasite resist- in the MHC, in the products of a small number ance in sheep (see ‘Genetics of resistance to of selected non-MHC loci, in nuclear satellite gastrointestinal parasites’), traits that distinguish DNA and in a Y-chromosome sequence. This N’Dama and Boran cattle trypanotolerance level of molecular and genetic characterization proved to be regulated by multiple, unlinked (Teale, 1993) provided what at the time was genes (Box 1.2). considered good differentiation of the major In parallel, the use of mouse models in- subspecies of cattle, providing a major transition spired by Alan Teale and the advanced intercross from neutral to functional markers. The two work led by ILRI collaborator Ariel Darvasi (Heb- marker types had previously been considered rew University of Jerusalem) led to the mouse F 2 totally separate – ‘fingerprinting’ versus gene (Kemp et al., 1997) and F6 (Iraqi et al., 2000) fine variants. It was anticipated that advances in mapping of trypanosomiasis resistance loci in molecular and computational technologies in murine. ILRI researchers identified here three the coming years would enable the examination areas of the genome that contributed to the con- and definition of genetic diversity beyond the trol of resistance to trypanosomiasis, a result subspecies level. that represented the first mapping of QTLs con- Informed by results of this early work, trolling resistance to a haemoparasitic disease of ILRAD and its collaborators (and other research major economic importance (see Chapter 2, this groups that then entered this research area) fo- volume). More recently, demonstration of tryp- cused on genome mapping to identify loci associ- anosomiasis resistance in transgenic mice laid ated with QTLs, such as Taylor et al. (1998), who a strong foundation for subsequent research on produced a genetic map of bovine chromosome 1, ILRI’s ‘Mzima cow’ project and related goat and spanning more than 160  cM, with five labora- chicken transgenics and biodiversity work. tories contributing 31,962 informative meioses It  awakened the possibility of using existing from 70 loci; Hanotte et al. (2003), who under- immunity to engineer new therapies and trans- took F 2 mapping work, and Noyes et al. (2011), genic livestock (Willyard, 2011) and inspired the who identified candidate genes in pathways concept of transgenic livestock in which the ILRI r esponding to T. congolense infection through cattle/mouse genome mapping group linked genetic analysis and expression analysis of tryp- up with human trypanosomiasis resistance anosusceptible Kenyan Boran and trypanotoler- r esearch. ant N’Dama cattle. Results of the F2 mapping All of the above results, over time, generated strongly supported a multi-locus model for the various levels of scientific impacts by enabling inheritance of trypanotolerance (Hanotte et al., the evaluation of performance under challenge 2003). Trypanotolerance QTLs were found on rather than reliance on indicator traits, an ap- 18 chromosomes, with an allele for resistance proach with potential to improve efficiency in present at nine N’Dama QTLs and five Boran genetic modification with selection in  situ. QTLs (Hanotte et  al., 2003). Multiple QTLs on Indeed, one of the original polymorphisms in several chromosomes were found to contribute N’Dama cattle is currently being selected for at to the three major tolerance indicators: a naemia, the ILRI research station at Kapiti, Kenya. 74 J.E.O. Rege, J. Ochieng and O. Hanotte Box 1.2. Mapping trypanotolerant QTLs in African cattle. To identify QTLs controlling resistance to trypanosomiasis in cattle, ILRI made an experimental cross between trypanotolerant African N’Dama (B. taurus) and trypanosusceptible improved Kenya Boran (Zebu) cattle (Hanotte et al., 2003). Sixteen phenotypic traits were defined describing anaemia, body weight and parasitaemia. In total, 177 F2 animals and their parents and grandparents were genotyped at 477 molecular marker loci covering all 29 cattle autosomes. Total genome coverage was 82%. Puta- tive QTLs were mapped to 18 autosomes at a genome-wide false discovery rate of less than 0.20. The results were consistent with a single QTL on 17 chromosomes and two QTLs on BTA16 (Bos taurus chromosome 16). Individual QTL effects ranged from approximately 6% to 20% of the phenotypic vari- ance of the trait. Excluding chromosomes with ambiguous or non-trypanotolerance effects, the allele for resistance to trypanosomiasis originated from the N’Dama parent at nine QTLs and from the Kenya Boran at five QTLs, and at four QTLs there was evidence of an overdominant mode of inheritance. These results suggested that selection for trypanotolerance within an F2 cross between N’Dama and Boran cattle could produce a synthetic breed with higher trypanotolerance levels than the parental breeds. Noyes et al. (2011) provided both the method and the pathway from QTL analysis to SNPs by combin- ing analytics pipelines. Researchers analysed the transcriptomes of trypanotolerant N’Dama and sus- ceptible Boran cattle after infection with T. congolense, followed by sequencing expressed sequenced tag libraries from these two breeds to identify polymorphisms that might underlie previously identified QTLs, and assessed QTL regions and candidate loci for evidence of selective sweeps. They identified a previously undescribed polymorphism in trypanotolerance QTLs that affected gene function in vitro and candidate genes by their expression profile and the pathways in which they participate (Box 1.3). Box 1.3. Candidate genes involved in the response to T. congolense infection. Scientists used three strategies to obtain short lists of candidate genes within QTLs that were previ- ously shown to regulate the response to infection (Hanotte et al., 2003). They analysed the transcrip- tomes of trypanotolerant N’Dama and susceptible Boran cattle after infection with T. congolense; they sequenced expressed sequenced tagged libraries from these two breeds to identify polymorphisms that might underlie previously identified QTLs and assessed QTL regions and candidate loci for evi- dence of selective sweeps (Noyes et al., 2011). The scan of the expressed sequence tags identified a previously undescribed polymorphism in the ARHGAP15 gene in the BTA2 trypanotolerance QTL. The polymorphism affects gene function in vitro and could contribute to observed differences in expression of the mitogen-activated protein kinases (MAPK) pathway in vivo. The expression data showed that Toll-like receptor (TLR) and MAPK pathways responded to infection and the former contained TLR adaptor molecule 1 (TICAM1) located within a QTL on BTA7. Genetic analyses showed that selective sweeps had occurred at the TICAM1 and ARHGAP15 loci in African taurine cattle, making them strong candidates for the genes underlying the QTL. Candidate QTL genes were identified in other QTLs by their expression profile and the pathways in which they participate. Genetics of resistance associated loss of production, the costs of an- to gastrointestinal parasites thelmintics and the death of infected animals are some of the major concerns that drew ILRI’s Helminthiasis is of considerable significance in attention in the early 1990s. In addition, there a wide range of agroclimatic zones in Africa and were (and still are) environmental concerns Asia, as well as in other developing regions influencing anthelmintic usage, including con- of the world. This disease constitutes one of the sumer demand for animal products and pastures world’s most important constraints to small- free of chemical residues. The high cost of anthel- ruminant production (ILCA, 1991; Over et  al., mintics, their uncertain availability, the i ncreasing 1992). The widespread occurrence of infection frequency of development of drug resistance and of grazing animals with internal parasites, the the limited scope in many communal pastoral Livestock Genetics and Breeding 75 systems for controlled grazing (Mwamachi et al., markers of resistance. By 1992, the initial stud- 1995; Waller, 1997) further limit options for ies in coastal Kenya that were started before the controlling this disease. In the 1990s, it was launch of the major continent-wide programme considered unlikely that new broad-spectrum (Reynolds et  al., 1992) were already showing anthelmintics and vaccines would be available clear signs that the local Red Maasai sheep were and accessible to smallholders. Utilization of more resistant to endoparasites than the intro- host genetic variation for resistance was con- duced Dorper sheep breed. sidered a more viable and sustainable approach Building on the Red Maasai and Dorper to the control of internal parasites. herds already established at the Diani Estate of ILRI conducted in-depth reviews (Baker Baobab Farm, located some 20  km south of et  al., 1992, 1994a,b; Baker, 1998) to develop Mombasa, ILRI scientists established several gen- strategies for this work. The best documented ex- etic groups to facilitate the planned experiments. amples of sheep breeds showing resistance to In addition to pure lines of the two breeds, a endoparasites in Africa were the Red Maasai double backcross population of Red Maasai and sheep of East Africa and the Djallonké sheep of Dorper sheep was also created (with the ultimate West Africa (Preston and Allonby, 1978, 1979; goal of identifying QTLs controlling resistance Baker et  al., 2003, 1994b; Baker, 1995). to gastrointestinal nematode parasites, mainly I nformed by this review, ILRI initiated an Haemonchus contortus). Under the same pro- ambitious research programme on the charac- gramme, a goat flock was also established and terization of African small ruminants for genetic used for a study from 1994 to 1996 to compare resistance to endoparasites. E vidence of genetic the resistance to naturally acquired infections variation in sheep and goats for resistance to, or of gastrointestinal nematodes (predominantly tolerance of, gastrointestinal nematodes had H. contortus) of the Galla and Small East African first been documented 40–50 years earlier and goats in the subhumid coastal region of Kenya. comprehensively reviewed by Gray (1991) and Three possible approaches for breeding for Gray and Woolaston (1991). Although the re- resistance are: (i) breeding for resistance (re- views provided evidence for a genetic basis of duced parasite numbers, as determined by faecal r esistance to endoparasites, both between and worm egg count); (ii) breeding for resilience within breeds of certain sheep and goat popula- (ability to produce under parasite challenge, as tions, almost all published reports available at the measured by packed cell volume); and (iii) breed- time suffered from inappropriate experimental ing for the number of treatments required dur- design, particularly in terms of small sample ing parasite infection (Woolaston and Baker, sizes (numbers of animals of each breed sam- 1996). The two traits used by the ILRI team in all pled) and lack of information on how the animals the studies were faecal worm egg count and had been sampled. The evidence for genetic vari- packed cell volume. ations in resistance to endoparasites within breeds was, at this point, more convincing, with Resistance to gastrointestinal nematodes heritabilities averaging about 0.35. in Red Maasai sheep With the objective of investigating and characterizing genetic resistance to endopara- Red Maasai sheep and three-quarter Red Maasai sites in indigenous African sheep and goat lambs were consistently more resistant (lower breeds, the programme (ILCA, 1991) studied faecal egg count) and more resilient (higher Dorper and Red Maasai sheep and Galla and packed cell volume) than the Dorper and Small East African goats in coastal Kenya; and three-quarter Dorper sheep. The difference be- Menz and Horro sheep breeds in the highlands tween the backcrosses for both faecal egg count of Ethiopia. It was envisaged that a second phase and packed cell volume was about half of the dif- of the programme would investigate – by exam- ference between the purebred Red Maasai and ining groups of animals that were most resistant Dorper lambs, indicating additive gene action. and most susceptible – the immunological and Moreover, even in artificial challenge experi- genetic mechanisms controlling resistance to ments where there were no significant genotype endoparasites. This work was expected to lead differences in faecal egg counts, the Red Maasai to identification of immunological or genetic and the three-quarter Red Maasai lambs had 76 J.E.O. Rege, J. Ochieng and O. Hanotte significantly fewer worms than the Dorper and Menz and Horro sheep were compared on- three-quarter Dorper animals. station in a highlands environment (Rege et al., An experiment was conducted in two envir- 1996; Haile et al., 2002; Rege et al., 2002). The onments – coastal subhumid and semi-arid – to predominant gastrointestinal parasites at this examine genotype–environment (G×E) inter- location are Longistrongylus and Trichostrongylus actions for both productivity and resistance to spp. There was no difference in resistance to gastrointestinal nematodes (Baker et  al., 2004). parasites between the Menz and Horro breeds as Here, too, the Red Maasai were consistently more reflected by faecal egg counts. While packed cell resistant and more resilient in both environments volume is a useful indicator of resistance when than the Dorper sheep, but there were significant H. contortus is the predominant gastrointestinal G×E interactions for ewe reproductive perform- parasite, this was not the case in this highland ance, ewe and lamb mortality rates, and live site in Ethiopia. Thus, although the Menz lambs weights. Overall, taking together live weights, re- had a significantly higher packed cell volume productive performance and mortality, the Red than Horro lambs, this could not be related to Maasai breed was considerably more efficient than their resistance to endoparasites and was in- the Dorper sheep in the subhumid environment, stead attributed to an adaptation of this breed to while in the semi-arid environment, there was a higher altitude (their original habitat). The Menz negligible breed difference in productive efficiency. lambs had significantly lower mortality than In a simulation study examining opportun- Horro lambs both pre- and post-weaning, which ities for using the Red Maasai breed resource for was also considered a reflection of their better cross-breeding versus pure breeding, König et al. adaptation to the high-altitude environment. (2017) concluded that cross-breeding with Dorper sheep was not recommended for harsh Small East African versus Galla goats environmental conditions due to the large breed differences expected in such an environment, Results of experiments examining gastrointes- while the breed could be improved through tinal nematode resistance in goats (Baker et  al., within-breed selection (e.g. applying a nucleus 1998a,b, 2001) showed that the Small East Afri- breeding scheme), can kids were more resistant to gastrointestinal Taken together, studies done on the Red nematode parasites than Galla kids, with the for- Maasai by the ILRI team and partners have mer having significantly lower faecal egg counts unequivocally demonstrated, on the basis of in- in the post-weaning period (8–14-month-old kids) dicators that collectively provide a reasonably and lower mortality from birth to 14 months of reliable picture of resistance – faecal egg count age. However, there was no significant breed ef- and packed cell volume – that the Red Maasai fect on packed cell volume, but Galla kids were perform significantly better than Dorper sheep significantly heavier at all measurement times under gastrointestinal nematode challenge. between birth and 14  months. However, herit- Moreover, the associated lowered mortality rates ability estimates were generally low at 0.18±0.08 in the breed lead to much faster flock growth (mean±se) for packed cell volume and 0.13±0.07 and productivity. Heritability estimates for packed for faecal egg counts. cell volume and faecal egg count (Baker et  al., 2003) point to the use of different traits for ILRI studies inspire and provide selection: within the Red Maasai breed, the focus benchmarks for national experiments should be on resilience (i.e. selection for high packed cell volume), while for the Dorper selec- The sheep breeds used in the various ILRI studies tion, it should be feasible for both improved re- have subsequently been widely used in other sistance (low faecal egg count) and resilience studies, and interpretations of results suggest (high packed cell volume). that these early studies by ILRI served as bench- marks or references for similar research in Menz versus Horro sheep of Ethiopia Africa. For example, using Dorper sheep in their experiments, Matika et  al. (2003) found Sabi Inspired by the Red Maasai results in Kenya, sheep of Zimbabwe to be more resistant than the similar studies were initiated in Ethiopia, where Dorper – the proportion of Dorper ewes needing Livestock Genetics and Breeding 77 treatment with antihelmintics was significantly investigate methods of blocking their spread; higher than that of Sabi ewes at all sampling (ii)  to assess genetic variation in resistance to dates, and this effect was particularly marked worms in indigenous and imported breeds of at 1 and 2 months post-lambing. Getachew et al. sheep and goats; and (iii) to disseminate infor- (2015) compared the Menz sheep breed with the mation about control of worms in sheep and Washera breed in the highlands of Ethiopia and goats in the tropics. Interventions investigated found that Menz sheep were better able to main- included the s trategic use of anthelmintics, the tain live weight during the parasite challenge, usefulness of ethnoveterinary therapies, bio- while Eguale et  al. (2009) compared the Arsi logical control, improved nutrition, grazing sheep breed with both Menz and Horro on resist- management and housing, and utilizing local in- ance to liver fluke infection and concluded that digenous breeds with at least some resistance to the Horro breed was more resistant than Menz worms. The results of this research are summar- and Arsi in that environment. Like the Red Maa- ized in Sani et al. (2004). This work established sai in eastern Africa, the Djallonké stands out in that importation of purebred Merinos was not various studies in West Africa as the breed with a good idea because of their susceptibility to the most compelling evidence of parasite resist- worms, but that imported St Croix sheep showed ance (e.g. Traoré et al., 2012; Soudré et al., 2018). resistance and adaptability to tropical conditions. Following publication of the early ILRI re- In goats, the native goats in the Philippines and sults on goats, the West African Dwarf goat drew imported Boer goats were somewhat more resist- attention; studies demonstrated that, like its ant than imported Anglo-Nubian and Saanen sheep counterpart the Djallonké, it was a prom- goats. A key achievement of this work was the ising candidate breed for genetic resistance to design and testing of integrated worm control gastrointestinal nematodes (e.g. Chiejina et  al., programmes in a number of different countries 2010; Behnke et al., 2011). and management systems in South-east Asia. Link between resistance and productivity QTLs for resistance to endoparasites in small ruminants In addition to evidence for between-breed differ- ences in resistance to endoparasites, these studies The existence of within-breed variation strongly also provided the first hard evidence of genetic suggested that selection for nematode resistance variation within breeds of sheep and goats in in sheep (and goats) in breeding schemes would A frica, i.e. significant heritability estimates strongly help to reduce the direct and indirect costs of pointing to opportunities for within-b reed im- parasitism to production. However, this is not provement of this trait through selection. Specif- widely practised because of the difficulty of ically, the results confirmed that Red Maasai measuring parasite resistance or correlated in- sheep and Small East African goats are more re- direct selection criteria. It was considered, there- sistant to gastrointestinal nematode parasites fore, that identifying genes or linked markers (predominantly H. contortus) than Dorper sheep with a significant association with the variance and Galla goats, respectively. The resistant Red of indicator traits of internal nematode resist- Maasai sheep and Small East African goats were ance in sheep would facilitate the inclusion of also two to three times more productive than the nematode resistance in breeding operations. susceptible Dorper sheep and Galla goats in the New technologies were already becoming avail- subhumid coastal Kenya environment. able that could be applied for breeding purposes With the formation of ILRI in 1995, some if the genes or QTLs involved in parasite resist- new research initiatives were undertaken in ance or markers that are closely linked to these South-east Asia, including work led by a small genes could be identified. ILRI team based in the Philippines and involving The hunt for QTLs (Silva et al., 2011; Mar- ten countries. The objectives of the Australian shall et al., 2013; Benavides et al., 2015) did not Centre for International Agricultural Research point to a definitive region as an obvious candi- (ACIAR)-supported project were: (i) to assess an- date for selection. Several genomic regions fea- thelmintic resistance in gastrointestinal nema- tured in the search for markers for nematode tode parasites (worms) of sheep and goats and to resistance in sheep, and genomic regions that 78 J.E.O. Rege, J. Ochieng and O. Hanotte have been identified as being significantly asso- breeding programme focusing on one or two ciated with indicator traits for nematode resist- genomic regions. Having realized that there was ance vary widely among studies. This could be no major gene (or small number of genomic re- due to differences in experimental protocols and gions with large effects) involved in conferring materials (studies differed in the sheep breeds resistance to gastrointestinal nematodes, and and nematode species, the indicator traits for in- that there were instead many genes/genomic re- ternal nematode resistance, and the challenge gions, each with small effects, a multi-trait selec- regimes), differences in the analytical ap- tion programme was initiated at ILRI’s Kapiti proaches, population stratification, complexity Ranch for both Red Maasai and the Dorper × of the physiological processes of nematode re- Red Maasai crosses, under natural and continu- sistance or a combination of factors. Sayers and ous challenge (see section on ‘Genetic improve- Sweeney (2005) reported similar results for ment of livestock’). the MHC genes, interferon-γ gene, IgE gene and microsatellites on chromosomes 1, 5 and 6. Crawford et al. (2006) reported a large number DAGRIS of suggestive QTLs (more than one per family per trait than would be expected by chance) and Information on the extent of diversity, character- concluded that most of the genes controlling istics and use of indigenous farm AnGR in devel- this trait are of relatively small effect. Marshall oping countries is the basis for their present as et al. (2009) suggested that where traits are con- well as future sustainable utilization. Published trolled by multiple QTLs (such as age and/or im- data on this diversity were disparate, uncoordin- mune status specificity in Merino sheep), a panel ated and largely inaccessible. In view of the lack of QTLs is required to achieve significant genetic of a systematic database on this information, gains through marker‐assisted selection. ILCA initiated development of the DAGRIS in Alba-Hurtado and Muñoz-Guzmán (2013) 1993 as a web-based electronic source of infor- noted that genetically resistant sheep have mation on selected indigenous farm AnGR (Rege non-specific mechanisms that block the initial et al., 2007). This was an important product of colonization by H. contortus larvae, and that genetic characterization work over the years and they also have an efficacious response of T-helper into the future. The working objectives of type 2 cells (e.g. increases in blood and tissue DAGRIS were to: (i) compile and organize infor- eosinophils, specific IgE class antibodies, mast mation on farm AnGR from all available sources; cells, interleukin-5, interleukin-13 and tumour (ii) maintain the integrity and validity of the in- necrosis factor) that protect them against the in- formation; and (iii) disseminate the information fection. The use of a mouse model to help the in a readily accessible way to key stakeholders. genome-wide search for QTLs influencing im- Today, DAGRIS has information on breeds/ munological responses to infection with the strains of eight species, namely, buffalo (139 gastrointestinal nematode (in this case the para- breeds/strains), cattle (189), chickens (126), site Heligmosomoides polygyrus) did not help ei- dromedary camels (10), goats (83), pigs (166), ther (Iraqi et  al., 2003; Menge et  al., 2003; sheep (179) and yak (30), with options to extend Behnke et  al., 2006). Here, too, many genomic it further to cover geese, turkey and ducks. Its regions with small effects were implicated. current geographical scope is Africa and selected Thus, although there are some genomic Asian countries, with an envisaged future cover- regions that have featured in multiple studies age of developing countries in Asia and Latin across breeds (e.g. the interferon-γ region on America and the Caribbean. chromosome 3 has come up as influencing a sig- The database contains information on the nificant proportion of the variance for nematode origins, distribution, diversity, present use and resistance traits by multiple authors; e.g. Coltman status of indigenous farm AnGR from past and et  al., 2001; Paterson et  al., 2001; Beh et  al., present research results, information that can be 2002), there really is no compelling convergence used to form the basis for designing breed im- on one or a few regions that explain the large provement as well as conservation programmes. variation in resistance. Currently, there is no evi- DAGRIS is available online (http://dagris.ilri. dence pointing to the possibility of establishing a cgiar.org; accessed 28 January 2020) as well as Livestock Genetics and Breeding 79 on CD-ROM, making it accessible even to those Resources’ to provide expert guidance on valu- without internet connectivity. The utility of this ing AnGR. In the face of strengthening laws for resource is demonstrated by increasing trends in intellectual property protection of germplasm website visits and citation indices. under the Convention on Biological Diversity, it was also becoming apparent that the past colle- gial system of free exchange of germplasm Biorepositories/biobank facilities among researchers was going to break down fast, requiring that some mechanism (e.g. The Azizi Biorepository is the long-term storage through mutually agreed terms) for genetic re- system and associated informatics tools that sources movement across borders would be comprise the biorepository at ILRI. The system needed, both for research and for commercial supports and has supported a number of activ- purposes. It was recognized that, because mar- ities and projects, including ‘Infectious Diseases kets exist only for finished or nearly finished of East African Livestock’, ‘Arbovirus Incidence commercial genetic resources, the value of un- and Diversity’, ‘People, Animals and their Zoo- improved materials and the value added that noses’, ‘Dynamic Drivers of Disease in Africa were not ‘valued by commercial interest’ in the Consortium’, ‘African Bovine Trypanosomiasis’, long process of breeding (including through ef- the ILRI livestock diversity collection and ILRI’s forts of indigenous communities) could not be unique collection of pathogen isolates. The core measured directly. Yet this was going to be crit- collection is approximately 450,000 samples in ical under the Convention on Biological Diver- vapour-phase liquid nitrogen with uniquely ro- sity regime, i.e. it was important that ‘parties’ bust, secure and well-monitored ultra-cold con- (e.g. communities and countries) had a way of ditions for long-term storage. Samples and determining the values of the AnGR involved associated data collection are based on strict in an ‘exchange’, whether within countries or protocols and procedures to ensure that they across borders. During the ILRI convening, Men- meet the required and acceptable minimum delsohn (1999) made at least three arguments standards. The repository provides a platform for for economic valuation of AnGR: (i) to inform researchers to mine samples and data that can breeding programmes; (ii) to guide choice of be used in their research. This in turn means breeds in conservation programmes and (iii) to that expensively obtained samples can be used facilitate benefit sharing (in the context of the and re-used for additional purposes and that the Convention on Biological Diversity). It was also metadata associated with each sample is con- considered that economic valuation would be tinually enriched, thus enhancing efficiency. instrumental in guiding resource allocation be- The repository approach also maximizes bio- tween biodiversity conservation and other so- logical understanding as it facilitates different cially valuable endeavours. Likewise, the results pieces of information about identical samples. In of valuation would be used in various types of addition, samples and data in the repository are genetic resource conservation, research and de- managed to ensure that they meet CGIAR re- velopment, including in the design of economic quirements of open access. ILRI’s Azizi Biorepos- incentives and institutional arrangements for itory is the only one of its kind in sub-Saharan farmers/genetic resource managers and breed- Africa for the products of long-term character- ers. For example, AnGR erosion could be under- ization work. It currently contains samples of a stood in terms of the replacement of breeds and wide range of livestock species and breeds, wild- strains, not only by substitution but also through life, insects and disease-causing organisms. cross-breeding and the extinction of livestock Samples are in various forms including blood, populations. serum, milk, faecal material, DNA and RNA. Following the 1999 ILRI/FAO joint ‘Eco- nomic Valuation of Animal Genetic Resources’ workshop, ILRI initiated a programme on ‘Eco- Economic valuation nomics of AnGR Conservation and Sustainable Use’, focusing on valuation methodologies. The In 1999, FAO and ILRI convened a workshop valuation dimensions covered included: risk of on ‘Economic Valuation of Animal Genetic extinction as an important consideration for 80 J.E.O. Rege, J. Ochieng and O. Hanotte a llocating conservation resources (Reist-Marti effect in terms of reduced extinction probability. et al., 2003; Simianer et al., 2003); farmer pref- The results indicated that, in some cases, highest erences (Jabbar and Diedhiou, 2003; Duguma priority could be given to breeds for which the et  al., 2011); trader preferences (Scarpa et  al., ‘conservation potential’ (i.e. the product of ex- 2003a); effects of subsidies of local and im- tinction probability and marginal diversity) is ported germplasm (Drucker et al., 2006); value maximum, and that these are not necessarily the and impact of cross-breeding (Karugia et  al., most endangered breeds. 2001; Ayalew et al., 2003); valuation in the con- Some of the results of this work have text of community-based management of AnGR pointed to the fact that conservation goals can (Drucker, 2003); specific genetic resources for be generally modest (FAO, 2013), and in view of specific environments (Scarpa et  al., 2003b; some of these findings, it has been estimated Omondi et  al., 2008); and influence on policy that the ‘not at risk’ status category (as applied (Drucker, 2010). An important goal of the eco- in the FAO recommended framework) requires nomic valuation of AnGR is the development of 2000 breeding females in species with high re- policies and strategies for conservation and sus- productive capacity and 6000 in species with tainable utilization of these resources. Economic low reproductive capacity. This suggests that the valuation tools were also seen as contributing to costs of conserving a priority portfolio of at-risk the development of policies and strategies for breeds may also be quite modest (Drucker, conservation and use by providing critical infor- 2010). An assessment of public willingness to mation for decision making in this domain. Of pay for conservation by Zander et al. (2013) and particular interest was the situation in develop- estimates of the support payments that would be ing countries, where, on the one hand, livestock required to achieve the stated conservation goals make important contributions to human liveli- suggest that such conservation costs may well hoods and food security, while on the other, gen- be both economically justifiable (benefits out- etic erosion was placing at risk many breeds weighing costs) and relatively low cost. Thus, adapted to the low-input agriculture and ex- conservation costs may be overestimated if treme environments typical of these countries. based only on conventional economic opportun- For example, although their productivity in ab- ity cost estimates, especially considering find- solute terms is lower than commercial breeds ings by Krishna et  al. (2013), which suggested under intensive production systems, indigenous that farmer willingness to participate in genetic AnGR are often the only option available for resources conservation activities for the public millions of farmers in the African agropastoral good may be more closely related to the con- systems, where exotic improved breeds under- sumption values of the genetic resources in perform (if they survive) in traditional manage- question than to their production opportunity ment systems. costs (which generally do not take into account the existence of farmers’ many non-market pref- erences and values). Tools for conservation decisions ILRI’s work in this area has received recog- nition. The first report on ‘The State of the Valuation studies undertaken by ILRI and part- World’s Animal Genetic Resources for Food and ners have provided tools and methods for valu- Agriculture’ (first SoW-AnGR; FAO, 2007) in- ation that can be broadly applied. Choice cluded a section on methods for economic valu- experiments were found to be a promising tool for ation with an overview of the various types of valuing phenotypic traits expressed by indigen- value (direct and indirect use values, option val- ous AnGR (Scarpa et al., 2003a), while the Weitz- ues, bequest values and existence values) and man approach was found to be a potentially described potential methods and tools for assess- effective methodology for determining conserva- ing them, with explicit reference to work done by tion strategies under highly varying circum- ILRI and examples of the use of these methods stances and for many species (Reist-Marti et  al., and tools and the findings obtained on the 2003). Simianer et  al. (2003) proposed a func- methods. The second SoW-AnGR (in 2015) fo- tional relationship between conservation funds cused on economics of use and conservation, re- spent in one population and the conservation flecting and explicitly recognizing advances that Livestock Genetics and Breeding 81 had been made on valuation methodologies The productivity potential of native live- during this period. It has been recognized, for stock breeds under various systems remains e xample, that accounting for total economic largely unexploited. This, plus the fact that value can determine, among other things, cross-breeding is considered to produce faster whether the benefits of intervention outweigh r esults, has informed the widespread practice of the costs, as well as appropriate intervention upgrading indigenous breeds through cross- strategies, i ncluding for situations in which spe- breeding using imported breeds, particularly for cific AnGR have little or no current market- milk and meat production. East Africa (espe- development potential. Where conservation cially Kenya but also Ethiopia, Rwanda, Tanza- funds are limited, understanding the ‘true’ (i.e. nia and Uganda to smaller extents) leads other total) economic value of different breeds and sub-Saharan African regions such as West their contribution to the public good can be an Africa in the use of cross-breeding, especially for important tool in priority setting and allocation generating ‘grade’ cows used in smallholder of funds (Fadlaoui et al., 2006). An understand- dairy systems. However, cross-breeding is also ing of the relative values of the different compo- widely applied in beef systems in southern Africa nents of total economic value can also be used to and in sheep and goats as well as in chickens provide insight into the viability of different use across the African continent. In addition to up- and conservation strategies, such as identifica- grading, systematic programmes aimed at creat- tion of the relevance of different types of eco- ing new synthetic or composite breeds apply a nomic value and associated ecosystems services combination of planned cross-breeding and se- to different stakeholder categories and their will- lection. Kenya, like South Africa, has had pro- ingness to pay for the services provided by the grammes aimed at developing dairy goats – the maintenance of breeds (Zander et al., 2013). In- Kenya dual-purpose goat and the Meru goat direct use values, such as cultural and landscape (Ojango et  al., 2010) – through systematic maintenance values, are likely to be of more cross-breeding. relevance to local communities, while option Other than its work on genetics of tolerance values are likely to be of relevance to a much to trypanosomiasis, ILRAD’s mandate did not in- broader range of stakeholders. To maximize clude genetic improvement of livestock. For its societal welfare, strategies for conservation part, despite a broader mandate on livestock pro- through active use need to be designed with a duction, ILCA’s first-generation programmes in view to maintaining the ongoing provision of the 1970s and 1980s had a more diagnostic the public-good breed-related functions that orientation – ‘understanding African livestock people value most (Martin-Collado et al., 2014). systems’. However, aspects of breeding were introduced through work started by John Trail and colleagues on quantifying performance in Genetic Improvement of Livestock smallholder and pastoral systems (e.g. Trail, 1981; Trail and Gregory, 1981, 1982; Trail et al., Cross-breeding of indigenous livestock with 1982, 1984, 1985) and by NARS visiting scien- exotic breeds has for a long time been considered tists and collaborators (Kiwuwa et  al., 1983; a faster way of making genetic improvements Tawonezvi et al., 1988a,b) recruited specifically and one that is simple because it does not require for that purpose. The ILCA Strategy and Long- the same level of performance and pedigree re- Term Plan from 1987 was more explicit in its at- cording as selection. However, inadequate nutri- tention to animal genetics and breeding. In this tion and the presence of significant biotic and plan, the Cattle Milk and Meat Thrust included a abiotic stresses, including poor management, sub-theme on ‘evaluation of the genetic potential have continued to limit the use of high-input of cattle breeds and their crosses for milk and exotic breeds of livestock and their crosses, and meat production’, while that of Small Ruminant is largely why many farmers in Africa still keep Meat and Milk included ‘evaluation of small the more resilient indigenous livestock breeds. r uminant genetic resources to identify appropri- There have been only a few success stories ate breeding strategies’. The Trypanotolerance of within-breed improvements of indigenous Thrust committed to the collection and analysis breeds. of data on the productivity of trypanotolerant 82 J.E.O. Rege, J. Ochieng and O. Hanotte breeds under various levels of trypanosomiasis et  al., 1997a,b), the latter being through an risk, and definition of selection criteria for trypa- Open Nucleus Breeding System (Box 1.4). ILRI notolerance (see Chapters 2 and 3, this volume), also contributed to the evaluation of two com- to devise optimum breeding programmes. posite breeds in Tanzania: the Mpwapwa cattle Through these activities, mostly implemented in (Kasonta, 1992) and the blended goat (Das et al., partnership with NARS, ILCA helped to analyse 1996). and publish data sets characterizing cattle and Apart from informing breeding policies small-ruminant populations in Africa on the within sub-Saharan African countries, some of basis of performance traits, and ILRAD and ILCA these studies also made methodological contri- led efforts in collaboration with NARS, ITC and butions as well as human capacity strengthen- CIRDES, among many partners, to document ing for the region. For example, the weight of trypanotolerance traits in cattle and small ru- calf weaned per 100 kg metabolic body weight minants. ILCA’s breed comparisons were done of cow exposed to the bull was examined as a principally by assisting NARS scientists to ana- possible approach for adjusting for differences in lyse on-farm and on-station data. In 1989 and maintenance requirements by different cow 1990, ILCA recruited visiting scientists1 and pro- genotypes, thus providing a ‘fair comparison’ be- vided other support to sub-Saharan African tween small-bodied indigenous breeds and countries in interpretation of breeding data (e.g. large-bodied exotic breeds and crosses (the belief Mwenya, 1990). that local breeds were small and not as good drove importations of exotic breeds); the effect of date of birth on growth performance in season- Analysis of long-term breeding ally bred beef cows was examined as a means of programmes determining the optimum breeding time in such systems (Rege and Moyo, 1993); and a method Much of this early work emphasized breed/ was proposed for estimating between-breed dif- genotype evaluations to provide information ferences in cross-breeding parameters and their on how different indigenous African livestock sampling variances in these systems (Rege et al., breeds and their crosses performed under differ- 1993a). Recommendations based on in-depth ent management systems (e.g. Rege et  al., analyses of the Sahiwal Stud of Kenya (Rege and 1993a,b,c; Tawah et  al., 1993; Thorpe et  al., Wakhungu, 1992) and the Kenya Dairy Cattle 1993; Moyo et al., 1996; Kurtu et al., 1999) or to Improvement Program using long-term data assess the effectiveness of long-term national provided evidence that informed changes in breeding programmes in terms of genetic pro- these breeding programmes, such as a decision gress made, for example, in cattle (e.g. Rege, to open the Sahiwal Stud herd (which until then 1991; Rege and Wakhungu, 1992; Tawah et al., had been a closed herd) to bring in new breeding 1994) and sheep (Yapi et al., 1994; Yapi-Gnaorè stock to reduce inbreeding levels and to broaden Box 1. 4. Open Nucleus Breeding System for sheep improvement in West Africa. A selection programme was set up in 1983 in Côte d’Ivoire to improve the growth and live weight of the indigenous Djallonké sheep using an Open Nucleus Breeding System (Yapi et al., 1997a,b). Selection was based on male individual weights at 80, 180 and 365 days of age. Multiple sires were used in farmers’ flocks. ILCA helped to analyse this programme in 1995/6. Individual animal models using an average numerator relationship was used to estimate breeding values from which genetic trends were derived. A total of 10,417 records of 80-day weights (WT80) of lambs born between 1984 and 1992 from 29 participating farmers, and 1978 and 849 records on 180-day weights (WT180) and 365-day weights (WT365), respectively, of lambs from the nucleus were analysed. Breeding values – as a measure of genetic trends – increased by 28, 11 and 14 g/year for WT80, WT180 and WT365, respectively. The results of the study indicated that genetic progress could be made in growth performance of indigenous sheep if reasonable levels of animal management as well as selection pressure are applied in the Open Nucleus Breeding System and suggested that community-based breeding schemes had potential for genetic improvement in small-ruminant populations in these systems. Livestock Genetics and Breeding 83 the genetic base, and a reorganization of the plus e vidence from genetic diversity research Kenya National Dairy Breeding Plan and its have i nduced a gradual recognition that ‘up- management. In all cases, the collaboration with grading towards exotic genotypes’ across all NARS scientists was in-built and used to provide systems is a mistake and that indigenous genet- training on aspects of design and analysis of live- ics should play an important role in breeding stock breeding programmes. programmes. The challenge to genetic improvement, however, goes well beyond the choice of breeds. Towards the optimization of breeding Institutional and organizational challenges are programmes in Africa as binding as genotype in many African coun- tries (Kosgey, 2004; Kosgey and Okeyo, 2007; ILCA scientists worked with NARS in the 1980s Rege et  al. 2011). Consequently, even cross- on breed comparison studies, which focused breeding practice, considered to be easier than on cattle (both dairy and beef) and small ru- within-breed selection, has generally not been minants. The finding by early researchers well-planned in Africa and outcomes have been (e.g. Kiwuwa et al., 1983) of similarity in per- inconsistent. Many cross-bred populations, in- formance of the 75% and 50% B. taurus crossed cluding the dairy herds in Kenya and other with African Zebu was among the first (based East African countries where cross-breeding for on a large data set covering 1968–1981) to sig- dairy production is considered a success, tend to nal that there was a ‘limit to upgrading’ and be a mix of genotypes (Aliloo et al., 2018; Mrode that this limit depended on the production sys- et al., 2018). The so-called ‘grade’ or cross-bred tem. Marshall et al. (2017) showed that exotic animals in smallholder s ystems in Kenya, for dairy cattle under good management do have e xample, are composed of a wide range of geno- higher milk yields than local × exotic crosses, types, and in the absence of pedigree records, but the additional cost of feed is such that local genotypes of individual cross-b red animals are × exotic crosses are more profitable. A similar generally unknown. project implemented in West Africa – the Sene- These challenges have informed ILRI’s re- gal Dairy Genetics project – resulted in strong cent work in ‘genetic improvement’. One problem recommendations on the best genotype in has been determination of optimum genotypes Senegal dairy systems, based on a trade-off per- for specific production systems. Figure 1.2 illus- spective (Salmon et  al., 2018). These studies trates levels of milk production by different 6000 5000 X 4000 E3 X XE3 E3 3000 XE3 2000 Y XE2E2 XE2 1000 XE1 XE1 O Y Indigenous Cross-bred Exotic Indigenous Cross-breds Synthetics ExoticsE1 Mixed rainfed Mixed rainfed Large-scale temperate/highland humid/subhumid commercial ranches Fig. 1. 2. Milk production by genotype in dairy production systems of eastern Africa. Light-coloured bars, minimum production; dark-coloured bars, maximum production; xi, differences in production due to ‘animal husbandry practices’; yi, differences in production due to ‘genotype.’ (From Mwacharo et al., 2009.) Milk yield per lactation (kg) 84 J.E.O. Rege, J. Ochieng and O. Hanotte genotypes – indigenous, cross-breds, composite Appropriate germplasm for smallholder breeds and exotics – in dairy production systems dairy farmers in eastern Africa. The figure shows that cross-breeding certainly has a role to play and The DGEA project was funded by the Bill & Me- also illustrates the critical need for matching linda Gates Foundation. Determining the size of genotypes with production environments, espe- breed differences in situ (linking farmers’ geno- cially considering that smallholder producers in types to productivity) was a core question of the these systems have little control over the many DGEA project. The answer was expected to allow environmental and production factors/stresses calculation of the net benefits of investment in that affect and constrain the productivity of their germplasm delivery. Specifically, the DGEA pro- livestock. This was the compelling documenta- ject set out to answer two broad questions (Rege tion of G×E presented in a way (yield gaps) that and Gibson, 2009): (i) what are the optimum development partners can easily understand. breed combinations for smallholders in different Subsequent data from ILRI’s own work (e.g. from production environments? and (ii) how can sus- the DGEA project) have not contradicted the gen- tainable germplasm supply systems be developed? eral patterns in genotype comparisons but have shown diminishing returns to upgrading under Optimum breed combinations smallholder systems (performance improvements beyond F 1 being rare in most situations). High-density SNP technology was used to opti- mize choice among cross-bred dairy cattle in Kenya and Uganda, primarily for in situ assess- Breeding programme design ment of the performance of existing genotypes (Ojango et al., 2014). While SNP data accurately There had been growing demand among NARS estimated breed composition, they only did so for ILRI engagement in: (i) non-ruminant live- under very careful construction analyses and, as stock research, specifically chickens; and (ii) sup- such, caution must be taken in applying their re- port for breeding programmes that went beyond sults in decision making. analysis of existing information. In ILRI’s long- term strategy to 2010 (ILRI, 2000), two decisions Sustainable germplasm supply were made: (i) to expand the genetic character- ization portfolio beyond cattle, sheep and goats to The activities under sustainable germplasm sup- include, for the first time, an explicit commitment ply involved the design, operation and support to ‘contribute to the characterization of indigen- of an innovation platform, bringing together ous swine, poultry, camel, buffalo and yak popu- agents in the germplasm supply chain to small- lations in Africa and Asia’; and (ii) to engage holders. The innovation platform led to identifica- more directly in ‘developing breeding strategies tion of business opportunities in dairy genotyping to improve utilization of diversity of indigenous (specifically heifers and artificial insemination). livestock to increase productivity in smallholder Several local farmers and traders established systems’. It is to be noted, however, that al- businesses that identified sellers of heifers, found though there had not been a formal programme buyers, and built seller and buyer capacity in as such in this area up to this point, individual heifer management. ILRI scientists were already contributing to breeding p rogramme design, mostly through Supporting on-farm selective breeding graduate student projects and fellowships by smallholders (e.g.  Kosgey, 2004; König et  al., 2017). Among the major activities initiated following the deci- The absence of breeding recording in developing sion to engage in breeding programme design, countries has been a major limiting factor to sys- four stand out: Two are in dairy – the DGEA and tematic selective breeding in Africa (Kosgey and the African Dairy Genetic Gains (ADGG) projects – Okeyo, 2007; Zonabend et  al., 2013). Other and two are in chickens – the Horro chicken im- limiting factors are poor breeding infrastructure provement and the African Chicken Genetic and lack of good analytical tools (Gizaw et  al., Gains (ACGG) projects. 2014; Mrode et al., 2018; FAO, 2016). It is now Livestock Genetics and Breeding 85 well recognized that, unless recording is owned Ethiopia. What started as a ‘proof of concept’ and run by industry (farmers themselves) with has become a major success, the Horro chicken sufficient incentives to make them participate, breed improvement programme in Ethiopia it  is unlikely that sustainable recording pro- (Box 1.5), and has demonstrated that within- grammes can be achieved. This explains why the breed selection is possible in indigenous chick- limited schemes remain restricted to the large ens. In November 2015, in recognition of these commercial farming sector where owners recog- achievements, the prime minister of Ethiopia nize and are willing to pay for recording. awarded the project the 6th National Science To address the recording challenge, ILRI, and Mathematics, Research and Innovation through the ADGG programme, in 2018/19 is Award of Ethiopia. supporting the establishment of farmer-driven The Horro chicken project, together with ICT-based on-farm pedigree and performance re- the community-based sheep and goat breeding cording in Ethiopia and Tanzania. The objectives program in Ethiopia (Box 1.6) and the Red of ADGG were to establish performance record- Maasai and Dorper sheep breeding programme ing and sampling systems (FAO, 2016) to pilot in Kenya (Box 1.7) are examples of success in es- farmer-feedback systems that help farmers im- tablishing within-breed genetic improvement prove their productivity, and to use the informa- programmes. The improved Horro chicken was tion and samples collected to develop systems for tested in the ACGG programme on-farm and selecting cross-bred bulls and cows of superior on-station (Box 1.8). The aim was to compare genetic merit for artificial insemination and nat- the performance of this improved chicken in re- ural mating (König et al., 2017). Informed by ex- gions of Ethiopia with large differences in alti- periences of the DGEA project, the main thrust tude, rainfall and temperature. The improved for this initiative is a digital platform for perform- Horro chicken in the ACGG comparisons ance data capture/analytics that combines util- reached 714 g at about 16 weeks of age. Unim- ization of genomics to demonstrate and inform proved chickens could not reach the same live implementation of expanded genomic evalu- weight, even at 20 weeks of age and under im- ation and better utilization of cross-bred cattle proved management conditions. Moreover, the populations (see https://www.ilri.org/research/ improved Horro chicken started egg laying at projects/african-dairy-genetic-gains). It is designed the age of 223 days compared with 256 days for to collect and use on-farm performance informa- unimproved indigenous chickens on-farm. tion and basic genomic data to identify and provide superior cross-bred bulls for artificial in- semination delivery and natural mating cows on Towards adapted and productive smallholder farms. While at the time of writing chickens for African smallholders the programme was still in the early stages, more than 82,000 smallholder herds and more than Many past efforts to make smallholder chickens 190,000 animals were already being recorded more productive in sub-Saharan Africa have on the platform in the two countries, with more not delivered optimal impact because they used than 5000 of these already genotyped and the high-producing commercial genotypes created results, together with the phenotypic records for intensive temperate feeding systems. In col- being used to generate genomic breeding value laboration with NARS in Ethiopia, Nigeria and predictions (Brown et al., 2016) in Tanzania and Tanzania and involving several other partners, later in Ethiopia. Farmers on the platform re- ILRI initiated a project (ACGG) in 2014 with ceived more than 6 million text messages in four the aim of identifying and delivering adapted languages to inform herd management decisions chickens to support productivity growth and and benefitted from productivity gains. i ncreased animal protein intake among rural people (Box 1.8). The key features of this project Demonstrating potential for within-breed were: (i) a focus on delivery of farmer-preferred selection chicken genotypes; (ii) use of innovation plat- forms to identify chicken value chain chal- In 2008, ILRI and Wageningen University started lenges and to develop solutions in participatory a pilot breeding project for Horro chickens in processes involving key value chain actors; 86 J.E.O. Rege, J. Ochieng and O. Hanotte Box 1.5. Horro chicken breed improvement programme in Ethiopia. The Horro chicken breeding programme implemented by ILRI in collaboration with Wageningen Univer- sity in Ethiopia started in 2008 as a PhD project of Nigussie Dana. The objective was to improve produc- tion of village chickens through participatory within-breed selection. The breeding objectives were identified using a participatory approach. The breeding programme aimed to develop a dual-purpose chicken through selective breeding. As a start, a survey was conducted to understand the production sys- tems and the needs and constraints of smallholder chicken farmers in 225 households. The breeding goal traits identified were egg production (number), body weight, (decreased) age at first egg and survival. The base population was established from 3000 eggs purchased from various locations in the Horro region and these were placed at the Ethiopian Institute of Agricultural Research in Debre Zeit, Ethiopia. Twenty cockerels and 260 hens were successfully hatched and raised, and these formed the parental population. Selection was based on individual performance (‘own performance’ or ‘mass selection’) until the eighth generation. In each generation, approximately 600 males and 600 females were produced as selection candidates and recorded for body weight and egg production. Females were selected based on own per- formance for body weight and egg production. Males were selected based on their performance for body weight. Selection pressure was 10–20% in the males and 50–60% in the females. Evaluation of the breeding programme was conducted when the programme was in generation 8. Breeding values were estimated for both cumulative egg numbers at 45 weeks of age and body weight at 16 weeks of age to evaluate the trend of changes over the generations. The genetic trends showed that by generation 8, survival had improved from less than 50% in the base generation to almost 100% in generation 8. Body weight per bird at 16 weeks had increased substantially from 550 g to 1100 g. Egg production tripled from 64 eggs per hen per year in the base generation to 172 eggs per hen per year by generation 8. Every generation of pure-line Horro birds was kept and used as parents for the coming generations. The nucleus flock established was kept at the Debre Zeit Agricultural Research Centre. The genetic change achieved through selection was monitored by comparing the unselected animals with the selected ones. This breeding programme achieved a large and significant improvement above unselected village chickens. While the performance of this population was lower than that of commer- cial chicken lines (the difference decreasing with each generation of selection), the improvement was impressive. And when adaptability, taste and likeability were taken into account, the improved indigen- ous birds were clearly superior to commercial chickens. Box 1.6. Community-based sheep and goat breeding programmes. A community-based breeding programme, implemented by the International Centre for Agricultural Re- search in the Dry Areas (ICARDA), ILRI, Boku University and the Ethiopian NARS, was started in 2009 in Ethiopia (ICARDA, 2018; Haile et al., 2019). The programme combined selective breeding of sheep and goats based on production parameters, such as body weight, lambing rate and survival. At the time of this assessment, 3200 households in 40 villages had benefitted with an average income increase of 20% in the programme sites of Bonga, Horro and Menz. Farmers had created 35 formal breeders’ co- operatives to participate in the programme, and it had been replicated in more than 40 programmes that sprang up based on inspiration and learnings from the original site. Most of the participating house- holds in Menz no longer needed assistance from government-run safety-net programmes that provided food; they were able to use their income from sheep sales to buy food. The breeding cooperatives were able to build capital from buying rams and bucks as well as from other investments. For example, the Bonga cooperative had a capital of around US$60,000. There was a high demand to breed rams and bucks from neighbouring communities and other governmental and non-governmental programmes. The government identified the community-based breeding programme as the strategy for genetic im- provement of small ruminants in the Ethiopia Master Plan and Growth and Transformation Plan II, and the programme is being replicated in Iran, Malawi, South Africa, Sudan, Tanzania and Uganda. (iii) developing sustainable public–private part- the centre of the project to ensure its success, nerships for improvement, multiplication and given that women are the primary owners, delivery of the identified genotypes and other managers and traders in chickens and chicken value chain services; and (iv) placing women at products in these countries. Livestock Genetics and Breeding 87 Box 1.7. Red Maasai and Dorper sheep breeding programmes. Starting with a sheep flock established in 1997 as an experimental flock comprising purebred Red Maasai and Dorper sheep and their crosses, a breeding programme was introduced in 2003 aimed at improving the growth performance and resilience of the main breeds and their crosses under range conditions. Following droughts and loss of animals by pastoral livestock keepers in 2008–2010, the ILRI flock of 1100 sheep became a main source of breeding animals for communities living within the surrounding rangelands in Kenya and neighbouring countries. Informed by the finding that many genes, each with small effects, rather than one major gene were involved in conferring resistance in the Red Maasai breed to endoparasites, ILRI established a multi-trait selection programme for both Red Maasai and Dorper × Red Maasai crosses under natural and continuous challenge, with the next-generation sires and dams identified from among the young rams and ewes selected on the basis of survival, growth rate and lambing intervals under parasite challenge. The net effect was that the Red Maasai and their crosses with Dorper sheep had, over more than 12 years of selection, improved, with their 9-month weight having almost doubled, the age at first lambing and lambing intervals having slightly decreased, and lambing rates having improved. The most important outcome of this work was that ILRI’s flock at its Kapiti Ranch remained one of the two major sources of improved Red Maasai rams in Kenya, with requests far above what could be supplied. Indeed, after the prolonged drought of 2008, ILRI provided many local farmers with replacements for Red Maasai sheep. With this experi- ence, Kapiti rams were used to initiate further Red Maasai genetic improvement under communi- ty-based set-ups in Nyando, Kisumu County, and in Trans Mara, Bomet County, Kenya. Inspired by these developments, a Red Maasai breed society was registered in the Kenya Livestock Breeders or- ganization and supported by a collaborative project of ILRI and the African Union–Interafrican Bureau for Animal Resources (AU-IBAR). Box 1.8. Project to identify and deliver adapted chickens for productivity. The vision of the ACGG project was to catalyse public–private partnerships for increasing smallholder chicken production and productivity growth as a pathway out of poverty in sub-Saharan Africa, with project sites in Ethiopia, Nigeria and Tanzania, countries with substantial poultry endowments and where the needs and opportunities for enhanced chicken productivity were considered among the greatest on the continent. The project was designed on the premise that having available and affordable brooded and pre-vaccinated chicks adapted to typical low-input systems in poor rural communities would greatly increase their chicken production and productivity and would reduce poverty, especially among poor women. Ten tropically adapted, low-input but productive chicken breeds from Africa and elsewhere were tested under on-farm conditions. Chicks were pre-vaccinated and brooded to 21 days old before distribution to households. The results showed that significant productivity gains could be made by testing and promoting chicken breeds that are more productive, tropically adapted and farmer preferred. The chicken strains that ACGG made available to farmers had significantly higher productivity in terms of both live body weight (an average of 200–300% higher than indigenous types) and egg pro- duction (100–200% higher) than the local chickens raised by more than 6000 farm households involved in the project. With the testing nearly completed in the three p roject has begun, as a next phase, to develop a project countries, the project in 2020 is focus- roadmap for long-term genetic gains to ensure ing on developing the private-sector-led institu- ongoing genetic improvement of the identified tional arrangements for delivery mechanisms breeds/types. Beyond the project countries that will ensure that the preferred chicken (Ethiopia, Nigeria and Tanzania), the germ- strains identified are available to smallholders plasm, data and k nowledge generated have the at competitive prices at the village level. For ex- potential to benefit millions of poor rural and ample, hatcheries partnering with ACGG peri-urban households in other countries started multiplication and delivery at scale to where backyard chicken production systems smallholder farmers. At the same time, the dominate. 88 J.E.O. Rege, J. Ochieng and O. Hanotte Reproductive technologies genetics of adaptation, specifically as part of the for smallholders efforts to explore the possibility of incorporating disease r esistance traits of some of Africa’s indi- In Africa, only Kenya and South Africa have had genous breeds into more productive breeds. active research in semen sexing and in vitro fer- Through these efforts, ILRI produced the first tilization. ILRI’s work in its facilities in Kenya African livestock cloned by somatic cell nu- has adapted and tested potential applications clear transfer using primary embryonic fibro- of selected technologies to address smallholder blasts. The only other successful cloning in challenges. For example, scientists from ILRI Africa was Futhi (Zulu for ‘replica’), a Holstein and the Department of Clinical Studies at the heifer born at the Artificial Insemination Centre University of Nairobi succeeded in producing at Brits, North West Province in South Africa, in Kenya’s first test-tube calf in 2009 using a tech- April 2003, through a collaboration between nique called in vitro embryo production (IVEP), scientists from South A frica and Denmark. Un- which makes it possible to rapidly multiply and like Tumaini, Futhi was derived from a single cell breed genetically superior cattle within a short taken from the ear of a donor cow, inserted into generation interval. IVEP eliminates the tedious an ovum and later implanted into a recipient steps of synchronizing donor cows and has the cow. Cloning of Tumaini was a proof of concept advantage of maximizing utilization of appro- for a technology that could be used to develop priate dam and sire genotypes by increasing the improved farm animals, carrying traits of eco- efficiency of multiplication in breeding, permit- nomic importance such as disease resistance ting determination of sex of the offspring and (e.g. trypanotolerance). facilitating the pre-testing of actual fertility sta- tus of the sire. IVEP can produce up to 300 off- spring per lifespan of a dam. ILRI and partners Strengthening NARS Capacity have also applied the IVEP technique in combin- in Livestock Genetics ation with sexed semen in what is called fixed- time artificial insemination. These protocols and Inadequate capacity is one of the major con- the accompanying capacity were the basis of straints to agricultural development in most of the programme for ILRI’s cloned calf (Yu et al., the developing world, especially in Africa and 2016). Asia. A recent study commissioned by FAO in Domestication of fixed-time artificial in- 2017 found that the major constraints to appli- semination (FTAI) protocols have enabled syn- cation of most agricultural technologies in chronization and have been used to extend Africa relate to inadequate human capacities, artificial insemination to nine counties in Kenya facilities, financial investments and institutional that were previously considered unsuitable for capacities. In most countries, there remains a high-yield dairying. This extension of FTAI has major lack of a critical mass of scientists in areas produced enough cross-bred heifers to support relevant for agricultural biotechnology, espe- related businesses, such as private artificial in- cially in the more advanced areas of modern bio- semination, and milk bulking and chilling ser- technology, such as molecular biology, genomics vices, in some of these counties. The Ethiopian and bioinformatics. application of FTAI has resulted in hundreds of In the livestock sector, working in partner- thousands of new dairy cross-bred cows and is ship with national and other international part- currently underpinning the country’s dairy ners, ILRI has contributed substantially in the improvement plan. research and application of medium- to high- Although the worldwide success rate of level biotechnologies for genetic characterization producing live cloned offspring from high- and improvement of livestock. These have been quality domestic livestock has improved in the implemented through four main streams: gradu- past two decades since the successful cloning of ate training, group training, individual short- Dolly the sheep in 1996, little has happened in term training and internships, and coaching and Africa in this regard. ILRI scientists have under- mentorship achieved through collaborative pro- taken research on cloning as a tool for support- jects. These are summarized below, with special ing its animal health and broader research on reference to animal genetics and breeding. Livestock Genetics and Breeding 89 Capacity development NARS scientists; (ii) to strengthen their commu- nication skills, to catalyse curriculum develop- ILRI has trained many MSc and PhD students ment and delivery; (iii) to develop computer-based in genetics and breeding, either through fellow- training resources, to stimulate contacts, collab- ships or in collaboration with NARS where ILRI- orations and networks; and (iv) to strengthen hosted fellows work. Since 2004, the facilities of the human base for work on AnGR in developing the Biosciences eastern and central Africa–ILRI countries. Hub (BecA-ILRI Hub), especially the Hub’s gen- During the ILRI-SLU project period (1999– omics and bioinformatics platforms, were a 2010), 195 scientists from 46 countries in major means of delivering high-quality training, A frica and Asia were trained in animal breeding including exposure to cutting-edge research fa- and genetics, including in design and implemen- cilities and approaches. As such, many of these tation of breeding strategies, and in teaching fellows subsequently became research leaders in and communication (Table 1.1) (Ojango et  al., their own institutions, and some have retained 2011). intensive collaboration with ILRI and other insti- In addition, the ILRI-SLU project developed tutions internationally, with these fellows using an electronic Animal Genetics Training Re- their relationships with ILRI to develop projects source (AGTR; http://agtr.ilri.cgiar.org; accessed and to provide opportunities for next-g eneration 28 January 2020), which is available online and graduate fellows. Indeed, many among the cur- on CD. The first version of this course was pro- rent generation of leading livestock geneticists duced in 2003, a second one in 2006, and an in African and, to a smaller extent, Asian NARS updated and expanded one (version 3) in 2011 have connections with ILRI, with a majority of on a fully web-enabled platform, which allows these having spent time in ILRI’s laboratories. direct online revisions of content. AGTR is a ILRI has trained an estimated 200 BSc, over unique, ‘one-stop’, user-friendly, and interactive 69  MSc and more than 66 PhD graduates, as multimedia resource targeted at researchers and well as over 35 postdoctoral fellows in livestock scientists teaching and supervising postgradu- genetics and breeding. ate students in animal breeding and genetics. It is a dynamic training resource designed to in- Group training form the design and implementation of breeding programmes and to provide information that Through its research programmes on cattle and empowers countries and institutions to under- small ruminants, ILCA organized, starting in take their own research applying the best avail- mid-1980 until its merger with ILRI in 1995, able information and knowledge. The AGTR annual courses for 25–30 NARS scientists in the course covers established and rapidly developing design and analysis of livestock breeding pro- areas, such as genetics-based technologies and grammes. These early courses focused on ana- their applications in livestock breeding pro- lysis and interpretation of data from on-station grammes. AGTR is essentially a platform that ex- and on-farm breeding programmes. tends the capacity strengthening work done by the ILRI-SLU project to reach more people over a ILRI-SLU project long-lasting period. One of the objectives of the project was to Starting in 1999, ILRI collaborated with the stimulate knowledge sharing and networks Swedish University of Agricultural Sciences within regions. By linking NARS and university (SLU) on a global project, ‘Capacity Building for lecturers from different countries in the training Sustainable Use of Animal Genetic Resources courses, three virtual regional networks were sub- in Developing Countries’, using a ‘training the sequently established: (i) Afrib Breeders in Africa; trainers’ model (Fig. 1.3). This capacity-building (ii) IAGRA in South-east Asia; and (iii) the South project directly targeted NARS scientists in de- Asia Genetics Group. These animal breeding and veloping countries who are responsible for re- genetics virtual networks were created by the search and training in animal breeding and project participants as tools for sharing know- genetics. The objectives of the project were: (i) to ledge and information and for facilitating the de- strengthen the subject knowledge and skills of velopment and review of collaborative proposals 90 J.E.O. Rege, J. Ochieng and O. Hanotte Sustainable utilization of animal genetic resources for enhanced food security and improved livelihoods Improved Increased Improved designs Stronger regional human access to of breeding networks of capacity resources and strategies for scientists base in AnGR information on AnGR conservation involved in AnGR utilization AnGR and use Increased Increased electronic regional Animal genetics Improved practical interest in AGB discussion training resources applications of the by students, groups on AnGR for use by NARS subject AnGR scientists and in developing developed policy makers countries Improved teaching More regional Case studies relevant methods and skills collborative to developing Improved and expanded research in higher-education projects for AnGR countries and related on AnGR institutions improvement breed information documented Increased Modules with Characterization, participation in primary information conservation and Improved university development for AnGR use and design of breeding curricula in the area of national and improvement schemes for of AnGR regional policies developed and improved livestock related to AnGR published Teaching Networking Communication Research Fig. 1. 3. Outcomes of the ILRI-SLU capacity development project. (Data from ILRI archives.) Table 1.1. ILRI-SLU training courses and trainees between 2000 and 2010. (Data from ILRI archives.) Year Region Countries Number of participants 2000 East and southern Africa 10 20 2001 West and Central Africa 10 18 2003 South-east Asia 9 18 2003 Sub-Saharan Africa 18 20 2005 South-east Asia 9 18 2006 South Asia 6 20 2006 Training request by ASARECA 9 19 2007 West and Central Africa 9 20 2008 East and Southern Africa 14 23 2009 South Asia 6 19 ASARECA, Association for Strengthening Agricultural Research in Eastern and Central Africa. Livestock Genetics and Breeding 91 on the characterization, conservation and design with operationalization of the BecA-ILRI Hub of livestock breeding programmes. In this way, platform in the early-2000s being a major driver; the project has helped to create ‘communities of large numbers of scientists from across Africa practice’ in these regions. sought space to extract and analyse DNA samples ILRI has organized and delivered many and to get help with their data analyses. The short-term group training courses in genetics, major constraint to ILRI’s hosting even larger genomics, bioinformatics and animal breeding, numbers of technical associates and research some in collaboration with the BecA-ILRI Hub fellows has been inadequate financial resources platform team. For example, between 2008 and to support them. 2018, ILRI scientists provided 2-week trainings ILRI’s livestock genetics activities have been every year in a European master’s degree pro- dependent on strong partnerships with NARS, gramme in animal genetics and breeding that both as providers/owners of samples and data focused on practical applications in developing and as direct collaborators in the research pro- countries. The programme initially supported cess, design, experimentation, analysis and re- students from developing countries to study in porting of research results. These collaborations Europe, after which they would return to imple- have been critical for mutual learning, with ILRI ment their learnings in their home countries. scientists benefiting by deepening their own The ILRI genetics team is a key resource for these understanding of the relevant livestock produc- training courses, providing topics for intern- tion systems and institutional contexts of re- ships and MSc thesis co-supervision, and host- search in NARS. NARS scientists, on the other ing students during their internships/MSc thesis hand, have benefitted from exposure to ad- projects. Other courses involving ILRI’s genetics vanced research facilities and methods that team include collaboration with Scotland’s many of them have no access to in their home Rural College, which since 2014 has been run- institutions. Very importantly, the exposure has ning a 1-week annual training course for 21 allowed many of these scientists to engage in ex- African scientists from 14 different countries in panded and productive networks of scientists, quantitative genetics and genomic selection. both in their own regions (Africa and Asia) and Through ‘technical associates’ and ‘research internationally – especially with ILRI’s many fellows’ mechanisms, ILRI’s livestock genetics traditional advanced research institute partners. team has trained a large number of scientists Through the direct coaching of these scientists from universities and national agricultural re- in the research process and their exposure to, search institutions. Technical associates are tech- and mentorship by, the ILRI research commu- nicians or scientific staff from NARS and other nity, as well as the expanded networks they can ILRI partner institutes who come to ILRI for up to engage in, ILRI has created a large number of 6 months for individual training at the request of global leaders in livestock genetics for develop- their employers; this category is oriented towards ment in Africa and Asia. persons already involved in research activities as- sociated with ILRI. Research fellows are univer- sity and national agricultural research institution Conclusions and the Future scientists undertaking work in research areas similar to those at ILRI. Designed to benefit the The efforts of ILRI and its partners in breed future research capability of the individuals and surveys, phenotypic and molecular character- their home institutions, research fellows spend a ization, breed evaluation studies, genetics of dis- maximum of 18  months at ILRI to undertake ease resistance, genetic improvement strategies, non-degree-related training in research method- development of associated tools and approaches, ologies, to discuss the design and planning of col- and the wider promotion and use of these tools laborative research, and to analyse and write up have collectively contributed to a deeper under- research results. While all areas of the livestock standing of African and Asian livestock popula- genetics programme have undertaken individual tions. The associated capacity development training through these mechanisms, the animal efforts have helped to broaden the application of genetic characterization (phenotypic and mo- these results and tools both in space and in time. lecular) team has hosted the largest numbers, Taken together, these investments have led to: 92 J.E.O. Rege, J. Ochieng and O. Hanotte (i) classification and characterization of indigen- proven breeding technologies such as cloning ous livestock breeds of Africa and Asia, using and associated embryo transfer technologies, phenotypic and molecular genetic information, and other tools needed to navigate new areas of facilitating a greater understanding of the livestock improvement such as developing underl ying genetic diversity in these regions; (ii) transgenic animals with resistance/tolerance databases on the geographical distribution and attributes. These tools, together with global ad- physical and performance characteristics of in- vancements in genetic technologies, including digenous livestock; (iii) contribution to the re- SNPs and gene drives as well as ICT platforms construction of genetic history and geography, that can be used to facilitate the creation of in- with links to human migration and settlement; tegrated large-scale data systems, especially (iv)  confirmation of the uniqueness of specific genomic data (and which can enhance the indigenous breeds and better quantification of democratization of data to ensure uncon- their unexploited potential such as disease re- strained access by NARS), promise new possi- sistance; (v) development and testing of tools for bilities for livestock improvement in the genetic characterization and conservation; and developing world. (vi) targeted application of functional genomics. In supporting the application of economic Recent and exciting findings include a study on valuation methodologies, ILRI’s work will focus the genome landscape of African livestock (Kim on understanding and applying the economics et  al., 2017) and the development of harmon- of conservation and use in decision-making ized phenotypic and genetic characterization contexts such as choice of traits and breeds/ tools and protocols led by AU-IBAR for national genotypes, public willingness to pay for services and regional gene banks (www.au-ibar.org/ (breeding and conservation) and incentive angr/432-regional-inception-workshops-for- mechanisms for conservation. ILRI’s work in animal- genetic-resources; accessed 28 January this domain (e.g. Muigai et al., 2009; Tarekegn 2020). et al., 2016) has demonstrated the link between The future of ILRI’s genetics research and knowledge of genetic diversity and conservation development programmes will be influenced by and use programmes. In future, a more compel- four factors: (i) the CGIAR global livestock ling link will need to be demonstrated, with agenda; (ii) ILRI’s long-term strategy; (iii) emer- practical examples across different species. ging and anticipated livestock challenges in ILRI will need to explore partnerships with coming years including new infections and dis- the private sector (the development of ICT-based ease variants spreading to new regions largely platforms, such as that envisaged in the ADGG driven by climate change, challenges associated project, can provide proof of concept) while en- with smallholder intensification, and challenges suring that the interests of smallholders are not imposed by the ever-decreasing farmland and compromised. Development of models for facili- limited forage resources, requiring more feed- tating the testing and delivery of appropriate efficient livestock; and (iv) partnership arrange- genotypes for smallholders and for ensuring on- ments that will require constant re-examination going genetic gains – such as what was piloted in of ILRI’s comparative advantage. the DGEA and envisaged in the ACGG projects – ILRI currently has a rich resource base, are examples. Indeed, ILRI has already shown collaborative arrangements and a wide-scale increased involvement in on-farm breeding pro- partnership model to push its livestock genetics grammes and extension services in some coun- agenda to higher levels and with greater im- tries in Africa and Asia, following enhanced pacts. These include the suites of techniques, partnership with NARS. It is envisaged that protocols and approaches for genetic improve- these endeavours will be pursued by an expand- ment of livestock as discussed in this chapter, ing network of global collaborators. Note 1Sam Okantah, from Ghana; Wilson Mwenya, from Zambia; and Ed Rege, from Kenya. Livestock Genetics and Breeding 93 References Abegaz, S., Solkner, J., Gizaw, S., Dessie, T., Haile, A. and Wurzinger, M. (2013) Description of production systems and morphological characteristics of Abergelle and Western lowland goat breeds in Ethiopia: implication for community-based breeding programmes. Animal Genetic Resources 52, 1–10. Aboagye, G.S., Tawah, C.L., and Rege, J.E.O. (1994) Shorthorn cattle of West and Central Africa III. Phys- ical, adaptive and special genetic characteristics. World Animal Review 78, 22–32. Alba-Hurtado, F. and Muñoz-Guzmán, M.A. (2013) Immune responses associated with resistance to haemonchosis in sheep. Immunology and Cell Biology of Parasitic Diseases 2013, 162158. Aliloo, H., Mrode, R., Okeyo, A.M., Ni, G., Goddard, M.E. and Gibson, J.P. (2018) The feasibility of using low-density marker panels for genotype imputation and genomic prediction of crossbred dairy cattle of East Africa. Journal of Dairy Science 101, 9108–9127. Almathen, F., Charruau, P., Mohandesan, E., Mwacharo, J.M., Orozco-terWengel, P., et al. (2016) Ancient and modern DNA reveal dynamics of domestication and cross-continental dispersal of the dromedary. Proceedings of the National Academy of Sciences USA 113, 6707–6712. Ayalew, W., Rischkowsky, B., King, J.M. and Bruns, E. (2003) Crossbreds did not create more net benefits than indigenous goats in Ethiopian smallholdings. Agricultural Systems 76, 1137–1156. Ayalew, W., van Dorland, A. and Rowlands, J. (eds) (2004) Design, execution and analysis of the livestock breed survey in Oromiya Regional State, Ethiopia. Oromiya Agricultural Development Bureau (OADB), Addis Ababa, Ethiopia, and ILRI, Nairobi, Kenya. Baker, R.L. (1995) Genetics of disease resistance in small ruminants in Africa. In: Gray, G.D., Woolaston, R.R. and Eaton, B.T. (eds) Breeding for Resistance to Infectious Diseases of Small Ruminants. ACIAR Monograph No. 34, Canberra, pp. 120–138. Baker, R.L. (1998) Genetic resistance to endoparasites in sheep and goats. A review of genetic resistance to gastrointestinal nematode parasites in sheep and goats in the tropics and evidence for resistance in some sheep and goat breeds in sub-humid coastal Kenya. Animal Genetic Resources Information 24, 13–30. Baker, R.L., Lahlou Kassi, A., Rege, J.E.O., Reynolds, L., Bekele, T., et al. (1992) A review of genetic resist- ance to endoparasites in small ruminants and an outline of ILCA’s research programme in this area. In: Proceedings of the 10th Scientific Workshop of the Small Ruminant Collaborative Research Sup- port Program. SR-CRSP, Nairobi, pp. 79–104. Baker, R.L., Mwamachi, D.M., Audho, J.O. and Thorpe, W. (1994a) Genetic resistance to gastrointestinal nematode parasites in Red Maasai sheep in Kenya. In: Proceedings of the 5th World Congress on Genetics Applied to Livestock Production, 7–12 August 1994, Guelph, Canada, Vol. 20, pp. 277–280. Baker, R.L., Mwamachi, D.M., Audho, J.O. and Thorpe, W. (1994b) Genetic resistance to gastrointestinal parasites in Red Maasai, Dorper and Red Maasai × Dorper ewes in coastal Kenya. In: Proceedings of the 12th Scientific Workshop of the Small Ruminant Collaborative Research Support Program. SR-CRSP, Nairobi, pp. 100–109. Baker, R.L., Mwamachi, D.M., Audho, J.O., Aduda, E.O. and Thorpe, W. (1998a) Resistance of Galla and Small East African goats in the sub-humid tropics to gastrointestinal nematode infections and the peri-parturient rise in faecal egg counts. Veterinary Parasitology 79, 53–64. Baker, R.L., Rege, J.E.O., Tembely, S., Mukasa-Mugerwa, E., Anindo, D., et al. (1998b) Genetic resistance to gastrointestinal nematode parasites in some indigenous breeds of sheep and goats in East Africa. In: Proceedings of the 6th World Congress on Genetics Applied to Livestock Production, Armidale, Australia, Vol. 25, pp. 269–272. Baker, R.L., Audho, J.O. Aduda, E.O. and Thorpe, W. (2001) Genetic resistance to gastro-intestinal nematode parasites in Galla and Small East African goats in the sub-humid tropics. Animal Science 73, 61–70. Baker, R.L., Nagda, S., Rodriguez-Zas, S.L., Southey, B.R., Audho, J.O., et al. (2003) Resistance and resili- ence to gastro-intestinal nematode parasites and relationships with productivity of Red Maasai, Dorp- er and Red Maasai × Dorper crossbred lambs in the sub-humid tropics. Animal Science 76, 119–136. Baker, R.L., Mugambi, J.M., Audho, J.O., Carles, A.B. and Thorpe, W. (2004) Genotype by environment interactions for productivity and resistance to gastro-intestinal nematode parasites in Red Maasai and Dorper sheep. Animal Science 79, 343–353. Barendse, W., Vaiman, D., Kemp, S.J., Sugimoto, Y., Armitage, S.M., et al. (1997) A medium-density genetic linkage map of the bovine genome. Mammalian Genome 8, 21–28. Beh, K.J., Hulme, D.J., Callaghan, M.J., Leish, Z., Lenane, I., et al. (2002) A genome scan for quantitative trait loci affecting resistance to Trichostrongylus colubriformis in sheep, Animal Genetics 33, 97–106. 94 J.E.O. Rege, J. Ochieng and O. Hanotte Behnke, J.M., Iraqi, F.A., Mugambi, J.M., Clifford, S., Nagda, S., et al. (2006) High resolution mapping of chromosomal regions controlling resistance to gastro-intestinal nematode infections in an advanced intercross line of mice. Mammalian Genome 17, 584–597. Behnke, J.M., Chiejina, S.N., Musongong, G.A., Nnadi, P.A., Ngongeh, L.A., et al. (2011) Resistance and resilience of traditionally managed West African Dwarf goats from the Savanna zone of northern Nigeria to naturally acquired trypanosome and gastrointestinal nematode infections. Journal of Hel- minths 85, 80–91. Benavides, M.V., Sonstegard, T.S., Kemp, S., Mugambi, J.M., Gibson, J.P. et al. (2015) Identification of novel loci associated with gastrointestinal parasite resistance in a Red Maasai × Dorper backcross population. PLoS One 10, e0122797. Bettridge, J.M., Psifidi, A., Terfa, Z.G., Desta, T.T., Lozano-Jaramillo, M., et al. (2018) The role of local adap- tation in sustainable production of village chickens. Nature Sustainability 1, 574–582. Blench, R.M. (1993) Ethnographic and linguistic evidence for the prehistory of African ruminant livestock, horses and ponies. In: Shaw, T., Sinclair, P., Andah, B. and Okpoko, A. (eds) The Archaeology of Africa: Food, Metals and Towns. Routledge, London, pp. 71–103. Blench, R.M. (1997) Neglected species, livelihoods and biodiversity in difficult areas: how should the public sector respond? Natural Resource Perspective Paper, No. 23. Overseas Development Institute, London. Bonfiglio, S., Ginja, C., De Gaetano, A., Achilli, A., Olivieri, A., et al. (2012) Origin and spread of Bos taurus: new clues from mitochondrial genomes belonging to haplogroup T1. PLoS One 7, e38601. Bovine HapMap Consortium (2009) Genome-wide survey of SNP variation uncovers the genetic structure of cattle breeds. Science 324, 528–532. Bradley, D.G., Machugh, D.G., Loftus, R.T., Sow, R.S., Hoste, C.H. and Cunningham, E.P. (1994) Zebu- taurine variation in Y chromosome DNA: a sensitive assay for introgression in West African trypanotol- erant cattle populations. Animal Genetics 25, 7–12. Brezinsky, L.S., Kemp, J. and Teale, A.J. (1993a) ILSTS005: a polymorphic bovine microsatellite. Animal Genetics 24, 73. Brezinsky, L.S., Kemp, J. and Teale, A.J. (1993b) ILSTS006: a polymorphic bovine microsatellite. Animal Genetics 24, 73. Brezinsky, L.S., Kemp, J. and Teale, A.J. (1993c) Five polymorphic bovine microsatellites (ILSTS010-014). Animal Genetics 24, 75–76. Brown, A., Ojango, J., Gibson, J., Coffey, M., Okeyo, M. and Mrode, R. (2016) Genomic selection in a cross- bred cattle population using data from the Dairy Genetics East Africa Project. Journal of Dairy Science 99, 7308–7312. Chenyambuga, S.W., Hanotte, O., Hirbo, J., Watts, P.C., Kemp, S.J., et al. (2004) Genetic characterization of indigenous goats of sub-Saharan Africa using microsatellite DNA markers. Asian-Australasian Journal of Animal Sciences 17, 445–452. Chiejina, S.N., Behnke, J.M., Musongong, G.A., Nnadi, P.A. and Ngongeh. L.A. (2010) Resistance and re- silience of West African Dwarf goats of the Nigerian savanna zone exposed to experimental escalating primary and challenge infections with Haemonchus contortus. Veterinary Parasitology 171, 81–90. CIRDES/ILRI/ITC (2000) Joint report of accomplishments and results (1993 –1999). Collaborative research programme on trypanosomosis and trypanotolerant livestock in West Africa, funded by the Commis- sion of the European Communities, Nairobi, Kenya. Clevers, H., MacHugh, N.D., Bensaid, A., Dunlap, S., Baldwin, C.L., et al. (1990) Identification of a bovine surface antigen uniquely expressed on CD4–CD8– T cell receptor γ/δ+ T lymphocytes. European Journal of Immunology 20, 809–817. Clutton-Brock, J. (1989) Cattle in ancient North Africa. In: Clutton-Brock, J. (ed.) The Walking Larder. Unwin Hyman, London, pp. 200–214. Coltman, D.W., Wilson, K., Pilkington, J.G., Stear M.J. and Pemberton, J.M. (2001) A microsatellite poly- morphism in the gamma interferon gene is associated with resistance to gastrointestinal nematodes in a naturally-parasitized population of Soay sheep. Parasitology 122, 571–582. Crawford, A.M., Paterson, K.A., Dodds, K.G., Tascon, C.D., Williamson, P.A., et al. (2006) Discovery of quantitative trait loci for resistance to parasitic nematode infection in sheep: I. Analysis of outcross pedigrees. BMC Genomics 7, 178–188. Dana, N., Dessie, T., van der Waaij, L.H. and van Arendonk, J.A.M. (2010) Morphological features of indi- genous chicken populations of Ethiopia. Animal Genetic Resources Information 46, 11–23. Das, M.S., Rege, J.E.O. and Mesfin, S. (1996) Phenotypic and genetic parameters of growth traits of blended goats at Malya, Tanzania. In: Proceedings of the Third Biennial Conference of the African Small Ruminant Research Network. UICC, Kampala, Uganda, p. 326. Livestock Genetics and Breeding 95 Decker, J.E., McKay, S.D., Rolf, M.M., Kim, J.W., Alcalá, A.M., et al. (2014) Worldwide patterns of ancestry, divergence, and admixture in domesticated cattle. PLoS Genetics 10, e1004254. Dessie, T., Dana, N., Ayalew, W. and Hanotte, O. (2012) Current state of knowledge on indigenous chicken genetic resources of the tropics: domestication, distribution and documentation of information on the genetic resources. World Poultry Science Journal 68, 11–19. Desta, T.T., Dessie, T., Bettridge, J., Lynch, S.E., Melese, K., et al. (2013) Signature of artificial selection and ecological landscape on morphological structures of Ethiopian village chickens. Animal Genetic Resources 52, 17–29. Dorji, T., Hanotte, O., Arbenz, M., Rege, J.E.O. and Roder, W. (2003) Genetic diversity of indigenous cattle populations in Bhutan: implications for conservation. Asian-Australasian Journal of Animal Sciences 16, 946–951. Doutressoulle, G. (1947) L’élevage en Afrique Occidentale Française. Larose, Paris. Drucker, A. (2003) Economic valuation of animal genetic resources: importance and application. In: Con- servation and Sustainable Use of Agricultural Biodiversity: A Sourcebook. CIP-UPWARD, Laguna, Philippines, pp. 416–424. Drucker, A.G. (2010) Where’s the beef? The economics of AnGR conservation and its influence on policy design and implementation. Animal Genetic Resources 47, 85–90. Drucker, A.G., Bergeron, E., Lemke, U., Thuy, L.T. and Zarate, A.V. (2006) Identification and quantification of subsidies relevant to the production of local and imported pig breeds in Vietnam. Tropical Animal Health and Production 38, 305–322. Duguma, G., Mirkena, T., Haile, A., Okeyo, A.M., Tibbo, M., et al. (2011) Identification of smallholder farmers and pastoralists’ preferences for sheep breeding traits: choice model approach. Animal 5, 1984–1992. Dunbar, G.S. (1970) African Ranches Ltd, 1914–1931: an ill-fated stock-raising enterprise in Northern Ni- geria. Annals of the Association of American Geographers 60, 102–123. Eguale, T., Mekonnen, G.A. and Chaka, H. (2009) Evaluation of variation in susceptibility of three Ethiopian sheep breeds to experimental infection with Fasciola hepatica. Small Ruminant Research 82, 7–12. Epstein, H. (1971) The Origin of the Domestic Animals of Africa. African Publishing Corporation, New York. Epstein, H. and Mason, I.L. (1984) Cattle. In: Masson, I.L. (ed.) Evolution of Domesticated Animals. Longman, New York, pp. 6–28. Fadlaoui, A., Roosen, J. and Baret, P.V. (2006). (2006) Setting priorities in farm animal conservation choices – expert opinion and revealed policy preferences. European Review of Agricultural Economics 33, 173–192. FAO (1993) Secondary guidelines for development of national farm animal genetic resources management plans. Measurement of domestic animal diversity (MoDAD): recommended microsatellite markers. Joint FAO/ISAG Standing Committee, FAO Animal Production and Health Division, Rome. FAO (2007) The state of the world’s animal genetic resources for food and agriculture. Edited by B. Rischkowsky and D. Pilling. FAO, Rome. Available at: www.fao.org/3/a-a1250e.pdf (accessed 18 January 2020). FAO (2013) In vivo conservation of animal genetic resources. FAO Animal Production and Health Guide- lines. No. 14. FAO, Rome. Available at www.fao.org/docrep/018/i3327e/i3327e.pdf (accessed 18 January 2020). FAO (2016) Development of integrated multipurpose animal recording systems. FAO Animal Production and Health Guidelines, No. 19. FAO, Rome. Farm Africa/ILRI (1996) Goat types of Ethiopia and Eritrea. Physical description and management systems. Farm Africa, London, and ILRI, Nairobi, Kenya. Freeman, A.R., Meghen, C.M., Machugh, D.E., Loftus, R.T., Achukwi, M.D., et al. (2004) Admixture and diversity in West African cattle populations. Molecular Ecology 13, 3477–3487. Getachew, T., Alemu, B., Sölkner, J., Gizaw, S., Haile, A., et al. (2015) Relative resistance of Menz and Washera sheep breeds to artificial infection with Haemonchus contortus in the highlands of Ethiopia. Tropical Animal Health and Production 47, 961–968. Gizaw, S., van Arendonk, J.A.M., Komen, H., Windig, J.J. and Hanotte, O. (2007) Population structure, genetic variation and morphological diversity in indigenous sheep of Ethiopia. Animal Genetics 38, 621–628. Gizaw, S., Goshme, S., Getachew, T., Haile, A., Rischkowsky, B., et al. (2014) Feasibility of pedigree record- ing and genetic selection in village sheep flocks of smallholder farmers. Tropical Animal Health and Production 46, 809–814. Gray, G.D. (1991) Breeding for resistance to trichostrongyle nematodes in sheep. In: Owen, J.B. and Axford, R.F.E. (eds) Breeding for Disease Resistance in Farm Animals. CAB International, Wallingford, UK, pp. 139–161. 96 J.E.O. Rege, J. Ochieng and O. Hanotte Gray, G.D. and Woolaston, R.R. (1991) Breeding for Disease Resistance in Sheep. Wool Research and Development Corporation, Melbourne, Australia. Haile, A., Tembely, S., Anindo, D.O., Mukasa-Mugerwa, E., Rege, J.E.O., et al. (2002) Effects of breed and dietary protein supplementation on the responses to gastrointestinal nematode infections in Ethiopian sheep. Small Ruminant Research 44, 247–261. Haile, A., Wurzinger, M., Mueller, J., Mirkena, T., Duguma, G., et al. (2019) Guidelines for Setting Up Com- munity-based Small Ruminants Breeding Programs, 2nd edn. ICARDA, Beirut. Halima, H., Neser, F.W., van Marle-Koster, E. and de Kock, A. (2007) Phenotypic variation of native chicken populations in northwest Ethiopia. Tropical Animal Health Production 39, 507–513. Hanotte, O., Tawah, C.L., Bradley, D.G., Okomo, M., Verjee,Y., et al. (2000a) Geographic distribution and frequency of a taurine (Bos taurus) and an indicine Zebu (Bos indicus) Y specific allele amongst sub-Saharan African cattle breeds. Molecular Ecology 9, 387–396. Hanotte, O., Verjee, Y., Ochieng, J. and Rege, J.E.O. (2000b) Cattle microsatellite markers for amplification of polymorphic loci in Asian bovidae. In: Shrestha, J.N.B. (ed.) Proceedings of the 4th Global Confer- ence on the Conservation of Domestic Animal Genetic Resources, Kathmandu, Nepal. Nepal Agricul- tural Research Council, Lalitpur, Nepal, and Rare Breeds International, Shropshire, UK, pp. 47–49. Hanotte, O., Bradley, D.G., Ochieng, J.W., Verjee, Y. Hill, E.W. and Rege. J.E.O. (2002) African pastoralism: genetic imprints of origins and migrations. Science 296, 336–339. Hanotte, O., Ronin, Y., Agaba, M., Nilsson, P., Gelhaus, A., et al. (2003) Mapping of quantitative trait loci controlling trypanotolerance in a cross of tolerant West African N’Dama and susceptible East African Boran cattle. Proceedings of the National Academy of Sciences USA 100, 7443–7448. Hassen, F., Bekele, E., Ayalew, W. and Dessie, T. (2007) Genetic variability of five indigenous Ethiopian cattle breeds using RAPD markers. African Journal of Biotechnology 6, 2274–2279. Herrera, M.B., Thomson, V.A., Wadley, J.J., Piper, P.J., Sulandari, S., et al. (2017) East African origins for Madagascan chickens as indicated by mitochondrial DNA. Royal Society Open Science 4, 160787. ICARDA (2018) Ethiopia: community-based breeding for genetic improvement of sheep and goats. ICARDA, Ethiopia. Available at: www.icarda.org/update/ethiopia-community-based-breeding-genetic- improvement- sheep-and-goats (accessed 28 January 2020). ILCA (1979) Small Ruminant Production in the Humid Tropics. ILCA, Addis Ababa. ILCA (1991) Proceedings of the Research Planning Workshop on Resistance to Endoparasites in Small Ruminants, 5–7 February 1991. ILCA, Addis Ababa. ILCA (1992) Genetic resistance to gastrointestinal parasitism in small ruminants. ILCA Project Proposal (2.2/01-02/01). ILRI (2000) ILRI Medium-term Plan 2001–2003: Making the Livestock Revolution Work for the Poor. ILRI, Nairobi. Iraqi, F., Clapcott, S.J., Kumari, P., Haley, C.S., Kemp, S.J. and Teale, A.J. (2000) Fine mapping of trypano- somiasis resistance loci in murine advanced intercross lines. Mammalian Genome 11, 645–648. Iraqi, F.A., Behnke, J.M., Menge, D., Lowe, A., Teale, A.J., et al. (2003) Chromosomal regions controlling resistance to gastro-intestinal nematode infections in mice. Mammalian Genome 14, 184–191. ISAG/FAO (2004). Secondary guidelines for development of national farm animal genetic resources man- agement plans. Measurement of domestic animal diversity (MoDAD): recommended microsatellite markers. FAO, Rome. Ishag, I.A., Reissmann, M., Peters, K.J., Musa, L.M.-A. and Ahmed, M.-K.A. (2010) Phenotypic and molecular characterization of six Sudanese camel breeds. South African Journal of Animal Science 40, 319–327. Jabbar, M.A. and Diedhiou, M. L. (2003) Does breed matter to cattle farmers and buyers? Evidence from West Africa. Ecological Economics 45, 461–472. Jianlin, H., Mburu, D., Ochieng, J., Rege, J.E.O. and Hanotte, O. (2000) Usefulness of New World Cameli- dae microsatellite primers for amplification of polymorphic loci in Old World camelids. Animal Genetics 31, 404–406. Jianlin, H., Ochieng, J.W., Rege, J.E.O. and Hanotte, O. (2002) Low level of cattle introgression in yak popu- lations from Bhutan and China: evidences from Y-specific microsatellites and mitochondrial DNA markers. In: Jianlin, H., Richard, C., Hanotte, O., McVeigh, C. and Rege, J.E.O. (eds) Yak Production in Central Asian Highlands. Yak and Camel Foundation of Germany, ILRI, Nairobi, pp. 190–196. Jianlin, H., Ochieng, J.W., Lkhagva, B. and Hanotte, O. (2004) Genetic diversity and relationship of domes- tic bactrian camels (Camelus bactrianus) in China and Mongolia. Journal of Camel Practice and Research 11, 97–99. Jordt, T. and Lorenzini, E. (1990) Multiple superovulations in N’Dama heifers. Tropical Animal Health and Production 22, 178–184. Livestock Genetics and Breeding 97 Karugia, J.T., Mwai, O.A., Kaitho, R., Drucker, A., Wollny, C.B.A. and Rege, J.E.O. (2001) Economic analysis of crossbreeding programmes in Sub-Saharan Africa: a conceptual framework and Kenyan case study. FEEM Working Paper No. 106. FEEM, Milan, Italy. Available at: http://dx.doi.org/10.2139/ ssrn.297058 (accessed 18 January 2020). Kasonta, J.S. (1992) Comparative growth and reproductive performance of Boran, Tanzania Shorthorn Zebu and Mpwapwa cattle breeds and various crosses. ILCA, Addis Ababa. Available at: http://agris. fao.org/agris-search/search.do?recordID=QM9400045 (accessed 18 January 2020). Kefena, E., Dessie, T., Tegegne, A., Beja-Pereira, A., Yusuf Kurtu, M., et al. (2014) Genetic diversity and matrilineal genetic signature of native Ethiopian donkeys (Equus asinus) inferred from mitochondrial DNA sequence polymorphism. Livestock Science 167, 73–79. Kemp, S.J., Brezinsky, L. and Teale, A.J. (1992) ILSTS002: a polymorphic bovine microsatellite. Animal Genetics 23, 184. Kemp, S.J., Hishida, O., Wambugu, J., Rink, A. Longeri, M.L., et al. (1995) A panel of polymorphic bovine, ovine and caprine microsatellite markers. Animal Genetics 26, 299–306. Kemp, S.J., Iraqi, F., Darvasi, A., Soller, M. and Teale, A.J. (1997) Localization of genes controlling resist- ance to trypanosomiasis in mice. Nature Genetics 16, 194–196. Kim, J., Hanotte, O., Mwai, O.A., Dessie, T., Bashir, S., et al. (2017) The genome landscape of indigenous African cattle. Genome Biology 18, 34. Kiwuwa, H., Trail, M., Kurtu, Y., Getachew, W., Anderson, F. and Durkin J. (1983) Crossbred dairy product- ivity in Arsi Region, Ethiopia. ILCA Research Report No. 11, ILCA, Addis Ababa. König, E.Z., Strandberg, E., Ojango, J.M.K., Mirkena, T., Okeyo, A.M. and Philipsson, J. (2017) Pure-breeding of Red Maasai and crossbreeding with Dorper sheep in different environments in Kenya. Animal Breeding and Genetics 134, 531–544. Kosgey, I.S. (2004) Breeding objectives and breeding strategies for small ruminant in the tropics. PhD the- sis, Wageningen University, Wageningen, The Netherlands. Kosgey, I.S. and Okeyo, A.M. (2007) Genetic improvement of small ruminants in low-input, smallholder pro- duction systems: technical and infrastructural issues. Small Ruminant Research 70, 76–88. Krishna, V., Drucker, A.G., Pascual, U., Raghu, P.T. and King, E.D.I.O., (2013) Estimating compensation payments for on-farm conservation of agricultural biodiversity in developing countries. Ecological Eco- nomics 87, 110–123. Kugonza, D.R., Jianlin, H., Nabasirye, M., Mpairwe, D., Kiwuwa, G.H., et al. (2011) Genetic diversity and differentiation of Ankole cattle populations in Uganda inferred from microsatellite data. Livestock Sci- ence 135, 135–147. Kurtu, M.Y., Tawah, C.L., Rege, J.E.O., Alemayehu, N. and Shibre, M. (1999) Lactation performance of purebred Arsi and Arsi × Friesian crosses under pre-partum and post-partum supplementary feeding regimes in Ethiopia. Animal Science 64, 625–633. Lawal, R.A., Al-Atiyat, R.M., Aljumaah, R.S., Silva, P., Mwacharo, J.M. and Hanotte O. (2018) Whole-ge- nome resequencing of red junglefowl and indigenous village chicken reveal new insights on the gen- ome dynamics of the species. Frontiers in Genetics 9, 264. Loftus, R.T., MacHugh, D.E., Bradley, D.G., Sharp, P.M. and Cunningham, P. (1994) Evidence for two independ- ent domestications of cattle. Proceedings of the National Academy of Sciences USA 91, 2757–2761. Lyimo, C.M., Weigend, A., Msoffe, P.L., Eding, H., Simianer, H. and Weigend, S. (2014) Global diversity and genetic contributions of chicken populations from African, Asian and European regions. Animal Gen- etics 45, 836–848. Marshall, K., Maddox, J.F., Lee, S.H., Zhang, Y., Kahn, L., et al. (2009) Genetic mapping of quantitative trait loci for resistance to Haemonchus contortus in sheep. Animal Genetics 40, 262–272. Marshall, K, Mugambi, J.M., Nagda, S., Sonstegard, T.S., van Tassell, C.P., Baker R.L. and Gibson J.P. (2013) Quantitative trait loci for resistance to Haemonchus contortus artificial challenge in Red Maasai and Dorper sheep of East Africa. Animal Genetics 44, 285–295. Marshall, K., Tebug, S., Juga, J., Tapio, M. and Missohou, A. (2016) Better dairy cattle breeds and better man- agement can improve the livelihoods of the rural poor in Senegal. ILRI Research Brief 65. ILRI, Nairobi. Marshall, K., Tebug, S., Salmon, G.R., Tapio, M., Juga, J. and Missohou, A. (2017) Improving dairy cattle productivity in Senegal. ILRI Policy Brief 22. ILRI, Nairobi. Martin-Collado, D., Diaz, C., Drucker, A.G., Carabaño, M.J. and Zander, K. K. (2014) Determination of non-market values to inform conservation strategies for the threatened Alistana-Sanabresa cattle breed. Animal 8, 1373–1381. Mason, I.L. (1988) A World Dictionary of Livestock Breeds, Types and Varieties, 3rd edn. CABI International, Wallingford, UK. 98 J.E.O. Rege, J. Ochieng and O. Hanotte Matika, O., Nyoni, S., van Wyk, J.B., Erasmus, G.J. and Baker, R.L. (2003) Resistance of Sabi and Dorper ewes to gastro-intestinal nematode infections in an African semi-arid environment. Small Ruminant Research 47, 95–102. Mburu, D.N., Ochieng, J.W., Kuria, S.G., Jianlin, H., Kaufmann, B., et al. (2003) Genetic diversity and rela- tionships of indigenous Kenyan camel (Camelus dromedarius) populations: implications for their clas- sification. Animal Genetics 34, 26–32. Mekonnen, A., Haile, A., Dessie, T. and Mekasha Y. (2012) On farm characterization of Horro cattle breed production systems in western Oromia, Ethiopia. Livestock Research for Rural Development 24, 100. Mendelsohn, R. (1999) Economic valuation of animal genetic resources: a synthesis report of the planning workshop. In: Economic Valuation of Animal Genetic Resources, Proceedings of an FAO/ILRI Work- shop, 15–17 March 1999, Rome. FAO, Rome, and ILRI, Nairobi, pp. 3–14. Menge, D.M., Behnke, J.M., Lowe, A., Gibson, J.P., Iraqi, F.A., et al. (2003) Mapping of chromosomal re- gions influencing immunological responses to gastrointestinal nematode infections in mice. Parasite Immunology 25, 341–349. Moyo, S., Swanepoel, F.J.C. and Rege, J.E.O. (1996) Evaluation of indigenous, exotic and crossbred cattle for beef production in a semi-arid environment: reproductive performance and cow productivity. Pro- ceedings of the Australian Society of Animal Production 21, 204–206. Mrode, R., Ojango, J.M. and Mwai, O. (2018) Developing innovative digital technology and genomic ap- proaches to livestock genetic improvement in developing countries. Journal of Animal Science 96, 507–507. Muigai, A.W.T. and Hanotte, O. (2013) The origin of African sheep: archaeological and genetic perspectives. African Archaeological Review 30, 39–50. Muigai, A.W.T., Okeyo, A.M., Kwallah, A.K., Mburu, D. and Hanotte, O. (2009) Characterization of sheep populations of Kenya using microsatellite markers: implications for conservation and management of indigenous sheep populations. South African Journal of Animal Science 39, 93–96. Muzzolini, A. (1983) L’ Art Rupestre du Sahara Central: Classification et Chronologie. Le Bœuf dans la Préhistoire Africaine [Rock Art of the Central Sahara: Classification and Chronology. Cattle in African Prehistory]. These de Troisième Cycle. University of Provence, Marseille, France. Mwacharo, J.M. and Rege, J.E.O. (2002) On-farm characterization of the indigenous small East African Zebu (SEAZ) cattle in the Southeast rangelands of Kenya. Animal Genetic Resources Information 32, 73–86. Mwacharo, J.M., Okeyo, A., Kamande, G. and Rege, J. (2006a) The small East African shorthorn zebu cows in Kenya. I: linear body measurements. Tropical Animal Health and Production 38, 65–74. Mwacharo, J.M., Jianlin, H. and Amano, T. (2006b) Native African chicken: valuable genetic resources for future breeding improvement. Journal of Animal Genetics 34, 63–69. Mwacharo, J.M., Nomura, K., Hanada, H., Jianlin, H., Hanotte, O. and Amano, T. (2007) Genetic relation- ship among Kenyan and other East African indigenous chickens. Animal Genetics 38, 485–490. Mwacharo, J.M., Ojango, J.M.K., Baltenweck, I., Wright, I., Staal, S., et al. (2009) Livestock product- ivity constraints and opportunities for investment in science and technology. Bill & Melinda Gates Foundation – ILRI Knowledge Generation Project Report, Nairobi, Kenya. Mwacharo, J.M., Bjørnstad, G., Mobegi, V., Nomura, K., Hanada, H., et al. (2011) Mitochondrial DNA reveals multiple introductions of domestic chicken in East Africa. Molecular Phylogenetics and Evolution 58, 374–382. Mwacharo, J.M., Bjørnstad, G., Han, J.L. and Hanotte, O. (2013a) The history of African village chickens: an archaeological and molecular perspective. African Archaeological Review 30, 97–114. Mwacharo, J.M., Nomura, K., Hanada, H., Han, J.L., Amano, T. and Hanotte, O. (2013b) Reconstructing the origin and dispersal patterns of village chickens across East Africa: insight from autosomal markers. Molecular Ecology 22, 2683–2697. Mwai, O., Hanotte, O., Kwon, Y.J. and Cho, S., (2015) African indigenous cattle: unique genetic resources in a rapidly changing world. Asian-Australasian Journal of Animal Sciences 28, 911–921. Mwamachi, D.M., Audho, J.O., Thorpe, W. and Baker, R.L. (1995) Evidence for multiple anthelmintic resist- ance in sheep and goats reared under the same management in coastal Kenya. Veterinary Parasit- ology 60, 303–313. Mwenya, W.N.M. (1990) Cattle breeding in SADCC countries 1. A review of cattle breeding policies and programmes. ILCA, Addis Ababa. Ndumu, D.B., Baumung, R., Hanotte, O., Wurzinger, M., Okeyo, M.A. et al. (2008) Genetic and morpho- logical characterization of the Ankole Longhorn cattle in the African Great Lakes region. Genetics, Selection and Evolution 40, 467–490. Livestock Genetics and Breeding 99 Nijman, I.J., Otsen, M., Verkaar, E.L.C., de Ruijter, C., Hanekamp, E., et al. (2003) Hybridization of banteng (Bos javanicus) and zebu (Bos indicus) revealed by mitochondrial DNA, satellite DNA, AFLP and microsatellites. Heredity 90, 10–16. Noyes, H., Brass, A., Obara, I., Anderson, S., Archibald, A.L., et al. (2011) Genetic and expression analysis of cattle identifies candidate genes in pathways responding to Trypanosoma congolense infection. Proceedings of the National Academy of Sciences USA 108, 9304–9309. Ojango, J.M., Ahuya, C., Okeyo Mwai, A. and Rege, J.E.O. (2010) The FARM-Africa dairy goat improvement project in Kenya: A case study. ILRI, Nairobi. Ojango, J.M.K., Malmfors, B., Mwai, O. and Philipsson, J. (2011) Training the trainers – an innovative and successful model for capacity building in animal genetic resource utilization in sub-Saharan Africa and Asia. ILRI and SLU, Nairobi. Ojango, J.M.K., Marete, A., Mujibi, D., Rao, J., Poole, J., et al. (2014) A novel use of high density SNP as- says to optimize choice of different crossbred dairy cattle genotypes in small-holder systems in East Africa. In: Proceedings of the 10th World Congress of Genetics Applied to Livestock Production, 17–22 August, Vancouver, Canada. American Society of Animal Science, Champaign, USA. Available at: www.asas.org/docs/default-source/wcgalp-posters/486_paper_9248_manuscript_560_0.pdf (ac- cessed 18 January 2020). Okomo, M.A., Rege, J.E.O., Teale, A. and Hanotte, O. (1998) Genetic characterisation of indigenous East African cattle breeds using microsatellite DNA markers. In: Proceedings of the 6th World Congress of Genetics Applied to Livestock Production. Vol. 27, pp. 243–246. Omondi, I., Baltenweck, I., Drucker, A.G., Obare, G. and Zander, K.K. (2008) Economic valuation of sheep genetic resources: implications for sustainable utilization in the Kenyan semi-arid tropics. Animal Health and Production 40, 615–626. Over, H.J., Jansen, J. and van Olm, P. (1992) Distribution and Impact of Helminth Diseases of Livestock in Developing Countries. FAO, Rome. Pain, A., Renauld, H., Berriman, M., Murphy, L., Yeats, C.A., et al. (2005) Genome of the host-cell trans- forming parasite Theileria annulata compared with T. parva. Science 309, 131–133. Park, W., Srikanth, K., Lim, D., Park, M., Hur, T., et al. (2018) Comparative transcriptome analysis of Ethiop- ian indigenous chickens from low and high altitudes under heat stress condition reveals differential immune response. Animal Genetics 50, 42–53. Paterson, K.A., McEwan, J.C., Dodds, K.G., Morris, C.A. and Crawford, A.M. (2001) Fine mapping a locus affecting host resistance to internal parasites in sheep. In: Proceedings of the Association for Ad- vancement in Animal Breeding and Genetics, 30 July–2 August, Queenstown, New Zealand, Vol. 13, pp. 91–94. Payne, W.J.A. (1970) Cattle Production in the Tropics, Vol. 1. General Introduction and Breeds and Breed- ing. Longman Group, London. Preston, J.M. and Allonby, E.W. (1978) The influence of breed on the susceptibility of sheep and goats to a single experimental infection with Haemonchus contortus. Veterinary Record 103, 509–512. Preston, J.M. and Allonby, E.W. (1979) The influence of breed on the susceptibility of sheep to Haemonchus contortus. Research in Veterinary Science 26, 134–139. Qi, X.B., Jianlin, H., Wang, G., Rege, J.E.O. and Hanotte, O. (2010) Assessment of cattle genetic introgres- sion into domestic yak populations using mitochondrial and microsatellite DNA markers. Animal Gen- etics 41, 242–252. Razafindraibe, H., Mobegi, V.A., Ommeh, S.C., Ralptondravao, M.L., Bjørnstad, G., et al. (2008) Mitochon- drial DNA origin of indigenous Malagasy chicken: implication for a functional polymorphism of the Mx gene. Annals of the New York Academy of Sciences 1149, 77–79. Rege, J.E.O. (1991) Genetic analysis of reproductive and productive performance of Friesian cattle in Kenya. II. Genetic and phenotypic trends. Journal Animal Breeding and Genetics 108, 424–433. Rege, J.E.O. (1992) Background to ILCA’s AGR characterization project, project objectives and agenda for the Research Planning Workshop. In: Rege, J.E.O. and Lipner, M. (eds) African Animal Genetic Re- sources: Their characterisation, conservation and utilisation. Proceedings of the Research Planning Workshop, 19–21 February 1992. ILCA, Addis Ababa. Rege, J.E.O. (ed.) (1999a) Economic valuation of farm animal genetic resources. In: Proceedings of an FAO/ILRI Workshop, 15–17 March, Rome. ILRI, Nairobi. Rege, J.E.O. (1999b) The state of African cattle genetic resources. I. Classification framework and identifi- cation of threatened and extinct breeds. Animal Genetic Resources Information 25, 1–25. Rege, J.E.O. (2008) Genetic diversity in livestock populations: trends, threats and required action. Journal of Tropical Agriculture Association No. 1 (Spring), 23–26. 100 J.E.O. Rege, J. Ochieng and O. Hanotte Rege, J.E.O. and Gibson, J.P. (2009) Identifying opportunities to improve the livelihoods of poor livestock keepers in sub-Saharan Africa through genetic improvement of livestock. A position paper prepared for the Bill & Melinda Gates Foundation (BMGF), Seattle, Washington, DC. Rege, J.E.O. and Lebbie, S.H.B. (2000) The goat resources of Africa: origin, distribution and contribution to the national economies. In: Proceedings of the International Conference on Goats (“Goat 2000”), 15–21 May, Tours, France, Vol. 2, pp. 927–931. Rege, J.E.O. and Lipner, M.E. (eds) (1992) African Animal Genetic Resources: Their Characterization, Con- servation and Utilization. Proceedings of the Research Planning Workshop, 19–21 February. Inter- national Livestock Centre for Africa, Addis Ababa. Rege, J.E.O. and Moyo, S. (1993) Effect of birth date on growth performance of calves of seasonally-bred range cattle. Journal of Animal Breeding and Genetics 110, 209–227. Rege, J.E.O. and Tawah, C.L. (1999) The state of African cattle genetic resources II. Geographical distribu- tion, characteristics and uses of present-day breeds and strains. Animal Genetic Resources Informa- tion 26, 1–25. Rege, J.E.O. and Wakhungu, J.W. (1992) An evaluation of a long term breeding programme in a closed Sahiwal herd in Kenya. II. Genetic and phenotypic trends and levels of inbreeding. Journal of Animal Breeding and Genetics 109, 374–384. Rege, J.E.O., Aboagye, G.S. and Akah, S. (1993a) A method for estimating between-breed differences in crossbreeding parameters and their sampling variances. Livestock Production Science 33, 327–340. Rege, J.E.O., von Kaufmann, R.R. and Mani, R.I. (1993b) On-farm performance of Bunaji (White Fulani) cattle I. Herd structures and cattle disposal. Animal Production 57, 199–210. Rege, J.E.O., von Kaufmann, R.R., Mwenya, W.N.M., Otchere, E.O. and Mani, R.I. (1993c) On-farm per- formance of Bunaji (White Fulani) cattle II. Growth, reproductive performance, milk offtake and mortal- ity. Animal Production 57, 211–220. Rege, J.E.O., Aboagye, G.S. and Tawah, C.L. (1994a) Shorthorn cattle of West and Central Africa I. Origin, distribution, classification and population statistics. World Animal Review 78, 2–13. Rege, J.E.O., Aboagye, G.S., and Tawah, C.L. (1994b) Shorthorn cattle of West and Central Africa II. Eco- logical settings, utility, management and production systems. World Animal Review 78, 14–21. Rege, J.E.O., Aboagye, G.S. and Tawah, C.L. (1994c) Shorthorn cattle of West and Central Africa IV. Pro- duction characteristics. World Animal Review 78, 33–48. Rege, J.E.O., Tembely, S., Mukasa, E., Sovani, S., Anindo, D., et al. (1996) Preliminary evidence for genetic resistance to endoparasites in Menz and Horro lambs in the highlands of Ethiopia. In: Lebbie, S.H.B. and Kagwini, E. (eds) Small Ruminant Research and Development in Africa. Proceedings of the Third Biennial Conference of the African Small Ruminant Research Network, 5–9 December 1994, UICC, Kampala, Uganda. ILRI, Nairobi. Rege, J.E.O., Kahi, A.K., Okomo-Adhiambo, M., Mwacharo, J. and Hanotte, O. (2001) Zebu cattle of Kenya: uses, performance, farmer preferences, measures of genetic diversity and strategies for improved use. ILRI, Nairobi. Rege, J.E.O., Tembely, S., Mukasa-Mugerwa, E., Sovani, S., Anindo, D., et al. (2002) Effect of breed and season on production and response to infections with gastro-intestinal nematode parasites in sheep in the highlands of Ethiopia. Livestock Production Science 78, 159–174. Rege, J., Hanotte, O., Mamo, Y., Asrat, B. and Dessie, T. (2007) Domestic Animal Genetic Resources Infor- mation System (DAGRIS). ILRI, Addis Ababa. Available at: www.ilri.org/research/projects/domes- tic-animal-genetic-resources-information-system-dagris (accessed 19 January 2020). Rege, J.E.O., Marshall, K., Notenbaert, A., Ojango, J.M.K. and Okeyo, A.M. (2011) Pro-poor animal im- provement and breeding – what can science do? Livestock Science 136, 15–28. Reist-Marti, S.B., Simianer, H., Gibson, J., Hanotte, O. and Rege, J.E.O. (2003) Weitzman’s approach and con- servation of breed diversity: an application to African cattle breeds. Conservation Biology 17, 1299–1311. Reynolds, L., Baker, R.L., Sherington, J. and Njubi, D. (1992) Resistance to gastro-intestinal parasites in Dorper and Red Maasai–Dorper crossbred sheep. In: Proceedings of the 10th Scientific Workshop of the Small Ruminant Collaborative Research Support Programme, Nairobi, Kenya, pp. 131–137. Rowlands, G.J., Nagda, S., Rege, J.E.O., Mhlanga, F., Dzama, K., et al. (2003) A report to FAO on ‘The design, execution, and analysis of livestock breed surveys – a case study in Zimbabwe.’ ILRI, Nairobi. Salmon, G., Teufel, N., Baltenweck, I., van Wijk, M., Claessens, L. and Marshall, K. (2018) Trade-offs in livestock development at farm level: different actors with different objectives. Global Food Security 17, 103–112. Sani, R.A., Gray, G.D. and Baker, R.L. (2004) Worm Control for Small Ruminants in Tropical Asia. ACIAR, Canberra. Livestock Genetics and Breeding 101 Sayers, G. and Sweeney, T. (2005) Gastrointestinal nematode infection in sheep – a review of the alterna- tives to anthelmintics in parasite control. Animal Health Research Reviews 6, 159–171. Scarpa, R., Kristjanson, P., Ruto, E., Radeny, M., Drucker, A. and Rege, J.E.O. (2003a) Valuing indigenous farm animal genetic resources in Africa: an empirical comparison of stated and revealed preference value estimates. Ecological Economics 45, 409–426. Scarpa, R., Drucker, A.G., Anderson, S., Ferraes-Ehuan, N., Gomez, V., et al. (2003b) Valuing genetic re- sources in peasant economies: the case of ‘hairless’ Creole pigs in Yucatan. Ecological Economics 45, 427–443. Shaw, A.P.M. and Hoste, C.H. (1987) Trypanotolerant cattle and livestock development in West and Central Africa, Vol. II. Country studies. Animal Production and Health Paper No. 67/2, FAO, Rome. Silva, M.V.B., Sonstegard, T.S., Hanotte. O., Mugambi, J.M., Garcia, J.F., et al. (2011) Identification of quan- titative trait loci affecting resistance to gastrointestinal parasites in a double backcross population of Red Maasai and Dorper sheep. Animal Genetics 43, 63–71. Simianer, H., Marti, S.B., Gibson, J.P., Hanotte, O. and Rege, J.E.O. (2003) An approach to the optimal al- location of conservation funds to minimize loss of genetic diversity between livestock breeds. Journal of Ecological Economics 45, 377–392. Tandoh, G., Gwaza, D. and Addass, P. (2018) Phenotypic characterization of camels (Camelus dromedar- ius) in selected herds of Katsina State. Journal of Applied Life Sciences International 18, 1–11. Tarekegn, G.M, Gizaw, S., Dessie, T., Mwai, O., Djikeng, A. and Tesfaye, K. (2016) A review on current know- ledge of genetic diversity of domestic goats (Capra hircus) identified by microsatellite loci: how those efforts are strong to support the breeding programs? Journal of Life Science and Biomedicine 6, 22–32. Tarekegn, G.M., Tesfaye, K., Mwai, O.A., Djikeng, A., Dessie, T., et al. (2018) Mitochondrial DNA variation reveals maternal origins and demographic dynamics of Ethiopian indigenous goats. Ecology and Evo- lution 8, 1543–1553. Tawah, C.L. and Rege, J.E.O. (1996a) Gudali cattle of West and Central Africa. Animal Genetic Resources Information 17, 159–178. Tawah, C.L. and Rege, J.E.O. (1996b) White Fulani cattle of West and Central Africa. Animal Genetic Resources Information 17, 137–158. Tawah, C.L., Mbah, D.A., Rege, J.E.O. and Oumate, H. (1993) Genetic evaluation of birth and weaning weight of Gudali and two-breed synthetic Wakwa beef cattle populations under selection in Cameroon: genetic and phenotypic parameters. Animal Production 57, 73–79. Tawah, C.L., Rege, J.E.O., Mbah, D.A. and Oumate, H. (1994) Genetic evaluation of birth and weaning weight of Gudali and two-breed synthetic Wakwa beef cattle populations under selection in Camer- oon: genetic and phenotypic trends. Animal Production 58, 25–34. Tawah, C.L., Rege, J.E.O. and Aboagye, G.S. (1997) A close look at a rare African breed – the Kuri cattle of Lake Chad Basin: origin, distribution, production and adaptive characteristics. South African Journal of Animal Science 27, 31–40. Tawonezvi, H.P.R., Ward, H.K., Trail, J.C.M. and Light, D. (1988a) Evaluation of beef breeds for rangeland weaner production in Zimbabwe 1. Productivity of purebred cows. Animal Science 47, 351–359. Tawonezvi, H.P.R., Ward, H.K., Trail, J.C.M. and Light, D. (1988b) Evaluation of beef breeds for rangeland weaner production in Zimbabwe 2. Productivity of crossbred cows and heterosis estimates. Animal Science 47, 361–367. Taye, M., Yilma, M., Rischkowsky, B., Dessie, T., Okeyo, M., et al. (2016) Morphological characteristics and linear body measurements of Doyogena sheep in Doyogena district of SNNPR, Ethiopia. African Jour- nal of Agricultural Research 11, 4873–4885. Taye, M., Lee, W., Caetano-Anolles, K., Dessie, T., Cho, S., et al. (2018) Exploring the genomes of East African Indicine cattle breeds reveals signature of selection for tropical environmental adaptation traits. Cogent Food and Agriculture 4, 1552552. Taylor, J.F., Brenneman, R.A., Burns, B.M., Davis, S.K., Yeh, C.-C., et al. (1998) Report of the first workshop on the genetic map of bovine chromosome 1. Animal Genetics 29, 228–235. Teale, A.J. (1993) Improving control of livestock disease: animal biotechnology in the Consultative Group for International Agricultural Research. Bioscience 43, 475–483. Thakur, M., Fernandes, M., Sathyakumar, S., Singh, S.K., Vijh, R. et al. (2018) Understanding the cryptic introgression and mixed ancestry of Red Junglefowl in India. PLoS One 13, e0204351. Thorpe, W., Kangethe, P., Rege, J.E.O., Mosi, R., Mwandotto, B.A.J. and Njuguna, P. (1993) Crossbreeding Ayrshire, Friesian and Sahiwal cattle for milk yield and preweaning traits of progeny in the semiarid tropics of Kenya. Journal of Dairy Science 76, 2001–2012. 102 J.E.O. Rege, J. Ochieng and O. Hanotte Trail, J.C.M. (1981) Merits and demerits of importing exotic cattle compared with the improvement of local breeds: cattle in Africa south of the Sahara. BSPA Occasional Publication 4, 191–231. Trail, J.C.M. and Gregory, K.E. (1981) Characterisation of the Boran and Sahiwal breeds of cattle for eco- nomic characters. Journal of Animal Science 52, 1286–1293. Trail, J.C.M. and Gregory, K.E. (1982) Production characters of the Sahiwal and Ayrshire breeds and their crosses in Kenya. Tropical Animal Health and Production 14, 45–57. Trail, J.C.M., Hoste, C.H., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979a) Trypanotolerant livestock in West and Central Africa. Vol. 1. General study. ILCA Monograph No. 2. ILCA, Addis Ababa. Trail, J.C.M., Hoste, C., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979b) Trypanotolerant livestock in West and Central Africa, Vol. 2. Country studies. ILCA Monograph No. 2. ILCA, Addis Ababa. Trail, J.C.M., Gregory, K.E., Marples, H.J.S. and Kakonge, J. (1982) Heterosis, additive maternal and addi- tive direct effects of the Red Poll and Boran breeds of cattle. Journal of Animal Science 54, 517–523. Trail, J.C.M., Gregory, K.E., Durkin, J. and Sandford, J. (1984) Crossbreeding cattle in beef production pro- grammes in Kenya. II. Comparison of purebred Boran crossed with the Red Poll and Santa Gertrudis breeds. Tropical Animal Health and Production 16, 191–200. Trail, J.C.M., Sones, K., Jibbo, J.M.C., Durkin, J., Light, D.E. and Murray, M. (1985) Productivity of Boran cattle maintained by chemoprophylaxis under trypanosomiasis risk. Research Report 9. ILCA, Addis Ababa. Traoré, A., Royo, L.J., Kaboré, A., Alavarez, I., Fernandez, I., et al. (2012) Preliminary assessment of pheno- typic and genotypic parameters associated with resistance to gastrointestinal parasites infestation in Burkina Faso sheep. European Journal of Experimental Biology 2, 1359–1366. Waller, P.J. (1997) Anthelmintic resistance. Veterinary Parasitology 72, 391–412. Wang, T., Sun, Y. and Qiu, H.-J. (2018) African swine fever: an unprecedented disaster and challenge to China. Infectious Diseases of Poverty 7, 111. Warui, H.M. (2008) Characterisation of sheep and goat genetic resources in their production system con- text in northern Kenya. PhD thesis, University of Hohenheim, Stuttgart, Germany. Willyard, C. (2011) Putting sleeping sickness to bed. Nature Medicine 17, 14–17. Woolaston, R.R. and Baker, R.L. (1996) Prospects of breeding small ruminants for resistance to internal parasites. International Journal for Parasitology 26, 845–855. Wragg D., Mwacharo J., Alcalde J., Hocking P. and Hanotte O. (2012) Analysis of genome-wide structure, diversity and fine mapping of Mendelian traits in traditional and village chickens. Heredity 109, 6–18. Xuebin, Q., Jianlin, H., Lkhagva, B., Chekarova, I., Badamdorj, D., et al. (2005) Genetic diversity and differ- entiation of Mongolian and Russian yak populations. Journal of Animal Breeding and Genetics 122, 117–126. Yapi, C.V., Oya, A. and Rege, J.E.O. (1994) Evaluation of an open nucleus breeding programme for growth of the Djallonké sheep in Côte d’Ivoire. In: Proceedings of the 5th World Congress on Genetics Ap- plied to Livestock Production, 7–12 August, Guelph, Canada. International Committee for World Con- gress on Genetics Applied to Livestock Production, Guelph, Ontario, Vol. 20, pp. 421–424. Yapi-Gnaorè, C.V., Oya, A., Rege, J.E.O. and Dagnogo, B. (1997a) Analysis of an open nucleus breeding programme for Djallonkè sheep in the Ivory Coast. 1. Examination of non-genetic factors. Animal Science 64, 291–300. Yapi-Gnaorè, C.V., Rege, J.E.O., Oya A. and Alemayehu, N. (1997b) Analysis of an open nucleus breeding programme for Djallonkè sheep in the Ivory Coast. 2. Response to selection on body weights. Animal Science 64, 301–307. Yu, M., Muteti, C., Ogugo, M., Ritchie, W., Raper, J. and Kemp, S. (2016) Cloning of the African indigenous cattle breed Kenyan Boran. Animal Genetics 47, 510–511. Zander, K.K., Signorello, G., de Salvo, M., Gandini, G. and Drucker, A.G. (2013) Assessing the total eco- nomic value of threatened livestock breeds in Italy: implications for conservation policy. Ecological Economics 93, 219–229. Zonabend, E., Okeyo, A.M., Ojango, J.M.K., Hoffmann, I., Moyo S. and Philipsson, J. (2013) Infrastructure for sustainable use of animal genetic resources in Southern and Eastern Africa. Animal Genetic Resources 53, 79–93. 2 Control of Pathogenesis in African Animal Trypanosomiasis: A Search for Answers at ILRAD, ILCA and ILRI, 1975–2018 Samuel J. Black University of Massachusetts at Amherst, Massachusetts, USA Contents Executive Summary 104 The problem 104 Scientific impacts 104 Development impacts 105 Economic impacts 105 Capacity building 106 Partnerships 106 Introduction 106 Scientific Challenges 107 African animal trypanosomiasis (AAT) causative agents and tsetse, circa 1970 107 Tsetse distribution and effects on cattle, circa 1970 108 Antigenic variation and recurring parasitaemia, circa 1970 108 AAT pathogenesis, circa 1970 109 ILRAD’s Initial AAT Mandate 110 ILRAD 1975–1979 111 Initial experimental questions on AAT vaccine development and control 111 AAT Research at ILRAD, ILCA and ILRI, 1979–2017 112 The search for AAT vaccine antigens 112 Metacyclic and first-generation bloodstream-stage trypomastigote VSGs 112 Candidate vaccine antigens other than VSGs 115 Other vaccine studies 116 Alternatives to an AAT vaccine 117 Improved AAT Management through Use of Trypanotolerant Stock 117 African Trypanotolerant Livestock Network 118 The dark-ground buffy coat phase-contrast diagnostic technique 119 Antigen enzyme-linked immunosorbent assays, isoenzymes and polymerase chain reaction 119 Ranching with chemoprophylaxis and chemotherapy: the N’Dama advantage 122 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 103 104 S.J. Black Parameters of trypanosome infection and animal health under natural tsetse/trypanosome challenge in N’Dama cattle 122 The cell and molecular biology of trypanotolerance in cattle 124 T. congolense cyclic infection and homologous cyclic reinfection of N’Damas and Borans 124 AAT-induced anaemia 125 AAT-induced lymphopenia and possible role of haemophagocytic syndrome in AAT 128 Antibody responses of trypanosome-infected N’Dama and Boran cattle 128 AAT-induced T-cell responses and immunosuppression 130 The genetic basis of trypanotolerance 132 QTLs and murine trypanotolerance 132 QTLs and bovine trypanotolerance 134 Conclusions 135 The Future 136 References 136 Executive Summary population with traps and insecticides, and in areas with a high population of trypanosome- The problem infected tsetse, animals are prophylactically administered antiparasitic drugs. To date, there African animal trypanosomiasis (AAT, also is no AAT vaccine available, as discussed below. known as ‘animal African trypanosomiasis’) is a serious disease of the tropics and subtropics, adversely affecting cattle production as animals Scientific impacts suffer from loss of condition, emaciation and anaemia, resulting in reduced meat and milk The International Livestock Research Institute production and draught power for agricultural (ILRI) and its two predecessors, the International production. Cattle mortality can reach 50–100% Livestock Centre for Africa (ILCA) and the Inter- within months of exposure. The disease is caused by national Laboratory for Research on Animal Dis- parasites that live in the host blood plasma, body eases (ILRAD), have made significant contributions tissue and interstitial fluids. Trypanosomes are to trypanosomiasis research since the early 1970s. transmitted to the host by a vector, the tsetse fly. The first contribution was the establishment of The parasite replicates within the tsetse fly and is protocols for cultivation in vitro of bloodstream transmitted through saliva when the fly feeds on stages of Trypanosoma brucei and subsequently the animals. While this disease predominantly oc- Trypanosoma congolense. This advance facilitated curs in sub-Saharan Africa, it has also been found downstream investigations of T. brucei and T. congo- in South America, where one AAT agent (Trypano- lense cell division cycles, endocytic processes, inter- soma vivax) has been established and tabanids action with antibodies against variable surface (biting flies) act as the mechanical vector. The glycoproteins (VSGs) and other trypanosome anti- most rigorous calculation of the cost of AAT in gens, interactions with trypanocidal drugs and sub-Saharan Africa dates from the late 1990s mechanisms of development of drug resistance. (Kristjanson et al., 1999). Kristjanson estimated the The second contribution was recognition that annual cost to be US$1.3 billion, excluding losses trypanosome strain complexity and surface coat from potential output in regions where trypano- antigenic variation precluded the development of somes prevent livestock production and excluding an effective conventional vaccine targeting the im- the costs of foregone power and manure output. munodominant coat antigens. Subsequent iden- Animal trypanosomiasis can be managed tification of required macromolecular growth by three strategies: (i) vector control/eradication; nutrients and uptake pathways also did not lead to (ii) use of trypanocides; and (iii) use of trypanotol- an effective vaccine because, under steady-state erant breeds of cattle (see Chapter 3, this volume). conditions, anti- receptor antibodies did not kill the Vector control includes reducing the tsetse fly parasites or prevent acquisition of the nutrients. Control of Pathogenesis in Animal African Trypanosomiasis 105 Although the vaccine approach proved unsuccess- in cattle proved greater than that found in ful, studies in this area at ILRAD advanced our crosses between inbred strains of mice; here, understanding of trypanosome biology. three regions of host DNA were shown to be The third contribution was a deepening associated with trypanotolerance and a candi- understanding of bovine immunology per se (see date resistance gene, Pram1 (PML-R ARA-regulated Chapter 4, this volume) and as it pertains to adaptor molecule 1), has been proposed for one responses against African trypanosomes and of these regions. other pathogens. A fifth contribution was the discovery of A fourth contribution was an understanding diagnostics and subsequent efforts to prevent of the biological and genetic basis of trypanotoler- and contain the development of trypanocidal ance. ILRAD, ILRI and their partners were able to drug resistance in the parasite. characterize N’Dama cattle, a small multi- purpose ILRAD and subsequently ILRI developed indigenous breed that can survive without computerized data management and analysis chemotherapy in some regions of sub-Saharan systems, geographic information systems and Africa where AAT kills susceptible breeds. digital georeferenced databases to address the Field research in the area of trypanotoler- distribution and dynamics of AAT in the contin- ance established the need for complementary ent. Combined with measures of productivity technologies, such as spraying to kill the tsetse during infection, such mapping tools have been vector and trypanocidal drugs. used to evaluate trypanotolerance, to define the Analysis of trypanosomiasis in trypa- effect of trypanosomiasis control on land use and notolerant N’Dama cattle and their crosses biodiversity, and to enhance decision making in with trypanosomiasis-susceptible Boran cattle livestock development programmes. ILRI’s epi- found that two major indicators of trypanotol- demiological research highlights include the de- erance, namely control of anaemia and para- velopment of a modelling technique to evaluate sitaemia, are unlinked. The work introduced control options of AAT, such as: (i)  chemother- an  additional indicator of trypanotolerance, apy, which remains the main parasite control namely the ability to generate IgG1 antibodies option; (ii) trypanotolerant cattle, which is an against buried VSG epitopes and epitopes on i mportant option if complementary chemo- many common trypanosome antigens, including prophylaxis is adequate; and (iii)  tsetse control congopain and heat-s hock protein 70 (Hsp70)/ methods, which are well established. binding immunoglobulin protein (BiP). Traits In terms of top-cited papers, ILRI has con- that distinguished between N’Dama and Boran tributed to 36% of the global research outputs on cattle trypanotolerance in cattle proved to be animal African trypanosomiasis and 64% of the regulated by multiple unlinked genes. Mapping global research outputs on trypanosomiasis re- of the trypanotolerance genes in cattle was an sistance, as shown in the Altmetric (www.altmet- integral, and initially a leading, component of ric.com/; accessed 5 February 2020) database. the bovine genome- mapping programme world- wide. The results of the study were consistent with a single quantitative trait locus (QTL) on Development impacts each of 17 chromosomes, and two on BTA16 (Bos taurus chromosome 16), with individual Economic impacts QTL effects ranging from about 6% to 20% of the phenotypic variance of the trait, weighing Estimates of economic losses to trypanosomiasis against use of markers for these QTLs in breed- in sub-Saharan Africa have been made several ing programmes to enrich for the trait. Archives times. Jahnke’s pioneering work (1976) showed of genomic and complementary DNA (cDNA) the potential output losses in East Africa. Several have been established from the bovine trypa- papers from ILCA/ILRAD (1988) showed positive notolerance studies carried out at ILRAD/ILRI economic returns to various forms of control, and are available for deeper analysis as technol- including spraying and trapping flies, trypan- ogy in this field undergoes further development ocides and use of trypanotolerant animals. to facilitate linking genes to disease-resistance Research has therefore had an economic impact traits. The genetic complexity of trypanotolerance outside the vaccine domain. 106 S.J. Black With respect to vaccine development, a brings together agents working on this disease, 1990s ex ante assessment showed that a vaccine ranging from rural communities to governments, against trypanosomiasis in Africa would have research institutes and development agencies. generated a real ex ante rate of return of 33% (Kristjanson et al., 1999). However, despite the known losses to trypanosomiasis and the poten- tial economic gains to reducing those losses, Introduction it has not been possible to estimate indirect eco- nomic impacts arising from the main scientific In 1970, Wiley published The African Trypanoso- impacts of trypanosomiasis research. In par- miases, edited by H.W. Mulligan and W.H. Potts. ticular, ex post development impacts of vaccine Most of the contributors to the book, including research were zero because decades of invest- A.R. (Ross) Gray, who served as Director General ment in vaccine research failed to produce an of ILRAD from 1982 to 1994, had gained expert- effective vaccine. ise on African trypanosomes, their vectors, tryp- anosomiasis pathogenic processes and disease Capacity building control strategies while working in Africa. As the first comprehensive work on these topics, the book ILRI has built capacity in both AAT control and was commissioned by the Trypanosomiasis Advis- farmer field training and information sharing in ory Panel of the Ministry of Overseas Develop- rational drug use. An estimated 258 scientists and ment of Great Britain and sponsored by the same students were trained through ILRAD, ILCA and ministry. The publication was in many respects a ILRI on trypanosomiasis, of which 61 were specif- legacy to independent Africa from expatriate sci- ically from the African Trypanotolerant Livestock entists who had worked to control the devastat- Network (ATLN). The institution has supported ing spread of human African trypanosomiasis 63 PhD students, 26 MSc students and four in- (HAT) and AAT during the colonial period, pos- terns working on trypanosomiasis research. sibly as a result of changing ecological dynamics associated with colonial conquest and manage- Partnerships ment (Ford, 1971; Scoones, 2014). HAT and AAT are caused by tsetse-transmitted Addressing AAT and the problems associated protozoan parasites that are endemic in the with trypanocide resistance has led to signifi- humid and semi-humid zones of sub-Saharan cant research collaborations, including: the Africa. As a landmark book, Mulligan and Potts Centre de Coopération Internationale en Re- (1970) set the baseline for trypanosomiasis re- cherche Agronomique pour le Développement/ search programmes at ILRAD and ILCA and later Département d’Elevage et de Médecine Vétéri- at ILRI. The African Trypanosomiases set a stand- naire ( CIRAD/EMVT, France); the Centre ard against which progress in the field since International de Recherche-Développement sur 1970, both fundamental and practical, can be l’Elevage en zone Subhumide, Direction Provin- measured, because it documents the scientific ciale des Resources Animales (DPRA, Burkina investment made during the colonial period in Faso); the Food and Agriculture Organization of West, Central and East Africa to control HAT, the United Nations (FAO, USA); Freie Universität and to a lesser extent, AAT. Indeed, the decision Berlin (FU-Berlin, Berlin); the International to publish The African Trypanosomiases attests to Trypanotolerance Centre (ITC, Gambia); Justus the fragility of trypanosomiasis control strat- Liebig University Giessen (Germany); the Labo- egies then in use, namely those based on tsetse ratoire Central Vétérinaire (LCV, Mali); Oxford control by environmental engineering, vector University (UK), Prince Leopold Institute of Trop- trapping and application of insecticides, and dis- ical Medicine (now the Institute of Tropical ease control by diagnosis and chemotherapy, all Medicine, Belgium); the University of Edinburgh of which can break down in the face of infra- (UK); the University of Glasgow (UK); and Uni- structure disturbance or parasite resistance to versity of Hannover (Germany). ILRI and its pre- trypanocidal drugs. decessors have contributed to the Programme The Consultative Group on International Against African Trypanosomiasis (PAAT), which Agricultural Research (CGIAR) was formed, in Control of Pathogenesis in Animal African Trypanosomiasis 107 1971, shortly after the publication of The African advances in trypanosome biology, biochemistry, Trypanosomiases. In support of its mandate, CGIAR immunology and immunopathology, including established two livestock research institutes, IL- new understanding of trypanosome virulence RAD, in Nairobi, Kenya, in 1973, and ILCA, in and host resistance mechanisms. Despite these Addis Ababa, Ethiopia, in 1974. ILRAD was to advances, the scientific community is still some ‘serve as a world center for research on ways and way from creating tools that would help farmers means of conquering, as quickly as possible, in sub-Saharan Africa to improve their livestock major animal diseases which seriously limit live- production in trypanosomiasis-endemic areas. stock industries in Africa and many other parts Examples of such tools would include an inex- of the world’ and to ‘concentrate initially on in- pensive, sensitive and specific AAT diagnostic tensive research concerning the immunological test to validate the need for, and efficacy of, and related aspects of controlling trypanosom- chemotherapy; new inexpensive multi-target iasis and theileriosis (mainly East Coast fever)’ trypanocidal drugs; a vaccine to prevent AAT or with the goal of decreasing the incidence and/or accelerate its cure in trypanosomiasis-sensitive severity of disease. ILCA was to ‘assist national livestock; and fully trypanotolerant livestock efforts which aim to effect a change in the with production traits more closely aligned to production and marketing systems of tropical those of improved breeds than of the smaller Africa so as to increase the sustained yield and multi-use N’Dama cattle. However, it is still rea- output of livestock products and improve the sonable to expect eventual success in developing quality of life of the people of this region’. at least some of these tools. ILRAD, ILCA and The mandates of these institutes were par- ILRI scientists have contributed to this work in tially merged in 1977 upon establishment of the areas of AAT diagnostics, understanding ATLN, which was based on the ILRAD campus mechanisms of AAT pathogenesis, defining the in Nairobi. The network investigated the use of trypanotolerance phenotype, identifying QTLs trypanotolerant cattle, primarily those of the that govern trypanotolerance and evaluating N’Dama breed, as a resource for productive live- putative vaccine antigens. These contributions stock farming in areas of sub-Saharan Africa are discussed after the following review of the where AAT is endemic; this was accomplished state of knowledge in 1970 regarding AAT, through analyses of animal health and produc- which outlines the problems tackled by ILRAD/ tion databases assembled from 13 countries in ILCA/ILRI. West, Central and East Africa (reviewed by d’Ieteren et al., 1998). N’Dama are West African taurine cattle that have been farmed in Africa in tsetse- endemic areas for several thousand years (Has- Scientific Challenges san, 2000). These multi-use livestock animals have been used to establish commercial herds AAT causative agents and tsetse, c.1970 and are undergoing selection for production traits. However, N’Dama cattle are still not popu- African animal trypanosomiasis (AAT) in live- lar in East Africa, where cattle producers favour stock is caused by infection with any of three spe- larger breeds, despite their susceptibility to AAT. cies of African trypanosomes, namely, T. brucei, Because of their trypanotolerant phenotype, T. congolense and T. vivax. The pathogenic proto- comparative analyses of N’Damas and less tryp- zoans are transmitted to their mammal hosts in anotolerant breeds with respect to immune the saliva of tsetse flies (genus Glossina) in which responses, infection-induced pathology and the the parasites undergo cyclic development and of genetic basis of disease control became a focus which there are more than 30 species and sub- of trypanosomiasis research at ILRAD and, with species, with eight playing a major role in tryp- respect to identification of markers for selective anosome transmission (Cecchi et al., 2015). breeding of trypanotolerant cattle, remain so African trypanosomes can also be transferred today at ILRI, which was established in 1995 by mechanically between hosts in blood held within merging ILRAD and ILCA. the proboscis of biting flies other than tsetse, In the 46 years since publication of The Afri- predominantly horse flies (family Tabanidae) can Trypanosomiases, there have been fundamental and stable flies (family Muscidae). Mechanical 108 S.J. Black transmission of the parasites by biting flies been estimated at various times since the found- causes spread of AAT within and across herds. ing of ILRAD to cover between 8.7 million and African trypanosomes live extracellularly 10.3 million km2 of the humid and semi-humid in the blood of their mammal hosts, and in the zones of sub-Saharan Africa. Ranching of cattle case of T. brucei and T. vivax, also in tissues. They is possible on the fringes of the tsetse habitat can be detected and distinguished from each where tsetse and trypanosome challenge are other by microscopic examination of wet, thin relatively low, but this requires support from or thick blood films, or dried, fixed and stained trypanocidal drugs and insecticides. Chemothera- thin blood films. This simple diagnostic test is peutic support is also required for ranching of effective when the level of parasitaemia is high the relatively trypanotolerant taurine breeds of but can be problematic at other times. Fiennes West Africa in trypanosomiasis-endemic areas, (1970) reported, ‘It was standard practice to although not to the same extent as that required examine 600 fields of thick smear preparations for trypanosomiasis-susceptible breeds. Use of (of bovine blood) but in some cases trypano- chemotherapy and vector control to manage somes were only detected after weeks or even AAT is expensive and only partially effective. months of daily searching.’ In contrast, cryptic In those areas where challenge is low enough to trypanosome infections could sometimes be permit ranching, calf mortality is still 6–10% r evealed by inoculation of putatively infected higher than in regions where trypanosomiasis blood into laboratory mice, although this was is not endemic, death in older animals is 2–8% often unsuccessful because not all trypanosomes higher, annual calving rates are 7% lower, milk grow in mice. Thus, definitive diagnosis of AAT yields are 2–26% lower and oxen are 38% less in the field was not always possible in the 1970s. efficient (Shaw, 2009). AAT-associated production losses in Africa were recently estimated to exceed US$4 billion a year (AU-IBAR, 2018). Tsetse distribution and effects on cattle, c.1970 Tsetse fly-infested areas of Africa extend from Antigenic variation and recurring the southern edge of the Sahara Desert to parasitaemia, c.1970 Angola, Zimbabwe and Mozambique. Of the three trypanosome species that cause AAT, only Gray (1970) reviewed the state of knowledge T. vivax is found in the western hemisphere, in of protective, but variable, antigens on African approximately ten countries in the Caribbean trypanosomes in Mulligan’s The African Trypano- and South and Central America. AAT is endemic somiases. Briefly, it was shown in the 1960s that in livestock maintained in tsetse-infested areas cell-free serum of infected animals, and wash of Africa. The disease is chronic and often fatal buffer of trypanosomes enriched by differential in cattle grazed in regions that are largely free of centrifugation, contained trypanosome material, wildlife species, although it sometimes resolves called exoantigen, that elicited trypanosome- without treatment. Cattle under chemothera- agglutinating antibodies, indicating that target peutic support can achieve immunity provided antigens were displayed on the surface of healthy they are exposed to restricted regional stocks of trypanosomes. Antibodies are disease-fighting T. brucei brucei, T. congolense and T. vivax. In con- proteins that are secreted by plasma cells, which trast, AAT is typically acute and fatal in regions are terminally differentiated B-lymphocytes or where livestock come into contact with trypano- B-cells. Antibodies against exoantigen were vari- somes that are transmitted from the sylvatic ant specific (i.e. reacted only with the exoantigen (wildlife) reservoir by tsetse, possibly reflecting to which they were raised) and protected against the intensity of challenge, trypanosome strain that variant but not others. diversity, and the presence of strains of the para- Results from several scientists showed that sites that are highly virulent (van den Bossche a single trypanosome could give rise to many dif- et al., 2011; Motloang et al., 2014). ferent antigenic types and Gray (1967, 1970) As a result of acute AAT, cattle are excluded raised the possibility that ‘the total number of from much of the tsetse fly habitat, which has antigens produced in one [trypanosome-infected] Control of Pathogenesis in Animal African Trypanosomiasis 109 host may be limited only by the time the animal disease, the spleen became small and atrophic, lives’. However, despite the very large repertoire showing cell depletion. Furthermore, during the of variable surface antigens expressed by blood- chronic stage of the disease, the red bone mar- stream-stage parasites, there was evidence, al- row of the shafts of the long bones disappeared. though far from convincing, that passage of a The infection-induced loss of erythropoiesis from trypanosome strain through a tsetse fly resulted the long bones may therefore affect the animal’s in expression of a basic antigenic type by the capacity to replace red blood cells. Fiennes also mammal-infective forms (metacyclic parasites), reported that fatty tissues throughout the body, and when transmitted to new hosts these gave especially around the heart and kidneys, showed rise to a set of bloodstream-stage parasites with degeneration, the lungs showed marbling due to a restricted set of predominant antigenic types dilation of lymphatic vessels and became oe- responsible for the first few waves of parasitaemia. dematous, the cardiac muscle also became flabby These findings raised hopes that a composite and oedematous, and exudates developed in the vaccine based on a combination of the common pleural and peritoneal cavities, all consistent and predominant exoantigens would be broadly with global inflammation. In addition, Fiennes protective. Despite this optimism, Gray also com- reported that, as the infection progressed, para- mented on the lack of knowledge of the struc- sitaemia often became cryptic, but aggregates of ture of exoantigens, mechanisms of antigenic dead trypanosomes and areas of necrosis could variation in mammals and tsetse, and the extent be found in tissues in the case of T. brucei, and in of variable surface antigen diversity, knowledge capillaries often associated with small focal that would certainly be needed to evaluate pos- necroses in the case of T. vivax and T. congolense. sible use of basic and predominant antigens in a There was also multiple organ and tissue degen- combinatorial vaccine. eration in which kidneys became necrotic, the liver showed enlargement accompanied by cen- tral lobular necrosis in the parenchyma, and AAT pathogenesis, c.1970 there was dilation of central veins and sinusoids and activation of phagocytic Kupffer cells. In A good deal of information had been assembled addition, lymph nodes became fibrotic and their on AAT pathogenesis before ILRAD was estab- follicles and germinal centres were depleted of lished. In Mulligan’s The African Trypanosomiases, mature lymphocytes, showing that AAT-induced Fiennes (1970) reported that infections of cattle destruction of secondary lymphoid organs was with T. brucei, T. congolense and T. vivax give a not restricted to the spleen. Thyroid and adrenal similar disease picture, suggesting ‘that the fun- glands were also severely affected in AAT, the damental processes of pathology in all forms of former filling with colloid before disintegration animal trypanosomiasis are possibly the same’. and the latter becoming necrotic and fibrotic. Fiennes reported that the cardinal signs of tryp- Anaemia is a consistent parameter of AAT anosomiasis consist of fever that spikes on clear- and was used productively by scientists at ILRAD/ ance of trypanosome parasitaemic waves but is ILCA/ILRI in comparative studies of AAT patho- later sustained, anaemia, cachexia/emaciation and genesis in infected N’Dama and Zebu cattle. In hypoproteinaemia/hypervolaemia. The severity 1970, little was known about molecular mech- of these signs of disease was observed to vary anisms of anaemia in AAT, although analyses of depending on the age of the infected bovid, the cattle infected with T. congolense or T. vivax led virulence of the infecting parasites and the stage Fiennes (1970) to propose four different courses of infection. Fiennes also noted that other patho- of AAT in which anaemia featured differently: gens ellicit similar signs of disease in cattle, e.g. (i) hyperacute, which was characterized by Babesia bigemina, which causes red-water fever; severe haemolysis and early death; (ii) acute, consequently, the clinical signs of AAT are not which was characterized by hydraemia followed pathognomonic. by dehydration, a haemolytic crisis and death With respect to morbid anatomy, Fiennes (hydraemia is an increase in blood volume reported that the spleen and lymph nodes of in- through water retention, resulting in a decrease fected cattle became greatly enlarged in the early in the specific gravity and protein concentration stages of AAT, but during the chronic stage of of blood plasma as well as a decrease in blood 110 S.J. Black packed cell volume (PCV) and red cell or haemo- of bloodstream parasites had only a few antigen- globin content per unit volume); (iii) chronic, ically distinct exoantigens, which might there- which was similar to acute but hydraemia per- fore serve as vaccine antigens. Little or nothing sisted without dehydration and the haemolytic was known about immune responses to parasite crisis was not fatal; and (iv) recovery, which was components other than the immunodominant rare and typically occurred in AAT with little or exoantigens. In addition, little or nothing was no hydraemic phase. Mechanisms of hydraemia known about trypanosome virulence factors or and dehydration were not defined, although Fi- host susceptibility/resistance factors that affected ennes and colleagues implicated a haemolysin in the severity of the pathological processes elicited trypanosome-induced anaemia showing that by AAT, or indeed the mechanisms of that path- blood plasma from infected animals sometimes ology. All of these problems were solved, to vari- lysed red blood cells of healthy animals in vitro at ous degrees, by the work of ILRAD. 37°C. The identity of the putative trypano- Major additions to the field between 1970 some-derived haemolysin was not established, and 1979 included isolation and partial charac- although it was shown to be inactivated by heat- terization of the organelle that defined the order ing at 56°C for 30 min, consistent with involve- Kinetoplastida to which the genera Trypanosoma ment of a heat-labile complement factor. and Leishmania belong, namely the kinetoplast Among the many pathological features of (Fairlamb et al., 1978); localization of blood- AAT, Fiennes drew attention to serum dilution stream-stage T. brucei glycolytic enzymes to a and accompanying hypoproteinaemia as being single microbody-like organelle called the gly- mainly responsible for the progressive decline of cosome (Opperdoes and Borst, 1977); isolation animals during the chronic stages of trypanosome and partial characterization of the variant- specific infections. Again, mechanisms of this path- surface antigen of T. brucei (Cross, 1975; Brid- ology were not identified, although kidney failure gen et al., 1976), which was previously called leading to retention of sodium and water is a exoantigen and is now called the variable surface likely candidate. Fiennes cited the work of glycoprotein (VSG); and cultivation of animal- several investigators implicating kinins, which infective bloodstream-stage T. brucei in vitro are inflammatory polypeptides, but it is un- (Hirumi et al., 1977). This last discovery was likely that inflammation alone would cause made at ILRAD. hypervolaemia. During the 1970s, the pace of discovery in In summary, by the start of the 1970s, it cell and molecular biology and immunology had was clear that AAT is a disease caused by three greatly accelerated through advances in the species of trypanosomes that undergo cyclic de- manipulation of DNA and RNA, nucleotide and velopment in biting flies of the genus Glossina protein sequencing, antigen epitope targeting (tsetse) and are transmitted to mammalian hosts with monoclonal antibodies (mAbs), cell popu- primarily in the saliva of these flies. Trypano- lation analysis using fluorescence-activated flow some isolates had been cryopreserved, shown to cytometry and cell sorting, and data processing retain infectivity for mammals and shown in ex- using personal computing, to name a few. Appli- perimental systems to induce host responses, cation of these technologies to the field of tryp- both pathological and protective, to the para- anosomiasis research at ILRAD/ILRI had major sites. These early studies showed that the chron- impacts on understanding the molecular biol- icity of infection was linked to possibly unlimited ogy of trypanosome antigenic variation and variation of trypanosome surface antigens/ the host immune response to the parasites, as exoantigens and escape from immune elimin- discussed below. ation. It had also been shown that exoantigens of bloodstream-stage trypanosomes elicited pro- tective antibody responses, but that protection was ILRAD’s Initial AAT Mandate restricted to homologous parasites. The diversity of bloodstream-stage trypanosome exoantigens The first researchers at ILRAD were challenged by suggested that they could not be combined as a the institute’s mandate to ‘conquer, as quickly as composite vaccine; however, there was some evi- possible, major animal diseases [that] seriously dence, albeit weak, that infective trypanosomes limit livestock industries in Africa’ and to make present in tsetse saliva and the first populations fundamental discoveries concerning the cell Control of Pathogenesis in Animal African Trypanosomiasis 111 and molecular biology of T. brucei, T. congolense, research focus groups with defined research T. vivax and Theileria parva, and their interactions goals. In addition, collaborations had been estab- with mammal hosts. lished among members of different disciplinary The acquisition of knowledge of host and and focus groups, thus promoting interdisciplin- pathogen biology has potential application to the ary research. control of disease, but should it take precedence In addition to assembling research teams over more applied approaches? This question was at ILRAD, a number of questions had been iden- posed by A.J.S. Davies and G.A.T. Targett in a lec- tified by 1979 as important to the mission to ture on ‘Some perspectives in parasitic disease’ develop AAT immunization and control strat- presented at the Inauguration Symposium on egies (Table 2.1). Current Trends in Immunology and Genetics and Their Implications for Parasitic Diseases, a meeting held on the ILRAD campus in 1978 Initial experimental questions on AAT (Davies and Targett, 1978: p. 69). They wrote, ‘Is vaccine development and control there any way that ILRAD can create an environ- ment in which serendipity can have its full force? By 1979, scientists had begun to answer the What shall be the balance between fundamental questions in Table 2.1. and field-level investigations? In any instance, do we know enough about disease in general, • T. brucei undergoes antigenic variation in and trypanosomiasis and East Coast fever in vitro in the absence of VSG-specific anti- particular, to adopt any specific approaches?’ bodies or other components of trypano- The ILRAD scientific community took a some-induced immune responses (Hirumi mixed view. It committed to fulfilling the man- et al., 1977; Doyle et al., 1980). Thus, im- date of the institute through the application of mune pressure was not required for anti- evidence-based research, while investigating the genic variation. basic biology, biochemistry and molecular biol- • Purified messenger RNA (mRNA) encoding ogy of African trypanosomes and their interaction a T. brucei VSG induced synthesis of VSG in with their hosts, with a view towards identifying an in vitro translation system (Williams vaccine and diagnostic antigens and increasing et al., 1978) and was an early step towards host resistance to AAT. characterizing the genetic basis of anti- genic variation. Work at ILRAD later showed that expression of a single VSG ILRAD, 1975–1979 gene could occur with or without duplica- tive transposition (Young et al., 1983a), a Under the leadership of Jim Henson, who process whereby a copy of a VSG gene was served as ILRAD Director General from 1974 inserted into an active VSG gene expression to 1978, ILRAD built a modern tsetse labora- site through recombination with repetitive tory and suites of modern open-plan laboratories, sequence in the barren region on one side research support and administrative buildings, of the VSG. at Kabete in Kenya. Over the next few years, • mAbs could distinguish between trypano- the institute was equipped with switchable some VSGs (Pearson et al., 1980). This tech- power, backup generators, facilities for safely nology opened the way to dissect and handling radioactive materials and storing compare VSGs and other antigens of blood- radioactive waste, animal management facil- stream-stage and metacyclic trypanosomes, ities, and a flow cytometry core including of different trypanosome species and highly trained staff who ran and maintained a isolates. Becton and Dickinson FACS II cell sorter. In • T. brucei and T. congolense VSGs have a com- short, in over 5 years, ILRI established stand- mon cross-reactive determinant (Barbet ards of operation equivalent to those in reput- and McGuire, 1978). However, immuno- able institutes worldwide. fluorescent staining of trypanosomes using By 1979, ILRAD had assembled an almost sera with activity against the cross-reactive full complement of scientific support and adminis- determinant showed that this determinant trative staff and, importantly, had established was not accessible on intact parasites. 112 S.J. Black Table 2.1. Experimental questions on AAT vaccine development and control. Is the genetic information for all VSGs present in each trypanosome? Do trypanosomes vary VSGs in response to immune pressure? Do VSGs of bloodstream-stage trypanosomes have a common determinant or determinants that can induce a protective immune response? Do mammal-infective, tsetse-derived (metacyclic) trypanosomes express a common VSG, and do the first bloodstream-stage trypanosomes derived from them express a limited set of predominant VSGs? Are there conserved trypanosome antigens that can induce protective immunity against AAT or that can be used in the development of diagnostic reagents? Do trypanotolerant animals mount protective immune responses against conserved components of VSGs and/or common trypanosome antigens? Is trypanotolerance a genetically acquired trait, and if so, how many genes are responsible? • Soluble T. brucei and T. congolense VSG activates (iv) the genetic basis of trypanotolerance. The complement in vitro (Musoke and Barbet, chapter is not a comprehensive review of all work 1977), which raised the possibility that ac- relevant to AAT carried out at ILRAD, ILCA and tivation of complement by VSG released by ILRI. For example, development of expertise in trypanosomes might be the cause of in- bovine immunology and of tools to dissect bovine flammation in AAT. immune responses, which were critical to AAT • Two trypanosomiasis resistance models were research, is discussed elsewhere (see Chapter 4, established: (i) inbred strains of mice differ in this volume). Important work at ILRAD on the their survival time after infection with T. con- physical nature of T. congolense (Rovis et al., 1978) golense, which was independent of the haplo- and T. vivax (Gardiner et al., 1987) VSGs, immuno- type of mouse major histocompatibility com- capture of mRNAs encoding VSGs (Shapiro and plex (MHC) expressed and which may be Young, 1981) and resolution of the puzzling dis- associated with the capacity to control para- parity between VSG expression with or without an sitaemia (Morrison et al., 1978); and (ii) expression-linked copy (Majiwa et al., 1982; Young N’Dama and Zebu cattle in The Gambia dif- et al., 1983a,b) are not discussed here. Similarly, fered in their capacity to control experimen- research at ILRAD and ILRI on trypanosome bio- tal infections with T. brucei and T. congolense chemistry, metabolism and death pathways (Lons- (Murray et al., 1977b,c). These models were dale-Eccles and Grab, 1987; Aboagye-Kwarteng used, and are still being used, to identify the et al., 1991; Bienen et al., 1991; Murphy and immunological, physiological and genetic Welburn, 1997; Welburn and Murphy, 1998; basis for resistance to trypanosomiasis. Welburn et al., 1999) and in vitro (Borowy et al., • T. congolense caused immune dysregulation 1985a, 1985b, 1985c; Borowy et al., 1988; Dube in mice by inducing cells, called suppressor et al., 1983) and in vivo (Peregrine et al., 1987, cells, that inhibit the responses of T lympho- 1991; Sutherland et al., 1991; Silayo et al., 1992; cytes in vitro (Pearson et al., 1979), introdu- Mamman et al., 1993; Burudi et al., 1994; Mam- cing a possible mechanism of immunosup- man and Peregrine, 1994; Mamman et al., 1994; pression in trypanosomiasis. Peregrine and Mamman, 1994; Waitumbi et al., 1994) assays of drug resistance in trypanosomes, AAT Research at ILRAD, which also fall outside the scope of the chapter, are not discussed. ILCA and ILRI, 1979–2017 This section traces the development of the research The search for AAT vaccine antigens themes outlined in Table 2.1. It encompasses work on AAT at ILRAD, ILCA and ILRI up to 2015 Metacyclic and first-generation and, where appropriate, references relevant work bloodstream-stage trypomastigote VSGs carried out at other institutions. This section ad- dresses: (i) AAT vaccine antigens; (ii) AAT diagnos- T. brucei, T. congolense and T. vivax infect and tics; (iii) mechanisms of AAT pathogenesis; and undergo specific developmental programmes in Control of Pathogenesis in Animal African Trypanosomiasis 113 tsetse, ultimately maturing to mammal-infective subset of specific VSGs (fewer than 28, which is stages at sites in the tsetse that allow deposition 1–2% of the total VSG repertoire) are expressed into mammal hosts in saliva. These mammal- by the T. brucei metacyclic population (Turner infective trypanosomes are called metacyclic try- and Barry, 1989). Similar data are unavailable pomastigotes and are diploid and non-dividing, for the metacyclic VSG repertoires of T. congolense which was shown for T. brucei by cell-cycle ana- and T. vivax, but recent genome analysis indicates lysis in the first publication from ILRAD that used VSG gene repertoire diversity in these species the flow cytometry core (Shapiro et al., 1984), (Jackson et al., 2012), and there is no reason to and was later shown by ILRAD scientists, in col- think that diversity of metacyclic VSGs expressed laboration with others, for T. congolense and in a tsetse infected with a clone of either parasite T. vivax metacyclics using culture-derived para- will be more limited than that of tsetse similarly sites and nuclear DNA microfluorimetry (Kooy infected with T. brucei, particularly in light of the et al., 1989). Upon deposition into the mammal in vivo infection-and-treatment studies reported host, the metacyclic trypomastigotes differenti- below. Although the repertoire of T. brucei VSGs ate to replicative bloodstream-stage trypomastig- expressed by metacyclic trypanosomes is smaller otes, thus establishing infection. Metacyclic and than that expressed on the bloodstream-stage bloodstream-stage trypomastigotes have been parasites, it is still substantial, is unstable over selected through evolution to express a more-or- time (Barry et al., 1983) and, as discussed below, less contiguous layer of VSG on the outer leaflet differs among different serodemes of T. brucei, of their plasma membrane. The VSG coat protects T. congolense and T. vivax, thus showing that the parasites from antibody-independent lysis by trypanosomes from different serodemes have dif- plasma complement factors (Devine et al., 1986), ferent metacyclic VSG repertoires. which are innate immune effector molecules that assemble to a lytic complex on some pathogens, induction of immunity by the infection-and- including uncoated trypanosomes, but not on treatment method The second line of investi- VSG-coated trypanosomes. As discussed earlier, gation involved induction of immunity by the bloodstream-stage trypanosomes generate diverse infection-and-treatment method. Cattle, goats VSGs by antigenic variation. However, investiga- and mice were subjected to cyclic infection using tions prior to the establishment of ILRAD (reviewed Glossina morsitans submorsitans infected with by Gray, 1970) had suggested that metacyclic T. brucei, T. congolense or T. vivax, cured by treat- and the first bloodstream-stage parasites might ment with diminazene aceturate (Berenil; Hoechst express only a few common and predominant AG, Frankfurt, Germany), and subsequently VSGs, which might therefore serve as vaccine exposed to tsetse infected with homologous or antigens. heterologous trypanosomes (Nantulya et al., 1980; Akol and Murray, 1983; Nantulya et al., 1984; direct analysis of vsgs expressed on metacyclic Akol and Murray, 1985; Nantulya et al., 1986; and bloodstream-stage trypanosomes Two Vos et al., 1988; Taiwo et al., 1990). These studies lines of investigation were initially pursued at showed that trypanocidal treatment of an estab- ILRAD to seek possible vaccine antigens. The lished infection resulted in protective immunity first was direct analysis of the VSGs expressed on against homologous but not heterologous cyclic metacyclic and bloodstream-stage trypanosomes. infection. Protection was associated with para- Analyses of expressed VSGs were performed us- site control at the level of the tsetse bite (i.e. in ing immune sera and mAbs, and were continued the skin), because a chancre that results from an with a variety of molecular genetic approaches immune response induced by parasites at the by colleagues in and outside of ILRAD. The stud- bite site did not develop in the immune animals. ies conducted at ILRAD clearly showed hetero- Immunity, at least in the case of infections geneity of expressed T. congolense metacyclic with homologous strains of T. brucei and T. con- VSGs (Nantulya et al., 1983) and substantial golense, correlated with the presence of neutral- heterogeneity of VSGs expressed by initial popu- izing serum antibodies specific for the VSGs of lations of bloodstream-stage T. vivax arising their metacyclic trypanosomes, whereas immun- from metacyclic parasites (Gardiner et al., 1986). ity in the case of infections with T. vivax resulted Studies in Scotland showed that only a small from accumulation of antibodies specific for 114 S.J. Black VSGs of homologous bloodstream-stage T. vivax. of AAT, it is possible that induction of type 1 Disappointingly, from the perspective of immu- T-helper cell responses against conserved com- noprotection, attempts to elicit protective im- ponents of these VSGs, such as the C-terminal- munity in individual animals against several domain conserved peptides of T. brucei and against serodemes of trypanosomes by simultaneous, other conserved antigens that are accessible in or sequential, cyclic infection of hosts followed trypanosomes present in the chancre and other by trypanocidal treatment were unsuccessful tissue sites, might expedite the development of (Dwinger et al., 1987). This resulted from failure protective innate immune responses in these to establish mixed infections in the hosts, or from regions. This approach is currently under inves- an inability of the infected and treated hosts to tigation (Black and Mansfield, 2016). develop protective immune responses against the wider range of metacyclic and bloodstream-stage VSGs of the mixed populations. the vsg cross-reactive determinant Rabbit One important observation made in sequen- antisera prepared at ILRAD against VSGs of an- tial cyclic infections with bloodstream-stage tigenically distinct clones of T. brucei and T. con- trypanosomes was that the superimposed heter- golense contained antibodies specific for epitopes ologous parasites did not establish an infection that were unique to the immunizing VSG as well (Morrison et al., 1982). Similarly, superimposed as antibodies specific for a cross-reacting deter- infections with metacyclic parasites did not elicit minant (CRD) common to all VSGs (Barbet and a chancre (Dwinger et al., 1986), indicating that McGuire, 1978). The latter antibodies did not the heterologous parasites were killed in the skin bind to intact trypanosomes. Thus, the CRD tar- or did not establish replicating bloodstream-stage get epitope was masked, or cryptic, on mem- parasites in the skin. This phenomenon was called brane-bound VSGs, suggesting that antibodies ‘interference’. Its maintenance required sustained against this epitope are unlikely to be host pro- infection, and it was not equally effective against tective. Subsequent studies at ILRAD showed all superinfecting trypanosomes (Dwinger et al., that the CRD was ‘located within oligosacchar- 1989), all suggesting against interference as a ides linked to the VSG through N-glycosidic and manageable strategy of AAT control. Interfer- other unidentified types of linkages’ (Rovis and ence was hypothesized to result from elevated Dube, 1981) and that these were added to VSGs microbicidal responses by innate effector cells in in the trans-Golgi region during export to the the skin of infected cattle, but the mechanism surface (Grab et al., 1984) and were present at was not studied at ILRAD. It is now well the C-terminal portion of the molecule and close e stablished that trypanosomes are killed by pro- to the trypanosome plasma membrane when at- inflammatory products of macrophages/mono- tached to the parasites. Subsequent studies at cytes, including reactive oxygen and nitrogen Cambridge University, UK, showed that the CRD species, amphiphilic peptides/cathelicidins, and, was exposed upon release of soluble VSG from in the case of some trypanosomes, the cytokine the trypanosome membrane as a result of cleav- tumour necrosis factor (TNF). It is also estab- age of the dimyristoyl glycosylphosphatidylin- lished that interferon-γ (IFN-γ), which is pro- ositol (GPI) lipid anchor, which attaches the VSG duced by type 1 T-helper cells, greatly increases to the plasma membrane, by an endogenous in trypanosome-infected animals and activates phospholipase C (PLC)-like hydrolase (Cardoso macrophages to produce these microbicidal de Almeida and Turner, 1983). Work at ILRAD products. showed that little or no soluble VSG is released by healthy bloodstream-stage trypanosomes a new vaccine approach It has recently been (Black et al., 1982), indicating that access of the proposed that activation of the innate defence PLC to the membrane-form VSG is tightly regu- system of the skin should be a strategy for AAT lated. The possibility of disturbing this regula- vaccination (Tabel et al., 2013). Thus, while im- tion to cause spontaneous release of VSG and munization and infection-and-t reatment regi- exposure of naked trypanosomes to destructive mens that induce immune responses against highly complement components in host plasma spurred variable components of metacyclic VSGs are con- further research on the location and regulation sidered unlikely to have an impact on the control of the VSG GPI-PLC. Control of Pathogenesis in Animal African Trypanosomiasis 115 Later work by Grab et al. (1987) at ILRAD (Knowles et al., 1987; Lonsdale-Eccles and Grab, showed that the VSG GPI-PLC was associated 1987; Authie et al., 1993a, 2001). with flagellar membrane fractions of disrupted trypanosomes, and more recently it has been conserved plasma membrane proteins The first shown at Trinity College Dublin and Cambridge immunization studies at ILRAD involved plasma University to be present as a linear (patchy) array membranes purified from bloodstream-stage on the face of the flagellar membrane, between T. brucei and also an 83-kDa protein that was the paraflagellar rod and the cell body and close present in lysates of T. brucei, T. congolense to the flagellar attachment zone (Hanrahan et al., and T. vivax. Immunization of rabbits and goats 2009; Sunter et al., 2013). Despite exposure on with these materials elicited high-titre antibody the outer leaflet of the flagellar membrane (Sunter responses but did not alter the course of the dis- et al., 2013), the GPI-PLC is not accessible on in- ease in the immunized animals following subse- tact trypanosomes to specific antibodies and is quent infection (Rovis et al., 1984), suggesting unlikely to serve as a vaccine antigen. Neverthe- that none of the target antigens was exposed at a less, VSG GPI-PLC may be the key to understand- high enough concentration on the surface of ing the pathogenesis of AAT. It was shown that trypanosomes to support antibody-mediated deletion of the gene encoding PLC from the tryp- killing, clearance or growth inhibition, and thus anosome genome did not prevent the capacity were not candidate vaccine antigens. of the parasite to infect and grow in tsetse or mammals but did substantially decrease tryp- anosome virulence (Webb et al., 1994, 1997), macromolecular nutrient receptors Groups implicating the PLC, or VSG GPI cleavage prod- at ILRAD and elsewhere turned their attention ucts, in dysregulation of immune control of the to receptor-mediated endocytosis in trypano- parasites. Thus, what began as a search for a somes, which might yield immunoprophylactic conserved, possible vaccine, component of VSG targets. Using axenic culture systems developed at ILRAD was part of the path to the discovery of at ILRAD, Black and colleagues obtained defini- the only known trypanosome virulence factor, tive proof that the parasites required LDL, IDL or the VSG GPI-PLC. HDL to progress through their cell division cycle, and transferrin for sustained replication (Black Candidate vaccine antigens and Vandeweerd, 1989; Morgan et al., 1993, other than VSGs 1996). Studies by Coppens and colleagues in Belgium implicated an 86-kDa T. brucei protein In addition to characterizing VSGs of the AAT in uptake of LDL, and suggested that uptake of parasites (Rovis et al., 1978; Gardiner et al., 1987), lipoproteins and parasite growth could be inhibited scientists at ILRAD began the isolation and char- by antibodies to this molecule (Coppens et al., acterization of four antigen systems that might 1988). However, collaborative studies (unpub- yield vaccine antigens: (i) conserved plasma lished data) between Coppens and Black carried membrane proteins (Rovis et al., 1984); (ii) recep- out at ILRAD showed that the antibodies did not tors for macromolecular nutrients (transferrin affect growth of culture-adapted bloodstream- and serum low-density (LDL), intermediate- stage trypanosomes, and further work on the density (IDL) and high-density (HDL) lipopro- putative LDL receptor did not yield a candidate teins) required by the parasites to grow (Black vaccine. The trypanosome receptors for bovine and Vandeweerd, 1989; Vandeweerd and Black, serum LDL, IDL and HDL have not yet been 1989; Grab et al., 1993); (iii) trypanosome endo- identified. somal compartments (Webster, 1989; Webster Studies at ILRAD to identify the T. brucei and Fish, 1989; Grab et al., 1992), including components that bind transferrin revealed a clathrin-coated endocytic vesicles (Shapiro and 90-kDa holotransferrin-binding T. brucei protein Webster, 1989; Shapiro, 1994), which would be (Grab et al., 1993). This protein was isolated expected to contain molecules involved in receptor- from parasites that had been grown in rats and mediated endocytosis; and (iv) peptidases that when injected into rats induced production of might be released from living or dying trypano- specific antibody that inhibited growth of blood- somes and hence might contribute to pathogenesis stream-stage T. brucei in axenic cultures; further 116 S.J. Black vaccine tests were not carried out in vivo. Isola- their growth in vitro or improve control of infec- tion of the T. brucei transferrin receptor (TbTfR) tion in the immunized rabbits (Shapiro, 1994). was subsequently achieved by Steverding et al. Similar results were obtained with endosomal (1995) in Germany, who showed that it was a proteins, purified by tomato lectin affinity chro- heterodimer of proteins encoded by VSG expression matography, reported below. site-associated gene 6 (ESAG6) and ESAG7 whose products had molecular masses of 50–60  kDa t. congolense cysteine peptidase (congopain) and 42 kDa, respectively, and thus were distinct A negative correlation was found at ILRAD be- from the material isolated at ILRAD. Interest- tween the titre of IgG1 antibodies specific for a ingly, uptake of transferrin by T. brucei was poor- T. congolense cysteine protease (CP), congopain, in ly inhibited by TbTfR-specific IgG but strongly post-infection bovine serum and the severity of inhibited by fragments of these antibodies com- AAT (Authie et al., 1993a). To determine whether prising their antigen-binding sites, which may these antibodies have a protective role in AAT, cat- have easier access than intact antibodies to the tle were induced to generate antibodies against the transferrin receptor that is embedded in VSGs of catalytic domains of two families of related T. the trypanosome flagellar pocket. The investigators congolense CPs, called CP1 and CP2, by immuniza- stated that attempts to protect mice from T. brucei tion with recombinant truncated proteins that had AAT by immunization with the receptor were been expressed in a baculovirus system and that unsuccessful. Subsequent studies showed that: lacked the trypanosome-specific C-terminal exten- (i) the different VSG expression sites in T. brucei sion (Authie et al., 2001). The immunized cattle contain different copies of ESAG6 and ESAG7; were subsequently infected with T. congolense by (ii) their products differ in binding affinity for dif- the bites of eight infected tsetse flies. Cattle sub- ferent mammal transferrins; and (iii) culture of jected to the same immunization regime with ov- bloodstream-stage trypanosomes in medium albumin served as controls. Immunization with containing transferrin to which their TbTfR had truncated CP1 or CP2 did not affect either the time low affinity resulted in selection of parasites that to patency or the levels of parasitaemia through- had switched to a VSG gene expression site with out infection, or the rapid decline in blood PCV ESAG6 and ESAG7 encoding a higher-affinity during the first 40 days after infection, but did ac- transferrin receptor for that transferrin (Gerrits celerate weight gain and recovery of blood PCV et al., 2002). Based on studies on the TbTfR to date, and of residual leukocyte counts following the ini- it seems unlikely that it will serve as a vaccine tial rapid decline in these parameters. Further- antigen. T. congolense also expresses ESAG6 and more, cattle immunized with CP2 mounted rapid ESAG7, but T. vivax lacks these genes (Jackson et al., and higher-titre IgG antibody responses against a 2013). In addition, work at ILRAD showed that the VSG (IL-C49) unrelated to that of the infecting T. vivax VSG is smaller than that of T. brucei and parasites. This last result is difficult to interpret be- the surface coat is more diffuse (Gardiner, 1989); cause the IL-C49 VSG-specific antibodies were not consequently, it would be of interest to determine assayed for cross-reactivity with the VSG, or any how T. vivax acquires iron and whether this mech- other antigens, of the infecting parasites. Despite anism could be blocked by a specific antibody to some difficulty in interpreting this study, it is quite the detriment of the parasite. clear that immunization with truncated CP2 alle- viated some aspects of AAT pathology. This work extracts of trypanosome endocytic vesicles was not continued at ILRAD, and there have been An attempt was made at ILRAD to induce anti- no reports (to mid-2018) on how the truncated bodies that interfere with endocytosis of essen- CP2 vaccine accelerates recovery from AAT. tial molecules by trypanosomes. Rabbits were immunized with proteins isolated from purified T. brucei clathrin-coated vesicles hypothesized to Other vaccine studies be ‘putative parasite receptors for adsorptive endocytosis’. The resulting antibodies recognized The search for AAT vaccine antigens was not re- many parasite proteins, including epitopes on stricted to ILRAD. Other investigators reported the parasite’s endocytic surface, but did not inducing partial protection (reviewed by La Greca stimulate in vitro lysis of the parasites, inhibit and Magez, 2011), achieved by immunizing with: Control of Pathogenesis in Animal African Trypanosomiasis 117 (i) trypanosome flagellar pocket fractions; (ii) DNA in the peritoneal cavity induced by the immun- encoding an invariant surface glycoprotein; ization regime. (iii) trypanosome cytoskeletal proteins; (iv) Trypa- nosoma evansi actin or tubulin; or (v) plasmid Alternatives to an AAT vaccine containing the catalytic and N-terminal domain of trans-sialidase. The primed animals were typ- As research at ILRAD, and subsequently ILRI, ically boosted with the priming antigen and in- on AAT vaccines waned, during the 1980s three fected shortly thereafter with a low number of AAT research themes emerged and were con- trypanosomes (500–1000). This regime resulted tinued at ILRI. These themes are summarized in in 40–60% of recipients showing sterile immun- Table 2.3 together with pros and cons (with ity (i.e. they did not become infected in contrast respect to ILRAD/ILCA/ILRI mandates) that to the uniform infection achieved in control un- were considered at the time. Perhaps it would have immunized animals). The opinion of the review also been reasonable to take a drug-discovery authors was that these immunization regimes approach (i.e. to interrogate axenic cultures of elicited short-lived innate immune responses pathogenic trypanosomes developed at ILRAD that killed infecting organisms. This interpretation with every compound made or extracted by hu- is consistent with data obtained in the author’s mankind in search of new effective trypanocidal laboratory in collaboration with Noel Murphy of compounds). While this could have been done ILRI and Derek Nolan of University College, in collaboration with Big Pharma, which was Dublin, on the vaccine potential of T. brucei tomato equipped with robotic testing centres and vast li- lectin binding (TL) antigens (Nolan et al., 1999), braries of testable compounds, ILRAD and ILRI which encompass most if not all trypanosome did not have the resources to discover new drugs endosomal proteins and macromolecular by brute force and that route was not followed. growth-factor receptors. Mice were immunized High-throughput in vitro drug-screening ap- via the peritoneal cavity with TL antigens in the proaches for African trypanosomiasis have been presence or absence of the antimitotic drug taken up by other researchers in recent years re- cyclophosphamide, which inhibits antibody pro- sulting in more than 20 papers cited in PubMed duction (Table 2.2). Irrespective of the gener- on this topic during the decade 2008–2018. ation of TL antigen-specific antibodies, 50–66% of the immunized mice resisted subsequent in- Improved AAT Management through fection by the intraperitoneal route. Subsequent Use of Trypanotolerant Stock studies showed that the protective response was lost 3 weeks after the last immunization and ILRAD’s interest in AAT diagnosis, treatment could not be boosted using soluble TL antigens. and management was linked to the decision to It was concluded that infecting parasites were develop a field/epidemiology programme that killed by a microbicidal innate immune response would strengthen the institute within Africa and Table 2.2. T. brucei TL antigens induce a protective innate immune response. (unpublished data, ILRAD). Cyclophosphamide TL antigen Ovalbumin (200 mg/kg body weight) Challenge Animals protected (%) + – – + 66 + – + + 50 – + – + 0 – + + + 0 Groups of mice (n = 6 per group) were primed by intraperitoneal injection of 20 μg of TL antigen (T. brucei clone ILtat 1.4; Nolan et al., 1999) or with 20 μg of ovalbumin, emulsified in Freund’s complete adjuvant, boosted after 1 month with the same amount of antigen emulsified in Freund’s incomplete adjuvant, and challenged with exponentially growing trypanosomes (500 T. brucei strain GUTat 3.1) 8 days later. Tail-blood parasitaemia was assayed at 4, 8 and 12 days post-challenge, and mice in which parasites were not seen on any occasion were designated as protected. Antibodies specific for TL antigens reached a titre of 1:1600 in a TL-antigen enzyme-linked immunosorbent assay, but were not detected in mice that were given TL antigens plus cyclophosphamide, which kills dividing B-cells, and were not detected in the control ovalbumin-immunized mice. TL, tomato lectin. 118 S.J. Black Table 2.3. AAT research at ILRAD/ILRI in the post-trypanosomiasis vaccine era. (Constructed by author.) Theme Pros Cons Improve productivity of N’Dama Helps producers Involves ILRAD in large- cattle in large herds under Involves ILRAD with ILCA in workforce scale epidemiology natural tsetse/trypanosome training and management of AAT in studies that address the challenge by diagnosis and N’Dama phenotype of chemotherapy Develops diagnostic tools to monitor AAT trypanotolerance, not the Identifies indicators of trypanotolerance mechanisms of expressed in a managed herd and trypanotolerance generates data on the heritability of some of these indicators Identify the cell and molecular Adds to our understanding of the bovine AAT pathogenesis might be basis of AAT pathogenesis immune system under AAT stress multifactorial and of by comparative analyses of and adds reagents to the bovine unfathomable biological innate and acquired immune immunology tool chest complexity responses of trypanotolerant May identify molecular triggers of and susceptible hosts pathology in AAT and thus contribute to the elucidation of the genetic basis of trypanotolerance and its exploitation Identify genes linked to Involves ILRAD in the global effort to Trypanotolerance may be a trypanotolerance through map the bovine genome polygenic trait with small comparative analysis of post- May identify host genes that control contributions from each of infection phenotypes and anaemia, parasitaemia and many unlinked genes, genome linkage maps of F2 mortality in AAT and expedite precluding mapping progeny of N’Dama × Boran marker-assisted selection to enrich crosses trypanotolerance that would address mechanisms of disease resist- (ILCA) and Max Murray (ILRAD) to identify indi- ance in the taurine breeds of cattle in West and cators of trypanotolerance in N’Dama cattle under Central Africa, namely the N’Dama and West natural tsetse/trypanosome challenge and to im- African Shorthorn. These taurine cattle, particu- prove management of N’Dama cattle under such larly N’Dama, were reputed to be trypanotoler- challenge. ATLN was an unusually collabora- ant, i.e. to have the capacity to survive and be tive effort involving the United Nations Environ- productive in tsetse-infested areas without ment Programme (UNEP), FAO, the ITC and the treatment (Camille-Isidore, 1906; Murray et al., Centre International de Recherche-Développement 1982). Trypanotolerance is thought to have sur l’Elevage en zone Subhumide (CIRDES), as arisen within the taurine breeds of West Africa well as national entities of Côte d’Ivoire, Ethiopia, during several thousand years of selection in Gabon, Kenya, Nigeria, The Gambia, Togo and tsetse- infested areas. The degree of trypanotol- what was then Zaire (now the Democratic Republic erance varies within the breed and may therefore of the Congo). be open to improvement through selective breed- Importation of N’Damas to ILRAD for on- ing. It may also be possible to introduce the trait site study, which had been (reasonably) discour- into other cattle breeds by marker-assisted selec- aged by the then-Director General of ILRI, Jim tion, creating a domestic bovid ideally suited to Henson, and by the Kenya Veterinary Service for African agriculture. fear of accidental disease spread, was eventually realized in 1984 when frozen N’Dama embryos were imported and implanted in surrogate Bo- African Trypanotolerant Livestock Network ran mothers, through the expertise of Ivan Mor- rison, Geoff Mahan and Torbin Jordt (Jordt et al., The African Trypanotolerant Livestock Network 1986a,b). In 1984, 10 N’Dama calves were (ATLN) was established in 1977 by John Trail born at ILRAD. These animals and their progeny Control of Pathogenesis in Animal African Trypanosomiasis 119 were used extensively in studies of mechanisms t esting, could be used in a proximal field labora- of resistance to AAT and in linkage genome- tory, provided an estimate of parasitaemia mapping studies reported later in the text. (number of trypanosomes/ml of blood) and al- ATLN had aimed to ‘improve livestock lowed identification of trypanosome species de- production in tsetse-infested areas of Africa termined by morphology. The test also provided by achieving a better understanding of genetic matched data on the blood PCV of the test [-ally acquired] resistance, environmental fac- animal. Briefly, blood was drawn by capillary tors that affect susceptibility and the efficacy of action into a heparinized haematocrit tube, present control measures, and by ensuring bet- centrifuged in a haematocrit centrifuge and ter application of existing knowledge and recent the blood PCV recorded as a percentage of total research findings’ (ILRAD, 1985). In 1985, the blood volume, as shown in Fig. 2.1. The tube was network was coordinating investigations at sites nicked with a diamond pen 1  mm beneath the in nine countries of West and Central Africa. buffy coat layer, which is composed of white Five of these study sites were well established, blood cells and trypanosomes and subsequently (Gabon, Ivory Coast, N igeria, Togo, Zaire), two snapped to remove the red cell layer. It is then were under development (Senegal, The Gambia) nicked 1 cm above the buffy coat (i.e. in the blood and two were under consideration (Benin, plasma layer) and snapped and the white blood Congo). cell/trypanosome plug is then expelled on to a ILRAD was to provide supervision of animal glass slide, covered with a 22 × 22 mm coverslip health, infection status, and tsetse evaluation, and the preparation scanned by microscopy for while ILCA was responsible for animal produc- the presence and prevalence of trypanosomes, tion, nutrition and data processing. Data collec- which can provide an estimate of parasitaemia. tion was rigorous. ‘Field operations involved the This technique was used by ILRAD to train le- simultaneous collection of data on infection, gions of veterinary technicians throughout Af- health and productivity. Staff at all sites recorded rica. The relationship of parasite prevalence in data on simple pre-printed forms for transmis- the buffy coat and estimated parasitaemia was sion to Nairobi every month. These were checked established by seeding samples of blood with for completeness, verified and entered into a known numbers of trypanosomes. computer file in Nairobi. Major analyses that With respect to animal health data, although were carried out on field data in Addis Ababa in- the AAT-associated decrease in PCV might arise volved the computation of productivity indices as a result of red blood cell lysis or phagocytosis, based on reproductive performance, viability, calf or through haemodilution, as discussed by growth, and cow weight, as well as the assessment Fiennes (1970), the decline in PCV denotes a of possible factors affecting animal performance detrimental change in subject health and there- at different sites (ILRAD, 1985). Data analysis fore informs on disease. With respect to infection programmes used for these studies had been status, one drawback of the DGBC technique is developed and tested using detailed records of that it provides an estimate of trypanosomes cir- matching animal health, animal productivity culating in the peripheral blood only, and hence and trypanocidal drug treatments kept for many may underestimate infections where trypano- years at two large ranches in East Africa, namely somes adhere to blood vessels (T. congolense) or Kilifi Plantations, Kenya, and Mkwaja Ranch, inhabit tissues (T. brucei and to a lesser extent Tanzania, with which ILRAD had a long-term T. vivax). Thus, the DGBC diagnostic technique involvement. was supplemented with antigen-detection tests as discussed below. The dark-ground buffy coat (DGBC) phase-contrast diagnostic technique Antigen enzyme-linked immunosorbent assays, isoenzymes and A dual anaemia/infection diagnostic test that polymerase chain reaction was developed by Murray et al. (1977a) played an important role in ATLN field studies. This test Expertise in mAb production at ILRAD, and was specific for blood-borne trypanosomes, and in molecular genetic techniques, including was sensitive, scalable for high-throughput the polymerase chain reaction (PCR) and 120 S.J. Black (a) (b) Number of trypanosomes Score Estimated parasitaemia tryps/ml Swarming > 100 per field 6+ >5 × 106 Blood > 10 per field 5+ >5 × 105 plasma 1 – 10 per field 4+ 104 – 5 × 104 Buffy coat = Packed 1 per 2 fields–1 per 10 fields 3+ 5 × 103 – 5 × 104 WBCs red and blood trypano- cells 1 – 10 per preparation 2+ 103 – 104 somes 1 per preparation 1+ 102 – 103 Microhaematocrit Microhaematocrit tube with spun tube with spun blood from cattle blood from with AAT healthy cattle Fig. 2.1. Microhaematocrit and estimation of parasites in the buffy coat. (a) The percentage of blood packed cell volume (PCV) is calculated as: PCV = packed red blood cells/(packed red blood cells + plasma) × 100. WBCs, white blood cells. (b) The blood buffy coat layer + 1 cm column of plasma, collected by cutting the microhaematocrit tube shown in (a) 1 mm below and 1 cm above the buffy coat, is expelled on to a glass slide, dispersed beneath a 22 × 22 mm cover slip, observed under dark-ground illumination with a Phaco2 NPL25/050 objective and a 10× eyepiece. Observations as shown in column 1 are scored as shown in column 2, providing an estimate of trypanosomes/ml blood as shown in column 3. The test can be used with all species of pathogenic African trypanosomes. (unpublished data from ILRAD.) orthogonal-field-alternation gel electrophoresis and minichromosomes in preparations from iso- (OFAGE), made it possible to design additional diag- lates and clones of trypanosomes, contributed to nostic tests based, respectively, on trypanosome anti- parasite characterization (Majiwa et al., 1985; gen capture from host blood or lysates of tsetse, and Masake et al., 1988; Kihurani et al., 2000), as did analysis of DNA extracted from the blood buffy coat, isoenzyme analysis, which detects polymorphisms or from tsetse or purified trypanosomes. These tests that result in changes in activity, molecular weight provide an estimate of parasite material present in and isoelectric point of a panel of enzymes detected the sample and thus provide evidence of both an on- by coupled coloured dye precipitation reactions after going infection and the intensity of that infection. gel electrophoresis or isoelectric focusing (Gibson The mAbs used were generated at ILRAD against et al., 1983; Knowles et al., 1988; Fasogbon et al., in vitro-propagated procyclic forms of T. congolense, 1990). These studies, performed at ILRAD and in T. brucei and T. brucei rhodesiense (the cause of East other institutions, contributed to the development African HAT) and T. vivax (Nantulya et al., 1987), of a molecular taxonomy of trypanosomes and and react with antigens in the corresponding blood- a better understanding of patterns of their co- stream-stage trypomastigotes. PCR primers designed evolution (Hamilton et  al., 2007; Adams et al., 2010). at ILRAD to amplify trypanosome species-specific The trypanosome antigen enzyme-linked nucleotide sequences (Kukla et al., 1987; ole-MoiYoi, immunosorbent assay (Ag-ELISA) is based on 1987; Dirie et al., 1993a,b; Masake et al., 1997) were capture of trypanosome antigen from blood or also powerful tools for parasite analysis, allowing in- another material using an antigen-specific mAb vestigators to dissect trypanosome diversity more immobilized on a microtitre plate. Unbound ma- fully than could be achieved by morphological terial is removed by washing, and bound antigen and mAb analysis. Similarly, OFAGE, which sep- is revealed by the addition of a second antibody arates DNA of different lengths and hence re- specific for the captured antigen, tagged with an solves the numbers and lengths of chromosomes enzyme so that it gives a colour reaction. This Control of Pathogenesis in Animal African Trypanosomiasis 121 reaction is proportional in intensity to the test and, when combined with parasitological amount of enzyme-tagged antibody bound and data generated using the latter, allows infection thus to the amount of captured antigen. This basic classification as follows: (i) parasitaemic (DGBC format was modified to facilitate tube capture positive); (ii) antigenaemic (Ag-ELISA positive); and agglutination reactions for field analyses of (iii) parasitaemic but not antigenaemic (DGBC infections. positive but Ag-ELISA negative); or (iv) antigenae- Vinand Nantulya was the driving force for mic but not parasitaemic (Ag-ELISA positive but developing, validating and field testing trypano- DGBC negative). These distinctions were of import- some Ag-ELISA diagnostic tests at ILRAD. These ance to field studies. For example, using a com- included tests for HAT caused by T. b. rhodesiense bination of the DGBC and Ag-ELISA tests on (Nantulya, 1988, 1997; Komba et al., 1992) samples from N’Dama, Zebu and N’Dama × and T. b. gambiense (Nantulya et al., 1992a; Nan- Zebu crosses subjected to sequential experimen- tulya, 1997); for AAT caused by T. brucei, T. con- tal T. congolense infections, colleagues at the ITC, golense or T. vivax in cattle, goats and horses in Banjul, The Gambia, found that during the (Nantulya et al., 1992b; Nantulya and Lind- course of the first parasitaemia wave, the overall qvist, 1989; Trail et al., 1991b; Kihurani et al., percentage of positive cases detected by DGBC in 1994; Masake et al., 1995); and for surra caused blood was higher (p<0.0001) than that obtained by T. evansi in camels (Nantulya et al., 1989; by Ag-ELISA in tested samples of the three cattle Waithanji et al., 1993; Nantulya, 1994). Many populations. However, although the Ag-ELISA of these tests were carried out in collaboration was less than 50% sensitive in detecting circulat- with scientists from institutes throughout Africa, ing antigens during the first 2 months of the pri- including the University of Nairobi, Kenya; the mary infection, during the second infection, the Kenya Trypanosomiasis Research Institute, overall number of infections detected by DGBC Kikuyu; the National Institute for Medical was lower in N’Dama (p<0.005) and the F 1 Research, Tabora, Tanzania; and the Noguchi population (p<0.001) than that obtained using Memorial Institute for Medical Research, Accra, the Ag-ELISA (Mattioli and Faye, 1996). Simi- Ghana; and with scientists from ATLN, dis- larly, Trail et al. (1991b), in ATLN, while study- cussed below. ing infection of N’Dama cattle in Gabon, Central As with the DGBC test, there are some oper- Africa, observed that of the animals detected ational limitations of the trypanosome Ag-ELISA. as parasitaemic, 90% were also positive by The trypanosome species-specific tests used the Ag-ELISA; however, 40% of the animals with same capture and detecting antibody; hence, their negative parasitological findings were found to sensitivity might be diminished by competition for be antigenaemic. Therefore, the tests should be the same antigenic epitopes. For these reasons, used in combination wherever possible. next-generation tests focused on generating cap- The trypanosome antigen-specific mAbs ture and detection antibodies that recognize dis- generated at ILRAD were also used to develop tinct epitopes on the same antigen. In addition, dot ELISAs for detecting trypanosome antigens the capture antibodies are competing with en- in extracts of tsetse, i.e. for diagnosing infections dogenous host antibodies for the same antigens of tsetse (Bosompem et al., 1995a,b, 1996). The and quite possibly the same antigenic epitopes (the assay was performed by lysing tsetse or tsetse tis- part of an antigen bound by an antibody), which sues in detergent (Triton X-100), adhering extract will also diminish test sensitivity. Next-generation to a localized spot on a nitrocellulose membrane tests are being developed with nanobodies – and treating with hydrogen peroxide to colour recombinant antigen-combining sites of single bleach the tsetse material, followed by membrane heavy-chain camelid antibodies (Muyldermans blocking with irrelevant protein to prevent et  al., 2009) – which, because of their antigen- non-specific sticking of the detection antibody to combining site and size, bind antigenic epitopes the membrane. An enzyme-tagged detection distinct from those recognized by conventional antibody was then added and an enzyme-linked antibodies and therefore do not compete with en- colour reaction was carried out to detect the dogenous antibody for antigen capture. bound antibody. Thus, over time, ILRAD/ILCA/ Despite these concerns, the Ag-ELISA provides ILRI developed tools to map trypanosome infec- additional information to the DGBC diagnostic tions of cattle and tsetse throughout Africa. 122 S.J. Black Combined with measures of productivity during for trypanosomes present in East and West Africa infection, such mapping tools could be used to (Clausen et al., 1992; d’Ieteren et al., 1997). evaluate trypanotolerance and hence the possi- bility that N’Dama cattle were an African solu- Parameters of trypanosome infection tion to an African problem. and animal health under natural tsetse/ trypanosome challenge in N’Dama cattle Ranching with chemoprophylaxis and chemotherapy: the N’Dama This section summarizes studies performed by advantage the ILCA/ILRAD ATLN that established param- eters of trypanotolerance under natural tsetse/ In the absence of chemoprophylaxis and chemo- trypanosome challenge, thus providing quality therapy, the impact of AAT is proportional to control for analysis of trypanotolerance in ex- tsetse challenge, with the killing of all trypano- perimental infections in the ILRAD laboratory somiasis-susceptible Boran cattle in areas of setting. In separate tests carried out in Gabon, high challenge (Blaser et al., 1979). In the ab- Central Africa, between 1987 and 1989, a total sence of chemoprophylaxis/chemotherapy, AAT of 436 1-year-old N’Dama cattle were main- also has a substantial impact on trypanotolerant tained for periods of 12, 18 or 24 weeks under a N’Dama cattle under natural tsetse/trypano- medium natural tsetse/trypanosome challenge, some challenge. In this respect, data from ATLN and infection and health parameters measured. showed that productivity of N’Dama herds in Infection prevalences were 25%, 31% and 9%, The Gambia decreases by 11% in areas of low respectively, and anaemia, measured by the aver- tsetse prevalence (determined by tsetse trapping), age packed red cell volume (PCV, measured as a by 42% in areas of moderate tsetse prevalence percentage) over the test period, or by the lowest and by 53% in areas of high tsetse prevalence, PCV reached, was found to be more closely asso- relative to that in areas without tsetse, based on ciated with animal performance than when analyses of the total weight of a 1-year-old calf measured by average PCV when detected as par- and the live weight equivalent of milk produced asitaemic (Trail et al., 1990). The possibility that by 100 kg of cow per year (Murray et al., 1984). maintenance of PCV under tsetse/trypanosome However, the presence of tsetse/trypanosome challenge might serve as a measure of trypa- challenge does not necessarily preclude product- notolerance was supported by studies of N’Da- ive ranching of cattle, even that of trypanoso- ma cows maintained for 3.5 years under a high miasis-susceptible Boran cattle. Thus, ATLN natural tsetse challenge in Zaire. Thus, a simul- analyses of 20,000 calving records for grade taneous evaluation of the effects of three pos- Boran beef cattle collected over 10 years at Mk- sible criteria of trypanotolerance – namely, time waja Ranch, located in a coastal region of Tan- detected as parasitaemic, parasitaemia score zania under high tsetse challenge, showed that and PCV – showed that the latter had the major these animals could be as productive as pure effect on reproductive performance, calf wean- Boran cattle maintained under trypanosomia- ing weight and cow productivity (Trail et al., sis-free conditions in Kenya as long as they were 1991a). Additional studies of N’Dama cattle in supported by a chemoprophylactic regime based Gabon under natural tsetse/trypanosome chal- on the use of isometamidium chloride (Samorin; lenge showed that ‘animals capable of maintaining May and Baker Ltd, London) and occasional use PCV values, even when [trypanosome antige- of the therapeutic drug diminazene aceturate naemic] on a high number of occasions, grew at (Berenil; Hoechst). In fact, the animals required the same rate as uninfected animals’ whereas an average of 4.4 treatments with Samorin and ‘animals that could not maintain PCV values when 0.6 treatments with Berenil each cow-year to infected had poorer growth’ (Trail et al., 1992), sustain their productivity (Trail et al., 1985). In supporting the hypothesis that maintenance of the absence of treatment, the level of tsetse chal- PCV during AAT is an indicator of trypanotoler- lenge was such that cattle could not survive ance and indicating variability in trypanotoler- (Blaser et al., 1979). Trypanosomiasis drug ance within members of an N’Dama herd. resistance did not pose a problem to this strategy Trail et al. (1993) evaluated the impact of at that time, although it has since been reported prophylactic and curative drug treatments on Control of Pathogenesis in Animal African Trypanosomiasis 123 the productivity of N’Dama cattle at the govern- 1985), which is much greater than the treat- ment ranch of the Office of Gabonais d’Amélio- ments required by N’Damas and is consistent ration de Production de Viande, in south-eastern with the lesser ability of the Borans to survive Gabon. In one study, curative drug treatments and be productive under natural tsetse/trypano- were given to 13.7% of cows over the calendar some challenge. year and to 40% of calves from birth to weaning. The above studies support the idea ‘that the The mean number of curative treatments was ability to control anaemia during infection as 0.16 per cow per year and 0.52 per calf from indicated by average PCV might be one reliable birth to weaning (where 1 signifies that curative criterion of trypanotolerance with which to iden- treatments administered are the same as the tify trypanotolerant animals’ (Trail et al., 1993). number of animals in the study, although not However, it is not the sole criterion. Thus, Trail necessarily that all animals receive one curative et al. (1994), working with data from 255 N’Dama treatment). Treatments were on the following under high natural tsetse challenge in Zaire over basis: cattle that were parasitaemic and had PCV a 2-year period from weaning at 10 months of values of less than 25% were routinely given a age, showed that the species of infecting trypano- curative treatment with Berenil, as were animals some, length of time parasitaemic, intensity of with PCV values below 20%, whether they were parasitaemia and anaemic condition had approxi- parasitaemic or not (antigenaemia was not meas- mately equal effects on the final performance ured). The kinetics and levels of recovery of PCV trait of daily live weight gain, and thus informa- were similar in both groups of animals, indicating tion on all of these is essential for assessing an that, even though parasites were not detected in animal’s overall trypanotolerance phenotype. blood, animals with low PCVs in this study were As evident from the number of animals studied nevertheless infected with trypanosomes. and the number of parameters measured, these Both calf and cow average PCVs were linked were demanding and comprehensive studies. to calf weaning weight, ‘there being a 0.91 kg The ATLN findings have been summarized increase for each 1% increase in calf average in a book (ILCA/ILRAD, 1988) and in several PCV, and a 0.95 kg increase in weaning weight review articles (Trail et al., 1989; Murray et al., for each 1% increase in cow average PCV’ (Trail 1990; d’Ieteren et al., 1998). Those findings et al., 1993). Similar studies by Trail et al. jointly validate the trypanotolerant status of (1991a) in Zaire where the prevalence of cow N’Dama and identify indicators of the trait under and calf infection was similar to those in the natural challenge. These studies show that tryp- Gabon study showed that a high average calf anotolerant cattle can be productive with low- PCV increased weaning weight by 9.9 kg (7.7% level chemoprophylactic/therapeutic intervention of total weaning weight) and a high average cow in areas of moderate to high tsetse/trypanosome PCV value increased calf weaning weight by challenge, thus establishing N’Dama as a gene 8.8 kg (6.1%). Hence, maintenance of N’Dama pool for use in tsetse-infested sub-Saharan PCV through curative drug treatment is an im- Africa. In the period of broad and systematic portant production strategy. Half of the cattle in studies of trypanotolerance, done jointly by the Gabon study (Trail et al., 1993) received a I LRAD and ILCA in the 1970s and continued by prophylactic treatment with Samorin at 6 and ILRI after 1994, the three institutions generated 18  months into the study to determine its im- more than 175 papers on various aspects of trypa- pact on prevalence of infection. This treatment notolerance, averaging four papers annually (24 regime decreased the number of curative treat- mean citations per paper) from 1975 to 2018. ments required in cows from an average of 0.25 Notable achievements were Murray et al. (1982) per cow-year in unprotected cows to 0.06 per on host susceptibility, which remains at the 99th cow-year in protected cows. Although there are percentile of citations in its field within Scopus; no data on the relative tsetse/trypanosome chal- the ILCA/ILRAD (1988) book on the achieve- lenge at the Gabon study site versus the Mkwaja ments of the ATLN; Trail et al. (1993, 1994) on Ranch in Tanzania cited earlier, it is noteworthy productivity; Hanotte et al. (2003) on genetic that grade Boran cattle on the Mkwaja Ranch mechanisms of trypanosomiasis; and the later required 4.4 prophylactic treatments and 0.6 r eview article of Naessens (2006) summarizing curative treatments per cow-year (Trail et al., more than 30 years of research on the subject. 124 S.J. Black The cell and molecular biology In addition, all N’Dama cattle suppressed the of trypanotolerance in cattle level of parasitaemia in an incremental manner until it became cryptic (less than 102 trypano- T. congolense cyclic infection somes/ml of blood), whereas those few infected and homologous cyclic reinfection Borans that did not require curative Berenil of N’Damas and Borans treatment were still presenting high levels of parasitaemia until all animals were treated with N’Dama cattle produced by embryo transplants a curative dose of Berenil to terminate infection. that were raised at ILRAD until 13  months of In a follow-up study, cattle that had been se- age were subjected to four sequential cyclic in- quentially exposed to four cyclic heterologous fections (Glossina morsitans centralis) with T. con- infections were cured after the last infection, golense derived from four different serodemes rested for 3  months and subjected to homolo- (Paling et al., 1991a). After each infection, the gous cyclic reinfection with the first infecting cattle were treated with a curative dose of Bere- serodeme, which was approximately 2  years nil and rested for 1 month before the next cyclic after their first exposure to this T. congolense sero- infection. Boran cattle, a trypanosomiasis- deme (Paling et al., 1991b). Although five of the susceptible breed of Bos indicus, were similarly eight Borans and all eight N’Damas had neutral- infected and these animals were treated with a izing anti-metacyclic VSG antibodies present in curative dose of Berenil as required. Irrespective their serum, determined by incubating the para- of the infecting T. congolense serodeme, there was sites in serum and their injection into mice, no significant difference between Boran and these antibodies did not prevent infection of the N’Dama groups in the number of chancres that cattle and presumably were present in serum at developed at tsetse bite sites or in the kinetics or much higher concentrations than in the skin. As magnitude of their development, indicating that with primary infections, neither the kinetics nor there was no pre-programmed (innate) compo- the size of the chancres that developed at tsetse nent in either host breed that prevented infec- bite sites following homologous reinfection dif- tion by killing parasites in the skin. fered significantly between N’Damas and Bo- All cattle became parasitaemic with similar rans. However, infected N’Damas subsequently kinetics irrespective of breed or trypanosome se- developed a single parasitaemic wave, which rodeme used for cyclic infection, indicating that was two orders of magnitude lower than that there was no pre-programmed response in either d eveloped on primary cyclic infection, and rap- breed that prevented migration of the parasites idly suppressed parasitaemia to a cryptic level, from the skin to the bloodstream, which occurs whereas the reinfected Borans developed mul- via the lymph (Akol and Murray, 1986). With tiple waves of parasitaemia, albeit of lesser mag- three of the four test serodemes, levels of parasit- nitude, than those developed on primary cyclic aemia and the kinetics of parasitaemic wave re- infection, and these decreased in amplitude as mission were similar in N’Dama and Boran infection progressed, thus tending towards cryp- cattle for 30 days or so after infection; with the tic infection. It is unclear whether these cattle, other serodeme, they were similar for the first and in particular the N’Damas, had serum anti- 15 days after infection. These data showed that bodies against bloodstream-stage T. congolense host protective immune responses were initially serodeme-1 VSGs that suppressed parasitaemia similar in the infected N’Damas and Borans. on homologous cyclic reinfection or had a recall Thereafter, breed-related differences in param- adaptive response against these VSGs or both. It eters of infection became evident, replicating, in is clear that Borans had a lesser advantage in the case of N’Dama cattle, indicators of trypa- this respect. Importantly, N’Dama cattle that notolerance observed with natural tsetse/tryp- were reinfected by homologous cyclic infection anosome challenge, as discussed above. Thus, did not show signs of anaemia, whereas simi- irrespective of the infecting serodeme, most larly reinfected Borans had an acute drop in PCV, infected Boran cattle, but no N’Dama cattle, rap- although less acute than Borans on the primary idly developed life-threatening anaemia (PCV of cyclic infection, and 50% of these animals re- 15% or less) by around 40  days after infection quired curative Berenil treatment by 50–70 days and were treated with a curative dose of Berenil. post-infection to avoid death. Control of Pathogenesis in Animal African Trypanosomiasis 125 Following curative treatment, the N’Dama 1995, 1996), and presumably in red bone mar- and Boran cattle were again subjected to hom- row elsewhere in the animals, which agrees with ologous cyclic reinfection, this time with tsetse conclusions from other investigators (Dargie transmitting T. congolense serodeme 2 (Williams et al., 1979). et al., 1991). The results were similar to those seen in the first homologous cyclic infection and, AAT-induced anaemia together with those for sequential cyclic infec- tions, are summarized in Table 2.4. Erythrophagocytosis by monocytes, macro- Thus, cyclic T. congolense infections carried phages and neutrophils has been demonstrated out at ILRAD showed that N’Dama cattle have a in cattle infected with African trypanosomes greater capacity to sustain blood PCV levels dur- (reviewed by Murray and Dexter, 1988). While ing infection than similarly infected Borans, in quantitative comparisons of this process were which PCVs dropped from a mean of 33% to not made in infected N’Damas and Borans at 16% over a period of about 40 days after infec- ILRAD, colleagues at the University of Ibadan, tion. Given that the lifespan of a bovine red blood Nigeria, who include ILRAD alumni Victor Taiwo cell is about 160 days, a drop in PCV of 50% over and Victor Anosa, showed, using an in vitro 40  days most likely reflects a substantial de- assay, that erythrocyte phagocytosis and lysis by struction of red blood cells. Thus, if the drop in splenic plastic adherent cells (predominantly PCV resulted solely from a shutdown of erythro- macrophages) from T. congolense-infected Borans poiesis and no other process, PCVs would have was greater than that of splenic macrophages dropped by only 25% during the 40-day period, from similarly infected N’Damas and correlated resulting in a PCV of more than 24% at 40 days dynamically with the degree of anaemia devel- after infection. Furthermore, a shutdown of oped by these animals (Taiwo and Anosa, 2000). erythropoiesis does not occur in T. congolense- In addition, researchers at ILRAD showed that infected Borans; it was shown at ILRAD that acute anaemia in T. vivax-infected cattle correl- T. congolense-infected N’Damas and Borans have ated with the production of TNF (Stijlemans an erythropoietic response characterized by et  al., 2016), a cytokine that increases the peaks above pre-infection levels of early and late phagocytic activity of bovine neutrophils, by erythroid progenitor cells (burst-forming units– blood monocytes (Rainard et al., 2000). In sup- erythroid and colony-forming units–erythroid, port of a role for erythrophagocytosis in AAT- respectively) in the bone marrow of the fourth, induced anaemia, studies using a mouse model fifth and sixth sternebrae (Andrianarivo et al., of T. brucei AAT showed that the acute drop in Table 2.4. Characteristics of experimental cyclic T. congolense infections in N’Dama and Boran cattle. (Data from Paling et al., 1991a,b; Williams et al., 1991.) Infection N’Damas Borans Primary cyclic infection Drop in PCV but not life-threatening Acute drop in PCV: 75–100% of (four serodemes) Parasitaemic waves decrease in animals require curative treatment by amplitude incrementally over time about 40 days post-infection to a cryptic level (<102 T. Repeating waves of parasitaemia do congolense/ml of blood), which is not decrease in amplitude over time reached by 100–130 days Slow recovery of PCV in those animals post-infection that did not need curative treatment Fast recovery of PCV as parasitaemia is suppressed Homologous cyclic No change in PCV Drop in PCV, although less so than in reinfection Rapidly suppress parasitaemia to a primary cyclic infection: 50% of cryptic level with occasional animals require curative treatment low-amplitude spikes of by 50–70 days post-infection parasitaemia Repeating waves of parasitaemia that decrease in amplitude over time in those animals that do not require curative treatment 126 S.J. Black PCV after infection results from phagocytosis of between cells of Bos taurus and B. indicus origin. erythrocytes by activated liver monocytic cells The development and severity of anaemia was and neutrophils and by splenic macrophages significantly less in intact N’Damas than in B orans, (Stijlemans et al., 2015) and is induced by IFN-γ chimeric Borans and chimeric N’Damas, all of produced by natural killer (NK) cells, NK T-cells which developed a similar degree of anaemia. and CD8+ T-cells. Furthermore, the drop in PCV Strikingly, levels of parasitaemia were lower in is prevented by the deletion of these cells and of the N’Damas, whether intact or chimeric, than the gene encoding the receptor for IFN-γ in the Borans, whether intact or chimeric. These (Cnops et al., 2015). results show that control of anaemia, which Dargie et al. (1979) noted a positive correl- in the chimeric N’Damas was similar to that of ation between the severity of anaemia in cattle infected Borans and is thus determined by the infected with T. congolense and the level and dur- genotype of the haematopoietic tissue, is quite ation of parasitaemia. While this may be the separate from control of peripheral blood para- case within a breed, it is not the case between sitaemia, which in the chimeric N’Damas resem- breeds. Thus, work at ILRAD showed that T. con- bled that of Borans and is thus determined by golense-infected Boran cattle developed anaemia non-haematopoietic elements although most more rapidly than similarly infected N’Dama likely acting on immune responses mediated by cattle at ILRAD, despite initially similar levels of progeny of haematopoietic stem cells. parasitaemia (Paling et al., 1991a). It is possible The haematopoietic chimeras were created that the number of T. congolense in the entire by implanting an N’Dama and a Boran embryo vascular system, rather than that of peripheral at the late-morula stage into the same surrogate blood only, will correlate with the level of an- Boran mother, which results in anastomoses of aemia. In this regard, earlier work at ILRAD chorioallantoic vessels in the placenta of the showed that T. congolense were not uniformly dis- twin foetuses in early fetal life and haematopoi- tributed in the host vasculature and that many etic chimerism in which the Boran component more parasites were present in the microcircula- dominates. Haematopoietic stem cells give rise to tion than were free in the cardiac blood, and erythroid, myeloid and lymphoid lineages, and some adhered to vessel walls (Banks, 1978). In hence, while it is clear that susceptibility to addition, Trail et al. (1993) showed in their stud- AAT-induced anaemia is a property of the Boran ies on N’Dama cattle in Gabon that some ani- haematopoietic system and not other aspects mals judged to be aparasitaemic by the DGBC of host physiology, it is unclear which compo- technique developed low PCVs that were re- nent(s) of the haematopoietic system predispose(s) stored by a curative dose of the trypanocidal to this pathology. drug Berenil. Production of erythrocytes from precursor Whether or not differences in parasite load cells in the bone marrow is stimulated by and distribution have an impact on AAT-in- erythropoietin, which is produced by kidney duced anaemia in N’Damas and Borans, it is cer- fibroblasts. Work at ILRAD showed that, while tainly not the only factor to do so; there is a clear the transcript for erythropoietin was similarly impact of the genotype of the haematopoietic increased in the kidneys of T. congolense-infected system. In a remarkable set of studies carried N’Damas and Borans at 35 days after infection, out at ILRAD, Naessens et al. (2003a) analysed there was higher expression of the transcript en- levels of anaemia and parasitaemia in infected coding the erythropoietin receptor (epoR) in the Borans and N’Damas exposed to cyclic infection bone marrow of the infected N’Damas than in with the T. congolense serodeme 1 (‘IL 1180’). the Borans (Suliman et al., 1999). Assuming a The cattle were either intact or modified experi- direct association between expressed gene and mentally to have a chimeric haematopoietic sys- protein, this finding suggests that haematopoi- tem, which in the case of N’Damas was 70–94% etic progenitor cells in the bone marrow of the of Boran origin, depending on the individual, N’Damas would be more responsive to ambient and in the case of Borans was from 30 to less erythropoietin than those of Borans, consistent than 5% of N’Dama origin, depending on the in- with their resistance to AAT-induced anaemia. dividual, based on analysis of a polymorphic However, unexpectedly, this did not result in a T-cell marker, CD5, using mAbs that distinguish stronger reticulocyte response, which is an Control of Pathogenesis in Animal African Trypanosomiasis 127 indicator of erythropoiesis, in the N’Damas. shown that AAT sometimes causes production Rather, the reticulocyte response of the infected of antibodies that react with antigens on healthy Borans during the early stages of infection was red blood cells, or against antigens exposed by greater than that of the N’Damas (Andrianarivo infection-related processes, or bound as immune et al., 1996), despite the more profound anaemia complexes with trypanosome VSG (Kobayashi that developed in the Borans. It is perhaps note- et al., 1976; Assoku and Gardiner, 1989; Rifkin worthy that the epoR transcript analyses were and Landsberger, 1990), and thus promotes performed with whole bone marrow from infected phagocytosis of the antibody-coated red blood N’Damas and Borans and not from purified cells. However, while this occurs in both N’Da- erythroid progenitor cells (Suliman et al., 1999); mas and Borans infected with a haemorrhagic hence, it is unclear which cell types in the former T. vivax, as shown at ILRAD (Assoku and Gar- have the increased epoR transcript. This may be diner, 1989; Williams et al., 1992), there is no important because erythropoietin is now known evidence that is the case in the T. congolense- to have pleiotropic effects on the immune sys- infected intact and chimeric N’Damas and Borans. tem, where it inhibits macrophage functions Indeed, Naessens and colleagues considered it (Nairz et al., 2012), among other processes. Con- unlikely that an adaptive immune response is sequently, higher expression levels of epoR in responsible for anaemia in the T. congolense- bone marrow macrophages of infected N’Damas infected cattle, because disruption of immune compared with infected Borans at an equal con- responses in N’Damas and Borans by complete centration of erythropoietin might result in a deletion of CD4+ T-cells, CD8+ T-cells and γδ lower level of activation and thus a lower level T-cells from the blood and peripheral organs of in situ phagocytosis of reticulocytes and using specific mAbs (Naessens et al., 2003b; erythroid progenitor cells consistent with hae- Sileghem and Naessens, 1995; Naessens, 2006) matopoietic-tissue intrinsic regulation of anaemia. did not affect their distinct levels of anaemia, Along similar lines, a significantly greater increase although the anti-CD4 treatment severely de- in transcript for the macrophage-activating cyto- creased host antibody responses (data not pub- kine IFN-γ, and the pyrogens interleukin (IL)-1α lished, but summarized in Naessens et al., 2002). and IL-1β, was found in the bone marrows of The pro-inflammatory cytokine TNF-α also infected Borans compared with those of the has a role in the induction of anaemia, at least N’Damas (Suliman et al., 1999), consistent with with some species/strains of African trypano- a greater potential for macrophage activation in somes. Thus, studies in mice have shown that the Borans. It is not known whether an N’Dama levels of anaemia were similar in T. congolense- background would affect these properties of the infected TNF-α knock-out mice and similarly in- Boran haematopoietic tissue; hence, we can only fected wild-type mice, suggesting that anaemia speculate that it would not, thus accounting for in this infection is TNF-α-independent. In con- the inability of N’Damas bearing the Boran trast, the anaemia induced by T. b. rhodesiense haematopoietic system to control anaemia. (Naessens et al., 2005) and T. b. brucei (Magez A second haematopoietic-system intrinsic et al., 1999) infection was significantly lower in mechanism that could lead to greater destruc- TNF-α knock-out compared with intact mice, tion of red blood cells in animals with the Boran suggesting that the infected intact mice have a haematopoietic system affects expression on red TNF-α-dependent mechanism that exacerbates blood cells of important surface antigens, in- anaemia. The highly virulent haemorrhagic cluding blood group antigens. Red blood cells of East African strain of T. vivax may also induce N’Damas have been shown to have significantly this TNF-α-dependent mechanism of anaemia in higher levels of sialic acids on their surface than infected cattle, including N’Damas, whereas red blood cells of Borans (Shugaba et al., 1994), T. congolense does not (Williams et al., 1992; and hence are less affected by trypanosome Sileghem et al., 1994). Furthermore, N’Dama trans-sialidases (Buratai et al., 2006), which cattle proved not to be tolerant against an cleave sialyl groups from surface glycoproteins infection with the haemorrhagic T. vivax, suffer- and glycolipids and directly promote phagocytosis ing as much as or more so than susceptible cattle. of the red blood cells (Nok and Balogun, 2003; Thus, N’Dama cattle may have evolved a trait Guegan et al., 2013). In addition, it has been that protects against TNF-α-independent 128 S.J. Black anaemia, but that is less, or not, effective against unresolved target cells in the liver and lymph TNF-α-d ependent anaemia. nodes. Perhaps this strangely aberrant behaviour of the NK cells is a futile attempt to dampen ex- AAT-induced lymphopenia and possible role cessive immune system activation, an excessive of haemophagocytic syndrome in AAT application of the very response that is defective and inactive in HPS. Analysis of NK and other Studies at ILRAD on peripheral blood leukocyte cytotoxic cell activation in AAT is warranted, as dynamics in N’Damas, Borans, chimeric N’Da- this might inform on the severe depletion of mas and chimeric Borans following cyclic infec- spleen and lymph node cells in late-stage infec- tion with T. congolense showed the same pattern tion, described by Fiennes (1970). as that of anaemia. Thus, while all groups of cat- tle had an initially similar acute decrease in num- Antibody responses of trypanosome-infected bers of total white blood cells, lymphocytes and N’Dama and Boran cattle neutrophils in the peripheral blood during devel- opment of the first parasitaemic wave (Ellis et al., Studies by the ATLN clearly showed that a cura- 1987; Williams et al., 1991; Naessens et al., tive dose of Berenil restored PCV, body weight 2003a), these values recovered quickly in intact and productivity in cattle with AAT, thus showing N’Damas but not in intact Borans, chimeric Borans that AAT-induced pathology is utterly depend- or chimeric N’Damas (Naessens et al., 2003a). ent on the continued presence of trypanosomes There was also a trend in gain of body weight to in affected hosts. The goal of ILRAD scientists suggest that it, too, followed the pattern of an- was to determine which aspects of the host im- aemia and leukopenia in the infected intact and mune response to trypanosomes promotes self- chimeric cattle, but these results were less clear cure and recovery versus sustained infection cut than anaemia and leukopenia. Nevertheless, and severe pathology by comparing immune re- the correlation between anaemia, leukopenia sponses that arise in similarly infected N’Dama and decrease in weight gain in intact and chi- and Boran cattle. Because African trypanosomes meric T. congolense-infected Borans, chimeric are extracellular parasites that are killed by VSG- B orans and chimeric N’Damas is consistent with specific antibody-dependent processes, researchers the possibility that these pathological processes addressed antibody responses. Two antigenic are co-regulated, leading Naessens to consider variants of T. congolense were cloned from the them as a pathogenic syndrome and to seek other first peak of parasitaemia arising in cattle that disease states in which these indicators of path- received the first cyclic infection (Paling et al., ology are similarly linked. The result was recog- 1991a). These clones were shown by VSG ana- nition that AAT and haemophagocytic syndrome lysis to be present in first-wave parasitaemias of (HPS) share many clinical and pathological fea- both N’Damas and Borans. Analysis of clone- tures (Naessens, 2006). HPS is ‘a severe and often specific antibodies in host serum was performed fatal syndrome resulting from potent and uncon- up to 35 days after infection, which corresponds trolled activation and proliferation of T lympho- to a major dip in parasitaemia in both host cytes, leading to excessive macrophage activa- breeds (Paling et al., 1991a). The studies showed tion and multiple deleterious effects. It is that the clone-specific antibodies comprised associated with defects in cytotoxic granule- both IgM and IgG classes and were similarly in- dependent cytotoxic activity of lymphocytes . . . creased in both breeds of cattle, reflecting the thus highlighting the determinant role of this similar patterns of parasitaemia in both breeds function in driving the immune system to a state up to 35  days after infection. Both classes of of equilibrium following infection’ (reviewed by antibody were shown to bind to VSG on intact Menasche et al., 2005). While little is known trypanosomes and mediate their attachment to about cytotoxic effector cells in AAT in cattle, the phagocytes (predominantly neutrophils and author’s laboratory has recently shown that NK monocytes) from N’Damas and Borans in vitro, cells are globally activated in a murine model of with attachment to adherent cells from N’Damas AAT, deleting splenic B2 B-cells (Frenkel et al., exceeding that of adherent cells from Borans, 2016) and CD8+ T-cells (D. Frenkel and S.J. Black and with IgG1 being the most efficient antibody unpublished data, 2019) in the spleen, and as yet at facilitating this attachment (Kamanga-Sollo Control of Pathogenesis in Animal African Trypanosomiasis 129 et al., 1991). Clearance of antibody-coated African (Williams et al., 1996), which had not been trypanosomes from the bloodstream in mouse assayed in the previous study. The kinetics models of AAT is mediated by phagocytic cells in and magnitude of T. congolense ILNat 3.1 VSG the liver, spleen and other organs (Macaskill et al., exposed-epitope-specific IgM, IgG2 and IgG1 1980; Black et al., 1985); consequently, the responses by N’Damas and Borans were close superior capacity of phagocytes from N’Damas to identical, as determined by fluorescent compared with Borans to bind antibody-coated a ntibody-binding assays on intact parasites, trypanosomes in vitro might be expected to be as- consistent with similar control of the first-wave sociated with more efficient clearance of the parasitaemia. However, this was not the case for parasites from infected N’Damas than from Bo- antibodies specific for buried VSG epitopes, rans in vivo and thus lower levels of parasitaemia against which Borans made IgM antibodies and in the former at equivalent levels of specific anti- little IgG1, while N’Damas made IgG1 antibodies bodies. Because this was not observed up to and little IgM. Analyses of splenic antibody- 35 days after infection of the cattle (Paling et al., secreting cells from these animals showed a 1991a), it is possible that a disparity in binding similar disparity in IgM:IgG ratios of plasma of antibody-coated parasites to phagocytes from cells in infected N’Damas and Borans secreting N’Damas and Borans does not arise in infected antibodies that bind to soluble VSG (Taylor et al., animals, perhaps as a result of activation of pha- 1996b). Unfortunately, we know very little about gocytes, which is known to increase expression how antibody responses against VSG on intact of IgG1 receptors on bovine monocytes (McGuire trypanosomes versus that of soluble VSG are et al., 1979). regulated (Black et al., 2010). However, it is pos- In a related study of the same serum sam- sible that antibody responses against exposed ples, an analysis was performed to determine the VSG epitopes on parasite-attached VSG, versus fine specificity of serum antibodies that recog- those against epitopes on VSG that are buried on nize T. congolense clone ILNat 3.1 VSG (Williams intact trypanosomes but accessible on soluble et al., 1996), a variant that arises in the first par- VSG, are made by different sets of B-cells that are asitaemic wave of infected N’Damas and Borans. differently regulated. Indeed, as a result of work This analysis is of interest because VSG is tightly at ILRAD, we know that a substantial portion of packed on the surface of intact trypanosomes; as the antibody response to VSG is made up of a result, only antibodies specific for antigenic low-affinity antibodies that react both with VSG epitopes present on the N-terminal domain of and with a variety of cross-reacting proteins, in- VSGs (exposed VSG epitopes) can bind to intact cluding β-galactosidase and autoantigens (Naes- trypanosomes. An antigenic epitope is that small sens and Williams, 1992; Williams et al., 1996). portion of the antigen that is bound by a specific These antibodies are predominantly produced by antibody; a single antigen can have several dif- a subset of B-lymphocytes that express the dif- ferent antigenic epitopes. Antibodies against ferentiation antigen CD5 (Williams et al., 1991) other epitopes on the same VSG molecule (bur- and, at least with respect to antibodies that react ied VSG epitopes) cannot bind to coated para- both with ILNat 3.1 VSG and β-galactosidase, sites but can bind once the VSG is released from are a feature of the response of infected Borans the parasites, for example as a result of cleavage but not of N’Damas (Williams et al., 1996). with the VSG GPI-PLC. Binding of antibody to The concentration of IgM and IgG2 anti- soluble VSG and its clearance by phagocytes bodies specific for VSG epitopes of T. congolense might be important because this material had ILNat 3.1, and of IgM antibodies specific for bur- been shown by investigators at ILRAD to acti- ied VSG epitopes present in the blood plasma of vate bovine complement (Musoke and Barbet, Borans and N’Damas declined to baseline by 1977), which can elicit a cascade of pro-inflam- 40  days after infection. In contrast, IgG1 anti- matory components and thus might contribute bodies in the blood plasma of the infected to systemic inflammation. This investigation N’Damas and Borans that was specific for both also showed that infected N’Damas and Borans exposed and buried VSG epitopes of ILNat 3.1 made IgM and IgG1 antibodies against exposed VSG remained at close to peak concentrations VSG epitopes on infecting parasites, and in add- 40 days after infection in both breeds of cattle. ition showed that IgG2 antibodies were made Data were not obtained for later time points. At 130 S.J. Black appropriate concentrations, antibodies specific mals (Ilemobade et al., 1982). While AAT can for exposed VSG epitopes neutralize trypanosomes dramatically affect primary and recall antibody expressing that VSG and therefore prevent re- responses to vaccines in cattle, it may not have currence of trypanosomes expressing that or a an equally profound effect on recall antibody re- cross-reacting VSG. In this regard, the accumu- sponses to early-arising VSGs, evidenced by oc- lation of trypanocidal/trypanostatic antibodies casionally recurring spikes of antibody against specific for exposed epitopes of multiple VSG types these VSGs during sustained infection in some in the serum of infected Cape buffalo (Syncerus infected Borans (Nantulya et al., 1979; Musoke caffer) correlates with, and may be responsible et al., 1981; Vos and Gardiner, 1990). Similarly, for, maintenance of cryptic parasitaemia in this processes that compromise antibody responses extremely trypanosomiasis-resistant bovid (Guir- to vaccine antigens do not appear to prevent the nalda et al., 2007). It would therefore be interest- immune responses that control newly arising ing to learn how long the ILNat 3.1-specific IgG1 trypanosome antigenic variants, reflected in re- remained in the plasma of the infected N’Damas peating peaks of parasitaemia, each of which is and Borans and whether similarly long-lived cleared by antibodies specific for exposed VSG IgG1 antibody responses arise against later VSGs epitopes. These observations suggest that im- in both breeds of cattle. This would inform the mune responses against exposed VSG epitopes relative effectiveness of host protective antibody differ in some important respect from primary responses over time in the trypanosomiasis-sus- and secondary immune responses against con- ceptible and trypanotolerant breeds. ventional antigens. Responses to conventional Infected N’Damas were also found to differ antigens are dependent on cognate interactions from Borans by producing IgG1 antibodies between B-cells, which produce the antibodies, against a greater number of proteins than are and CD4+ T-cells, which provide critical stimuli common to different trypanosome serodemes (by cell contact and by secretion of helper cyto- (Shapiro and Murray, 1982; Authie et al., kines) that direct B-cell proliferation and differ- 1993b), including the 33-kDa congopain (Authie entiation to antibody-secreting cells. To determine et al., 1993a), which was discussed earlier with whether T-cells are required for antibody responses respect to its possible contribution to AAT patho- to exposed and buried VSG epitopes, Naessens genesis and as a possible target for an anti-AAT deleted CD4+ T-cells from N’Dama and Boran vaccine. As with the response against buried calves with CD4-specific mAbs prior to their cyclic VSG epitopes considered above, the more diverse infection with T. congolense (unpublished data, antibody response against trypanosome-com- summarized in Naessens, 2006). The antibody mon antigens and the lesser production of anti- responses ‘were found to be markedly reduced bodies against irrelevant antigens in infected and delayed in the depleted animals. . . . This was N’Damas compared with Borans (Williams et al., the case for IgG and IgM antibodies to surface- 1996) suggests functional differences in their exposed and internal trypanosome epitopes, as immune responses. The possibility that this well as for natural IgM antibodies that react with reflects differences in infection-induced T-cell- non-trypanosome antigens.’ Thus, help from CD4+ dependent B-cell responses in the infected cattle T-cells was required for these responses. is considered next. T-cells provide help for B-cell responses to two types of antigens, called T-cell-dependent AAT-induced T-cell responses (TD) and T-cell-independent type 2 (TI-2). TD and immunosuppression antigens are soluble proteins without multiple repeating (identical) epitopes on each molecule. Infection with T. congolense and T. vivax com- TI-2 antigens are usually polysaccharides that promises primary and recall vaccine responses lack direct mitogenic activity and have multiple in Boran cattle (Rurangirwa et al., 1978, 1979, repeating epitopes. The B-cell response to TD 1980, 1983; Whitelaw et al., 1979; Ilemobade antigens requires a direct interaction between: et al., 1982; Sharpe et al., 1982). This manifests as a (i) B-cells that have endocytosed the antigen, substantial decrease in production of vaccine- processed it in an endosome and placed peptides specific IgM and IgG antibodies in the infected derived from the antigen on their surface in compared with the uninfected vaccinated ani- complex with MHC class II (MHC-II); and Control of Pathogenesis in Animal African Trypanosomiasis 131 (ii) T-cells with receptors specific for the MHC-II anosome and common trypanosome antigens, antigen peptide complex. This response develops although these decreased somewhat with time in germinal centres of lymphoid follicles and after infection, and despite repeated curative generates B- and T-memory cells, as well as treatments and reinfections, retained memory antibody-p roducing plasma cells and memory T-cells against these antigens for years after in- plasma cells. The memory cells mount rapid im- fection. mune responses upon re-encountering the same Infections with AAT parasites in mice are antigen. The plasma cells secrete lots of anti- known to induce unresponsiveness of lymph- body specific for the antigen that simulated the node T-cells to mitogens in vitro, which results B-cell, and the long-lived memory plasma cells from inhibition of secretion of the T-cell growth migrate to niches in the bone marrow and con- factor IL-2 and its receptor on T-cells, is medi- tinue to secrete this specific antibody, often for ated by nitric oxide and prostaglandin produc- years. In contrast, the B-cell response to TI-2 tion by macrophages, which is stimulated by antigens occurs in the absence of MHC-II- trypanosome components, IFN-γ and TNF-α restricted T-cell help, although it can be facili- (Sileghem et al., 1991; Darji et al., 1992; Schleifer tated by cytokines produced by activated T-cells and Mansfield, 1993; Sternberg and Mabbott, (Mond et al., 1995), does not require a germinal 1996; Gomez-Rodriguez et al., 2009). Work at centre, and yields short-lived antibody-secreting ILRAD showed that suppressive monocytes/ plasma cells but not memory cells or memory macrophages that inhibited T-cell proliferation plasma cells. Responses against TI-2 antigens, in response to a mitogen (concanavalin A) because they do not require complex interactions in  vitro arose in the peripheral blood, lymph of T- and B-cells, arise faster than those against nodes and spleen of T. congolense-infected cattle TD antigens and play an important role in pro- (Flynn and Sileghem, 1991) and simultaneously tecting against pathogens (Vos et al., 2000). suppressed production of IL-2 and IL-2 receptor Trypanosome-common antigens such as expression (Sileghem and Flynn, 1992b) by the congopain and Hsp70/BiP are most likely TD T-cells but did not suppress their production of antigens, and soluble VSG has formally been IFN-γ (Sileghem and Flynn, 1992a). Unlike the shown to be a TD antigen in cattle, stimulating mouse AAT-induced suppressor cells, those in T-cells that require antigen-presenting cells cattle were not inhibited by indomethacin, (McKeever et al., 1994). VSG epitopes exposed on which prevents production of prostaglandins the surface of intact trypanosomes could jointly (Flynn and Sileghem, 1991), and did not pro- be considered as TI-2 antigens. Thus, although duce nitric oxide in response to IFN-γ (Taylor each independent VSG molecule lacks repeating et al., 1996a), possibly because of a potent IL-10 antigenic epitopes, the assembly of VSGs on the response (Taylor et al., 1998; O’Gorman et al., parasite surface is an array of repeating, anti- 2006). The mechanism through which macro- genically identical epitopes. Therefore, the devel- phages from the infected cattle inhibit T-cell pro- opment of IgG antibody responses against liferative responses in vitro was not resolved nor congopain, Hsp70/BiP and buried VSG epi- was a relationship explored between this in vitro topes in T. congolense-infected N’Damas but not T-cell suppressive response and the impaired re- Borans, discussed above, suggests that the cap- sponses of infected Borans and to a lesser extent acity to mount immune responses against TD N’Damas to TD antigen in vivo. antigens is, or becomes, compromised in Borans A loss in confidence that immunological during infection but not, or to a lesser extent, in differences between infected N’Damas and Bo- N’Damas. In support of this, Flynn and Sileghem rans could be resolved in any way that would im- (1991) at ILRAD showed that T. congolense- prove productivity of either breed under natural infected Borans were unable to generate T-cell tsetse/trypanosome challenge was perhaps a responses to soluble VSG of an infecting tryp- factor in the decision by ILRI to withdraw from anosome (cloned from the first parasitaemic this area of investigation and to diversify into wave) and developed only short-lived responses studies of other pathogens. Nevertheless, it is to other trypanosome antigens, whereas T. congo- noteworthy that this programme showed that two lense-infected N’Damas generated long-lived major indicators of trypanotolerance – control of T-cell responses to both VSG of an infecting tryp- anaemia and parasitaemia – are unlinked, and 132 S.J. Black introduced an additional indicator of trypa- trypanotolerance following natural or experi- notolerance, namely the ability to generate IgG1 mental infection (Dolan, 1987) or, in the absence antibodies against buried VSG epitopes and epi- of infection, using markers with high fidelity to topes on many common trypanosome antigens, the trypanotolerance trait, such as polymorphic including congopain and Hsp70/BiP. One pos- loci that affect or are responsible for trypanotol- sibly practical finding from the programme was erance, if such were available. that infected N’Damas retain responsiveness to An ILRAD/ILRI Trypanotolerance Gene TD antigens throughout infection, whereas Mapping Programme was established in the B orans do not. This suggests that N’Damas, more hope of identifying markers of trypanotolerance so than Borans, would benefit from priming with for marker-assisted selection and also of providing putative anti-AAT vaccine antigens such as con- insights into biological processes that control gopain (Lalmanach et al., 2002) and conserved parasitaemia, anaemia and the efficacy of immune peptides of VSGs (Black and Mansfield, 2016). responses. Figure 2.2 (Ullmann et al., 2005) A second possibly practical outcome is that the outlines the approach taken and explains the magnitude and diversity of IgG1 responses against mapping process, assuming that a single gene trypanosome-common antigens and VSG buried controls the trait; the legend additionally considers epitopes should be used together with data on traits governed by expression of many genes in anaemia and body weight gain in infection-based combination or multiple unlinked genes. screening for trypanotolerance. QTLs and murine trypanotolerance ILRAD started a Bovine Trypanotolerance Map- The genetic basis of trypanotolerance ping Programme in 1990. Its leaders, A.J. Teale and S.J. Kemp, found that while the principle Selection for trypanotolerance has occurred in behind the mapping strategy is simple, the practice certain breeds of B. taurus (humpless) and B. in- is extremely demanding (Kemp and Teale, 1998). dicus (humped) cattle (Dolan, 1987), such as the Mapping required: (i) multiple ovulation families N’Dama, a small humpless long-horn from West of full-sibling F2 N’Dama × Boran; (ii) challenge Africa, and the Orma Boran, the smallest breed of about 200 F2 animals with T. congolense under of the humped short-horned Zebu cattle from carefully defined conditions of challenge and East Africa. Obviously, trypanotolerance is a de- monitoring of responses; (iii) development of a sirable trait in cattle to be ranched in tsetse-in- genomic map of high marker density (with in- fested areas, which includes most of sub-Saharan put from the global scientific community); and Africa. Despite data from ATLN showing that (iv) genotyping of the three generations of cattle N’Dama cattle can be productive under low (Kemp and Teale, 1998) tsetse/trypanosome challenge without any treat- Clearly, this was a major undertaking, par- ment and can be productive under medium tse- ticularly given the absence, at the time of initiation, tse/trypanosome challenge when managed by of a high-marker-density bovine genome map, diagnosis-based application of prophylactic and the age of cows at first calving (about 3 years) curative drugs, there is reluctance among pro- and calving intervals (time between birth of a ducers in East Africa to adopt this breed because calf and the birth of a second calf from the same it is small, feisty and has a low milk yield. Produ- mother – about 12–14  months). Given these cers would prefer the trypanotolerant phenotype constraints, it would take several years to gener- to be introduced by introgressive hybridization ate the F2s. Mapping studies were also conducted into larger breeds of cattle, including improved in crosses of inbred mouse strains that had been European breeds, which would be straightfor- shown at ILRAD to differ in susceptibility to in- ward if disease resistance was due to the pres- fections with T. congolense following needle chal- ence of one or a few genes only, but considerably lenge (Morrison et al., 1978). Survival of inbred less so if a large number of loci are controlling mice after infection with T. congolense might not the trait. Selection of breeds with high expres- be analogous to trypanotolerance in N’Dama sion of trypanotolerance within a herd could be cattle but nevertheless provided a robust pheno- based on their high expression of indicators of type for mapping. Thus, infected C57BL/6 mice Control of Pathogenesis in Animal African Trypanosomiasis 133 Genome of resistant strain Genome of susceptible strain 'Grandparent' F1 F2 Fig. 2.2. Linkage mapping to identify genes responsible for trypanotolerance in N’Dama cattle. A single gene is used as an example. N’Damas are homozygous for a resistant allele (open triangle) and Borans for a susceptible allele (filled triangle). Their F1 offspring obtain the resistant allele from the N’Dama parent and a susceptible allele from the Boran parent. When the F1s are crossed, the F2 generation contains animals with all possible combinations of alleles of the trypanotolerance gene. When infected with trypanosomes, individuals are expected to differ in expression of indicators of trypanotolerance. A correlation of indicators of trypanotolerance with areas of the grandparents’ genomes in each animal allows mapping of areas of the grandparents’ genome responsible for the trypanotolerance trait. The principle is the same for creating a linkage map for a trait that is controlled by many genes if these are arranged as clusters. Traits that are controlled by multiple unlinked genes that are dispersed throughout the genome are not good candidates for linkage mapping. (Constructed by author.) survived for 110.2 days after infection and simi- a disease is linked to a genetic marker is determined larly infected BALB/c mice survived for 49.5 by a statistical method, which yields a LOD score days, while few infected A/J mice survived longer (logarithm of the odds to the base 10). This as- than 20  days after infection. Inbred strains of sesses the probability that a pedigree where the mice have several advantages over cattle with re- disease and the marker are co-segregating is due spect to mapping studies. In each inbred strain, to the existence of linkage or to chance. For ex- all genes are homozygous, a genome map had ample, an LOD score of 3 means the odds are been developed using microsatellite markers (Diet- 1000:1 in favour of genetic linkage. Regions on rich et al., 1996), the age of mice at first pupping mouse chromosomes 5 (maximum LOD score is about 11  weeks, the mice have several off- 4.6) and 17 (maximum LOD score 11), within spring from each mating and there is a short large genomic intervals of 20–40 cM (1 cM con- interval between births. It is therefore unsurpris- tains on average 50 genes), were found to be im- ing that mapping studies progressed somewhat portant in determining resistance in both crosses, faster in crosses between inbred strains of mice while a region on chromosome 1 (maximum than in N’Dama × Boran crosses. LOD score 5.7) showed evidence of involvement Loci (sites of genes) controlling the dur- in only the cross of C57BL/6 × BALB/c (Kemp ation of survival of mice after needle challenge et al., 1997). The impact of these loci on survival with T. congolense were mapped using C57BL/6, times after infection was of large effect, account- BALB/c and A/J grandparents and F gener- ing for most of the genetic variation in both F2 2 ations of the C57BL/6 × A/J F1 and the C57BL/6 populations. The three loci on chromosomes 17, × BALB/c F1 crosses. Survival and marker data 5 and 1 were designated, respectively, Tir1, Tir2 and were subjected to parametric linkage analysis, Tir3 (for trypanosome infection response). Tir1 whereby the probability that a gene important for represents the major trypanotolerance QTL in 134 S.J. Black mouse with an additive effect of 31 days on sur- QTLs and bovine trypanotolerance vival time. Following these initial QTL mapping results, an advanced intercross line approach Twenty-three groups of N’Dama × Boran F2 ani- was taken, which involves random and sequential mals, each containing between three and 13 intercrossing of F s, infection to assess pheno- calves, together with parental groups of N’Damas 2 type and linkage mapping. Using F crosses, Tir1 and Borans, were challenged with T. congolense- 6 was mapped to a 95% confidence interval of infected tsetse flies. Infections were followed for 1.3 cM, Tir2 was mapped to a 12 cM region and 150  days with respect to 16 phenotypic traits Tir3 resolved into three QTLs (Tir3a, 10  cM; (Hanotte et al., 2003). Twenty-eight of the F2 off- Tir3b, 1.8 cM, and Tir3c, 8 cM) (Iraqi et al., 2000), spring needed curative treatment (minimum day necessitating higher resolution mapping for pos- 14, maximum day 146), and the last value taken itional cloning of genes underlying the QTL. An for these animals for all traits studied was taken as attempt to do this using F12 generations fixed for the value of the traits for the remainder of the the susceptibility or resistance alleles at Tir1 challenge period. The traits measured informed mapped Tir2 to less than 1 cM but was less suc- infection-induced decreases in PCV, recovery of cessful in narrowing the positions of Tir3a, Tir3b PCV, infection-induced loss in body weight, recov- and Tir3c (Nganga et al., 2010). An issue arising ery of body weight, levels of parasitaemia between from F12 generation mapping was that the map days 11 and 150 post-infection and the number of positions of Tir2 and Tir3a, Tir3b and Tir3c were times an individual is parasitaemic by the DGBC different from the F mapping study. Further technique, and thus jointly informed QTLs for 6 mapping of congenic mice carrying the C57BL/6 parasitaemia, body weight and anaemia. Tir1, Tir2 and Tir3 resistance alleles on the A/J The animals were genotyped at 477 molecu- background partially resolved this disparity, lar marker loci covering all 29 cattle autosomes supporting the F location of Tir2 and the F lo- (i.e. any chromosome that is not a sex chromo-6 12 cation of Tir3a. These studies, conducted by an some), covering 82% of the bovine genome. international research team including ILRI sci- Putative QTLs were mapped to 18 autosomes. entists, showed that survival after infection was The results were consistent with a single QTL on increased in Tir1 and Tir2 but not Tir3 congenics each of 17 chromosomes, and two on BTA16. (mice that are genetically identical except for the Individual QTL effects ranged from about 6% to loci of interest) and that survival was negatively 20% of the phenotypic variance of the trait. Al- correlated with parasitaemia (i.e. mice with leles for resistance to trypanosomiasis originated lower parasitaemia survived longer) but posi- from the N’Dama parent at nine QTLs and from tively correlated with alanine aminotransferase the Kenyan Boran at five QTLs, and at four QTLs levels in serum, suggesting that inflammatory there was evidence of an overdominant mode of responses in the liver were beneficial (Rathkolb inheritance, when the heterozygote lies outside et al., 2009). Using a wide variety of techniques the phenotypic ranges of the parents (Hanotte to identify candidate genes for murine tolerance et al., 2003). to infections with T. congolense, it was postulated Noyes et al. (2011) tried to obtain short lists that Pram1 (an adaptor protein used in T-cell re- of candidate genes by focusing on polymor- ceptor signalling) was the most plausible candi- phisms within the bovine trypanotolerance QTL date QTL gene in Tir1 and that Cd244 (NK cell by transcriptome analysis of gene expression in receptor 2B4) was a strong candidate QTL gene the liver, spleen and precrural lymph node of at the Tir3c locus (Goodhead et al., 2010). Thus, N’Damas and Borans after infection with T. con- contrary to concerns that trypanotolerance in golense. In this work, they assessed QTL regions mice might be regulated by multiple unlinked and candidate loci for evidence of selective sweeps genes, this proved not to be the case, at least (the reduction or elimination of variation among with respect to events that control early mor- the nucleotides in neighbouring DNA of a muta- tality in T. congolense-infected mice. There have tion as the result of the recent fixation of a bene- been no further publications in this area, so a ficial allele due to strong positive natural selection). contribution of a Pram1 polymorphism to infec- The gene expression data showed that Toll-like tion-induced early mortality remains to be receptors and mitogen-activated protein kinase confirmed. pathways responded to infection, and the former Control of Pathogenesis in Animal African Trypanosomiasis 135 contained TICAM1 (TIR domain-containing termining the genetic basis for trypanotolerance adapter molecule 1), which is within a trypa- in cattle, archives of genomic and cDNA have notolerance QTL on BTA7. Genetic analysis been established from the studies reported above showed that selective sweeps had occurred at the and are available for deeper analysis as technol- TICAM1 and ARHGAP15 (Rho GTPase-activat- ogy in this field undergoes further development to ing protein 15) loci in the N’Damas, making facilitate linking genes to disease-r esistance traits. these strong candidates for genes underlying the QTL. TICAM1 (also known as TRIF; TIR domain- containing adapter protein inducing IFN-β) is an adapter for a few Toll-like receptor signalling Conclusions cascades, while ARHGAP15 regulates signal transduction in the immune system. These pro- From 1975 to 2015, scientists at ILRAD/ILCA/ teins are involved in inflammatory and other im- ILRI tried, with sustained focus and ingenuity, to mune responses. Field studies using 192 cattle identify trypanosome vaccine antigens and key produced from (N’Dama × Kenya Boran) × aspects of trypanosomiasis pathogenesis that Kenya Boran subjected to natural challenge, and could be targeted immunologically or genetic- using a scoring system expanded to take into ac- ally to increase the productivity of cattle under count avoidance of infection and genotyping us- tsetse/trypanosome challenge. Their research ing 35 microsatellite markers spanning five bovine efforts, although logical, incremental and pains- chromosomes that were found in the above stud- taking, did not achieve this goal. ies to contain trypanotolerance QTLs, showed Infection and treatment, a process com- that trypanotolerance was expressed in propor- monly used to induce protective immunity to tion to N’Dama origin marker alleles (Orenge other pathogens, was shown to have limited suc- et  al., 2011) and additionally showed the im- cess with African trypanosomes because of their portance of sex and local environment condi- huge antigenic diversity. Subsequent investiga- tions in determining the response to challenge tions of host-derived macromolecular nutrients (Orenge et al., 2012). Given that trypanotoler- that drive/sustain trypanosome cell-cycle pro- ance QTLs are present on 18 chromosomes, with gression were successful but did not have prac- an allele for resistance present at nine N’Dama tical value because targeting their conserved QTLs and five Boran QTLs, the relevant genes, or receptors with lytic or blocking antibodies was markers tightly linked to these, would need to be ineffective. Although work at ILRAD identified resolved before the QTLs can be used in marker- low-, intermediate- and high-density serum lipo- assisted selection. Thus, traits that distinguish proteins and transferrin as required trypanosome between N’Dama and Boran cattle trypanotoler- growth factors, antisera against receptors for ance in cattle proved to be regulated by multiple these molecules had little or no impact on tryp- unlinked genes. There have been no further anosome infectivity or growth in mammalian publications on identifying candidate genes for hosts. Despite the lack of translational impact of trypanotolerance in bovids. these investigations, the feeder layer-dependent Mapping of trypanotolerance genes has been and axenic culture systems, macronutrient an enormous undertaking and, as with so many bioassays and trypanosome cell-cycle analysis aspects of trypanosomiasis research, it has not methodology developed at ILRAD for the above been rewarded with leaps in understanding. studies were forerunners of the high-throughput Nevertheless, as a result of these studies we are drug screens, genetic manipulation, and cell- cycle now fully aware of the genetic complexity of the and differentiation pathway analyses that com- trait in cattle and its distinct and lesser c omplexity prise much of the current basic research on the in mice, at least with respect to survival after African trypanosomes. infection. Indeed, given that a candidate gene for Investigations at ILRAD/ILCA/ILRI to eluci- Tir1 has been proposed, it would certainly be worth- date mechanisms of trypanotolerance using Cape while to determine how expression of this gene, buffalo and N’Dama and Boran cattle models were Pram1, affects infection-induced mortality and also scientifically rewarding but of little transla- changes in immune parameters that accompany tional value. These studies showed that superior this (Morrison et al., 1978). With respect to de- control of trypanosomiasis in Cape buffalo and, to a 136 S.J. Black lesser extent, in N’Dama cattle, was associated with S.J. Black and colleagues continue to seek a key the ability to suppress trypanosome parasitaemia initiator pathway of trypanosomiasis pathogen- and to sustain PCV, body weight and immune com- esis and have recently shown (Frenkel et al., 2016) petence throughout infection, the latter evidenced that T. brucei, as well as T. congolense (S.J. Black by production of IgG antibodies specific for VSGs of and D. Frenkel, unpublished data, 2019), infection newly arising trypanosome variable antigen types co-opts host NK cells to delete splenic B2 cells and and other trypanosome antigens. However, the thus prevent sustained TD antibody responses physiological mechanisms underlying this product- against VSG and other trypanosome antigens. ive response, and its absence in trypanosomiasis- Strikingly, when NK cell killing is prevented, susceptible hosts, were not identified. Furthermore, trypanosome-infected mice suppress parasit- analysis of the genetic basis of trypanotolerance in aemia, do not develop anaemia or wasting, and cattle showed this to involve 18 QTLs, all of which do not succumb to infection-induced early mor- make relatively minor (10% or less) contributions tality. Thus, elucidation of the trypanosome to the trait and therefore are of little value for se- component that activates NK cells might provide lective breeding to improve trypanotolerance of Bo- a novel target for AAT immunoprophylaxis or a rans and other livestock. While disappointing with probe to identify cattle that do not respond to it respect to AAT control, studies of AAT pathogen- and hence may be trypanotolerant. Finally, other esis at ILRAD/ILRI did identify the definitive ques- events might supersede the need to selectively tion for immunological research on AAT, namely, control AAT. S.J. Kemp and colleagues at ILRI how do trypanosomes eliminate TD antibody re- have initiated a multinational project in Africa sponses in trypanosomiasis-susceptible mammals? to identify those rare cattle that remain healthy In addition, the work at ILRI on the genetic basis of and productive under ambient disease and en- trypanotolerance contributed a high-density single- vironmental stress. This will entail animal prod- nucleotide polymorphism (SNP) map of the bovine uctivity reporting by legions of smallholder genome that has intrinsic value for analysis of QTLs farmers in disease-endemic regions, SNP ana- that control other traits, including susceptibility to lyses of blood leukocytes from cattle with a other diseases. high-productivity phenotype, selective breeding and validation. The Future ILRI and ILRAD alumni and their global Acknowledgements partners continue to seek solutions to AAT. Jayne Raper and colleagues are exploring genetic The author thanks Max Murray, Jan Naessens engineering of cattle to express a trypanosome- and Stuart Shapiro for sharing material, lytic factor derived from primates that are resist- helpful discussions and critical reviews of the ant to trypanosome infection (MacMillan, 2016). manuscript. References Aboagye-Kwarteng, T., ole-MoiYoi, O.K. and Lonsdale-Eccles, J.D. (1991) Phosphorylation differences among proteins of bloodstream developmental stages of Trypanosoma brucei brucei. Biochemistry Journal 275, 7–14. Adams, E.R., Hamilton, P.B. and Gibson, W.C. (2010) African trypanosomes: celebrating diversity. Trends in Parasitology 26, 324–328. Akol, G.W. and Murray, M. (1983) Trypanosoma congolense: susceptibility of cattle to cyclical challenge. Experimental Parasitology 55, 386–393. Akol, G.W. and Murray, M. (1985) Induction of protective immunity in cattle by tsetse-transmitted cloned isolates of Trypanosoma congolense. Annals of Tropical Medicine and Parasitology 79, 617–627. Control of Pathogenesis in Animal African Trypanosomiasis 137 Akol, G.W. and Murray, M. (1986) Parasite kinetics and immune responses in efferent prefemoral lymph draining skin reactions induced by tsetse-transmitted Trypanosoma congolense. Veterinary Parasit- ology 19, 281–293. Andrianarivo, A.G., Muiya, P., Opollo, M. and Logan-Henfrey, L.L. (1995) Trypanosoma congolense: com- parative effects of a primary infection on bone marrow progenitor cells from N’Dama and Boran cattle. Experimental Parasitology 80, 407–418. Andrianarivo, A.G., Muiya, P. and Logan-Henfrey, L.L. (1996) Trypanosoma congolense: high erythropoietic potential in infected yearling cattle during the acute phase of the anemia. Experimental Parasitology 82, 104–111. Assoku, R.K. and Gardiner, P.R. (1989) Detection of antibodies to platelets and erythrocytes during infec- tion with haemorrhage-causing Trypanosoma vivax in Ayrshire cattle. Veterinary Parasitology 31, 199–216. AU-IBAR (2018) The International Scientific Council for Trypanosomosis Research and Control holds its 34th conference to review and recommend control approaches for sleeping sickness and nagana. Available at: http://www.au-ibar.org/2012-10-01-13-08-42/press-releases/332-isctrc/1174-the-international- scientific-council-for-trypanosomosis-research-and-control-holds-it-s-34th-conference-to- review-and- recommend-control-approaches-for-sleeping-sickness-and-nagana (accessed 31 January 2020). Authie, E., Duvallet, G., Robertson, C. and Williams, D.J. (1993a) Antibody responses to a 33 kDa cysteine protease of Trypanosoma congolense: relationship to ‘trypanotolerance’ in cattle. Parasite Immun- ology 15, 465–474. Authie, E., Muteti, D.K. and Williams, D.J. (1993b) Antibody responses to invariant antigens of Trypanosoma congolense in cattle of differing susceptibility to trypanosomiasis. Parasite Immunology 15, 101–111. Authie, E., Boulange, A., Muteti, D., Lalmanach, G., Gauthier, F. and Musoke, A.J. (2001) Immunisation of cattle with cysteine proteinases of Trypanosoma congolense: targetting the disease rather than the parasite. International Journal for Parasitology 31, 1429–1433. Banks, K.L. (1978) Binding of Trypanosoma congolense to the walls of small blood vessels. Journal of Protozoology 25, 241–245. Barbet, A.F. and McGuire, T.C. (1978) Crossreacting determinants in variant-specific surface antigens of African trypanosomes. Proceedings of the National Academy of Sciences USA 75, 1989–1993. Barry, J.D., Crowe, J.S. and Vickerman, K. (1983) Instability of the Trypanosoma brucei rhodesiense meta- cyclic variable antigen repertoire. Nature 306, 699–701. Bienen, E.J., Webster, P. and Fish, W.R. (1991) Trypanosoma (Nannomonas) congolense: changes in re- spiratory metabolism during the life cycle. Experimental Parasitology 73, 403–412. Black, S. and Mansfield, J. (2016) Prospects for vaccination against pathogenic African trypanosomes. Parasite Immunology 38, 735–743. Black, S. and Vandeweerd, V. (1989) Serum lipoproteins are required for multiplication of Trypanosoma brucei brucei under axenic culture conditions. Molecular and Biochemical Parasitology 37, 65–72. Black, S.J., Hewett, R.S. and Sendashonga, C.N. (1982) Trypanosoma brucei variable surface antigen is released by degenerating parasites but not by actively dividing parasites. Parasite Immunology 4, 233–244. Black, S.J., Sendashonga, C.N., O’Brien, C., Borowy, N.K., Naessens, M., et al. (1985) Regulation of para- sitaemia in mice infected with Trypanosoma brucei. Current Topics in Microbiology and Immunology 117, 93–118. Black, S.J., Guirnalda, P., Frenkel, D., Haynes, C. and Bockstal, V. (2010) Induction and regulation of Tryp- anosoma brucei VSG-specific antibody responses. Parasitology 137, 2041–2049. Blaser, E., Jibbo, J. and McIntyre, W. (1979) A field trial of the protective effect of Samorin and Berenil in Zebu cattle under ranching conditions in Tanzania. In: Report of the 15th Meeting of the Internations Scientific Counil for Trypanosomiasis research and Control, Banjul, The Gambia, 1977. OAU/STRC, Nairobi, pp. 383–386. Borowy, N.K., Fink, E. and Hirumi, H. (1985a) In vitro activity of the trypanocidal diamidine DAPI on animal- infective Trypanosoma brucei brucei. Acta Tropica 42, 287–298. Borowy, N.K., Fink, E. and Hirumi, H. (1985b) Trypanosoma brucei: five commonly used trypanocides as- sayed in vitro with a mammalian feeder layer system for cultivation of bloodstream forms. Experimen- tal Parasitology 60, 323–330. Borowy, N.K., Hirumi, H., Waithaka, H.K. and Mkoji, G. (1985c) An assay for screening drugs against ani- mal-infective bloodstream forms of Trypanosoma brucei brucei in vitro. Drugs under Experimental and Clinical Research 11, 155–161. 138 S.J. Black Borowy, N.K., Nelson, R.T., Hirumi, H., Brun, R., Waithaka, H.K., et al. (1988) Ro 15-0216: a nitroimidazole compound active in vitro against human and animal pathogenic African trypanosomes. Annals of Tropical Medicine and Parasitology 82, 13–19. Bosompem, K.M., Assoku, R.K. and Nantulya, V.M. (1995a) Hydrogen peroxide destaining: a new method for removing non-specific stains in nitrocellulose membrane-based dot-ELISA for the detection of tryp- anosomes in tsetse flies (Glossina spp. ). Journal of Immunological Methods 187, 23–31. Bosompem, K.M., Moloo, S.K., Assoku, R.K. and Nantulya, V.M. (1995b) Detection and differentiation between trypanosome species in experimentally infected tsetse flies (Glossina spp.) using dot-ELISA. Acta Tropica 60, 81–96. Bosompem, K.M., Masake, R.A., Assoku, R.K., Opiyo, E.A. and Nantulya, V.M. (1996) Field evaluation of a dot-ELISA for the detection and differentiation of trypanosome species in infected tsetse flies (Glossi- na spp.). Parasitology 112, 205–211. Bridgen, P.J., Cross, G.A. and Bridgen, J. (1976) N-terminal amino acid sequences of variant-specific surface antigens from Trypanosoma brucei. Nature 263, 613–614. Buratai, L.B., Nok, A.J., Ibrahim, S., Umar, I.A. and Esievo, K.A. (2006) Characterization of sialidase from bloodstream forms of Trypanosoma vivax. Cell Biochemistry & Function 24, 71–77. Burudi, E.M., Peregrine, A.S., Majiwa, P.A., Mbiuki, S.M. and Murphy, N.B. (1994) Response of diminazene-re- sistant and diminazene-susceptible Trypanosoma congolense to treatment with diminazene when oc- curring as a mixed infection in goats. Annals of Tropical Medicine and Parasitology 88, 595–606. Camille-Isidore, P. (1906) L’elevage dans l’Afrique Occidentale Francaise. Gouvernement General de l’Afri- que Occidentale Francaise, Paris. Cardoso de Almeida, M.L. and Turner, M.J. (1983) The membrane form of variant surface glycoproteins of Trypanosoma brucei. Nature 302, 349–352. Cecchi, G., Paone, M., Argiles Herrero, R., Vreysen, M.J. and Mattioli, R.C. (2015) Developing a continental atlas of the distribution and trypanosomal infection of tsetse flies (Glossina species). Parasites & Vectors 8, 284. Clausen, P.H., Sidibe, I., Kabore, I. and Bauer, B. (1992) Development of multiple drug resistance of Trypa- nosoma congolense in Zebu cattle under high natural tsetse fly challenge in the pastoral zone of Samorogouan, Burkina Faso. Acta Tropica 51, 229–236. Cnops, J., de Trez, C., Stijlemans, B., Keirsse, J., Kauffmann, F., et al. (2015) NK-, NKT- and CD8-derived IFNα drives myeloid cell activation and erythrophagocytosis, resulting in trypanosomosis-associated acute anemia. PLoS Pathogens 11, e1004964. Coppens, I., Baudhuin, P., Opperdoes, F.R. and Courtoy, P.J. (1988) Receptors for the host low density lipo- proteins on the hemoflagellate Trypanosoma brucei: purification and involvement in the growth of the parasite. Proceedings of the National Academy of Sciences USA 85, 6753–6757. Cross, G.A. (1975) Identification, purification and properties of clone-specific glycoprotein antigens consti- tuting the surface coat of Trypanosoma brucei. Parasitology 71, 393–417. Dargie, J.D., Murray, P.K., Murray, M., Grimshaw, W.R. and McIntyre, W.I. (1979) Bovine trypanosomiasis: the red cell kinetics of N’Dama and Zebu cattle infected with Trypanosoma congolense. Parasitology 78, 271–286. Darji, A., Lucas, R., Magez, S., Torreele, E., Palacios, J., et al. (1992) Mechanisms underlying trypanosome- elicited immunosuppression. Annales de la Société Belge de Médecine Tropicale 72 (Suppl. 1), 27–38. Davies, A.J.S. and Targett, G.A.T. (1978) Some Perspectives in Parasitic Disease. in ILRAD Inauguration Symposium on Current Trends in Immunology and Genetics and their Implications for Parasitic Dis- eases. ILRAD. Nairobi. 1978. pp. 65–87. Devine, D.V., Falk, R.J. and Balber, A.E. (1986) Restriction of the alternative pathway of human complement by intact Trypanosoma brucei subsp. gambiense. Infection and Immunity 52, 223–229. d’Ieteren, G., Coulibaly, L., Atse, P., Hecker, P., Krebs, H., et al. (1997) Trypanocidal drug resistance in four regions of Cote d’Ivoire. Importance and possible impact on sustainability of integrated strategies for trypanosomiasis control. In: Proceedings of the 23rd Meeting of the International Scientific Council for Trypanosomiasis Research and Control, 11–15 September 1995, Banul, The Gambia. OAU/STRC, Nairobi, pp. 233–247. d’Ieteren, G.D., Authie, E., Wissocq, N. and Murray, M. (1998) Trypanotolerance, an option for sustainable livestock production in areas at risk from trypanosomosis. Revue Scientifique et Technique 17, 154–175. Dietrich, W.F., Miller, J., Steen, R., Merchant, M.A., Damron-Boles, D., et al. (1996) A comprehensive genetic map of the mouse genome. Nature 380, 149–152. Control of Pathogenesis in Animal African Trypanosomiasis 139 Dirie, M.F., Murphy, N.B. and Gardiner, P.R. (1993a) DNA fingerprinting of Trypanosoma vivax isolates rapidly identifies intraspecific relationships. Journal of Eukaryotic Microbiology 40, 132–134. Dirie, M.F., Otte, M.J., Thatthi, R. and Gardiner, P.R. (1993b) Comparative studies of Trypanosoma (Duttonella) vivax isolates from Colombia. Parasitology 106, 21–29. Dolan, R.B. (1987) Genetics and trypanotolerance. Parasitology Today 3, 137–143. Doyle, J.J., Hirumi, H., Hirumi, K., Lupton, E.N. and Cross, G.A. (1980) Antigenic variation in clones of animal-infective Trypanosoma brucei derived and maintained in vitro. Parasitology 80, 359–369. Dube, D.K., Mpimbaza, G., Allison, A.C., Lederer, E. and Rovis, L. (1983) Antitrypanosomal activity of sinefungin. American Journal of Tropical Medicine and Hygiene 32, 31–33. Dwinger, R.H., Luckins, A.G., Murray, M., Rae, P. and Moloo, S.K. (1986) Interference between different serodemes of Trypanosoma congolense in the establishment of superinfections in goats following transmission by tsetse. Parasite Immunology 8, 293–305. Dwinger, R.H., Murray, M. and Moloo, S.K. (1987) Potential value of localized skin reactions (chancres) induced by Trypanosoma congolense transmitted by Glossina morsitans centralis for the analysis of metacyclic trypanosome populations. Parasite Immunology 9, 353–362. Dwinger, R.H., Murray, M., Luckins, A.G., Rae, P.F. and Moloo, S.K. (1989) Interference in the establishment of tsetse-transmitted Trypanosoma congolense, T. brucei or T. vivax superinfections in goats already infected with T. congolense or T. vivax. Veterinary Parasitology 30, 177–189. Ellis, J.A., Scott, J.R., Machugh, N.D., Gettinby, G. and Davis, W.C. (1987) Peripheral blood leucocytes subpopulation dynamics during Trypanosoma congolense infection in Boran and N’Dama cattle: an analysis using monoclonal antibodies and flow cytometry. Parasite Immunology 9, 363–378. Fairlamb, A.H., Weislogel, P.O., Hoeijmakers, J.H. and Borst, P. (1978) Isolation and characterization of kinetoplast DNA from bloodstream form of Trypanosoma brucei. Journal of Cell Biology 76, 293–309. Fasogbon, A.I., Knowles, G. and Gardiner, P.R. (1990) A comparison of the isoenzymes of Trypanosoma (Dut- tonella) vivax isolates from East and West Africa. International Journal for Parasitology 20, 389–394. Fiennes, R. (1970) Pathogenesis and pathology of animal trypanosomiases. In: Mulligan, H.W. (ed.) The African Trypanosomiases. Wiley, London, pp. 729–750. Flynn, J.N. and Sileghem, M. (1991) The role of the macrophage in induction of immunosuppression in Trypanosoma congolense-infected cattle. Immunology 74, 310–316. Ford, J. (1971) The Role of trypanosomiases in African Ecology. A Study of the Tsetse Fly problem. Clarendon Press. Oxford. Frenkel, D., Zhang, F., Guirnalda, P., Haynes, C., Bockstal, V., et al. (2016) Trypanosoma brucei co-opts NK cells to kill splenic B2 B cells. PLoS Pathogens 12, e1005733. Gardiner, P.R. (1989) Recent studies of the biology of Trypanosoma vivax. Advances in Parasitology 28, 229–317. Gardiner, P.R., Thatthi, R., Gathuo, H., Nelson, R. and Moloo, S.K. (1986) Further studies of cyclical transmission and antigenic variation of the ILDar 1 serodeme of Trypanosoma vivax. Parasitology 92, 581–593. Gardiner, P.R., Pearson, T.W., Clarke, M.W. and Mutharia, L.M. (1987) Identification and isolation of a variant surface glycoprotein from Trypanosoma vivax. Science 235, 774–777. Gerrits, H., Mussmann, R., Bitter, W., Kieft, R. and Borst, P. (2002) The physiological significance of transferrin receptor variations in Trypanosoma brucei. Molecular and Biochemical Parasitology 119, 237–247. Gibson, W.C., Wilson, A.J. and Moloo, S.K. (1983) Characterisation of Trypanosoma (Trypanozoon) evansi from camels in Kenya using isoenzyme electrophoresis. Research in Veterinary Science 34, 114–118. Gomez-Rodriguez, J., Stijlemans, B., de Muylder, G., Korf, H., Brys, L., et al. (2009) Identification of a para- sitic immunomodulatory protein triggering the development of suppressive M1 macrophages during African trypanosomiasis. Journal of Infectious Diseases 200, 1849–1860. Goodhead, I., Archibald, A., Amwayi, P., Brass, A., Gibson, J., et al. (2010) A comprehensive genetic analysis of candidate genes regulating response to Trypanosoma congolense infection in mice. PLoS Neglected Tropical Diseases 4, e880. Grab, D.J., Webster, P. and Verjee, Y. (1984) The intracellular pathway and assembly of newly formed vari- able surface glycoprotein of Trypanosoma brucei. Proceedings of the National Academy of Sciences USA 81, 7703–7707. Grab, D.J., Webster, P., Ito, S., Fish, W.R., Verjee, Y. and Lonsdale-Eccles, J.D. (1987) Subcellular localization of a variable surface glycoprotein phosphatidylinositol-specific phospholipase-C in African trypanosomes. Journal of Cell Biology 105, 737–746. 140 S.J. Black Grab, D.J., Wells, C.W., Shaw, M.K., Webster, P. and Russo, D.C. (1992) Endocytosed transferrin in African trypanosomes is delivered to lysosomes and may not be recycled. European Journal of Cell Biology 59, 398–404. Grab, D.J., Shaw, M.K., Wells, C.W., Verjee, Y., Russo, D.C., et al. (1993) The transferrin receptor in African trypanosomes: identification, partial characterization and subcellular localization. European Journal of Cell Biology 62, 114–126. Gray, A.R. (1967) Some principles of the immunology of trypanosomiasis. Bulletin of the World Health Or- ganization 37, 177–193. Gray, A.R. (1970) Antigenic variation In: Mulligan, H.W. (ed.) The African Trypanosomiases. Wiley, London, pp. 113–116. Guegan, F., Plazolles, N., Baltz, T. and Coustou, V. (2013) Erythrophagocytosis of desialylated red blood cells is responsible for anaemia during Trypanosoma vivax infection. Cellular Microbiology 15, 1285–1303. Guirnalda, P., Murphy, N.B., Nolan, D. and Black, S.J. (2007) Anti-Trypanosoma brucei activity in Cape buffalo serum during the cryptic phase of parasitemia is mediated by antibodies. International Journal for Parasitology 37, 1391–1399. Hamilton, P.B., Gibson, W.C. and Stevens, J.R. (2007) Patterns of co-evolution between trypanosomes and their hosts deduced from ribosomal RNA and protein-coding gene phylogenies. Molecular Phyloge- netics and Evolution 44, 15–25. Hanotte, O., Ronin, Y., Agaba, M., Nilsson, P., Gelhaus, A., et al. (2003) Mapping of quantitative trait loci controlling trypanotolerance in a cross of tolerant West African N’Dama and susceptible East African Boran cattle. Proceedings of the National Academy of Sciences USA 100, 7443–7448. Hanrahan, O., Webb, H., O’Byrne, R., Brabazon, E., Treumann, A., et al. (2009) The glycosylphosphatidy- linositol-PLC in Trypanosoma brucei forms a linear array on the exterior of the flagellar membrane before and after activation. PLoS Pathogens 5, e1000468. Hassan, F.A. (2000) Climate and cattle in North Africa: a first approximation. In: Blench, R. and MacDonald, K. (eds) The Origin and Development of African Livestock: Archeology, Genetics, Linguistics and Ethnography. UCL Press, London, pp. 61–86. Hirumi, H., Doyle, J.J. and Hirumi, K. (1977) African trypanosomes: cultivation of animal-infective Trypano- soma brucei in vitro. Science 196, 992–994. Ilemobade, A.A., Adegboye, D.S., Onoviran, O. and Chima, J.C. (1982) Immunodepressive effects of trypa- nosomal infection in cattle immunized against contagious bovine pleuropneumonia. Parasite Immun- ology 4, 273–282. ILCA/ILRAD (1988) Livestock Production in Tsetse Affected Areas of Africa. Proceedings of a meeting held in Nairobi, Kenya from the 23rd to 27th November 1987. ILRAD, Nairobi, and ILCA Addis Ababa. ILRAD (1985) Scientists explore resistance to trypanosomes. ILRAD Reports, Volume 3, number 2.ILRAD, Nairobi. Iraqi, F., Clapcott, S.J., Kumari, P., Haley, C.S., Kemp, S.J. and Teale, A.J. (2000) Fine mapping of trypano- somiasis resistance loci in murine advanced intercross lines. Mammalian Genome 11, 645–648. Jackson, A.P., Berry, A., Aslett, M., Allison, H.C., Burton, P., et al. (2012) Antigenic diversity is generated by distinct evolutionary mechanisms in African trypanosome species. Proceedings of the National Acad- emy of Sciences USA 109, 3416–3421. Jackson, A.P., Allison, H.C., Barry, J.D., Field, M.C., Hertz-Fowler, C. and Berriman, M. (2013) A cell-surface phylome for African trypanosomes. PLoS Neglected Tropical Diseases 7, e2121. Jahnke, H.E. (1976) Tsetse Flies and Livestock Development in East Africa: A Study in Environmental Eco- nomics. Weltforum Verlag. Munich, Germany. Jordt, T., Mahon, G.D., Touray, B.N., Ngulo, W.K., Morrison, W.I., et al. (1986a) Successful transfer of frozen N’Dama embryos from the Gambia to Kenya. Tropical Animal Health and Production 18, 65–75. Jordt, T., Mahon, G.D., Touray, B.N., Ngulo, W.K., Morrison, W.I., et al. (1986b) Successful transfer of N’Da- ma embryos into Boran recipients. Veterinary Record 119, 246–247. Kamanga-Sollo, E.I., Musoke, A.J., Nantulya, V.M., Rurangirwa, F.R. and Masake, R.A. (1991) Differences between N’Dama and Boran cattle in the ability of their peripheral blood leucocytes to bind anti- body-coated trypanosomes. Acta Tropica 49, 109–117. Kemp, S.J. and Teale, A.J. (1998) Genetic basis of trypanotolerance in cattle and mice. Parasitology Today 14, 450–454. Kemp, S.J., Iraqi, F., Darvasi, A., Soller, M. and Teale, A.J. (1997) Localization of genes controlling resist- ance to trypanosomiasis in mice. Nature Genetics 16, 194–196. Control of Pathogenesis in Animal African Trypanosomiasis 141 Kihurani, D.O., Nantulya, V.M., Mbiuki, S.M., Mogoa, E., Nguhiu-Mwangi, J. and Mbithi, P.M. (1994) Trypa- nosoma brucei, T. congolense and T. vivax infections in horses on a farm in Kenya. Tropical Animal Health and Production 26, 95–101. Kihurani, D.O., Masake, R.A., Nantulya, V.M. and Mbiuki, S.M. (2000) Characterization of trypanosome isolates from naturally infected horses on a farm in Kenya. Veterinary Parasitology 89, 173–185. Knowles, G., Black, S.J. and Whitelaw, D.D. (1987) Peptidase in the plasma of mice infected with Trypano- soma brucei brucei. Parasitology 95, 291–300. Knowles, G., Betschart, B., Kukla, B.A., Scott, J.R. and Majiwa, P.A. (1988) Genetically discrete populations of Trypanosoma congolense from livestock on the Kenyan coast. Parasitology 96, 461–474. Kobayashi, A., Tizard, I.R. and Woo, P.T. (1976) Studies on the anemia in experimental African trypanosom- iasis. II. The pathogenesis of the anemia in calves infected with Trypanosoma congolense. American Journal of Tropical Medicine and Hygiene 25, 401–406. Komba, E., Odiit, M., Mbulamberi, D.B., Chimfwembe, E.C. and Nantulya, V.M. (1992) Multicentre evalu- ation of an antigen-detection ELISA for the diagnosis of Trypanosoma brucei rhodesiense sleeping sickness. Bulletin of the World Health Organization 70, 57–61. Kooy, R.F., Hirumi, H., Moloo, S.K., Nantulya, V.M., Dukes, P., et al. (1989) Evidence for diploidy in metacyclic forms of African trypanosomes. Proceedings of the National Academy of Sciences USA 86, 5469–5472. Kristjanson, P.M., Swallow, B.M., Rowlands, G.J., Kruska, R.L. and de Leeuw, P. (1999) Measuring the costs of African animal trypanosomosis, the potential benefits of control and returns to research. Agri- cultural Systems 59, 79–98. Kukla, B.A., Majiwa, P.A., Young, J.R., Moloo, S.K. and ole-MoiYoi, O.K. (1987) Use of species-specific DNA probes for detection and identification of trypanosome infection in tsetse flies. Parasitology 95, 1–16. La Greca, F. and Magez, S. (2011) Vaccination against trypanosomiasis: can it be done or is the trypano- some truly the ultimate immune destroyer and escape artist? Human Vaccines 7, 1225–1233. Lalmanach, G., Boulange, A., Serveau, C., Lecaille, F., Scharfstein, J., et al. (2002) Congopain from Tryp- anosoma congolense: drug target and vaccine candidate. Biological Chemistry 383, 739–749. Lonsdale-Eccles, J.D. and Grab, D.J. (1987) Lysosomal and non-lysosomal peptidyl hydrolases of the bloodstream forms of Trypanosoma brucei brucei. European Journal of Biochemistry 169, 467–475. Macaskill, J.A., Holmes, P.H., Whitelaw, D.D., McConnell, I., Jennings, F.W. and Urquhart, G.M. (1980) Immunological clearance of 75Se-labelled Trypanosoma brucei in mice. II. Mechanisms in immune animals. Immunology 40, 629–635. MacMillan, S. (2016) Cloned bull could contribute to development of disease-resistant African cattle. ILRI News. ILRI, Nairobi. Available at: https://news.ilri.org/2016/09/05/cloned-bull-could-contribute-to- development-of-disease-resistant-african-cattle/ (accessed 3 February 2020). Magez, S., Radwanska, M., Beschin, A., Sekikawa, K. and de Baetselier, P. (1999) Tumor necrosis factor alpha is a key mediator in the regulation of experimental Trypanosoma brucei infections. Infection and Immunity 67, 3128–3132. Majiwa, P.A., Young, J.R., Englund, P.T., Shapiro, S.Z. and Williams, R.O. (1982) Two distinct forms of sur- face antigen gene rearrangement in Trypanosoma brucei. Nature 297, 514–516. Majiwa, P.A., Masake, R.A., Nantulya, V.M., Hamers, R. and Matthyssens, G. (1985) Trypanosoma (Nanno- monas) congolense: identification of two karyotypic groups. EMBO Journal 4, 3307–3313. Mamman, M. and Peregrine, A.S. (1994) Pharmacokinetics of diminazene in plasma and cerebrospinal fluid of goats. Research in Veterinary Science 57, 253–255. Mamman, M., Katende, J., Moloo, S.K. and Peregrine, A.S. (1993) Variation in sensitivity of Trypanosoma congolense to diminazene during the early phase of tsetse-transmitted infection in goats. Veterinary Parasitology 50, 1–14. Mamman, M., Moloo, S.K. and Peregrine, A.S. (1994) Relapse of Trypanosoma congolense infection in goats after diminazene aceturate is not a result of invasion of the central nervous system. Annals of Tropical Medicine and Parasitology 88, 87–88. Masake, R.A., Nyambati, V.M., Nantulya, V.M., Majiwa, P.A., Moloo, S.K. and Musoke, A.J. (1988) The chromosome profiles of Trypanosoma congolense isolates from Kilifi, Kenya and their relationship to serodeme identity. Molecular and Biochemical Parasitology 30, 105–112. Masake, R.A., Moloo, S.K., Nantulya, V.M., Minja, S.H., Makau, J.M. and Njuguna, J.T. (1995) Comparative sensitivity of antigen-detection enzyme immunosorbent assay and the microhaematocrit centrifugation technique in the diagnosis of Trypanosoma brucei infections in cattle. Veterinary Parasitology 56, 37–46. Masake, R.A., Majiwa, P.A., Moloo, S.K., Makau, J.M., Njuguna, J.T., et al. (1997) Sensitive and specific detec- tion of Trypanosoma vivax using the polymerase chain reaction. Experimental Parasitology 85, 193–205. 142 S.J. Black Mattioli, R.C. and Faye, J.A. (1996) A comparative study of the parasitological buffy coat technique and an antigen enzyme immunoassay for trypanosome diagnosis in sequential Trypanosoma congolense infections in N’Dama, Gobra zebu and N’Dama × Gobra crossbred cattle. Acta Tropica 62, 71–81. McGuire, T.C., Musoke, A.J. and Kurtti, T. (1979) Functional properties of bovine IgG1 and IgG2: interaction with complement, macrophages, neutrophils and skin. Immunology 38, 249–256. McKeever, D.J., Awino, E., Kairo, A., Gobright, E. and Nene, V. (1994) Bovine T cells specific for Trypanoso- ma brucei brucei variant surface glycoprotein recognize nonconserved areas of the molecule. Infec- tion and Immunity 62, 3348–3353. Menasche, G., Feldmann, J., Fischer, A. and de Saint Basile, G. (2005) Primary hemophagocytic syndromes point to a direct link between lymphocyte cytotoxicity and homeostasis. Immunological Reviews 203, 165–179. Mond, J.J., Lees, A. and Snapper, C.M. (1995) T cell-independent antigens type 2. Annual Review of Immunology 13, 655–692. Morgan, G.A., Laufman, H.B., Otieno-Omondi, F.P. and Black, S.J. (1993) Control of G1 to S cell cycle progres- sion of Trypanosoma brucei S427cl1 organisms under axenic conditions. Molecular and Biochemical Parasitology 57, 241–252. Morgan, G.A., Hamilton, E.A. and Black, S.J. (1996) The requirements for G1 checkpoint progression of Trypanosoma brucei S 427 clone 1. Molecular and Biochemical Parasitology 78, 195–207. Morrison, W.I., Roelants, G.E., Mayor-Withey, K.S. and Murray, M. (1978) Susceptibility of inbred strains of mice to Trypanosoma congolense: correlation with changes in spleen lymphocyte populations. Clinical & Experimental Immunology 32, 25–40. Morrison, W.I., Wells, P.W., Moloo, S.K., Paris, J. and Murray, M. (1982) Interference in the establishment of superinfections with Trypanosoma congolense in cattle. Journal of Parasitology 68, 755–764. Motloang, M.Y., Masumu, J., Mans, B.J. and Latif, A.A. (2014) Virulence of Trypanosoma congolense strains isolated from cattle and African buffaloes (Syncerus caffer) in KwaZulu-Natal, South Africa. Onderstepoort Journal of Veterinary Research 81, e1–e7. Mulligan, H.W. and Potts, W.H. (1970) The African Trypanosomiases. Wiley, London. Murphy, N.B. and Welburn, S.C. (1997) Programmed cell death in procyclic Trypanosoma brucei rhode- siense is associated with differential expression of mRNAs. Cell Death & Differentiation 4, 365–370. Murray, M. and Dexter, T.M. (1988) Anaemia in bovine African trypanosomiasis. A review. Acta Tropica 45, 389–432. Murray, M., Murray, P.K. and McIntyre, W.I. (1977a) An improved parasitological technique for the diagnosis of African trypanosomiasis. Transactions of the Royal Society of Tropical Medicine and Hygiene 71, 325–326. Murray, P.K., Murray, M., Wallace, M., Morrison, W.I. and McIntyre, W.I.M. (1977b) Trypanosomiasis in Nda- ma and Zebu Cattle. I. An experimental investigation of susceptibility to Trypanosoma brucei, T. con- golense and mixed infections. In: Report of the 15th Meeting of the Internations Scientific Counil for Trypanosomiasis research and Control, Banjul, The Gambia, 1977. OAU/STRC, Nairobi, pp. 470–481. Murray, P.K., Murray, M., Wallace, M., Morrison, W.I. and McIntyre, W.I.M. (1977c) Trypanosomiasis in Nda- ma and Zebu cattle. II. The influence of weight of infection on the severity of the diseases. In: Report of the 15th Meeting of the Internations Scientific Counil for Trypanosomiasis research and Control, Banjul, The Gambia, 1977. OAU/STRC, Nairobi, pp. 482–487. Murray, M., Morrison, W.I. and Whitelaw, D.D. (1982) Host susceptibility to African trypanosomiasis: trypanotolerance. Advances in Parasitology 21, 1–68. Murray, M., Trail, J. and Grootenhuis, J. (1984) Trypanotolerant livestock: potential and future exploitation. Outlook on Agriculture 13, 43–51. Murray, M., Trail, J.C. and D’Ieteren, G.D. (1990) Trypanotolerance in cattle and prospects for the control of trypanosomiasis by selective breeding. Revue Scientifique et Technique 9, 369–386. Musoke, A.J. and Barbet, A.F. (1977) Activation of complement by variant-specific surface antigen of Trypanosoma brucei. Nature 270, 438–440. Musoke, A.J., Nantulya, V.M., Barbet, A.F., Kironde, F. and McGuire, T.C. (1981) Bovine immune response to African trypanosomes: specific antibodies to variable surface glycoproteins of Trypanosoma brucei. Parasite Immunology 3, 97–106. Muyldermans, S., Baral, T.N., Retamozzo, V.C., de Baetselier, P., de Genst, E., et al. (2009) Camelid immuno- globulins and nanobody technology. Veterinary Immunology and Immunopathology 128, 178–183. Naessens, J. (2006) Bovine trypanotolerance: a natural ability to prevent severe anaemia and haemo- phagocytic syndrome? International Journal for Parasitology 36, 521–528. Control of Pathogenesis in Animal African Trypanosomiasis 143 Naessens, J. and Williams, D.J. (1992) Characterization and measurement of CD5+ B cells in normal and Trypanosoma congolense-infected cattle. European Journal of Immunology 22, 1713–1718. Naessens, J., Teale, A.J. and Sileghem, M. (2002) Identification of mechanisms of natural resistance to African trypanosomiasis in cattle. Veterinary Immunology and Immunopathology 87, 187–194. Naessens, J., Leak, S.G., Kennedy, D.J., Kemp, S.J. and Teale, A.J. (2003a) Responses of bovine chimae- ras combining trypanosomosis resistant and susceptible genotypes to experimental infection with Trypanosoma congolense. Veterinary Parasitology 111, 125–142. Naessens, J., Mwangi, D.M., Buza, J. and Moloo, S.K. (2003b) Local skin reaction (chancre) induced fol- lowing inoculation of metacyclic trypanosomes in cattle by tsetse flies is dependent on CD4 T lympho- cytes. Parasite Immunology 25, 413–419. Naessens, J., Kitani, H., Nakamura, Y., Yagi, Y., Sekikawa, K. and Iraqi, F. (2005) TNF-α mediates the develop- ment of anaemia in a murine Trypanosoma brucei rhodesiense infection, but not the anaemia associated with a murine Trypanosoma congolense infection. Clinical & Experimental Immunology 139, 405–410. Nairz, M., Sonnweber, T., Schroll, A., Theurl, I. and Weiss, G. (2012) The pleiotropic effects of erythropoietin in infection and inflammation. Microbes and Infection 14, 238–246. Nantulya, V.M. (1988) Immunodiagnosis of rhodesiense sleeping sickness: detection of circulating trypano- somal antigens in sera and cerebrospinal fluid by enzyme immunoassay using a monoclonal antibody. Bulletin de la Société de Pathologie Exotique et de ses Filiales 81, 511–512. Nantulya, V.M. (1994) Suratex: a simple latex agglutination antigen test for diagnosis of Trypanosoma evan- si infections (surra). Tropical Medicine and Parasitology 45, 9–12. Nantulya, V.M. (1997) TrypTect CIATT – a card indirect agglutination trypanosomiasis test for diagnosis of Trypanosoma brucei gambiense and T. b. rhodesiense infections. Transactions of the Royal Society of Tropical Medicine and Hygiene 91, 551–553. Nantulya, V.M. and Lindqvist, K.J. (1989) Antigen-detection enzyme immunoassays for the diagnosis of Trypanosoma vivax, T. congolense and T. brucei infections in cattle. Tropical Medicine and Parasit- ology 40, 267–272. Nantulya, V.M., Musoke, A.J., Barbet, A.F. and Roelants, G.E. (1979) Evidence for reappearance of Trypa- nosoma brucei variable antigen types in relapse populations. Journal of Parasitology 65, 673–679. Nantulya, V.M., Doyle, J.J. and Jenni, L. (1980) Studies on Trypanosoma (nannomonas) congolense IV. Experimental immunization of mice against tsetse fly challenge. Parasitology 80, 133–137. Nantulya, V.M., Musoke, A.J., Moloo, S.K. and Ngaira, J.M. (1983) Analysis of the variable antigen composition of Trypanosoma brucei brucei metacyclic trypanosomes using monoclonal antibodies. Acta Tropica 40, 19–24. Nantulya, V.M., Musoke, A.J., Rurangirwa, F.R. and Moloo, S.K. (1984) Resistance of cattle to tsetse-trans- mitted challenge with Trypanosoma brucei or Trypanosoma congolense after spontaneous recovery from syringe-passaged infections. Infection and Immunity 43, 735–738. Nantulya, V.M., Musoke, A.J. and Moloo, S.K. (1986) Apparent exhaustion of the variable antigen reper- toires of Trypanosoma vivax in infected cattle. Infection and Immunity 54, 444–447. Nantulya, V.M., Musoke, A.J., Rurangirwa, F.R., Saigar, N. and Minja, S.H. (1987) Monoclonal antibodies that distinguish Trypanosoma congolense, T. vivax and T. brucei. Parasite Immunology 9, 421–431. Nantulya, V.M., Lindqvist, K.J., Diall, O. and Olaho, M. (1989) Two simple antigen-detection enzyme im- munoassays for the diagnosis of Trypanosoma evansi infections in the dromedary camel (Camelus dromedarius). Tropical Medicine and Parasitology 40, 415–418. Nantulya, V.M., Doua, F. and Molisho, S. (1992a) Diagnosis of Trypanosoma brucei gambiense sleeping sickness using an antigen detection enzyme-linked immunosorbent assay. Transactions of the Royal Society of Tropical Medicine and Hygiene 86, 42–45. Nantulya, V.M., Lindqvist, K.J., Stevenson, P. and Mwangi, E.K. (1992b) Application of a monoclonal anti- body-based antigen detection enzyme-linked immunosorbent assay (antigen ELISA) for field diagnosis of bovine trypanosomiasis at Nguruman, Kenya. Annals of Tropical Medicine and Parasitology 86, 225–230. Nganga, J.K., Soller, M. and Iraqi, F.A. (2010) High resolution mapping of trypanosomosis resistance loci Tir2 and Tir3 using F12 advanced intercross lines with major locus Tir1 fixed for the susceptible allele. BMC Genomics 11, 394. Nok, A.J. and Balogun, E.O. (2003) A bloodstream Trypanosoma congolense sialidase could be involved in anemia during experimental trypanosomiasis. Journal of Biochemistry 133, 725–730. Nolan, D.P., Geuskens, M. and Pays, E. (1999) N-linked glycans containing linear poly-N-acetyllactosamine as sorting signals in endocytosis in Trypanosoma brucei. Current Biology 9, 1169–1172. 144 S.J. Black Noyes, H., Brass, A., Obara, I., Anderson, S., Archibald, A.L., et al. (2011) Genetic and expression analysis of cattle identifies candidate genes in pathways responding to Trypanosoma congolense infection. Proceedings of the National Academy of Sciences USA 108, 9304–9309. O’Gorman, G.M., Park, S.D., Hill, E.W., Meade, K.G., Mitchell, L.C., et al. (2006) Cytokine mRNA profiling of peripheral blood mononuclear cells from trypanotolerant and trypanosusceptible cattle infected with Trypanosoma congolense. Physiological Genomics 28, 53–61. ole-MoiYoi, O.K. (1987) Trypanosome species-specific DNA probes to detect infection in tsetse flies. Para- sitology Today 3, 371–374. Opperdoes, F.R. and Borst, P. (1977) Localization of nine glycolytic enzymes in a microbody-like organelle in Trypanosoma brucei: the glycosome. FEBS Letters 80, 360–364. Orenge, C., Munga, L., Kimwele, C., Kemp, S., Korol, A., et al. (2011) Expression of trypanotolerance in N’Dama × Boran crosses under field challenge in relation to N’Dama genome content. BMC Proceed- ings 5 (Suppl. 4), S23. Orenge, C.O., Munga, L., Kimwele, C.N., Kemp, S., Korol, A., et al. (2012) Trypanotolerance in N’Dama × Boran crosses under natural trypanosome challenge: effect of test-year environment, gender, and breed composition. BMC Genetics 13, 87. Paling, R.W., Moloo, S.K., Scott, J.R., Gettinby, G., McOdimba, F.A. and Murray, M. (1991a) Susceptibility of N’Dama and Boran cattle to sequential challenges with tsetse-transmitted clones of Trypanosoma congolense. Parasite Immunology 13, 427–445. Paling, R.W., Moloo, S.K., Scott, J.R., McOdimba, F.A., Logan-Henfrey, L.L., et al. (1991b) Susceptibility of N’Dama and Boran cattle to tsetse-transmitted primary and rechallenge infections with a homologous serodeme of Trypanosoma congolense. Parasite Immunology 13, 413–425. Pearson, T.W., Roelants, G.E., Pinder, M., Lundin, L.B. and Mayor-Withey, K.S. (1979) Immune depression in trypanosome-infected mice. III. suppressor cells. European Journal of Immunology 9, 200–204. Pearson, T.W., Pinder, M., Roelants, G.E., Kar, S.K., Lundin, L.B., et al. (1980) Methods for derivation and detection of anti-parasite monoclonal antibodies. Journal of Immunological Methods 34, 141–154. Peregrine, A.S. and Mamman, M. (1994) A simple method for repeated sampling of lumbar cerebrospinal fluid in goats. Laboratory Animals 28, 391–396. Peregrine, A.S., Moloo, S.K. and Whitelaw, D.D. (1987) Therapeutic and prophylactic activity of isometamid- ium chloride in Boran cattle against Trypanosoma vivax transmitted by Glossina morsitans centralis. Research in Veterinary Science 43, 268–270. Peregrine, A.S., Moloo, S.K. and Whitelaw, D.D. (1991) Differences in sensitivity of Kenyan Trypanosoma vivax populations to the prophylactic and therapeutic actions of isometamidium chloride in Boran cat- tle. Tropical Animal Health and Production 23, 29–38. Rainard, P., Riollet, C., Poutrel, B. and Paape, M.J. (2000) Phagocytosis and killing of Staphylococcus aur- eus by bovine neutrophils after priming by tumor necrosis factor-α and the des-arginine derivative of C5a. American Journal of Veterinary Research 61, 951–959. Rathkolb, B., Noyes, H.A., Brass, A., Dark, P., Fuchs, H., et al. (2009) Clinical chemistry of congenic mice with quantitative trait loci for predicted responses to Trypanosoma congolense infection. Infection and Immunity 77, 3948–3957. Rifkin, M.R. and Landsberger, F.R. (1990) Trypanosome variant surface glycoprotein transfer to target membranes: a model for the pathogenesis of trypanosomiasis. Proceedings of the National Academy of Sciences USA 87, 801–805. Rovis, L. and Dube, D.K. (1981) Studies on the biosynthesis of the variant surface glycoproteins of Trypa- nosoma brucei: sequence of glycosylation. Molecular and Biochemical Parasitology 4, 77–93. Rovis, L., Barbet, A.F. and Williams, R.O. (1978) Characterisation of the surface coat of Trypanosoma con- golense. Nature 271, 654–656. Rovis, L., Musoke, A.J. and Moloo, S.K. (1984) Failure of trypanosomal membrane antigens to induce protection against tsetse-transmitted Trypanosoma vivax or T. brucei in goats and rabbits. Acta Tropica 41, 227–236. Rurangirwa, F.R., Tabel, H., Losos, G., Masiga, W.N. and Mwambu, P. (1978) Immunosuppressive effect of Trypanosoma congolense and Trypanosoma vivax on the secondary immune response of cattle to Mycoplasma mycoides subsp. mycoides. Research in Veterinary Science 25, 395–397. Rurangirwa, F.R., Tabel, H., Losos, G.J. and Tizard, I.R. (1979) Suppression of antibody response to Leptospi- ra biflexa and Brucella abortus and recovery from immunosuppression after Berenil treatment. Infection and Immunity 26, 822–826. Rurangirwa, F.R., Tabel, H., Losos, G.J. and Tizard, I.R. (1980) Immunosuppression in bovine trypanosom- iasis. The establishment of ‘memory’ in cattle infected with T. congolense and the effect of post infection Control of Pathogenesis in Animal African Trypanosomiasis 145 serum on in vitro (3H)-thymidine uptake by lymphocytes and on leucocyte migration. Tropenmedizin und Parasitologie 31, 105–110. Rurangirwa, F.R., Musoke, A.J., Nantulya, V.M. and Tabel, H. (1983) Immune depression in bovine trypano- somiasis: effects of acute and chronic Trypanosoma congolense and chronic Trypanosoma vivax in- fections on antibody response to Brucella abortus vaccine. Parasite Immunology 5, 267–276. Schleifer, K.W. and Mansfield, J.M. (1993) Suppressor macrophages in African trypanosomiasis inhibit T cell proliferative responses by nitric oxide and prostaglandins. Journal of Immunology 151, 5492–5503. Scoones, I. (2014) The politics of trypanosomiasis control in Africa. STEPS Working Paper 57. STEPS Cen- ter, Brighton, UK. Shapiro, S.Z. (1994) Failure of immunization with trypanosome endocytotic vesicle membrane proteins to pro- vide nonvariant immunoprotection against Trypanosoma brucei. Parasitology Research 80, 240–244. Shapiro, S.Z. and Murray, M. (1982) African trypanosome antigens recognized during the course of infec- tion in N’dama and Zebu cattle. Infection and Immunity 35, 410–416. Shapiro, S.Z. and Webster, P. (1989) Coated vesicles from the protozoan parasite Trypanosoma brucei: purification and characterization. Journal of Protozoology 36, 344–349. Shapiro, S.Z. and Young, J.R. (1981) An immunochemical method for mRNA purification. Application to messenger RNA encoding trypanosome variable surface antigen. Journal of Biological Chemistry 256, 1495–1498. Shapiro, S.Z., Naessens, J., Liesegang, B., Moloo, S.K. and Magondu, J. (1984) Analysis by flow cytometry of DNA synthesis during the life cycle of African trypanosomes. Acta Tropica 41, 313–323. Sharpe, R.T., Langley, A.M., Mowat, G.N., Macaskill, J.A. and Holmes, P.H. (1982) Immunosuppression in bovine trypanosomiasis: response of cattle infected with Trypanosoma congolense to foot-and-mouth disease vaccination and subsequent live virus challenge. Research in Veterinary Science 32, 289–293. Shaw, A.P. (2009) Assessing the economics of animal trypanosomosis in Africa – history and current perspectives. Onderstepoort Journal of Veterinary Research 76, 27–32. Shugaba, A., Umar, I., Omage, J., Ibrahim, N.D., Andrews, J., et al. (1994) Biochemical differences (O-acetyl and glycolyl groups) in erythrocyte surface sialic acids of trypanotolerant N’dama and trypanosuscep- tible Zebu cattle. Journal of Comparative Pathology 110, 91–95. Silayo, R.S., Mamman, M., Moloo, S.K., Aliu, Y.O., Gray, M.A. and Peregrine, A.S. (1992) Response of Trypanosoma congolense in goats to single and double treatment with diminazene aceturate. Research in Veterinary Science 53, 98–105. Sileghem, M. and Flynn, J.N. (1992a) Suppression of interleukin 2 secretion and interleukin 2 receptor expression during tsetse-transmitted trypanosomiasis in cattle. European Journal of Immunology 22, 767–773. Sileghem, M. and Flynn, J.N. (1992b) Suppression of T-cell responsiveness during tsetse-transmitted trypanosomiasis in cattle. Scandinavian Journal of Immunology 11, 37–40. Sileghem, M. and Naessens, J. (1995) Are CD8 T cells involved in control of African trypanosomiasis in a natural host environment? European Journal of Immunology 25, 1965–1971. Sileghem, M., Darji, A. and de Baetselier, P. (1991) In vitro simulation of immunosuppression caused by Trypanosoma brucei. Immunology 73, 246–248. Sileghem, M., Flynn, J.N., Logan-Henfrey, L. and Ellis, J. (1994) Tumour necrosis factor production by monocytes from cattle infected with Trypanosoma (Duttonella) vivax and Trypanosoma (Nannomonas) congolense: possible association with severity of anaemia associated with the disease. Parasite Immunology 16, 51–54. Sternberg, M.J. and Mabbott, N.A. (1996) Nitric oxide-mediated suppression of T cell responses during Trypanosoma brucei infection: soluble trypanosome products and interferon-α are synergistic inducers of nitric oxide synthase. European Journal of Immunology 26, 539–543. Steverding, D., Stierhof, Y.D., Fuchs, H., Tauber, R. and Overath, P. (1995) Transferrin-binding protein complex is the receptor for transferrin uptake in Trypanosoma brucei. Journal of Cell Biology 131, 1173–1182. Stijlemans, B., Cnops, J., Naniima, P., Vaast, A., Bockstal, V., et al. (2015) Development of a pHrodo-based assay for the assessment of in vitro and in vivo erythrophagocytosis during experimental trypanoso- mosis. PLoS Neglected Tropical Diseases 9, e0003561. Stijlemans, B., Brys, L., Korf, H., Bieniasz-Krzywiec, P., Sparkes, A., et al. (2016) MIF-mediated hemodilu- tion promotes pathogenic anemia in experimental African trypanosomosis. PLoS Pathogens 12, e1005862. Suliman, H.B., Logan-Henfrey, L., Majiwa, P.A., ole-MoiYoi, O. and Feldman, B.F. (1999) Analysis of erythro- poietin and erythropoietin receptor genes expression in cattle during acute infection with Trypanosoma congolense. Experimental Hematology 27, 37–45. 146 S.J. Black Sunter, J., Webb, H. and Carrington, M. (2013) Determinants of GPI-PLC localisation to the flagellum and access to GPI-anchored substrates in trypanosomes. PLoS Pathogens 9, e1003566. Sutherland, I.A., Moloo, S.K., Holmes, P.H. and Peregrine, A.S. (1991) Therapeutic and prophylactic activity of isometamidium chloride against a tsetse-transmitted drug-resistant clone of Trypanosoma congo- lense in Boran cattle. Acta Tropica 49, 57–64. Tabel, H., Wei, G. and Bull, H.J. (2013) Immunosuppression: cause for failures of vaccines against African Trypanosomiases. PLoS Neglected Tropical Diseases 7, e2090. Taiwo, V.O. and Anosa, V.O. (2000) In vitro erythrophagocytosis by cultured macrophages stimulated with extraneous substances and those isolated from the blood, spleen and bone marrow of Boran and N’Dama cattle infected with Trypanosoma congolense and Trypanosoma vivax. Onderstepoort Jour- nal of Veterinary Research 67, 273–287. Taiwo, V.O., Nantulya, V.M., Moloo, S.K. and Ikede, B.O. (1990) Role of the chancre in induction of immun- ity to tsetse-transmitted Trypanosoma (Nannomonas) congolense in goats. Veterinary Immunology and Immunopathology 26, 59–70. Taylor, K., Lutje, V. and Mertens, B. (1996a) Nitric oxide synthesis is depressed in Bos indicus cattle in- fected with Trypanosoma congolense and Trypanosoma vivax and does not mediate T-cell suppres- sion. Infection and Immunity 64, 4115–4122. Taylor, K.A., Lutje, V., Kennedy, D., Authie, E., Boulange, A., et al. (1996b) Trypanosoma congolense: B-lymphocyte responses differ between trypanotolerant and trypanosusceptible cattle. Experimental Parasitology 83, 106–116. Taylor, K., Mertens, B., Lutje, V. and Saya, R. (1998) Trypanosoma congolense infection of trypanotolerant N’Dama (Bos taurus) cattle is associated with decreased secretion of nitric oxide by interferon-α-acti- vated monocytes and increased transcription of interleukin-10. Parasite Immunology 20, 421–429. Trail, J., Murray, M., Sones, K., Jibbo, J., Durkin, J. and Light, D. (1985) Boran cattle maintained by chemo- prophylaxis under trypanosomiasis risk. Journal of Agricultural Science 105, 147–166. Trail, J.C., d’Ieteren, G.D. and Teale, A.J. (1989) Trypanotolerance and the value of conserving livestock genetic resources. Genome 31, 805–812. Trail, J.C., d’Ieteren, G.D., Colardelle, C., Maille, J.C., Ordner, G., et al. (1990) Evaluation of a field test for trypanotolerance in young N’Dama cattle. Acta Tropica 48, 47–57. Trail, J., d’Ieteren, G.D., Feron, A., Kakiese, O., NMulungo, M. and Pelo, M. (1991a) Effect of trypanosome infection, control of parasitemia and control of anemia development in productivity of N’Dama cattle. Acta Tropica 48, 37–45. Trail, J.C., d’Ieteren, G.D., Maille, J.C., Yangari, G. and Nantulya, V.M. (1991b) Use of antigen-detection enzyme immunoassays in assessment of trypanotolerance in N’Dama cattle. Acta Tropica 50, 11–18. Trail, J.C., d’Ieteren, G.D., Viviani, P., Yangari, G. and Nantulya, V.M. (1992) Relationships between trypano- some infection measured by antigen detection enzyme immunoassays, anaemia and growth in trypa- notolerant N’Dama cattle. Veterinary Parasitology 42, 213–223. Trail, J.C., d’Ieteren, G.D., Murray, M., Ordner, G., Yangari, G., et al. (1993) Measurement of trypanotoler- ance criteria and their effect on reproductive performance of N’Dama cattle. Veterinary Parasitology 45, 241–255. Trail, J.C., Wissocq, N., d’Ieteren, G.D., Kakiese, O. and Murray, M. (1994) Quantitative phenotyping of N’Dama cattle for aspects of trypanotolerance under field tsetse challenge. Veterinary Parasitology 55, 185–195. Turner, C.M. and Barry, J.D. (1989) High frequency of antigenic variation in Trypanosoma brucei rhode- siense infections. Parasitology 99, 67–75. Ullmann, A.J., Lima, C.M.R., Guerrero, F.D., Piesman, J. and Black, W.C. (2005) Genome size and organ- ization in the black-legged tick, Ixodes scapularis, and the southern cattle tick, Boophilus microplus. Insect Molecular Biology 14, 217–222. van den Bossche, P., Chitanga, S., Masumu, J., Marcotty, T. and Delespaux, V. (2011) Virulence in Trypano- soma congolense Savannah subgroup. A comparison between strains and transmission cycles. Para- site Immunology 33, 456–460. Vandeweerd, V. and Black, S.J. (1989) Serum lipoprotein and Trypanosoma brucei brucei interactions in vitro. Molecular and Biochemical Parasitology 37, 201–211. Vos, G.J. and Gardiner, P.R. (1990) Parasite-specific antibody responses of ruminants infected with Trypanosoma vivax. Parasitology 100, 93–100. Vos, G.J., Moloo, S.K., Nelson, R.T. and Gardiner, P.R. (1988) Attempts to protect goats against challenge with Trypanosoma vivax by initiation of primary infections with large numbers of metacyclic trypano- somes. Parasitology 97, 383–392. Control of Pathogenesis in Animal African Trypanosomiasis 147 Vos, Q., Lees, A., Wu, Z.Q., Snapper, C.M. and Mond, J.J. (2000) B-cell activation by T-cell-independent type 2 antigens as an integral part of the humoral immune response to pathogenic microorganisms. Immunological Reviews 176, 154–170. Waithanji, E.M., Nantulya, V.M. and Mbiuki, S.M. (1993) Use of antigen capture tube enzyme-linked immunosorbent assay for the diagnosis of Trypanosoma evansi infections in dromedary camels (Camelus dromedarius). Revue Scientifique et Technique 12, 665–672. Waitumbi, J.N., Murphy, N.B. and Peregrine, A.S. (1994) Genotype and drug-resistance phenotype of Tryp- anosoma evansi isolated from camels in northern Kenya. Annals of Tropical Medicine and Parasitology 88, 677–683. Webb, H., Carnall, N. and Carrington, M. (1994) The role of GPI-PLC in Trypanosoma brucei. Brazilian Journal of Medical and Biological Research 27, 349–356. Webb, H., Carnall, N., Vanhamme, L., Rolin, S., van den Abbeele, J., et al. (1997) The GPI-phospholipase C of Trypanosoma brucei is nonessential but influences parasitemia in mice. Journal of Cell Biology 139, 103–114. Webster, P. (1989) Endocytosis by African trypanosomes. I. Three-dimensional structure of the endocytic organelles in Trypanosoma brucei and T. congolense. European Journal of Cell Biology 49, 295–302. Webster, P. and Fish, W.R. (1989) Endocytosis by African trypanosomes. II. Occurrence in different life-cycle stages and intracellular sorting. European Journal of Cell Biology 49, 303–310. Welburn, S.C. and Murphy, N.B. (1998) Prohibitin and RACK homologues are up-regulated in trypano- somes induced to undergo apoptosis and in naturally occurring terminally differentiated forms. Cell Death & Differentiation 5, 615–622. Welburn, S.C., Lillico, S. and Murphy, N.B. (1999) Programmed cell death in procyclic form Trypanosoma brucei rhodesiense – identification of differentially expressed genes during con A induced death. Memórias do Instituto Oswaldo Cruz 94, 229–234. Whitelaw, D.D., Scott, J.M., Reid, H.W., Holmes, P.H., Jennings, F.W. and Urquhart, G.M. (1979) Immuno- suppression in bovine trypanosomiasis: studies with louping-ill vaccine. Research in Veterinary Sci- ence 26, 102–107. Williams, D.J., Naessens, J., Scott, J.R. and McOdimba, F.A. (1991) Analysis of peripheral leucocyte popu- lations in N’Dama and Boran cattle following a rechallenge infection with Trypanosoma congolense. Parasite Immunology 13, 171–185. Williams, D.J., Logan-Henfrey, L.L., Authie, E., Seely, C. and McOdimba, F. (1992) Experimental infection with a haemorrhage-causing Trypanosoma vivax in N’Dama and Boran cattle. Scandinavian Journal of Immunology Suppl 11, 34–36. Williams, D.J., Taylor, K., Newson, J., Gichuki, B. and Naessens, J. (1996) The role of anti-variable surface glycoprotein antibody responses in bovine trypanotolerance. Parasite Immunology 18, 209–218. Williams, R.O., Marcu, K.B., Young, J.R., Rovis, L. and Williams, S.C. (1978) A characterization of mRNA activites and their sequence complexities in Trypanosoma brucei: partial purification and properties of the VSSA mRNA. Nucleic Acids Research 5, 3171–3182. Young, J.R., Miller, E.N., Williams, R.O. and Turner, M.J. (1983a) Are there two classes of VSG gene in Trypanosoma brucei? Nature 306, 196–198. Young, J.R., Shah, J.S., Matthyssens, G. and Williams, R.O. (1983b) Relationship between multiple copies of a T. brucei variable surface glycoprotein gene whose expression is not controlled by duplication. Cell 32, 1149–1159. 3 Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia: ILRI’s Role in the Field Delia Grace1, Ekta Patel2 and Thomas Fitz Randolph2 1International Livestock Research Institute, Nairobi, Kenya and University of Greenwich, UK; 2International Livestock Research Institute, Nairobi, Kenya Contents Introduction 148 Field Work on African Animal Trypanosomiasis 149 Trypanotolerance 149 Community-based tsetse control 151 Trypanocide resistance 151 ILRI impacts on AAT in Kenya 152 ILRI impacts in Ethiopia 152 ILRI impacts in Uganda 154 ILRI impacts in West Africa 154 AAT economic and environmental studies 158 Other AAT research and impacts 159 Conclusions and the Future 160 References 161 Introduction 10–40% (Swallow, 2000). Estimated total losses due to trypanosomiasis range widely depending African animal trypanosomiasis (AAT) occurs on the methods used, assumptions made and where the tsetse fly vector exists in sub-Saharan types of losses estimated (ILRAD, 1994; Budd, Africa, between the latitudes 15°N and 29°S 1999; Kristjanson et al., 1999; Swallow, 2000). (Randolph et al., 2003). It affects ruminants, The upper range of published estimates would camels, horses and pigs, and constrains cattle make annual losses from trypanosomiasis equal production over an area variously estimated to to one-third of the estimated livestock gross be between 8 and 10 million km2 in Africa. domestic product in sub-Saharan Africa. Around 67 million cattle live in tsetse-infested The disease is complex, involving three spe- areas out of a total of 253 million in Africa1. In cies of parasite, namely Trypanosoma congolense, some situations, around one in 20 of the cattle T. vivax and T. brucei. T. vivax can also be trans- at risk die each year, and the productivity of the mitted mechanically by biting flies, and thus is survivors in terms of draught power, milk pro- also found in parts of Africa free or cleared of duction, growth and birth rate are lowered by tsetse and in parts of Central and South America. © International Livestock Research Institute 2020. The Impact of the International 148 Livestock Research Institute (eds J. McIntire and D. Grace) Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 149 Two related parasites, T. brucei subsp. gambiense like other government services, became increas- and T. brucei rhodesiense, cause human African ingly underfunded and dysfunctional, and donors trypanosomiasis (HAT), also known as sleeping were less willing to pay for large-scale, long-term sickness. While HAT is generally considered a control. Moreover, control efforts before the disease of people, livestock and wildlife act as 1980s usually relied on ground and aerial spray- reservoirs for T. b. rhodienense and possibly for ing of insecticide, which was increasingly seen as T. b. gambiense, complicating the control of HAT. expensive in economic and environmental terms. Knowledge of the biology of tsetse and trypano- One study (Meyer et al., 2016) argues that tsetse somiasis was greatly advanced by Mulligan and control operations covered barely 1.3% of the Potts (1970) and by Maudlin et al. (2004). estimated infested area in sub-Saharan Africa. Trypanosomes have the ability to undergo As a result of this decline in the funding antigenic variation (Cross, 1975), enabling available for control, an upsurge of AAT cases host invasion and allowing them to establish occurred during the post-independence period. persistent infections. In addition, each species In some countries, this was exacerbated by the comprises an unknown number of strains, all movement of people and livestock into formerly capable of elaborating a different repertoire of sparsely inhabited tsetse-infested areas. AAT variable antigen types. Hence, no vaccine is cur- incidence increased rapidly in the 1980s and rently available (the problem of antigenic vari- 1990s, and cattle herds almost disappeared in ation and the history of vaccine development some areas, such as the Ghibe Valley of Ethiopia are discussed in Chapter 2, this volume). and the Yale agropastoral zone of Burkina Faso. The initial mission of the International The 1980s also saw a major epidemic of HAT in Laboratory for Research on Animal Disease Sudan and Uganda, which spread into Kenya. As (ILRAD) was to tackle AAT and East Coast fever, a result of these upsurges in human and animal two of the most serious intractable African live- disease, there was renewed interest in the field stock diseases. As a result, a large body of re- control of AAT, several new initiatives were es- search on AAT was conducted over 30  years: tablished involving ILRAD and the International genetics, breeding and immunology research Livestock Centre for Africa (ILCA). are discussed in Chapters 1, 2 and 4 of this vol- Apart from efforts to understand and man- ume, respectively. This chapter reviews the earl- age AAT, ILRI focused on three major areas: ier field work of ILRAD followed by that of (i) the promotion of trypanotolerant cattle; International Livestock Research Institute (ILRI) (ii) community-based control of tsetse in Ethiopia after 1994 in East and West Africa, including and later Burkina Faso; and (iii) managing the engagement of those institutions with re- resistance to trypanocides. ILRI did not engage gional and global initiatives. with the attempt to eradicate tsetse from Africa, which scientists believed was infeasible. The areas of ILRI activity are first briefly outlined, and the rest of the chapter focuses on the work Field Work on African Animal done and the impacts achieved in different coun- Trypanosomiasis tries. Table 3.1 summarizes ILCA/ILRI research since about 1990 on field aspects of trypano- AAT has long been regarded as the single most somiasis control. important cattle disease in Africa, and between 1905 and 1960 it commanded the attention of five imperial powers, dedicated government ser- Trypanotolerance vices, and national, regional and international research institutions. It has been estimated that The challenges of tsetse control and the ongoing 25% of colonial research spending went to tryp- cost of trypanosomiasis control by veterinary anosomiasis (Rogers and Randolph, 2002). drugs were recognized early in the history of ILRI. However, the final decade of the last century saw One option that gained attention early on was to practical control of trypanosomiasis. With the exploit the genetic tolerance to trypanosomiasis onset of the Great African Depression in the 1980s infection found in several African cattle breeds, (Leonard and Straus, 2003), tsetse programmes, such as the taurine N’Dama and West African 150 D. Grace, E. Patel and T. Randolph Table 3.1. Research on tsetse and trypanosomiasis control in West Africa and East Africa since 1990 by ILRI and partners. (Data from ILCA and ILRI annual reports, various years.) Tsetse reduction Reduction of Institution Location Period Objectives Interventions from baseline AAT prevalence Difficulties Sustainability ILCA Ghibe River 1990–1992 Vector control ITC, ITT, TRY 74–81% 85% Theft of traps, Unsustained. Valley, depending on socio-political Tsetse Ethiopia species disturbances reinvasion, second phase from 1993 mixed results ILCA Ghibe River 1991–1993 Vector control ITC, TRY 0–93% 60% Slow decline in Sustained in the Valley, depending tsetse densities short-term. Ethiopia on species Control was in place 5 years later. Full cost recovery from farmers ILRI evaluation Burkina Faso 1990–2000 Vector control ITC, ITT 98.4% 80% Farmers could not Epidemics on CIRDES meet costs to controlled; projects continue control however, tsetse reinvasion and return of AAT after campaign ILRI evaluation Kenya 2002–2006 Vector control ITC Not measured 22% Low effectiveness No business case of ICIPE for innovation project ILRI Burkina Faso, 2002–2006 Integrated ITC, ITT, TRY, Significant Significant Communal action Only TRY proved Mali, control TTC difficult; disabling sustainable Guinea policy CIRDES, Centre International de Recherche-Développement sur l’Elevage en zone Subhumide; ICIPE, International Centre of Insect Physiology and Ecology; ITC, insecticide-treated cattle; ITT, insecticide-treated traps and targets; TRY, trypanocidal drugs; TTC, trypanotolerant cattle. Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 151 Shorthorn breeds of West and Central Africa. large larva and have only up to 12 offspring in a These breeds had evolved over millennia to ac- typical lifetime. Their low reproductive rate makes quire a significant degree of innate resistance to them vulnerable to interventions that cause even trypanosomiasis. Trypanotolerant cattle were small increases in mortality. Historically, area- therefore identified as a promising approach to wide and aerial spraying was successfully used to livestock keeping in tsetse-infested areas (Murray control tsetse. However, the flies reinvaded after et al., 1982). control, and repeated campaigns proved expen- There is evidence that cross-breeding occurs sive and raised environmental concerns. between bos indicus and bos taurus in West and A more targeted approach is the use of Central Africa and that this may enhance cloth traps or targets that are soaked with in- r esistance to trypanosomiasis (Traoré et al., 2017 secticide, with or without baits. In the early for southern Mali). The African Trypanotolerant 1900s, sticky traps worn by plantation workers Livestock Network (ATLN) was established in were successfully deployed on the Island of Prin- 1977 as a joint venture between ILRAD and cipe to eradicate Glossina palpalis. In the mid- ILCA. ILRAD was to investigate animal health, 20th century, great advances were made in the infection status and vector behaviour, while ILCA design of traps and targets, making them highly was responsible for animal production, nutrition effective at reducing tsetse. The live-bait tech- and data processing. This research fell into four nique involves treating cattle with appropriate areas: (i)  the epidemiology of AAT; (ii) the cri- insecticide formulations, usually by means of teria of trypanotolerance; (iii) the genetics of cattle dips, or as pour-on, spot-on or spray-on trypanotolerance; and (iv) the health and prod- veterinary formulations. These are highly effect- uctivity benefits of interventions. Field sites were ive against tsetse and have the additional advan- located in Côte d’Ivoire, Ethiopia, Gabon, Zaire, tage of controlling other flies and cattle ticks. The Gambia and Senegal. Along with the Food The availability of these simple, cheap and effect- and Agriculture Organization of the United ive technologies led to interest in community- Nations (FAO), ILCA surveyed the number and based approaches to tsetse control. distribution of trypanotolerant cattle in 1977 and updated those figures in 1985. Two ILCA/ILRAD monographs established the Trypanocide resistance state of scientific knowledge at the outset of the ATLN (Trail et al., 1979a,b). A long compendium In the absence of a vaccine (Chapter 2, this vol- (ILCA/ILRAD, 1988) summarized the work of the ume), the principal control strategies are redu- first decade of the ATLN, findings that were extended cing or eliminating the tsetse vector, applying by Rowlands and Teale (1994) and Itty (1992). The trypanocides and keeping trypanotolerant stock. work of the ATLN included the following: Less than 1% of the infested area is under vector • Establishment of a research network to im- control and fewer than 20% of the cattle at risk prove livestock production by ensuring op- are trypanotolerant, but around the majority of timal application of existing knowledge and cattle at risk receive trypanocidal drugs making recent research. this the most popular control option and one • Understanding the vectorial potential of that is both effective at scale and sustainable tsetse. without continued external support. Current • Supporting multiplication of herds of tryp- trypanocides have been in use for over 40 years anotolerant animals. and new drugs have not been developed because • Maintaining experimental cattle and small drug development is expensive. One older survey ruminant herds. of drug research and development costs derived • Evaluating the performance of livestock an estimate of ‘nearly’ US$900 million per new under AAT risk. compound across a range of human drugs (DiMasi et al., 2003). The market for trypano- cides in smaller African markets is estimated to Community-based tsetse control be US$20 million per year (Sones, 2001). Hence, it will probably be unprofitable for private firms Tsetse (Glossina spp.) are unusual flies. The fe- to develop new trypanocides for the African mar- males do not lay eggs but instead produce a single, ket. Given this context, threats to the efficacy of 152 D. Grace, E. Patel and T. Randolph the older trypanocides undermine the most efficacy and subsequently cost–benefit of using widely used strategy for control. prophylactic versus curative control. (Prophy- Drug resistance is the heritable loss of sen- lactic drugs are given to prevent the animals sitivity of a microorganism to an antimicrobial from becoming sick; curative drugs are given to to which it had been sensitive. Modern cattle animals when they become sick.) Prophylactic trypanocides were introduced in the 1950s, and treatment was more expensive but more profit- the first cases of resistance were reported in the able, mainly because lactation loss was avoided. next decade. The emergence of resistance fur- When disease fell below a certain level, prophy- ther complicated the breakdown in testse con- laxis was less profitable (Itty et al., 1988). trol that had led to reinvasions of tsetse. For example, ILRI research did contribute ILRAD scientists had a major role in this to the growing interest in using pour-ons for emerging research area. Much of this work was control of tsetse. A study in Kwale found inci- laboratory based, focused on understanding the dence rates of trypanosomiasis in the biweekly- mechanisms of resistance and developing tests treated (28.2%) and monthly-treated (38.6%) and assays for resistance. In addition, drug-re- animals were statistically lower than in the bi- sistant strains isolated from the field were char- monthly (63.9%) and control (72.6%) groups acterized at ILRAD. However, ILRAD and ILCA (Muraguri et al., 2003). Similarly, a study in a scientists were also involved in some of the early trypanosomiasis endemic area in Teso District, studies on field resistance and were among the western Kenya, found that when the tsetse dens- first to assess resistance in Ethiopia, Uganda, ity was very low, control of trypanosomiasis in Kenya, Guinea, Mali, Ghana and other countries. the Orma-Teso Zebu offspring in western Kenya ILRI scientists were also instrumental in required targeting of individual affected animals developing field tests for trypanocide resistance. in the dry seasons (Gachohi et al., 2009). These involved testing cattle for the presence of One of the most important studies in Kenya trypanosomes, treating them with trypanocides was a combined epidemiological and economic and then examining blood at intervals over investigation into a promising novel technology: 98 days to see if the trypanosomes were still pre- a synthetic tsetse repellent (Bett et al., 2010). sent despite treatment (an indication of resist- The evaluation involved 2000 cattle: 1000 head ance). These methods were widely applied by in a control group and another 1000 animals other researchers and subsequently refined by treated with tsetse-repellent dispensers sus- shortening the follow-up period to 28 days. This pended from neck collars. The effectiveness of abbreviated methodology for evaluating the the repellent was monitored for 16 months. The presence of resistance lowers the cost signifi- trial results showed that the treatment reduced cantly so that it can be used more readily by trypanosomiasis infection rates by between 18% national agencies as an initial protocol for and 23% compared with the control levels, indi- screening (Mungube et al., 2012). The abbrevi- cating that the repellent was not ‘a viable alter- ated trypanocide resistance tool has subse- native to existing control techniques’ (Irungu quently been used by other researchers. et al., 2007). ILRI impacts on AAT in Kenya ILRI impacts in Ethiopia Given that ILRAD’s mandate focused on vaccine In the 1970s and 1980s in Ethiopia, there were development, there was relatively little field epi- substantial movements of people and their live- demiology in Kenya. Moreover, East Coast fever stock from the densely populated northern high- was seen by most stakeholders as of higher prior- lands to the scarcely populated, tsetse-infested ity in Kenya. Small studies addressed aspects of south-western region. These livestock were vul- control resulting in useful recommendations, al- nerable to AAT, and losses were originally high. though it is not clear to what extent recommenda- Concern over the situation in Ghibe Valley, tions were taken up or what impact they had. south-west Ethiopia, prompted ILCA to select it An early study, conducted between 1982 as the Ethiopia study site for the ATLN. This net- and 1986 in coastal Kenya, evaluated the work comprised a set of 11 research sites located Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 153 in different trypanosomiasis-risk areas through- poured on them. This method was first tried with out tropical Africa in order to study the complex limited success in the 1940s using dichlorodiphe- interactions that affect trypanotolerance, to pro- nyltrichloroethane (DDT). However, development vide baseline data for livestock development in of pour-on formulations using pyrethroids, a safe tsetse-affected areas and to evaluate different and biodegradable insecticide, led to successful methods for controlling trypanosomiasis (see control of tsetse in trials in Zambia and elsewhere. Map 4, p. xx). In 1991, a control scheme was started in Ghibe Studies on the prevalence of trypanosome Valley using insecticides poured directly on to cat- infections in East African Zebu cattle and the tse- tle. This was considered to be more sustainable, as tse challenge that such animals are exposed to farmers directly benefited and the transaction started in January 1986. An average of 840 East costs in maintaining screens or traps were African Zebu cattle (non-trypanotolerant) from avoided. For the first 2  years, insecticides were around ten herds in the Ghibe Valley were moni- given free of charge. The control was very effect- tored from January 1986 to April 1990. The ive and promoted widely by ILCA, and as a result, studies provided information on the epidemi- many families migrated into the area. ology and transmission dynamics of AAT, which An economic assessment quantified the were subsequently used in disease models and in benefits (Omamo et al., 2002). Following con- planning control. Moreover, they resulted in the trol, the apparent density of tsetse and biting detection of drug resistance for the first time in flies in the region fell by 95%. This reduction in Ethiopia (Rowlands et al., 1993). Although sick tsetse challenge led to a decrease in trypano- animals were treated with trypanocides, the some prevalence in cattle of over 61%, despite a prevalence of AAT increased yearly, raising the high level of resistance to trypanocides. The suspicion of drug resistance, which was con- number of curative drug treatments per animal firmed by laboratory analysis of trypanosome fell by 50%. Mean calf growth rate increased by isolates collected in 1989. 20%, while mean calf mortality and abortion It was decided to test an intervention based decreased by 57%. Average cow body weight on tsetse control using cloth screens impregnated was boosted by 4%, the cow:calf ratio increased with insecticides. This approach had been first by 49% and adult male body weight increased by used in 1914, when tsetse was eradicated from 8%. Between 1995 and 1997, expenditure on the Island of Principe, in the Gulf of Guinea, by trypanocidal drugs fell by US$39,000, which killing the flies using ‘sticky traps’ – wooden more than offset the US$16,000 cost of the boards coated with sticky material strapped to the pour-on. The value of increased output of meat backs of plantation workers. Serious interest in (40%) and milk (30%) implied a benefit:cost the use of vector capture was revived in the 1970s ratio of 11.6 over 2  years and of 9.3 over with the development of more effective cloth 10  years, with increases in annual household traps and screens and the discovery that particu- income between 10% and 34%. lar colours and odours attract tsetse flies. Much of After 2 years of free treatments, a more sus- the foundational research was conducted in tainable model was investigated with the impos- Zimbabwe, with another focus in West Africa. ition of a charge of US$0.6 per treatment. An Both traps and targets (cloths or screens) economic study found that, while many farmers were developed and successfully piloted in South, continued to pay for treatments, there was a re- East and West Africa, but had not been used for duction in demand (from 97% of farmers to the control of trypanosomiasis in cattle in situ- around 60%) and that poorer farmers dropped ations where resistance to trypanocides occurs. out disproportionately (Swallow and Woudy- The trial found that use of targets reduced vector alew, 1994). There were also efforts to hand populations and infections, even in the presence management of the scheme over to the commu- of drug resistance, but did not appear to be sus- nity. By 2000, ILRI was no longer able to con- tainable due to widespread thefts of targets tinue supporting the scheme; it appears that the (Leak et al., 1996). treatments stopped and the tsetse increased Another approach to tsetse control is to turn (Wilson, 2003). In a subsequent project from cattle into mobile targets. The animals are dipped 2008 to 2011, ILRI helped to establish 13 or sprayed with insecticide or the c ompound is trypanosomiasis cooperatives to link private 154 D. Grace, E. Patel and T. Randolph veterinary drug suppliers to the remote commu- 2003 (see Chapter 8, this volume). Surveys nities to ensure a supply of trypanocides to and found that reporting was strongly biased to to reduce dependence on the public supply sys- areas near health centres with good capacity, tem. This promising delivery approach has not and models suggested that the major impact of been systematically evaluated. sleeping sickness was deaths in people who were An analysis of land-use change by ILRI in never treated or reported. A livestock survey 2003 revealed huge increases in human and helped identify hotspots. It linked cattle move- animal populations, as well as land cultivated ment to the introduction of sleeping sickness to in the Ghibe Valley (Wilson, 2003). This analysis new areas and suggested that mass treatment of concluded that the control of tsetse had a role cattle might prevent the spread of sleeping sick- in attracting migrants, but this was difficult to ness. The treatment of 30,000 cattle in Kamuli quantify because other climatic and demographic was highly effective at removing human-infective factors influenced population movements. HAT parasites from the cattle reservoir and con- tributed to a significant decrease in human HAT cases (Fyfe et al., 2016). This work noted rela- ILRI impacts in Uganda tively little interest in trypanosomiasis as a stand-alone issue and concluded that commu- Trypanosomiasis has long been a concern in nity approaches would need external support. Uganda. Both the Gambian and Rhodesian forms In 2002, ILRI undertook research in Uganda of HAT are found in Uganda, and one of the lar- to develop decision-support tools for improving gest recorded human epidemics occurred in the trypanosomiasis control. This objective was to early 19th century, when over 500,000 people are enable technicians working with tsetse and estimated to have died. After the political unrest of sleeping sickness in Uganda to learn techniques the 1980s, tsetse invaded new areas of Uganda, for conducting geographical information system and various control campaigns were implemented (GIS) analysis to provide decision makers with (Okello-Onen et al., 1994; Okoth, 1999). information to facilitate targeting for control of ILRI was a partner in one of the research the disease. The project successfully established projects, which aimed to better understand the GIS capacity within Uganda. infection dynamics and drug sensitivities of the trypanosome parasites prevalent in dairy sys- tems in peri-urban Mukono District, Uganda, be- ILRI impacts in West Africa tween 1995 and 1999. This demonstrated a low prevalence and a low pathogenicity of trypano- Field work in West Africa initially started under some infections in cattle. An economic study by ATLN. These conducted ground-breaking re- ILRI found that, while trypanocides constituted search into understanding the presence, preva- up to 7.2% of health costs, they decreased profit- lence and impacts of tsetse and HAT. For ability on farms only by 1%. The dairy farms example, herds were monitored monthly for sev- were profitable, and the authors suggested tryp- eral years to understand production economics anosomiasis was less important than previously under the risk of trypanosomiasis. The ATLN thought, demonstrating the importance of eco- work concluded, in seven papers in the Demo- nomic analysis to justify interventions (Thorn- cratic Republic of the Congo (formerly Zaire), ton and Odero, 1998). Togo, Ethiopia and The Gambia (summarized As a result, the research consortium de- by Thornton and Odero, 1998, pp. 8–11), that cided that the methodologies developed in the trypanotolerant cattle were ‘economically justi- project should be applied to other areas of Africa fiable’ in a variety of situations with respect to where the disease was thought to have a signifi- tsetse challenge, control methods and experi- cant detrimental effect on livestock production, ence with trypanotolerance. particularly in areas where drug resistance was Following the Ugandan and ATLN work, suspected to occur, and Burkina Faso was se- ILRI collaborated and later led a research pro- lected for the next phase. gramme on trypanocide resistance in West ILRI contributed to research on the control A frica. The recommended control strategy was of HAT in South-east Uganda from 2000 to an integrated approach. This comprised vector Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 155 suppression in epidemiological hot spots, disease In the next stage, four ‘best-bet’ strategies management at the herd level and strategic use were evaluated: (i) trypanotolerant cattle; of trypanocides, combined with keeping local (ii)  community-based trypanosomiasis control; tolerant breeds. The first activity, from 1998 to (iii) training farmers and paravets in integrated 1999, was to assess AAT prevalence, tsetse chal- AAT control; and (iv) rational drug use infor- lenge and drug use. Resistance of trypanosomes mation for farmers and service providers. to isometamidium and diminazene was also Community-based control using insecticide-treated demonstrated by both in vivo and in vitro methods targets and insecticidal spraying decreased cat- (McDermott et al., 2003; Knoppe et al., 2006). tle mortality (71%), decreased abortion (66%), This led to a more ambitious programme increased traction (95%) and decreased farmer that sought to ensure the efficacy of trypano- expenditures on drugs (50%), while milk pro- cides as a component of integrated control of duction increased by 10%. The intervention was trypanosomiasis in the cotton zone of West deployed in a highly participative way, testing A frica. The first objective was to assess the pres- the hypothesis that previous attempts at com- ence and level of resistance across three coun- munity-based control had failed because they tries (Burkina Faso, Mali and Guinea) in the were top down and insufficiently participatory cotton zone of West Africa. This entailed an (Meyer et al., 2016). (A review of previous vector- enormous survey, eventually covering more control projects in Burkina Faso showed that than 30,000 cattle. The results showed a community-based control was in all cases effect- marked gradient in land cover, tsetse species, ive and in no case continued after the project cattle breed, resistance and productivity, from withdrew; Grace, 2003). However, while efforts more extensive, low-input, low-output, low- were more sustainable than attained previously, disease systems in east Guinea to more intensive, most of the activities were eventually aban- productive and high-disease systems in west doned. Researchers concluded that community Burkina Faso. The pattern of tsetse distribu- tsetse control does work but does not continue tion, trypanosomiasis prevalence and trypano- without external support because of high trans- somiasis risk varied predictably, with intensified action costs of setting up and maintaining the agriculture apparently driving change from a community-level institutional innovations needed situation of low disease and little resistance to to sustain control efforts (Box 3.1). This substan- one of high disease and high resistance. tiated earlier theoretical work that also argued Tailored recommendations were made for opti- that community-based control was not sustain- mal control across the region (Clausen et al., able because of its ‘prisoners dilemma’ nature 2010). (McCarthy et al., 2003). An epidemiological model describing the As part of the capacity building, several ap- transmission dynamics of trypanosomiasis was propriate technologies were developed, adapted developed and used to explain the trends identi- and tested. Anaemia is an important sign of fied in the spatial analysis for West Africa (Grace, trypanosomiasis and one that farmers were poor 2006). The model suggested that a change in at detecting. Researchers tested two field de- cattle breed is driving the emergence of resist- vices: a colour chart developed for diagnosing ance through the increase in drug use by farm- anaemia in sheep and a blood prick test designed ers to maintain trypanosusceptible Zebu – the for diagnosing anaemia in pregnant women. main driver is hence the change in preferred cat- Both were effective, but the colour chart system tle breed rather than drug use. Moreover, the was simpler and cheaper (Grace et al., 2007). situation in West Africa appears to be on the ac- The project also calibrated a weight band for the celerating portion of the sigmoidal resistance cattle population in the cotton zone. This is a curve, implying that prevalence, morbidity and tape measure that converts girth into body mortality, and resistance are likely to deteriorate weight and hence allows more appropriate dos- considerably from the present levels (Grace, ing. Several decision-support tools to aid differ- 2006). The optimistic conclusion from model- ential diagnosis were also tested with varying ling is that vector control, even if only resulting success. in small increases in tsetse mortality, may rap- Trypanotolerant cattle keeping was evalu- idly reverse the trend of emerging resistance. ated and it was found that, while farmers regard 156 D. Grace, E. Patel and T. Randolph Box 3.1. Community-based trypanosomiasis control. An ILRI initiative to improve control of AAT in Burkina Faso started with an evaluation of previous pro- jects (Grace, 2003). Research projects are rarely revisited after initial implementation, and this threw light on the ‘life after project’ scenario by looking at eight completed projects in Burkina Faso, six of which had a major participatory component. The projects took place over 25 years and encompassed a wide range of farming systems, institutions, partners and social conditions. They incorporated community participation and used low-cost and effective strategies that resulted in the rapid and total resolution of trypanosomiasis problems in all project areas. Participation and long-term viability issues (including cost recovery) were incorporated from the onset, and benefits to farmers were major, obvious and acknowledged. Yet, despite these substantial successes, none of the communities has continued with the strategies; tsetse reinvasion has occurred in all cases, and after years of investment in partici- pative trypanosomiasis control, farmers are once more ‘on their own’ and experiencing substantial losses from trypanosomiasis. The review identified some factors that blocked sustainability. While farmers learned technical skills for tsetse control, they did not learn management or business skills in implementing these. Although control was relatively cheap, the small financial requirements were a barrier. Farmers showed much more hypothetical willingness to pay than actual. None of the projects used delivery systems that were capable of delivering control after the withdrawal of the project. The systems used were either of known low sustainability (e.g. government, project or existing village groups) or new institutions whose sustainability and appropriateness for village conditions was un- proven (e.g. private vets). Communities encountered unexpectedly high transaction costs in avoiding free riders. the productivity (and disease resistance) of tryp- was reduced substantially. In villages with para- anotolerant cattle quite highly, they also cite vets, the expenditure was reduced by 36%, while their undesirable features: unpredictable tem- in villages with both vector control and paravets, perament, low working speed, short legs render- expenditure was reduced by 58%. Integration of ing them liable to damage the crop while trypanosomiasis control strategies is often ad- weeding, slow growth and slow weight gain, vised, both because each strategy has limitations smaller overall size, and low sale price and slower and because perverse effects are possible (suc- sale. The project confirmed other findings that cessful vector control in the absence of training trypanotolerant cattle were gradually being re- and information on drug use, for example, has placed as farmers switched to more productive resulted in paradoxical increased use of trypan- breeds, and concluded that encouraging trypa- ocides; Kamuanga, 2001a,b). Researchers con- notolerant cattle was not a viable strategy, at cluded that the intervention was successful but least while development, wealth and market in- the cost of training farmers and paravets was a tegration were on an upwards trend. barrier to widespread use. Moreover, in some The second intervention combined training situations, paravets are not allowed to operate. on vector control, diagnosis and treatment of The project was the first to explore rational trypanosomiasis, and diagnosis and treatment drug use in a context of ‘harm reduction’ as an of other common diseases, as well as use of trad- option for trypanosomiasis control. This was a itional medicines and nutrition. Scientists found concept borrowed from human health work large and significant differences in knowledge with prostitutes and drug users. It assumes that and skill both in farmers (in Burkina Faso) and if people are highly motivated to do things that in paravets (in Guinea) before and after training, are not recommended by authorities, punishing and between those trained and their peers who them will lead to worse outcomes and they had not been trained (Grace et al., 2008). Ten should instead be supported to undertake the ac- months after training, there was no significant tivities as safety as possible. Although official decrease in the level of knowledge and skill. policy is that veterinary treatments should only Moreover, there was a clear synergistic effect be- be given by veterinarians or paraprofessionals tween training and vector control. For example, under their direct supervision, the project had in the villages with intervention, the annual ex- shown that most veterinary treatments were penditure on trypanocides per farm household given by farmers and that it was neither feasible Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 157 nor economically viable for treatments to be given fever). The simple hygiene information that was by veterinarians. In this context, the project set provided resulted in an important decrease in out to see whether farmers who were going to side effects. High levels of complications are as- treat their animals anyway could be persuaded sociated with unhygienic administration. The to treat them more rationally. researchers conclude this was an effective, Much effort was spent in developing mes- cheap, sustainable and scalable option. The sages that could be understood by illiterate main challenge was the reluctance of national farmers. These addressed the problems identi- authorities to countenance farmers giving fied as being most likely to lead to treatment fail- treatments. ure, specifically misdiagnosis, underdosage and The researchers saw the need for additional poor injection technique. The intervention was action to promote the uptake of successful strat- evaluated through a large cluster-randomized, egies. Customized informational brochures and double-blinded, controlled trial, which at the training materials were designed to support ‘Ra- time was one of the most rigorous evaluations tional drug use’ (Fig. 3.1). Prototype materials, conducted by ILRI. It found that rational drug including a visually appealing, cartoon-style information improves farmer practice and re- booklet with an engaging storyline were de- duces drug underdosage: the improvement in signed for targeting adult literacy programmes dosage was particularly encouraging, given the and primary-school children. A communication highly significant relationship between under- strategy for dissemination of rational drug use dosage and treatment failure. In the medium messages, including using radio messages, was term, there was a 35% increase in farmers giv- also formulated. The project formulated recom- ing the correct dosage and an 8% increase in mendations for how drug companies could im- the appropriate use of isometamidium as a prove the quality of the information provided prophylaxis in the test group versus the control. with their products that would promote ‘Ra- In addition, there were better clinical outcomes tional drug use’. While company representatives and fewer treatment complications with ra- were in favour of the suggestions, they felt that tional drug use information: improvements in the current low returns to investment in the clinical parameters were significant only for a trypanocide market did not merit the added cost decrease in animal temperature (indicating less of adopting the recommendations. Fig. 3.1. An example of extension material to promote rational drug use. 158 D. Grace, E. Patel and T. Randolph A subsequent study assessed farmers’ know- trypanocides in response to resistance and these ledge and management of trypanosomiasis. In economic findings suggest why the messages the absence of a clear control group, propensity might not have high uptake. score matching was used (Liebenehm et al., 2009). Using three different matching algorithms, signifi- cant and robust differences between matched AAT economic and environmental studies participants and non-participants regarding cattle farmers’ knowledge were identified. A methodological innovation at ILRI was com- Hence, it can be concluded that the gain in farm- bining economic analysis with herd simulation ers’ knowledge is attributable directly to partici- models. Von Kaufmann et al. (1990) developed a pation in the research intervention. The bioeconomic herd model (called here the ATLN strongest effect of the research intervention is model) that was used to quantify the costs of on the curative knowledge of AAT and subse- AAT, to assess control strategies and to evaluate quent adequate control decisions. Moreover, sig- the benefits of trypanotolerant livestock (Itty, nificant advancements in preventative strategies 1992; Itty and Swallow, 1994). Some estimates were also observed. Overall, the research project of these costs are as follows: was effective in increasing farmers’ knowledge of good practices and contributed significantly The Gambia: 37% of the national herd in to improved livestock and farm productivity (Af- • The Gambia were at risk annually from fognon et al., 2009). trypanosomiasis (ILRAD, 1993). The annual ILRI found widespread concern among farm- economic costs of trypanosomiasis were ers and market agents about a proliferation of fake estimated to be less than 1 % of the annual and substandard trypanocide drugs. However, value of the total cattle herd and nearly all market studies using ‘dummy customer’ or ‘secret of the costs of trypanosomiasis were attrib- shopper’ methods found no evidence of counter- uted to production losses. feit drugs, and the results of drug quality tests con- Zimbabwe: due to extensive tsetse control ducted in 2005 on samples of drugs taken from • campaigns, only 4% of Zimbabwe’s cattle both formal (veterinary pharmacies) and informal population were at risk in the late 1980s and (black markets) sources revealed no major diffe- early 1990s (ILRAD, 1993). The annual cost rence in quality of the products. of trypanosomiasis control in Zimbabwe Economic and policy analysis was under- was largely attributable to tsetse control by taken in parallel with the epidemiological stud- spraying, not to production losses, which ies. Affognon (2007) investigated the short-term were small as a percentage of the value of productivity effect of drugs in controlling AAT the national cattle herd. under increasing drug resistance. For the first Côte d’Ivoire: the annual cost of trypanosom- time, a damage-control framework was applied • iasis was estimated to be 90% from produc- to animal disease control in Africa. The results tion losses and 10% from control costs. showed that the marginal value product (MVP) Cameroon: in Adamawa Province, tsetse of isometamidium in all epidemiological condi- • control led to substantial reductions in mor- tions and the MVP of diminazene in conditions tality rather than to significant increases in of high disease prevalence and high drug resist- cattle numbers. Changes in land use involved ance revealed a suboptimal (underuse) of these a shift to mixed farming among previously two major trypanocide molecules, not taking pure pastoralists. into account the externality of resistance. This means that, even in the face of increasing drug Subsequent economic analyses estimated resistance, trypanocidal drugs remain econom- that AAT in West Africa (Burkina Faso, Mali and ically attractive. Moreover, at the current sub- Ghana) was estimated to cause annual losses of optimal level of isometamidium use for the US$450 million in the 1990s (Itty, 1992). The use epidemiological conditions, investing in more of trypanocides in those three countries was drug use would be more than compensated for thought to protect some 17 million head of cattle by avoided production losses. For decades, the from the disease. However, the general use of tryp- veterinary experts had promoted reduced use of anocides was inducing resistance to trypanocides Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 159 and though the older studies mentioned the cost from from less than US$500/km2 to more than of resistance they did not quantify it. US$10,000/km2. The greatest potential benefits The economics of trypanosomiasis control are to Ethiopia, because of its high livestock dens- using trypanotolerant cattle were investigated ities and the historical importance of animal under the auspices of ATLN in Kenya, Ethiopia, traction, followed by regions of Kenya and The Gambia, the Democratic Republic of the Uganda (Shaw et al., 2014). Congo, Togo and Côte d’Ivoire (Itty, 1992). Itty A related study built on these findings to and Swallow (1994) showed that tsetse control evaluate the cost:benefit ratios as profitability appeared appropriate in situations with higher measures for various control methods (Shaw disease risk and that imports of trypanotolerant et al., 2015). Trypanocide prophylaxis is the only stock (in the Democratic Republic of the Congo profitable approach at low cattle density. Where and Togo) were not necessarily profitable. Stud- cattle densities are higher, the use of insecti- ies were undertaken to determine farmers’ cide-treated cattle is the most consistently profit- willingness to participate in vector-control pro- able method, with benefit:cost ratios greater grammes. These so-called ‘contingent valuation than 5. In areas of high potential for mixed surveys’ have been conducted in Burkina Faso, farming using oxen in Western Ethiopia, the fer- Ethiopia and Kenya to assess willingness to tile areas north of Lake Victoria and the dairying contribute labour and money to vector control areas of western and central Kenya, all control (Thornton and Odero, 1998, pp. 69–71, sum- methods achieve benefit:cost ratios from 2 to marizing three studies). Contingent valuation over 15, and for elimination strategies, ratios methods suggested that farmers were willing to from 5 to over 20. The costs of interventions pay around US$0.5–1 per treatment. In several against tsetse exceed benefits where cattle dens- countries, this was compared with the revealed ities are less than 20/km2. willingness to pay during campaigns: in general, McCarthy et al. (2003) developed a theoret- the observed willingness to pay correlated with ical model to explain why farmers might use dif- the estimates derived from contingent valuation ferent methods of trypanosomiasis control. studies but was less and, in at least one country, They argued that the public goods nature of declined with time (Kamuanga et  al., 2000, traps and targets, combined with an underlying 2001b). incentive structure that may resemble a prison- The most recent work on the economic er’s dilemma, explained the well-documented benefits of intervening against AAT is a study of failure of community-based control with traps Ethiopia, Kenya, Somalia, South Sudan, Sudan and targets, a failure that was most commonly and Uganda. Using a map of cattle production attributed to a lack of community participation. systems, herd models for each system were devel- ILRI also developed guidelines on methods oped for scenarios with or without AAT. The herd and tools for conducting impact assessments of models were based on estimated parameters of tsetse/trypanosomiasis interventions on the en- cattle productivity (fertility, mortality, yields, vironmental, social and economic systems and sales) from which growth of cattle populations on approaches for the integrated impact assess- and income were estimated over a 20-year period. ment of the interventions (Maitima et al., A spatial expansion model was adapted to esti- 2007)2. mate how cattle populations might migrate to new areas when maximum stocking rates are exceeded in older production areas. Last, differ- Other AAT research and impacts ences in income between the with and without scenarios were mapped, giving a measure of the ILRI examined possible changes in distribution potential benefits that could be obtained from of the three groups of tsetse in relation to chan- intervening against tsetse and trypanosomiasis ging climate, human population density and (Shaw et al., 2014). The estimated net present expected disease control activities (Reid et al., value of benefits to livestock keepers for the entire 2000; Coleman et al., 2001; Grace, 2014). The study area is nearly US$2.5 billion, at a discount key findings of these studies were that climate rate of 10% over 20 years – is approximately change is indeed likely to change the distribu- US$3300/km2 of tsetse-infested area – varying tional potential of tsetse but that anthropogenic 160 D. Grace, E. Patel and T. Randolph changes resulting directly from population expan- Trypanotolerant cattle have many advantages sion would be more important in determining in terms of disease resistance, but they are not actual changes in tsetse distributions. With re- preferred by farmers who are increasingly fo- spect to population density, it was estimated cused on productivity. However, trypanotolerant that human population growth after 2000 cattle populations have been declining for dec- would reduce tsetse-infested areas from roughly ades relative to Zebu crosses, which are preferred 8 million km2 to between 5 and 6.5 million km2 for their production characteristics. One study in 2040 (Reid et al., 2000, p. 231). (Agyemang and Rege, 2004) found that the In the early 2000s ILRI developed a deter- shares of trypanotolerant stock in all cattle in ministic mathematical model for trypanosomia- west and central Africa were falling in the late sis transmission. This was initially used to 1990s 2004). compare the effectiveness of different control Trypanotolerant stock are likely to persist, strategies (McDermott and Coleman, 2001). The without the need for much external support, in relative rankings of the effect of control strat- areas where, because of poor market access and egies on reducing disease prevalence were vector other constraints, low-input, low-output sys- control, vaccination and drug use, in that order. tems remain attractive. In other areas, trypa- Epidemiological modelling was used in several notolerant cattle will probably continue to be other projects, mainly to inform design and re- replaced by higher value and more productive search questions. Zebu crosses. Control of tsetse flies by insecti- cides has been much promoted and has gener- ated many scientific innovations and successful Conclusions and the Future pilots. However, the high cost and high level of coordination and management needed means it AAT has long been regarded as the single most has never proven sustainable outside ranches important disease of the single most important and externally supported projects. The use of livestock species in Africa. Early research at ILRI trypanocidal drugs, in contrast, is both sustain- focused on developing a vaccine. As it became able and scalable. Farmers are willing to buy clear that this was no easy endeavour, and as and use curative and, although to a consider- AAT worsened or became more obvious in many ably less extent, preventative drugs. Drugs are African countries, ILRI stepped up to address on the whole wisely used, but the lack of infor- our understanding and control of AAT in ad- mation on correct treatment certainly leads to vance of a vaccine. some irrational use and hastens the develop- Work started in East Africa and then ex- ment of resistance. ILRI research showed how tended to West Africa. There was emphasis on providing simple information to farmers can understanding the disease and its impacts slow this. and on testing solutions. Over four decades Looking to the future, AAT is likely to re- of  field research on AAT, an evolution is main a priority constraint for African livestock. e vident: researchers went from publishing in We now have approaches that are highly effect- proceedings to publishing in high-impact- ive at reducing the impact of AAT, either singly factor journals, from focusing on the efficacy or in combination. We also understand better of solutions to their uptake, and from forth- the challenges of adoption of even economically right advocates of favoured strategies to more attractive strategies and how the changing dy- nuanced critiques of the challenges of AAT namics of AAT may lead to future opportunities control. for optimized control. ILRI participated in important networks, most notably ATLN, and did not participate in important failed attempts to control AAT. It Acknowledgements i nvestigated all three of the major strategies for controlling AAT and found two of them want- The authors thank Peter-Henning Clausen for ing, at least without continued external support. reviewing the chapter. Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 161 Notes 1 William Wint, personal communication, April 2019. 2 See Introduction, Box I.1 on ‘Aspects of the Economic Burden of Trypanosomiasis’ for mention of ex-ante modelling of a hypothetical trypanosomiasis vaccine. References Affognon, H. (2007) Economic analysis of trypanocide use in villages under risk of drug resistance in West Africa. PhD dissertation, University of Hanover, Germany. Affognon, H., Coulibaly, M., Diall, O., Grace, D., Randolph, T. and Waibel, H. (2009) Étude des politiques relatives aux stratégies de gestion de la chimiorésistance dans le cadre de la lutte contre la trypano- somose en Afrique de l’Ouest: cas du Mali. ILRI Research Report 17. ILRI, Nairobi. Agyemang, K., and Rege, J.E.O. (2004) Trends in genetic composition of livestock with respect to trypa- notolerance and desirable economic traits with special emphasis on mixed farming systems in west and central Africa. In: Williams, T.O., Tarawali, S.A, Hiernaux, P. and Fernandez-Rivera, S. (Eds). Sustainable crop-livestock production for improved livelihoods and natural resource management in west Africa. Proceedings of an international conference held at the International Institute of Tropical Agriculture (IITA) Ibadan, Nigeria, 19–20th November 2001. ILRI, Nairobi; and Technical Centre for Agriculture and Rural Co-operation (CTA), Wageningen, The Netherlands. Bett, B., Randolph, T.F., Irungu, P., Nyamwaro, S., Kitala, P., et al. (2010) Field trial of a synthetic tsetse- repellent technology developed for the control of bovine trypanosomosis in Kenya. Preventive Veterin- ary Medicine 97, 220–227. Budd, L. (1999) DFID-funded Tsetse and Trypanosome Research and Development Since 1980. Depart- ment for International Development, London, UK. Clausen, P.H., Bauer, B., Zessin, K.H., Diall, O., Bocoum, Z., et al. (2010) Preventing and containing trypanocide resistance in the cotton zone of West Africa. Transboundary and Emerging Diseases 57, 28–32. Coleman, P.G., Perry, B.D. and Woolhouse, M.E.G. (2001) Endemic stability: a veterinary idea applied to human public health. Lancet 357, 1284–1286. Cross, G. (1975) Identification, purification and properties of clone-specific glycoprotein antigens constitut- ing the surface coat of Trypanosoma brucei. Parasitology 71, 393–417. DiMasi, J.A., Hansen, R.W. and Grabowski, H.G. (2003) The price of innovation: new estimates of drug development costs. Health Economics 22, 151–185. Fyfe, J., Picozzi, K., Waiswa, C., Bardosh, K.L. and Welburn, S. (2016) Impact of mass chemotherapy in domestic livestock for control of zoonotic T. b. rhodesiense human African trypanosomiasis in Eastern Uganda. Acta Tropica 165, 216–229. Gachohi, J.M., Bett, B. and Murill, G.A. (2009) Factors influencing the prevalence of trypanosomosis in Orma Boran trypanotolerant and Teso Zebu (trypanosusceptible) cattle crosses in Teso District, West- ern Kenya. Livestock Research for Rural Development 21, 216. Grace, D. (2003) Participative trypanosomosis control in Burkina Faso: lessons learned, way forward. Working Paper No 2. ILRI, Nairobi. Grace, D. (2006) Epidemiology and control of cattle trypanosomosis in villages under risk of trypanocide resistance in West Africa. PhD thesis, Freie Universitat, Berlin. Grace, D. (2014) The business case for One Health. Onderstepoort Journal of Veterinary Research 81, a725. Grace, D., Himstedt, H., Sidibe, I., Randolph, T. and Clausen, P. (2007) Comparing FAMACHA eye color chart and Hemoglobin Color Scale tests for detecting anaemia and improving treatment of bovine trypanosomosis in West Africa. Veterinary Parasitology 147, 26–39. Grace, D., Randolph, T., Diall, O. and Clausen, P.-H. (2008) Training farmers in rational drug-use improves their management of cattle trypanosomosis: a cluster-randomised trial in south Mali. Preventive Veter- inary Medicine 83, 83–97. ILCA/ILRAD. (1988) Livestock Production in Tsetse Affected Areas of Africa. Proceedings of a meeting held in Nairobi, Kenya, 23–27 November 1987. ILCA, Addis Ababa, and ILRAD, Nairobi. 162 D. Grace, E. Patel and T. Randolph ILRAD. (1993) Estimating the costs of animal trypanosomiasis in Africa. ILRAD Reports 11. Available at: https://cgspace.cgiar.org/bitstream/handle/10568/16354/ilrad_reports_1993_11_2.pdf.pdf? sequence=1&isAllowed=y (accessed 25 January 2020). ILRAD. (1994) Trypanosomiasis. ILRAD Reports 12. Available at: https://cgspace.cgiar.org/bitstream/ handle/10568/16367/ilrad_reports_1994_12_1%262.pdf.pdf?sequence=1&isAllowed=y (accessed 25 January 2020). Irungu, P., Bett, B., Mbogoh, S.G., Nyamwaro, S.O. and Randolph, T.F. (2007) Evaluating farmers’ prefer- ence for the attributes of a new tsetse repellent in Kenya: a conjoint analysis approach. Selected poster presented at the 29th International Scientific Council for Trypanosomiasis Research and Con- trol Conference, 1–6 October, Luanda, Angola. ILRI, Nairobi. Itty, P. (1992) Economics of Village Cattle Production in Tsetse Affected Areas of Africa: A Study of Trypano- somiasis Controlling Trypanotolerant Cattle and Chemotherapy in Ethiopia, Kenya, Côte d’Ivoire, the Gambia, Zaire and Togo.Hartung-Gorre Verlag, Konstanz, Germany. Itty, P. and Swallow, B.M. (1994) The economics of trypanotolerant cattle production in regions of origin and areas of introduction. In: Rowlands, G.J. and Teale, A.J., (eds) Proceedings of a Workshop organized by International Laboratory for Research on Animal Diseases and the International Live- stock Centre for Africa, ILRAD, Nairobi, Kenya. ILRAD/ILCA, Nairobi. Itty, P., Chema, S., d’Ieteren, G.D.M., Durkin, J.W., Leak, S.G.A., et al. (1988) Economic aspects of cattle production and of chemoprophylaxis for control of trypanosomiasis in village East African Zebu cattle in Kenya. In: Livestock Production in Tsetse Affected Areas of Africa. Proceedings of a meeting held in Nairobi, Kenya, 23–27 November 1987. ILRAD and ILCA, Nairobi, pp. 360–376. Kamuanga, M., Swallow, B.M., Hamade, S. and Bauer, B. (2000) Evaluating contingent and actual contri- butions to a local public good: tsetse control in the Yale agro-pastoral zone, Burkina Faso. Ecological Economics 39, 115–130. Kamuanga, E.K., Antoine, C., Brasselle, A.S., Swallow, B., d’Ieteren, G. and Bauer, B. (2001a) Impact of tsetse control on immigration, livestock population, cropping practices and farmer-herder conflicts in the Mouhoun Valley of southern Burkina Faso. In: Proceedings of the 25th Meeting of the ISCTRC, Mombasa, Kenya, 1999. OAU/STRC Publication No. 12. ISCTRC, Nairobi, pp. 239–253. Kamuanga, M., Sigué, H., Swallow, B., Bauer, B. and d’Ieteren, G. (2001b) Farmers’ perceptions of the impacts of tsetse and trypanosomosis control on livestock production: evidence from southern Burkina Faso. Tropical Animal Health and Production 33, 141–153. Knoppe, T., Bauer, B., McDermott, J.J., Peregrine, A.S., Mehlitz, D. and Clausen, P.H. (2006) Isometamid- ium sensitivity of Trypanosoma congolense stocks from cattle in West Africa tested in mice and the drug incubation infectivity test. Acta Tropica 97, 108–116. Kristjanson, P.M., Swallow, B.M., Rowlands, G.J., Kruska, R.L. and de Leeuw, P. (1999) Measuring the costs of African animal trypanosomosis, the potential benefits of control and returns to research. Agri- cultural Systems 59, 79–98. Leak, S.G., Peregrine, A.S., Mulatu, W., Rowlands, G.J. and d’Ieteren, G. (1996) Use of insecticide- impregnated targets for the control of tsetse flies (Glossina spp.) and trypanosomiasis occurring in cattle in an area of south-west Ethiopia with a high prevalence of drug-resistant trypanosomes. Trop- ical Medicine & International Health 1, 599–609. Leonard, D.K. and Straus, S. (2003) Africa’s stalled development: international causes and cures. Lynne Rienner Publishers, Boulder, Colorado. Liebenehm, S., Affognon, H.L. and Waibel, H. (2009) Assessing the impact of agricultural research on farmer’s knowledge about African animal trypanosomiasis: an application of the propensity score matching approach. In: 49th Annual Conference, 30 September–2 October, Kiel, Germany. German Association of Agricultural Economists (GEWISOLA). Maitima, J.M., Rodriguez, L.C., Kshatriya, M. and Mugatha, S. (2007) Guidelines for assessing environ- mental and socio-economic impacts of tsetse and trypanosomiasis interventions. ILRI Manuals and Guides No. 5. ILRI, Nairobi. Maudlin, I., Holmes, P.H. and Miles, M.A. (eds) (2004) The Trypanosomiases. CAB International, Wallingford, UK. McCarthy, N., McDermott, J. and Coleman, P. (2003) Animal health and the role of communities: an ex- ample of trypanasomosis control options in Uganda. EPTD Discussion Paper No. 103. International Food Policy Research Institute, Washington, DC. McDermott, J.J. and Coleman, P.G. (2001) Comparing apples and oranges – model-based assessment of different tsetse-transmitted trypanosomiasis control strategies. International Journal for Parasitology 31, 603–609. McDermott, J., Woitag, T., Sidibé, I., Bauer, B., Diarra, B., et al. (2003) Field studies of drug-resistant c attle trypanosomes in Kenedougou Province, Burkina Faso. Acta Tropica 86, 93–103. Tsetse and Trypanosomiasis Control in West Africa, Uganda and Ethiopia 163 Meyer, A., Holt, H.R., Selby, R. and Guitan, J. (2016) Past and ongoing tsetse and animal trypanosomiasis control operations in five African countries: a systematic review. PLOS Neglected Tropical Diseases 10, e0005247. Mulligan, H.W. and Potts, W.H. (1970) The African Trypanosomiases. Wiley-Interscience, London. Mungube, E.O., Vitouley, H.S., Allegye-Cudjoe, E., Diall, O., Boucoum, Z., et al. (2012) Detection of multiple drug-resistant Trypanosoma congolense populations in village cattle of south-east Mali. Parasite Vectors 5, 155. Muraguri, G., McLeod, A. and McDermott, J.J. (2003) Efficacy of a deltamethrin-based pour-on in the con- trol of tick-borne diseases and trypanosomosis in Kwale District, Kenya. International Journal of Trop- ical Insect Science 23, 69–74. Murray, M., Morrison, W.I. and Whitelaw, D.D. (1982) Host susceptibility to African trypanosomiasis: trypa- notolerance. Advances in Parasitology 21, 1–68. Okello-Onen, J., Heinonen, R., Ssekitto, C.M., Mwayi, W.T., Kakaire, D. and Kabarema, M. (1994) Control of tsetse flies in Uganda by dipping cattle in deltamethrin. Tropical Animal Health and Production 26, 21–27. Okoth, J.O. (1999) Tsetse and trypanosomiasis control problems in south-eastern Uganda: past, present, and alternative strategies. Schweizerische Medizinische Wochenschrift 129, 1091–1098. Omamo, S.W., Kagwanja, J., Reid, R., d’Ieteren, G., Ndjwa, N., et al. (2002) Agricultural extension reform in Africa: insights and lessons from livestock disease control in South-West Ethiopia. Socio-economics and Policy Research Working Paper 46. ILRI, Nairobi. Randolph, T.F., Saatkamp, H. and Perry, B.D. (2003) A short history of animal health economics. In: ISVEE 10: Proceedings of the 10th Symposium of the International Society for Veterinary Epidemiology and Economics, Vina del Mar, Chile, p. 827. Reid, R.S., Kruska, R.L., Deichmann, U., Thornton, P.K. and Leak, S.G.A. (2000) Human population growth and the extinction of the tsetse fly. Agriculture, Ecosystems & Environment 77, 227–236. Rogers, D.J. and Randolph, S.E. (2002) A response to the aim of eradicating tsetse from Africa. Trends in Parasitology 18, 534–536. Rowlands, G.J. and Teale, A. (1994) Towards increased use of trypanotolerance: current research and fu- ture directions. Proceedings of a Workshop organized by the International Laboratory for Research on Animal Diseases and the International Livestock Centre for Africa held at ILRAD, Nairobi, Kenya. Rowlands, G.J., Mulatu, W., Authie, E., d’Ieteren, G., Leak, S., et al. (1993) Epidemiology of bovine trypano- somiasis in the Ghibe valley, south West Ethiopia. 2. Factors associated with variations in trypano- some prevalence, incidence of new infections and prevalence of recurrent infections. Acta Tropica 53, 135–150. Shaw, A.P.M., Cecchi, G., Wint, G.R.W., Mattioli, R.C. and Robinson, T.P. (2014) Mapping the economic benefits to livestock keepers from intervening against bovine trypanosomosis in Eastern Africa. Preventive Veterinary Medicine 113, 197–210. Shaw, A.P.M., Wint, G.R., Cecchi, G., Torr, S.J., Mattioli, R.C. and Robinson, T.P. (2015) Mapping the bene- fit–cost ratios of interventions against bovine trypanosomosis in Eastern Africa. Preventive Veterinary Medicine 122, 406–416. Sones, K. (2001) Pharmaceutical companies: partners or enemies? Newsletter on integrated control of pathogenic trypanosomes and their vectors 3, 19–21. Swallow, B.M. (2000) Impacts of trypanosomiasis on African agriculture. PAAT Technical Scientific Series. FAO, Rome. Swallow, B.M. and Woudyalew, M. (1994) Evaluating willingness to contribute to a local public good: application of contingent valuation to tsetse control in Ethiopia. Ecological Economics 11, 153–161. Thornton, P.K. and Odero, A.N. (eds) (1998) Compendium of ILRI Research Impact and Adoption, 1975–98. ILRI, Nairobi. Trail, J.C.M., Hoste, C.H., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979a) Trypanotolerant Livestock in West and Central Africa. Vol. 1. General Study. ILCA Monograph No. 2. ILCA, Addis Ababa. Trail, J.C.M., Hoste, C.H., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979b) Trypanotolerant Livestock in West and Central Africa. Vol. 2. Country Studies. ILCA Monograph No. 2. ILCA, Addis Ababa. Traoré, S.A., Markemann, A., Reiber, C., Piepho, H.P. and Zárate, A.V. (2017) Production objectives, trait and breed preferences of farmers keeping N’Dama, Fulani Zebu and crossbred cattle and implications for breeding programs. Animal 11(4), pp. 687–695. von Kaufmann, R., McIntire, J. and Itty, P. (1990) ILCA bio-economic herd model for micro-computer. Tech- nical Reference Guide. ILCA, Addis Ababa. Wilson, C. (2003) Land-use/land-cover change analysis for Ghibe valley - Ethiopia: 1993–2003. Farming in Tsetse Controlled Areas (FITCA), Environmental Monitoring and Management Component, Project No. 7.ACP.RP.R.578. 4 Impact Assessment of Immunology and Immunoparasitology Research at ILRAD and ILRI Samuel J. Black and Cynthia L. Baldwin University of Massachusetts at Amherst, Massachusetts, USA Contents Executive Summary 165 The problem 165 ILRI’s contribution in the global context 165 Impacts of ILRI research 166 Scientific impacts 166 Development impacts 166 Capacity building and partnerships 167 Introduction 167 Scientific impact measurements defined 167 Traditional bibliographic measurements 168 Influence measurements 168 Practical outcomes 168 Historic Overview 168 State of knowledge of immune systems at ILRAD’s founding 169 Technology developed at ILRAD for studying bovine immunology 170 FACS at ILRAD 170 mAb production at ILRAD 170 BoLA typing at ILRAD 170 T-cell cloning at ILRAD 171 Measurements of scientific impact 171 Quantity of manuscripts 171 Citation analyses 171 Impact of highly cited manuscripts 172 The h-index of scientists 172 Cost comparators 175 Historic overview of fundamental and translational immunological research 177 Bovine immune system 177 Immunoparasitology of theileriosis 183 Immunoparasitology of trypanosomiasis 187 © International Livestock Research Institute 2020. The Impact of the International 164 Livestock Research Institute (eds J. McIntire and D. Grace) Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 165 Beyond Nairobi and Kabete: outreach to other diseases and places 194 Analysis of work environment and subsequent influence 194 Applications of tools and techniques 195 Additional disease research that benefitted 195 Conclusions 195 The Future 198 References 198 Executive Summary d epleting assets and reducing incomes, all of which are detrimental to household food and nu- The problem tritional security. The classic form of ECF control has been to spray acaricides to kill the ticks. Be- African animal trypanosomiasis (AAT, also ginning in the 1970s, and advancing over the known as ‘animal African trypanosomiasis’) is a years, an immunization procedure was developed parasitic disease caused by flagellated protozoans against ECF, which involved inoculation with live belonging to the family Trypanosomatidae, genus T. parva sporozoite forms and simultaneous treat- Trypanosoma, section Salivaria. The parasites are ment with a dose of the antibiotic oxytetracyc- transmitted by biting flies known as tsetse flies line. The development of a vaccine to protect (family Glossinidae, genus Glossina), which in- cattle against ECF was one of the founding aims habit much of tropical and subtropical Africa. The of the International Laboratory for Research on parasites live in the blood plasma, body tissue and Animal D iseases (ILRAD), a forerunner of the fluids of their host. In tropical Africa, AAT consti- International Livestock R esearch Institute (ILRI). tutes a major obstacle to the development of ani- mal production, causing major economic losses as the animals suffer from loss of condition, emaci- ILRI’s contribution in the global context ation and anaemia, resulting in reduced meat and milk production and draught power. While the an- This chapter assesses the research on bovine im- nual cost of the disease v aries among agroecolog- munology and immunoparasitology conducted ical zones, mortality in cattle can reach a rate of over 42 years, from 1973 to 2015, first at ILRAD 50–100% within months of exposure and there (1973–1994) and subsequently at ILRI, which are substantial losses from the exclusion of cattle was formed by merging ILRAD and the Inter- from regions where the disease prevents cattle national Livestock Centre for Africa (ILCA) in production and from the loss of animal draught 1995. This assessment covers the approaches power in areas where mixed farming is feasible. taken, the performance of research teams, the Theileriosis, commonly known as East Coast scientific truths uncovered, the cost-effectiveness fever (ECF), is another parasitic disease. Fatal to of the research undertaken and the practical cattle, this disease is also caused by a protozoan outcomes achieved, notably, the development of parasite, Theileria parva. The disease occurs in monoclonal antibodies (mAbs) and other tools 12 countries in eastern, central and southern to better define the bovine immune system. The Africa, where the tick vectors of this parasite are chapter makes extensive use of citation data found. ECF causes major economic losses by af- along with the personal reflections of scientists fecting both dairy cattle and young Zebu cattle in who participated in the research and surveys of pastoralist systems and ranches. It is among the opinion leaders in the field. most serious constraints to cattle productivity in The specific scientific goals and achieve- the countries where it is found. Some of the direct ments of ILRI and its predecessors were as follows: losses due to ECF include cattle mortality, the stunting of calves and reduced milk production. • Making a substantive contribution to bovine Indirect losses include the lack of adoption of immunology was realistic and has been more productive breeds of cattle and the costs as- substantially achieved. sociated with ECF prevention and control. ECF • Measuring the diversity of strains of Thei- affects households by reducing the milk supply, leria parva, Trypanosoma brucei, Trypanosoma 166 S.J. Black and C.L. Baldwin vivax and Trypanosoma congolense was realis- While the ILRAD/ILRI T. parva vaccine tic and has been substantially achieved. programme has not yet resulted in full disease • Identifying mechanisms of immunity that control, contributing scientists have identified, kill parasites or limit the growth of the cloned and expressed recombinant T. parva pro- above parasites was realistic and has been teins that stimulate protective T-cells and anti- substantially achieved. body responses (Nene et  al., 1992, 1995; • Developing an effective subunit vaccine Graham et  al., 2006). They have further pro- against any of the parasites was an ambitious vided evidence that a subunit vaccine, perhaps goal and so far has not been achieved. comprising a sporozoite p67 subunit that elicits sporozoite-neutralizing antibodies to- gether with schizont epitopes that stimulate Impacts of ILRI research appropriate CD4+ helper T-cells and CD8+ cyto- toxic T-cells in cattle, is a realistic possibility. ILRI and its predecessors led global research in While research towards a T. parva vaccine at trypanosomiasis (see Chapters 2 and 3, this vol- ILRAD/ILRI continues to progress, this was not ume) and in ECF (theileriosis) (see Chapters 5 and 6, the case with research to develop a vaccine effect- this volume). A vital part of work on both diseases ive against AAT, where high variation of pro- was in immunology and immunoparasitology. tective parasite antigens blocked the vaccine research programme from the outset and led to Scientific impacts its closure in 2001. Studies of mechanisms of trypanosomiasis susceptibility and tolerance at ILRAD/ILRI covered a wide field and left a leg- The scientific impacts of the immunology and acy of infection-induced immune response data immunoparasitology work started at ILRAD and in cattle but did not identify candidate vaccine continued at ILRI have been substantial. These antigens or definitively identify resistance studies resulted in a substantive body of work, as genes for introgression into the gene pool of de- shown by the well-cited 68 peer-reviewed publi- sirable Boran livestock. Nevertheless, this work cations (see Table 4.6). made notable contributions to our understand- • First, the immunology programme developed ing of trypanosome biology and immunology. a comprehensive suite of mAbs that identi- fied the major bovine T-cell subpopulations, immunoglobulin isotypes, phagocytic cell Development impacts populations, several bovine lymphocyte antigens (BoLAs; also called major histo- Many of the aims of the T. parva immunology compatibility (MHC) antigens) and the cir- programme at ILRAD/ILRI were achieved, but cumsporozoite coat of T. parva sporozoites development of an effective subunit T. parva vac- used in host-cell invasion. cine remains a challenge. The immunology and • Second, the team developed and main- immunoparasitology research at ILRAD and tained in vitro clones of bovine CD4+ and later ILRI had high scientific and technical im- CD8+ T-cells that facilitated analysis of mAb pacts on a moderate budget. Outputs from this specificity, investigation of T. parva anti- research were primarily publications and mAbs genic diversity and T. parva peptide–BoLA specific for bovine leukocytes and immuno- complexes that induce T. parva strain and globulins, which are still being used in many host MHC-specific protective immunity. laboratories throughout the world. Despite its • Third, the team used these tools to investi- academic/scientific successes, this immunology/ gate, and to a large extent elucidate, host immunoparasitology research did not affect the immune responses that control the fre- incidence or economic impact of theileriosis or quency, magnitude and duration of African trypanosomiasis in any substantial way. How- trypanosome and T. parva parasitaemic epi- ever, the authors remain optimistic that the fun- sodes in cattle and Cape buffalo, the latter damental research on bovine immunology and of which are disease-resistant reservoir the immunoparasitology of theileriosis and hosts of the parasites. trypanosomiasis carried out at ILRAD/ILRI will Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 167 ultimately have development outcomes. Devel- immune responses would lead to strategies for dis- opment impacts of the ECF work eventually ease control, with fundamental research tailing emerged through use of the infection-and- off over time and improved control of these dis- treatment method (ITM) (see Chapter 6, this eases eventually dominating the research efforts. volume). While the authors have no doubt that this will eventually prove to be the case, it should be appre- ciated that investment in the immunological Capacity building and partnerships control of challenging diseases, especially those caused by complex, protozoan, parasites, is not for The immunology work at ILRAD and ILRI had the impatient. Clear evidence of this is provided major capacity development effects. First, it con- not only by ILRAD’s long experience with its tar- tributed to the achievements of scientists who get diseases but also by the many other organiza- went on to have distinguished careers in im- tions, programmes and consortia conducting munology and related fields. Second, it contrib- similar long-term disease research on such im- uted to the development of scientific methods portant disease problems as cancer, human im- applied to immunology and other fields and to munodeficiency virus (HIV)/acquired immune diseases other than ECF and trypanosomiasis, deficiency syndrome (AIDS) and malaria. notably the identification of bovine leukocyte populations and subpopulations using mAbs, Scientific impact measurements defined cloning and long-term propagation of bovine T-cells, and the use of cloned bovine T-cells to identify host-protective parasite antigens. The evaluation in this chapter looks at both ‘out- puts’ (the quality and amount of research prod- ucts produced) and ‘outcomes’ (the changes in behaviour or influence brought about due to the Introduction research) as measures for assessing the impacts of the research. Assessments of how the re- ILRAD was established in 1973 with the man- search altered hunger, nutrition and poverty – date to conduct basic research to develop safe, the ultimate aims of this research – is not at- effective and economical control measures for tempted here. Although there is a strong push to livestock diseases that seriously limit world food measure impacts in development work in terms production. of how work changes a given situation for the Theileriosis and AAT (Randolph et  al., better, it is often unclear how to measure such 2003) were chosen as target diseases because changes stemming from development pro- they severely constrain livestock production in grammes and interventions. One way to meas- sub-Saharan Africa. Prior research on their ure impact is to address the question, ‘Would the causative agents, which are, respectively, tick- same changes have occurred if this investment transmitted apicomplexans of the Theileria parva had not been made?’, but this approach is ham- species and tsetse-transmitted kinetoplastids, pered by the fact that a negative control for this namely, Trypanosoma brucei, T. congolense and question is commonly missing. (This is particu- T. vivax, had shown that it was possible to immunize larly true regarding achievements in the basic livestock so that they were protected against re- scientific research required to develop interven- infection with the immunizing strain. However, this tions such as vaccines.) immunity broke down when animals were exposed Another measure of research impact, and to other stains of T. parva or trypanosomes. Thus, it the one taken here, is to determine whether was important to establish the nature of protective what was found and assumed to be ‘true’ at the immunity against the parasites and the mechan- time of publication of research results remained isms leading to breakdown of immunity upon true over the following years and was taken up heterologous strain challenge as steps towards or built upon by other scientists. This evaluation identifying and implementing immunologically aims to determine whether the right experimen- based disease control strategies. It was envisaged tal directions were chosen and to measure the that knowledge of the pathogens and of effective performance of ILRAD/ILRI immunology and 168 S.J. Black and C.L. Baldwin immunoparasitology research with respect both ILRI. To assess this behaviour change, the fol- to the outputs (e.g. the scientific truths un- lowing were evaluated: covered and new tools generated) and to the cost-effectiveness of that work. Regarding the • The Hirsch index (h-index) of individual latter, to assess the magnitude of investment scientists (discussed later), which measures needed to achieve those ‘scientific truths’, the their productivity and influence on others only possible comparators are the kinds and over their careers up to 2015. levels of investments made for similar diseases • Results of a survey to obtain personal having the same goals – of preventing infection r eflections on the influence that working or reducing pathology – as those made for ECF at ILRAD/ILRI had on post-ILRAD/ILRI and AAT at ILRAD and more recently at ILRI. career choices and research directions of The performance indicators and formal surveys scientists. outlined below were used to measure the impacts • A survey of opinion leaders in relevant on the field of veterinary – more specifically, fields on ILRAD/ILRI impacts. bovine – immunology and the application of • Examples of former ILRAD/ILRI scientists knowledge gained for controlling infectious dis- continuing to address infectious diseases eases in ruminants. that affect world food production. Traditional bibliographic measurements Practical outcomes Traditional bibliographic measurements of the Practical outcomes are defined here as commer- output of scientific research in the form of peer- cialized or shared tools or methods and add- reviewed manuscripts include measurements of itional training of individuals who are product- both quality and quantity. While it is important ive in the field. Tools are physical entities such as that research papers are published in ‘top-tier’ mAbs but also methods and protocols developed scientific journals to be broadly read by the sci- such as bovine MHC typing and lymphocyte entific community, it is equally important to cloning. Commercialized products include mAbs, have a body of work that includes papers in spe- vaccines and diagnostic kits. An outcome of ciality journals and on social media. trainees means that such individuals stay in the This chapter thus presents data on: (i) quan- field and are productive rather than simply a tities of manuscripts, including citation indices number listed under ‘output’ (i.e. trained but did and the Web of Science index for each; and (ii) the not get a job in that field). To determine practical number of papers of a collective body of manu- outcomes, the following were assessed by survey scripts with ten or more citations, used to calcu- responses: late an individual’s i10 index (the number of • Tools generated at ILRAD/ILRI and used by papers with more than ten citations) to assess others outside of that environment; this in- both the quantity and quality of the collection. cludes an analysis of the diseases to which such tools were applied and the geograph- Influence measurements ical distribution of that application. • Former ILRAD/ILRI scientists who had Research influences are used here to show how trainees who remained in their scientific research affects behaviour and discourse. Re- fields. search that has influence has the ‘capacity to • Commercialization of scientific tools. produce an effect on the advancement of scien- tific knowledge’ (Donaldson and Cooke, 2013). This includes how a researcher or research group influences other groups or directions of Historic Overview research. The ‘change in behaviour’ concept here includes whether ILRAD/ILRI influenced a An historic overview of the goals and achieve- given field of research or the approach taken by ments of the research was made from the individual researchers regarding solving scien- scientific literature. (Note that Altmetric (www. tific problems at institutions other than ILRAD/ altmetric.com/; accessed 5 February 2020), Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 169 which counts Twitter re-tweets, Facebook ‘likes’, As shown by seminal work from Hozumi and comments in journals, electronic views and Tonegawa (1976), these antigen-specific recep- downloads, and posts on other social media, was tors are encoded by multiple immunoglobulin not considered here due to the historic nature of gene segments that rearrange during B-cell many of the publications.) This includes descrip- development. tions of the bovine immune system and research In addition, several functionally distinct landscape at the inception of ILRAD and how classes of antibody had been identified and there ILRAD/ILRI research helped that landscape de- was evidence from immunoglobulin heavy- velop and mature in parallel with other groups chain allotype expression in mice that the genes addressing fundamental immunological ques- encoding these proteins were arranged in a sin- tions in a variety of host species or addressing gle genetic locus. Although the T-cell anti- diseases and pathogens closely related to those gen-specific receptor had not been identified, it targeted by ILRAD/ILRI. The questions ad- was known from the work of McDevitt et  al. dressed in these narratives are: (1972) that genes encoded in the immune • response region (I region) of the mouse MHC Was the right research direction taken to could control antibody responses. It was also fulfil the mandate of ILRAD/ILRI? • known from Zinkernagel and Doherty (1974) Was the team successful, i.e. was the re- that genes encoded in the H-2 region of the search reflective and creative? • mouse MHC restrict the specificity of cytotoxic Was there a substantive body of work that T-cells reactive with virus-infected cells. This led showed the step-by-step progression of Zinkernagel and Doherty to propose that cyto- knowledge in addition to key papers in top- toxic T-cells recognize MHC proteins that are tier journals? • changed by virus infection, a hypothesis that Were the results published in a timely was later proven by receptor isolation and manner? co-crystallization with ligand–MHC complex. It was also known that macrophages were re- quired for inducing immune responses in vitro State of knowledge of immune systems and that activated T-cells secrete materials that at ILRAD’s founding control other leukocytes, but these molecules, which were later to be called cytokines, had not When ILRAD started, it was known from studies been isolated. in mice and humans that lymphocytes could be Despite elegant work done on immune sys- divided into two major groups, B-cells and T-cells. tem function carried out first in chickens to de- The former were known to be precursors of anti- lineate the Bursa as an organelle required for body-producing cells and the latter to be com- B-lymphocyte development (thus Bursa-derived prised of at least three subpopulations, which or B-cell) and the thymus as an organelle re- express distinct functions, namely: (i) cytotoxic quired for T-lymphocyte development (thy- T-lymphocytes, now generally called CD8+ mus-derived or T-cell) (Cooper et al., 1966), and T-cells, which kill infected host cells; (ii) T-helper later in ruminants to characterize cell compo- lymphocytes, now called CD4+ T-cells, which fa- nents of the peripheral lymph (Smith et  al., cilitate the development of antibody responses, 1970; Scollay et al., 1976), in 1978 our under- delayed-type hypersensitivity, and phagocyte re- standing of the immune systems of birds and ru- cruitment and activation; and (iii) suppressor minants lagged behind that of humans and T-lymphocytes, which decrease the magnitude mice. Different classes of bovine and sheep im- of immune responses, now known as regulatory munoglobulins and their functions had been T-cells (Tregs). It was also known that antibody- identified, some complement factors had been producing B-lymphocytes had differentiated isolated and systems to evaluate T-cell proliferative from B-cells that expressed membrane- bound, responses in vitro and in vivo had been developed; cell-surface, antigen-specific receptors com- the latter was known as skin testing. However, posed of immunoglobulin heavy and light the description of leukocyte differentiation anti- chains and that those of a single cell could bind gens and cellular functions, including those of only one or a restricted group of antigens. antigen-specific T- and B-cells, was hindered by a 170 S.J. Black and C.L. Baldwin paucity of reagents to identify, isolate and char- the Medical Research Council Laboratory of acterize specific cell types. The establishment M olecular Biology, in Cambridge, UK, and was first of a well-funded and well-equipped veterinary described in 1975 (Kohler and Milstein, 1975). immunology research programme at ILRAD, as Briefly, antigen- stimulated lymphocytes were described below, helped to change this. immortalized by fusion to mutant B-lymphocyte tumour cells (myeloma cells) with an enzyme de- ficiency that enables hybrid selection. During Technology developed at ILRAD replication of selected hybrids, called myeloma for studying bovine immunology hybrids or hybridomas, chromosome assortment results in some hybrid progeny that express the ILRAD began operations at Nairobi’s Kenyatta genes encoding the immunoglobulin heavy and Hospital in 1975, and its research and support light chains derived from the lymphocyte part- facilities were inaugurated in 1978 on a 70-ha ner together with appropriate housekeeping site near Nairobi. In support of its commitment genes and the transforming gene from the mye- to immunological research, ILRAD recruited loma fusion partner. Hybridomas are cloned and faculty members with expertise in what were grown in vitro, and those producing mAbs of the then cutting-edge technologies – including fluo- desired specificity are identified by screening cul- rescence-activated cell sorting (FACS), the pro- ture supernatants for the presence of antibodies duction of mAbs, ruminant tissue typing and reactive with the target antigen. Terry Pearson, T-cell cloning – with a view to applying these one of the founding faculty members of ILRAD, techniques for dissecting the ruminant immune had worked with Kohler and Milstein in Cam- system in health and disease. The histories of bridge and brought the mAb technology directly these technologies and their implementation at to ILRAD. Samuel Black, who joined ILRAD in ILRAD are briefly summarized below. 1979, introduced complementary techniques for rapid cloning of hybridomas and screening of leukocyte-specific mAbs using FACS. The first IL- FACS at ILRAD RAD mAbs were made against trypanosome and Fluorescence-activated cell sorting (FACS) was T. parva antigens and described by Pearson et al. invented in the late 1960s by a team at Stanford (1980) in a general methodology paper. University led by Leonard Herzenberg. Its pur- pose was to analyse and isolate viable cells using BoLA typing at ILRAD their light-scattering properties and fluorescence emitted by attached fluorochrome-conjugated Histocompatibility antigens are cell-surface antibodies or other materials. Although a FACS glycoproteins that have been selected through machine was commercially available from Becton evolution to present antigens to T-cells. They Dickinson Immunocytometry Systems in the early vary from individual to individual and were ini- 1970s, at the time of ILRAD’s inception, only a few tially typed in humans using serum from trans- laboratories worldwide had FACS facilities. John J. plant-rejection patients and multiparous women. (Jack) Doyle, one of ILRAD’s founding faculty mem- The antigens responsible for transplant rejection bers, established the FACS facility at ILRAD in were named human leukocyte antigens (Row- 1977. The facility became fully operative in 1981, lands et al., 2003) by a nomenclature committee and the first publication from ILRAD using FACS of the World Health Organization (WHO) in addressed the replicative cycle of bloodstream-stage 1968 and are encoded by genes in the MHC T. brucei and T. vivax (Shapiro et al., 1984). locus. Bovine MHC molecules are called bovine lymphocyte antigens (BoLA). BoLA typing was mAb production at ILRAD started in Kenya by a small team led by Alan Teale. This group was initially based at the Kenya Monoclonal antibodies (mAbs) are antibodies of Agricultural Research Institute (KARI), in Mu- a single antigenic specificity that are produced guga, and used serological methods to identify by a single B-lymphocyte and its clonal progeny. the MHC antigens of African cattle as part of a The technology to generate immortal cell lines study aimed at understanding the problems that produce mAbs against specific antigens was a ssociated with a T. parva-infected cell-line-based developed by George Kohler and Cesar Milstein at vaccine for ECF (see Chapter 6, this volume). The Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 171 group moved to ILRAD in 1984, where Steve vine counterparts were found for the most part Kemp soon joined. A strong partnership devel- to require T-cell growth factor and periodic anti- oped with the immunologists working in both gen stimulation. T-cell cloning facilitated func- trypanosomiasis and theileriosis to develop a tional dissection of the bovine immune system panel of reagents to characterize the MHC diver- and analysis of the specificity of T-cell responses sity of African cattle in order to support vaccine to T. parva-infected bovine lymphocytes. The development and to look for associations be- T-cell clones also expedited screening of bovine tween tissue type and resistance or susceptibility leukocyte-specific mAbs to identify those that to disease. This group made important technical react with functionally distinct populations of contributions to the typing technology and con- leukocytes. tributed substantially to the international BoLA classification system. BoLA typing was of critical importance in mapping the specificity of cyto- Measurements of scientific impact toxic T-cells against bovine cells infected with T. parva, as exemplified by the 1987 paper dem- Quantity of manuscripts onstrating both MHC and T. parva strain restriction (Morrison et al., 1987). In addition, attempts to The number of published manuscripts considered associate MHC type with disease susceptibility for the topic as defined was 400. The manu- developed into a genome-wide association study, scripts address bovine immunology and immu- which in turn provided an important impetus to noparasitology of theileriosis and AAT and were the international bovine genome mapping effort selected from all papers listed in ILRAD and and to the more recent discovery of gene vari- ILRI annual reports from 1975 to 2013. The ants associated with resistance to trypanosom- manuscripts were grouped initially as those iasis in mice (Goodhead et al., 2010). from ILRAD and ILRI that concerned bovine immunology, with a second group concerning T-cell cloning at ILRAD bovine immunoparasitology and a third group comprising manuscripts on the mouse model of T-cell clones are used to dissect the specificity African trypanosomiasis. This mouse model was and functions of individual T-cells participating included because it was used at ILRAD and in cell-mediated immune responses. The cloned subsequently ILRI to inform research in bovine cells can be grown as long-term cultures that re- trypanosomiasis. tain both phenotype and function. Key to the de- velopment of T-cell clones was the finding in Citation analyses 1976 by Morgan et  al. (1976) that interleukin (IL)-2 has human T-cell growth-inducing prop- Citation data are frequently used for evaluating erties. In 1977, Gillis and Smith (1977) reported the scientific productivity of individuals and insti- that mouse cytotoxic T-cell clones could be tutions. Not everyone agrees that this is a useful grown in vitro supported by the addition of me- measure, and much has been written on the dium containing IL-2, which in culture super- topic. However, few would argue against the no- natants of stimulated cells was referred to as tion that the absence of citations of a published T-cell growth factor. It was subsequently re- work that is accessible on PubMed and has been ported in 1981 by Spits et  al. (1981) that ‘out’ for some years is a reasonable indicator of long-term culture of cloned human T-cells re- low/no influence, while a high citation index is a quires frequent stimulation with their cognate reasonable indicator of substantial influence. antigen in addition to provision of T-cell growth Nevertheless, citation averages vary among fields, factor. In 1982, Wendy Brown was recruited to and field parameters should be taken into account ILRAD to generate bovine T-cell growth factor when considering the citation data. For example, and establish clones of bovine T-cells. This was citation averages are much higher in the field of achieved in 1985 by Brown and Grab (1985) us- molecular biology than in mathematics (field ing lectin-stimulated bovine T-cells and in 1986 averages, respectively, of 10.8 and 3.5 citations by Teale et al. (1986) using T-cells that proliferate during the period 2000–2010, based on journal in response to mismatched BoLA molecules. Like articles indexed by Thompson Reuters in its Es- antigen-specific human T-cell clones, their bo- sential Science Indicators (ESI) database). Hence, 172 S.J. Black and C.L. Baldwin interpretation of citation data is nuanced. With from other academic institutions, there are a respect to agricultural sciences and animal sci- few highly cited publications (more than 100 ence, ESI field averages over the 2000–2010 times) and many publications with fewer cit- period were, respectively, 7.1 and 7.7, whereas ations. This is a reasonable citation profile for that for immunology was 21.8. It is not clear how the agricultural sciences and animal science work on ruminant immunology and immuno- based on average rates of citation by field as dis- parasitology should be classified because much cussed above. It is noteworthy that papers with of this is not considered to be mainstream im- fundamental immunology content tended to be munology. Hence, while the field average com- more highly cited (Fig. 4.1) than other more parator for these topics may be imperfect, we specialist papers, and this is in line with overall consider it worthwhile tabulating citations of expectations for this field. the ILRAD/ILRI literature in these areas as an indicator of influence. Impact of highly cited manuscripts Citations (until 2018) from selected IL- RAD/ILRI publications from 1975 to 2018 The most highly cited manuscripts are shown in were obtained from Scopus. For bovine immun- Tables 4.1–4.3 for three thematic research groups. ology and immunoparasitology research, 74% An abbreviated explanation of their impact on of papers from ILRAD and 56% from ILRI have the field is given. These publications address im- been cited more than ten times, and 56% of all munology and immunoparasitology work from papers focused on the mouse model of African ILRAD/ILRI in journals with broader science trypanosomiasis have been cited more than ten readership and high impact factors. times. The lower percentage of papers from ILRI cited more than ten times is partly attribut- The h-index of scientists able to the more recent dates of publications included in that group (those published in the The h-index (in June 2015) of 16 key scientists last 2 years have been cited less often). All of having worked at ILRAD and/or ILRI, and in these would be counted in the i10 index of sci- some cases still working at ILRI, was obtained. entists. As would be expected with publications These h-indices were compared with those from 40 35 ILRAD ILRI Mice-tryps 30 25 20 15 10 5 0 >100 50–100 25–50 10–24 1–9 Number of citations Fig. 4.1. Percentage of manuscripts by citation groups. Mice-tryps trypanosomiasis studies in mice. (Data from www.scopus.com). Percentages of manuscripts Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 173 Table 4.1. The most highly cited studies on basic bovine immunology. Scopus citation Year of percentile publication Publication Impact on the field 94 1996 Davis, W.C., Brown, W.C., Hamilton, M.J., Wyatt, Still one of the only C.R., Orden, J.A., et al. (1996). Analysis of mAbs that reacts monoclonal antibodies specific for the γδ TcR. with T-cell receptor Veterinary Immunology and Immunopathology (TCR) chains of 52, 275–283. livestock species 97 1979 McGuire, T.C., Musoke, A.J. and Kurtti, T. (1979) First description of the Functional properties of bovine IgG1 and functional properties IgG2: interaction with complement, of bovine IgG1 and macrophages, neutrophils and skin. IgG2 Immunology 38, 249–256. 97 1990 Clevers, H., MacHugh, N.D., Bensaid, A., Study showing that the Dunlap, S., Baldwin, C.L., et al. (1990) bovine cells Identification of a bovine surface antigen recognized by the uniquely expressed on CD4– CD8– T cell mAb to WC1 were γδ receptor γδ+ T lymphocytes. European Journal T-cells of Immunology 20, 809–817. 98 1986 Baldwin, C.L., Teale, A.J., Naessens, J.G., First mAb to bovine Goddeeris, B.M., MacHugh, N.D. and CD4 and Morrison, W.I. (1986) Characterization of a demonstration that subset of bovine T lymphocytes that express these cells have a BoT4 by monoclonal antibodies and function: distinct function from similarity to lymphocytes defined by human T4 CD8+ T-cells and murine L3T4. Journal of Immunology 136, 4385–4391. 95 1986 Ellis, J.A., Baldwin, C.L., MacHugh, N.D., First mAb to bovine Bensaid, A., Teale, A.J., et al. (1986) CD8 and Characterization by a monoclonal antibody demonstration that and functional analysis of a subset of bovine these cells have T lymphocytes that express BoT8, a molecule distinct function from analogous to human CD8. Immunology 58, CD4 T-cells 351–358. 92 1986 Goddeeris, B.M., Baldwin, C.L., ole-MoiYoi, O. A crucial method for and Morrison, W.I. (1986) Improved methods both purifying bovine for purification and depletion of monocytes monocytes from from bovine peripheral blood mononuclear blood and depleting cells. Functional evaluation of monocytes in them from responses to lectins. Journal of Immunological lymphocyte Methods 89, 165–173. populations 96 1986 Lalor, P.A., Morrison, W.I., Goddeeris, B.M., First mAbs made at Jack, R.M. and Black, S.J. (1986) Monoclonal ILRAD to T-cell and antibodies identify phenotypically and macrophage functionally distinct cell types in the bovine populations and lymphoid system. Veterinary Immunology and some of the first ever Immunopathology 13, 121–140. made for bovine research An additional manuscript had 108 citations but is a workshop report and so was not included in the table; however, it demonstrates the importance of the cluster-defined (CD) workshops: Howard, C.J. and Naessens, J. (1993) Summary of workshop findings for cattle (tables 1 and 2). Veterinary Immunology and Immunopathology 39, 25–47. 174 S.J. Black and C.L. Baldwin Table 4.2. The most highly cited studies on the bovine immune responses to T. parva. Scopus citation Year of percentile publication Publication Impact on the field 93 1980 Pearson, T.W., Pinder, M., Roelants, G.E., Kar, First paper on mAbs S.K., Lundin, L.B., et al. (1980) Methods for produced at ILRAD derivation and detection of anti-parasite monoclonal antibodies. Journal of Immunological Methods 34, 141–154. 88 1992 Musoke, A., Morzaria, S., Nkonge, C., Jones, E. A major contribution and Nene, V. (1992) A recombinant sporozoite towards developing a surface antigen of Theileria parva induces sporozoite-neutralizing protection in cattle. Proceedings of the National vaccine Academy of Sciences USA 89, 514–518. 60 1981 Eugui, E.M. and Emery, D.L. (1981) Genetically First paper to show that restricted cell-mediated cytotoxicity in cattle T. parva-specific cytotoxic immune to Theileria parva. Nature 290, T-cells were restricted 251–254. by host genotype 86 1986 Goddeeris, B.M., Morrison, W.I., Teale, A.J., The first paper to show Bensaid, A. and Baldwin, C.L. (1986) Bovine that MHC class I cytotoxic T-cell clones specific for cells antigens restrict the infected with the protozoan parasite Theileria specificity of T. parva parva: parasite strain specificity and class I parasite strain-specific major histocompatibility complex restriction. cytotoxic T-cells Proceedings of the National Academy of Sciences USA 83, 5238–5242. 93 1979 Pearson, T.W., Lundin, L.B., Dolan, T.T. and First paper to show that Stagg, D.A. (1979) Cell-mediated immunity to T. parva-infected Theileria-transformed cell lines. Nature 281, lymphocytes induce 678–680. cytotoxic cells 88 1988 Baldwin, C.L., Black, S.J., Brown, W.C., Study showing that all Conrad, P.A., Goddeeris, B.M., et al. (1988) three major Bovine T cells, B cells, and null cells are lymphocyte transformed by the protozoan parasite populations in cattle Theileria parva. Infection and Immunity 56, (B-cells, CD4+ and 462–467. CD8+ αβ T-cells, and γδ T-cells) could become infected with T. parva 86 1994 McKeever, D.J., Taracha, E.L., Innes, E.L., Proof that CD8 T-cells MacHugh, N.D., Awino, E., et al. (1994) mediate protective Adoptive transfer of immunity to Theileria immunity to T. parva parva in the CD8+ fraction of responding efferent lymph. Proceedings of the National Academy of Sciences USA 91, 1959–1963. 99 1977 Murray, M., Murray, P.K. and McIntyre, W.I. Simple and highly (1977) An improved parasitological technique sensitive test for the for the diagnosis of African trypanosomiasis. presence of Transactions of the Royal Society of Tropical trypanosomes in blood Medicine and Hygiene 71, 325–326. 96 1984 Nantulya, V.M., Musoke, A.J., Rurangirwa, F.R. Study showing that it is and Moloo, S.K. (1984) Resistance of cattle possible to get to tsetse-transmitted challenge with trypanosome Trypanosoma brucei or Trypanosoma serodeme-specific congolense after spontaneous recovery from immunity in cattle syringe-passaged infections. Infection and Immunity 43, 735–738. Continued Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 175 Table 4.2. Continued. Scopus citation Year of percentile publication Publication Impact on the field 93 1978 Barbet, A.F. and McGuire, T.C. (1978) Identification of Crossreacting determinants in variant- conserved variable specific surface antigens of African surface glycoprotein trypanosomes. Proceedings of the National elements that induce Academy of Sciences USA 75, 1989–1993. antibodies and are not exposed on intact trypanosomes 97 1998 Nantulya, V.M. and Lindqvist, K.J. (1989) An important step Antigen-detection enzyme immunoassays towards the for the diagnosis of Trypanosoma vivax, development of a T. congolense and T. brucei infections in diagnostic test cattle. Annals of Tropical Medicine and Parasitology 40, 267–272. 86 1993 Authié, E., Duvallet, G., Robertson, C., and Study showing the Williams, D.J. (1993) Antibody responses presence of to invariant antigens of Trypanosoma wide-spectrum IgG congolense in cattle of differing susceptibility production against to trypanosomiasis. Parasite Immunology trypanosome antigens 15, 101–111. in resistant animals 93 1982 Akol, G.W. and Murray, M. (1982) Early events Description of the following challenge of cattle with tsetse chancre that develops infected with Trypanosoma congolense: in skin development of the local skin reaction. Veterinary Record 110, 295–302. 82 1992 Naessens, J. and Williams, D.J. (1992) Study showing that the Characterization and measurement of CD5+ CD5+ population of B cells in normal and Trypanosoma B-cells is uniquely high congolense-infected cattle. European in the spleen of cattle Journal of Immunology 22, 1713–1718. and especially during infection a cohort of 16 accomplished scientists in similar causative pathogens vary their antigens recog- or parallel fields who worked in Europe and the nized by the host’s immune system. In addition, USA during the same period and who represent both HIV and T. parva infect T-cells as their major the same stages of career development as those target. This results in an inability of the host from ILRAD/ILRI (Table 4.4). The h-index is based immune system to respond effectively and/or on a formula that yields the highest integer h efficiently to the pathogen once the infection is such that h among the investigator’s published established. Plasmodium spp. and T. parva are papers (N ) have collected at least h citations, related phylogenetically as well, belonging to the p while the remaining papers (N – h) have fewer phylum Apicomplexa, and have similar life cycles: p than h citations each. Thus, the h-index reflects both are transmitted by sporozoites delivered in productivity as well as influence in the field. the saliva of ectoparasites and live inside their mammalian host cells as schizonts that develop Cost comparators into merozoites, which are released and infect red blood cells for vector infection with the next To assess the investment in related fields, rough ectoparasite blood meal. cost comparisons were made with research on Funding for research on African trypano- HIV and malaria parasites (Plasmodium spp.). somiasis and theileriosis has been very limited in The similarities of AIDS and malaria to ECF and comparison with funding on HIV/AIDS and trypanosomiasis include the fact that all four malaria. The funding for HIV research by the 176 S.J. Black and C.L. Baldwin Table 4.3. The most highly cited studies on the immune responses to African trypanosomiasis in the mouse model. Scopus citation Year of percentile publication Publication Impact on the field 92 1978 Morrison, W.I., Roelants, G.E., Mayor- Established susceptible/ Withey, K.S. and Murray, M. (1978) resistant strains of mice and Susceptibility of inbred strains of mice linked this trait to immune to Trypanosoma congolense: correlation responsiveness with changes in spleen lymphocyte populations. Clinical and Experimental Immunology 32, 25–40. 82 1985 Black, S.J., Sendashonga, C.N., O’Brien, Integrated parasite C., Borowy, N.K., Naessens, M., et al. differentiation to non- (1985) Regulation of parasitaemia in dividing forms and mice infected with Trypanosoma brucei. development of parasite- Current Topics in Microbiology and specific immune responses Immunology 117, 93–118. 93 1982 Black, S.J., Hewett, R.S. and Control of T. brucei growth Sendashonga, C.N. (1982) (e.g. by a quorum-sensing Trypanosoma brucei variable surface mechanism or an innate antigen is released by degenerating immune response) is a parasites but not actively dividing prerequisite for parasites. Parasite Immunology 4, development of an adaptive 233–244. immune response; 91 1982 Sendashonga, C.N. and Black, S.J. antibodies against the (1982). Humoral immune responses variable surface against Trypanosoma brucei variable glycoprotein (VSG) surface antigen are induced by distinguish between degenerating parasites. Parasite exposed and buried VSG Immunology 4, 245–257. epitopes 88 1979 Morrison, W.I. and Murray, M. (1979) Study showing that resistance Trypanosoma congolense: inheritance to trypanosomiasis is of susceptibility to infection in inbred genetically pre-determined strains of mice. Experimental Parasitology 48, 364–374. Not available 1978 Pearson, T.W., Roelants, G.E., Lundin, Early description of L.B. and Mayor-Withey, K.S. (1978) T-regulatory cells (now Immune depression in trypanosome- known as Tregs) potentially infected mice. I. Depressed contributing to chronicity of T-lymphocyte responses. European infection Journal of Immunology 8, 723–727. Table 4.4. Estimated h-indices of scientists as a measure of research productivity and impact. (Data from author interviews with ILRI scientists.) Scientist category Range of h-indices Median h-index ILRAD/ILRI 14–39 26 Comparators 17–44 29 USA’s National Institutes of Health (NIH) alone not include AIDS research spending by other was approximately US$2.6 billion in 2016 and in governments and international organizations. 2017 (www.hiv.gov/federal-response/funding/ A global mapping of research funding found budget; accessed 6 February 2020), and this does that malaria funding averaged US$1.8 billion Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 177 annually over the period 2006–2010 (Pigott 3. Does the published work show step-by-step et al., 2012). progression of knowledge through key papers in The funding research on bovine immun- top journals? ology and immunoparasitology and on African 4. Were the results published in a timely trypanosomiasis and theileriosis is estimated as manner? 50% of the annual budget of ILRAD and 20% of the annual budget of ILRI (see Introduction, this volume). In the mid-1980s, the budgets for the Bovine immune system ILRAD laboratories working on immunology and Fundamental knowledge of the bovine immune the flow cytometry facility were not more than system was needed to unravel the immune US$2.5 million per year per disease, which in to- responses to bovine theileriosis and AAT so that day’s terms would still be only US$20 million protective and non-protective immune responses per year or less than 1% of the current HIV or could be defined. Such knowledge is needed to malaria yearly budgets. craft vaccines based on logic rather than on trial There is still no effective vaccine for malaria, and error. Thus, the task of describing the bovine HIV or African trypanosomiasis (Table 4.5). For immune system started coincidently with evalu- ECF, there is the original multi-strain ITM vaccine ating the immune responses to the two diseases. known as the Muguga cocktail, which was devel- ILRAD research started when our under- oped at the Kenya Agriculture Research Institute, standing of the mammalian immune system was at Muguga, Kenya (see Chapter 6, this volume). rudimentary compared with current knowledge. This was refined and characterized at ILRAD and At the inception of ILRAD, several fundamental subsequently ILRI, the latter in partnership with the advances were made in immunology. These Global Alliance for Livestock Veterinary Medicines were the invention of the FACS and the develop- (GALVmed). The vaccine is now marketed as a stop- ment of mAb technology followed closely by the gap theileriosis control measure with anticipation ability to grow individual T-cells into popula- of a more effective subunit vaccine in the future. tions (a  method known as ‘cell cloning’). This was coincident with the burgeoning field of mo- lecular biology, from which techniques were soon Historic overview of fundamental and applied to immunology, allowing cloning of genes translational immunological research that encoded cell-surface differentiation antigens of lymphocytes and monocyte/macrophages and genes encoding products of these cells, such This section asks the following questions: as antibodies of several classes or isotypes. 1. Was the original research direction appro- These four technologies allowed the replace- priate? ment of older, clumsier methods for studying 2. Was the original research successful? the roles of individual lymphocyte populations Table 4.5. Comparisons of diseases for which a vaccine is sought. Disease characteristic HIV/AIDS Malaria Trypanosomiasis Theileriosis Antigenic variation ✓ ✓ ✓ (Antigenic diversity) Apicomplexa × ✓ × ✓ Lives inside ✓ ✓ × ✓ host cells Lives in ✓ × × ✓ T-lymphocytes Taken up by × ✓ × ✓ ectoparasites Research investment US$2400 million at US$2550 million US$2.5 million a US$2.5 million a to develop a NIH in 2014 globally in 2010 year at ILRAD in year at ILRAD in vaccine 1980s 1980s 178 S.J. Black and C.L. Baldwin and their products with more precise methods to mid-1980s, with ruminant immunology closely that allowed identification and isolation of indi- following and occasionally preceding the human vidual lymphocyte populations and evaluation and mouse work (Fig. 4.2). In 1980, mAbs of the roles of their molecules (antibodies and specific for bovine immunoglobulin classes were cytokines) in protective immune responses. produced (McGuire and Musoke, 1981) (Fig. 4.2 To understand the importance of the new and Table 4.6) and these were used to identify technologies for improving the ability to study bovine B-cells and their antibody products. bovine cells and their products, we need to look at These different classes of antibodies were shown immunology in the late 1970s. As an example, to have different functions in 1979–1980 at in the 1970s, bovine lymphocytes were separ- ILRAD and by others elsewhere in the world. ated into T- and B-cell populations by passing The T-cell receptor complex that interacts with lymphocytes over nylon wool packed in plastic foreign antigen was defined at this time, along columns such that the T-cells flowed through with the encoding genes. While there was some and the B-cells and monocytes stuck to the fibres. understanding that T-lymphocyte subpopulations This technique was first applied to separation of exhibit a variety of functions (cytotoxic T-cells bovine lymphocytes in 1974 (Rouse and Babiuk, lyse infected host target cells; helper T-cells assist 1974). Bovine T-cells were identified by their B-cells in making and secreting antibodies), it ability to interact with sheep red blood cells in a became apparent that T-cell subsets with these method known as erythrocyte (or E) rosetting, functions could be largely divided into cells that first applied to bovine cells in 1976 (Grewal bore the CD8 and CD4 markers, respectively. In et al., 1976). The red blood cells interacted with addition, their ability to respond to antigen was a T-cell surface molecule that was later desig- dictated by the type of MHC on the presenting nated CD2 and to which mAbs were made at cells or target cells that they interacted with. ILRAD (Baldwin et al., 1988b). B-cells were iden- This was defined in humans in 1982 after prior tified by their ability to interact with sheep red work in mice. As can be seen in Fig. 4.2, this blood cells coated with antibodies and complement understanding was emulated by work at ILRAD in procedures known as erythrocyte–antibody on the bovine immune system published in 1986 rosetting and erythrocyte–antibody–complement (Baldwin et al., 1986; Ellis et al., 1986; Goddeeris rosetting in 1977 (Takashima et al., 1977) and et al., 1986a,b,c). This advance was made possible using polyclonal antiserum that contained by the ability to clone bovine T-cells, as reported antibodies to surface immunoglobulins (Takashima at ILRAD the preceding year (Brown and Grab, et al., 1977). 1985), and by the generation of mAbs at ILRAD These methods of identifying bovine T- and that reacted with bovine CD4 or bovine CD8 B-cells lent themselves to crude separation tech- (Baldwin et  al., 1986; Ellis et  al., 1986). Func- niques, such as density-gradient separation of tional studies were performed on these CD4+ and the rosetted and non-rosetted cells or ‘panning’ CD8+ cell populations using the FACS to purify by coating antibodies specific for surface immuno- the cells, eventually confirming that those bear- globulins on to plastic dishes or beads and allow- ing the CD8 molecule were those that had been ing the reacting cells to adhere while other cells identified as having the ability to kill parasite- were washed away (Usinger and Splitter, 1981). infected host cells in 1982 (Fig. 4.2) (Emery A variety of lectins were employed in 1979 to et al., 1982). stimulate the T- and B-cells to grow in culture The mAbs described in the preceding para- (Pearson et al., 1979b). In addition, in 1981, peanut graph also allowed scientists at ILRAD and elsewhere agglutinin was conjugated to a fluorescent tag to identify a large population of lymphocytes in and used to identify T-cells, i.e. those cells that ruminant (cattle and sheep) blood that did not fit bound peanut agglutinin were considered T-cells into either of these T-cell subpopulations (CD4+ (Usinger and Splitter, 1981); this was a harbin- or CD8+), nor were they B-cells, and thus were ger of the fluorescently tagged mAbs that would called null cells (Baldwin et al., 1988a). This ob- soon be developed to allow efficient identifica- servation was made simultaneously with the tion and isolation of cell populations. discovery of a second type of T-cell antigen- Details of the human and mouse immune specific receptor (TCR) composed of γ and δ chains systems were just being worked out in the early in the mouse and human systems (Fig. 4.2). Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 179 Flow cytometry invented late 1960s – at ILRAD in 1977 mAbs developed1975 – at ILRAD in 1978 T-cell clones derived 1977 – at ILRAD in 1985 1970 1979/80 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1992 TCR gene locus; 4th TCR gene E-rosetting CD5 found (CD2) to identify T-cells CD4+ cells react with Cells expressing MHC class II-bearing TCR complex with cells; CD3 and TCR interact; CD3, TCR  genes + used by sheep andCD8 cells react with TCR  and  genes; δ heterodimer MHC class I; novel TCR gene (TCR ); cattle T-cells CD6 identified; mAb to human CD2 The nature of the δ protein, mAb to CD25-Tac δ molecule is a second TCR Suggested that Third major T3 (later CD3) is mAb to human CD3; T lymphocyte population associated with TCR peptide mapping described in sheep an as yet Caroline comes unknown to ILRI antigen-specific receptor Bovine WC1+ cells identified; Bovine IgG1 and IgG2 mAb to bovine CD2 and CD6; functions; Bovine Bovine IL-2; mAbs to mAb to bovine MHCbovine CD1 characterized mAb to bovine IgM cytotoxic T-cells bovine T-cell bovine CD4 + and class I’s; clones CD8+ and genes in 2nd bovine monocytes Bovine MHC MHC class I locus Adoptive transfer of T-cells class II types between calves shows with T-cell protective role against clones Theileria Clumsy separation techniques for bovine lymphocytes employed: Key: 1974 nylon wool columns to isolate T-cells Historical for cattle 1976–7 rosetting for T-and B-cell identification Human 1981 panning with anti-sIg and gradients for separating rosetted cells, Ruminants fluorescent lectins to identify T-cells Bovine developments from ILRAD Fig. 4.2. Timeline of fundamental developments in T-cell immunology. TCR, T-cell receptor; sIG, surface immunoglobulin. (Constructed by authors from ILRAD archives). 180 S.J. Black and C.L. Baldwin Table 4.6. Milestones in understanding the bovine immune system. Year Milestones Reference(s) 1979 Functions of bovine IgG1 and IgG2 McGuire et al. (1979); Cytotoxoic T-cells to T. parva-infected cells Pearson et al. (1979a,b) Survey of lectins that stimulated bovine T- and B-cells 1980 Anti-IgM reagent made Pinder et al. (1980a,b) mAbs that react with cattle and other Bovidae 1981 Adoptive transfer of thoracic-duct lymphocytes to twin calves to Emery (1981); Emery and prove their role in immunity to T. parva Moloo (1981); Masake and T-cells responsible for immunity to T. parva Morrison (1981); Masake Plasma cells from trypanosome-infected cattle make IgM et al. (1981); Pinder et al. Trypanosome infection suppresses lymphocyte response to (1981a,b) lectins Possible first observation of γδ T-cells responding in autologous mixed-lymphocyte reaction A subset of T-cells transformed by T. parva 1982 Antibodies do not react with cell membrane antigens of Creemers (1982); Emery T. parva-infected cells et al. (1982); Musoke et al. Bovine antibodies do react with trypanosomes (1982); Pearson et al. Bovine IgG2 antibodies block sporozoite infectivity (1982); Wells et al. (1982) Cytotoxic T-cells to T. parva-infected cells Cattle immunized with trypanosome glycoprotein and measurement of antibodies and cellular immunity as demonstrated by lymphocyte proliferation 1983 Phagocytosis of antibody-opsonized trypanosomes using IgM Ngaira et al. (1983) and IgG1 1985 Bovine IL-2 characterized Brown and Grab (1985); T-cell clones maintained with IL-2 for Naessens et al. (1985); 6 months Teale et al. (1985) mAb to T. parva-infected cells made MHC-restricted alloreactive cytotoxic T-cell lines generated 1986 Anti-CD4 mAb made and T-cell helper function demonstrated Baldwin et al. (1986); Ellis Anti-CD8 mAb made and cytotoxic function showed et al. (1986); Goddeeris Monocyte isolation methods et al. (1986a,c); Lalor et al. Cytotoxic clones are MHC restricted (1986); Teale et al. (1986) Alloreactive T-cell clones are CD8+ and MHC restricted Anti-monocyte mAbs made 1987 CD4+ T-cell clones specific for T. parva-infected lymphocytes Baldwin et al. (1987); Baldwin T. parva-infected T-cell clones retain function and MHC and Teale (1987); Ellis specificity et al. (1987a); Emery et al. Monocytes stimulate autologous and allogeneic mixed- (1987); Goddeeris et al. lymphocyte reaction (1987); Teale et al. (1987); Non-lymphoid cells in afferent lymph present antigen to T-cells Teale and Kemp (1987) γδ T-cells respond in the autologous mixed-lymphocyte reaction MHC class II molecules typed with CD4+ T-cell clones 1988 mAb (IL-A29) to a unique surface-determinant WC1, Baldwin et al. (1988a,b); on γδ T-cells Bensaid et al. (1988); Ellis Bovine CD4+, CD8+ and γδ T-cells and B-cells are infected et al. (1988); MacHugh et al. with T. parva (1988); Naessens et al. (1988); Anti-CD2 and anti-CD6 mAb made Brown et al. (1989a,b) BoLA class I characterized on cells CD1 characterized Ig allotypes Continued Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 181 Table 4.6. Continued. Year Milestones Reference(s) 1989 T-cell lines and clones used to characterize T. parva antigens Brown et al. (1989a,b); Alleles of CD5 Howard et al. (1989); Olobo In vitro activation of bovine B-cells and Black (1989) WC1+ γδ T-cells described mAb IL-A25 characterized Haemopoietic stem cell cultures 1990 mAb to polymorphic determinants of BoLA class I MHC Clevers et al. (1990); molecule Dobbelaere et al. (1990); Young calves resist T. parva Fritsch and Nelson (1990); mAb to mature B-cell antigen made Kemp et al. (1990); Koch Second MHC class I locus gene cloned et al. (1990); Naessens Panel of mAbs reactive with monomorphic and polymorphic et al. (1990); Toye et al. antibody epitopes (1990); Williams et al. (1990); Bensaid et al. (1991) 1991 γδ T-cells kill T. parva-infected cells Goddeeris et al. (1991); Distinction of naïve and memory CD4+ T-cells Howard et al. (1991); Lutje CD8+ and γδ T-cells suppress CD4+ T-cell responses shown and Black (1991); Veiled cells present antigen efficiently McKeever et al. (1991) 1992 Anti-tumour necrosis factor (TNF)-α mAbs made that neutralize Naessens and Williams TNF (1992); Sileghem et al. Limiting dilution analysis for cytotoxic T-cells established (1992); Taracha et al. Response by CD5+ B-cells (B1 B-cells) in trypanosome-infected (1992) cattle 1993 WC1 gene family and mAbs analysed MacHugh et al. (1993); Bovine B1 cells defined Naessens (1993) 1994 WC1+ γδ T-cells respond to invariant trypanosomiasis antigen Flynn and Sileghem (1994) 1997 Role of CD5+ B-cells in trypanosome-infected cattle Buza et al. (1997) 1999 γδ T-cells respond to T. parva-infected host cells Daubenberger et al. (1999) 2007 The adjuvant CpG enhances CD4+ T-cell responses Graham et al. (2007) 2011 The adjuvant Flt3L + granulocyte–macrophage colony- Mwangi et al. (2011) stimulating factor as adjuvant for CD4+ T-cell responses It  was fortuitous that lambs and calves have shops for mAbs to human antigens gave rise to such a large proportion of γδ T-cells in their per- the cluster-defined (CD) antigens of humans. ipheral blood mononuclear cell (PBMC) popula- These workshops were held at intervals of several tions, making the likely relevance of this new years to designate mAbs to human cell-surface type of TCR to ruminant immunology immedi- antigens, while those studying other species ately apparent. followed suit, conducting species-specific work- It was postulated that the large ‘null’ cell shops and naming their genes in parallel to the population of lymphocytes within the PBMC human CD system.) populations were cells expressing this newly The presence of mAbs defining various discovered TCR (i.e. γδ T-cells). The ‘null cell’- members of this family of γδ T-cell-unique cell- specific mAb T19 of sheep and another for cattle surface antigens facilitated the isolation of γδ made at ILRAD precipitated molecules ranging T-cells. The mAbs include those pan-reactive from 180 to 240 kDa (Mackay et al., 1989). The with all or most WC1 family members as well as molecules turned out to be part of a multigene others that are specific for a subgroup of WC1 family whose members were eventually designated gene products and which thereby define subsets Workshop Cluster 1 (WC1) by an international or subpopulations of WC1+ γδ T-cells (Chen et al., workshop held to group mAbs according to simi- 2009). This was followed by the formal demon- lar reactivity with bovine antigens (Morrison stration that cells reactive with the WC1-specific and Davis, 1991; Sopp et al., 1991). (Such work- mAbs were in fact T-cells expressing the γ and δ 182 S.J. Black and C.L. Baldwin TCRs in both sheep and cattle from studies at generated was crucial for advancing institu- ILRAD and others (Mackay and Hein, 1989; tional research in the early stages and defining Mackay et al., 1989; Clevers et al., 1990). Further the fundamentals of the bovine immune system studies showed that not all γδ T-cells bear this (e.g. CD4+ and CD8+ T-cells, immunoglubulin lineage-specific marker. Thus, in organs such as functions, monocytes and efficient antigen- the spleen, the majority of γδ T-cells do not express presenting cells), as well as bovine peculiarities WC1, i.e. are WC1– (MacHugh et al., 1997), and relative to mouse and human systems (WC1+ γδ in the uterus no γδ T-cells express WC1 (Tuo T-cells). Simultaneous with studies at ILRAD, et al., 1999). In contrast, as few as 1% of the per- the ruminant immune system, including that of ipheral blood γδ T-cells in cattle are WC1– (Bald- cattle, was being evaluated at the Basel Institute win et al., 2000), although this can change with of Immunology, the Walter and Eliza Hall Insti- the conditions under which the cattle are held. tute in Melbourne, the John Curtin School of For example, in one study, it was shown that the Medicine in Canberra, and Washington State proportion of the WC1+ relative to the WC1– popu- University, among other institutions. As part of lation decreased when the cattle were moved this work, many more mAbs were developed. from open grazing at the ILRI Kapiti Ranch to a Findings by the various groups regarding mAbs Biosafety Level 2 (BSL2) facility on the ILRI cam- and their targets coalesced in a number of inter- pus (Baldwin et  al., 2000), suggesting that the national workshops in which ILRAD scientists WC1+ population was maintained at a high level played fundamental and leadership roles. ILRAD/ by stimulants such as ticks and undefined envir- ILRI scientists were key leaders and organizers as onmental antigens not found in a BSL2 facility. well as active participants in ‘mAb workshops’ Such studies provide models for understand- that were held every few years to compare re- ing expansion of particular γδ T-cell populations agents against bovine immune system molecules in otherwise apparently healthy animals and (Morrison and Davis, 1991; Howard and Naes- provide a useful comparison to the response of sens, 1993; Naessens and Howard, 1993). These γδ T-cells to both T. parva-infected host cells and workshops involved scientists exchanging re- trypanosome components, discussed in later sec- agents and then analysing them in a pre-agreed tions of this chapter. The simultaneous progress manner in a number of assays and comparing in understanding basic bovine immunology their results. ILRAD hybridomas (the transformed and immune responses to disease is noted in cell lines that produce mAbs) were also gener- Table 4.6. At times, the advances obtained using ously distributed around the world. the infection systems preceded our definition of the As determined by the literature review here, immune system; in other cases, the elements immunological research at ILRAD/ILRI has had defined in the basic immunological studies were a substantial impact worldwide. The right ap- subsequently applied to infection pathology and proach was taken, and appropriate tools were immunity. developed to allow further understanding of the Because these studies showed that the bo- ruminant immune system in support of vaccine vine immune system was fundamentally similar development and efficacy testing according to to that of humans, knowledge from one is reason- the mandate of ILRAD. Moreover, the research ably applicable to the other (i.e. one can assume team that undertook these studies was reflective with optimistic caution that the same principles and creative in generating a vast number of apply to both). While only a limited repertoire of tools to study the bovine immune system using tools may be required to develop the ‘big picture’ the most modern technologies and following in the context of disease and vaccine efficacy in closely the discoveries informing the fundamen- livestock species, tools that identify unique aspects tals of the murine and human immune systems. of the ruminant immune systems relative to that In addition, in some cases, knowledge about the of mice and humans are also needed, as these components and function of the bovine system species-specific immune response elements may preceded that in either mice or humans. These be important in vaccine design and efficacy studies resulted in a substantive body of work, as (e.g. to include the responses of γδ T-cells). shown by the examples in Table 4.6 of dozens of The participation of ILRAD scientists with peer-reviewed publications. The timeline in scientists from around the world in characterizing Fig. 4.2 indicates that key findings during the the bovine immune system through the mAbs nascent period of molecular immunology were Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 183 published in a timely manner, being interwoven the importation into the USA (again to Plum with those in humans and mice or occurring in Island) of a large collection of ILRAD/ILRI mAbs, the same scientific era. Most of this work was where they are being tested for safety. published in international peer-reviewed journals of immunological scientific societies as follows: Immunoparasitology of theileriosis Journal of Immunology (American Association of Immunologists), Infection and Immunity (Ameri- Before ILRAD was established, research at KARI, can Association of Microbiology), Immunology at Muguga, some 15 km from ILRAD, had shown (British Society of Immunology), European that T. parva sporozoites derived from infected Journal of Immunology (European Federation tick salivary glands can be cryopreserved (Cun- of Immunological Societies) and Veterinary ningham et al., 1973), that the sporozoites can Immunology and Immunopathology (American transform bovine lymphoid cells in vitro (Brown Association of Veterinary Immunologists). So, et al., 1973) and in vivo, and that cattle can be while not necessarily published in the high- immunized by infection with these parasites in est-impact journals, in part because of its itera- combination with long-acting tetracycline or tive nature, the work was nevertheless published other drugs (Radley et al., 1975; McHardy et al., in well-respected journals. To a lesser extent, IL- 1976). Immunization by simultaneous infection RAD/ILRI contributed to making reagents or and chemotherapeutic treatment was effective tools to measure the products of macrophages against homologous challenge, and a cocktail of and T-cells (i.e. cytokines). The production of parasites had been established, called the ‘Mu- mAbs to tumour necrosis factor (TNF)-α at ILRI guga cocktail’, which protected against several was notable, especially since this particular cyto- strains of T. parva that are endemic in East Africa kine is associated with anaemia, which is a char- but not against all strains of T. parva and not acteristic of AAT in cattle. Generating anti- against many strains of Cape buffalo-derived cytokine mAbs for cattle and other ruminants T. parva (subsp. lawrencei). Knowing that cattle was addressed more thoroughly by groups at could be immunized against T. parva subsp. parva Compton (UK), Namur (Belgium) and Mel- strains, ILRAD set out to identify the protective bourne (Australia). immune response, the nature and diversity of The measurable outputs and impacts from host-protective T. parva antigens and the best the generation of reagents to study the bovine administration of these immunogens to induce immune system can be found in the global broad host protection. The seeming modesty of e xports of the mAbs to other parts of the world these aims was belied by the concomitant need including the USA and Europe. In 1990, a group to define cells and function in the bovine immune of US-based scientists working in bovine im- system, as discussed above, to determine the munology paid for safety testing of the ILRAD respective roles of antibody-mediated immunity mAbs to allow their importation into the USA. and cell-mediated immune responses in protec- This meant that the hybridoma cell lines had to tion. Scientists needed to determine not only the be sent to a US government containment labora- identity of the protective parasite antigens but tory (Plum Island, New York) and tested in live- also their antigenic stability and their utility in stock before being allowed on the USA mainland. vaccination protocols. The group of hybridomas included those that se- In addition to these goals, fundamental crete antibodies specific for MHC class I and II gaps in knowledge cried out for resolution. What (IL-A19 and IL-A21), CD4+ and CD8+ T-cells types of lymphocytes are infected and transformed (IL-A11, IL-A12 and IL-A51), macrophages by T. parva? Does the cell type infected affect (IL-A15) and others. Some of these imported the induction of protective immunity, i.e. do the mAbs were sold by the Washington State-based infected cells retain functional specialization company Veterinary Medical Research and De- and release molecules that misdirect the im- velopment (VMRD), which continued to offer mune responses? What prevents infected cattle these until a few years ago. Serotec sells ILRAD/ from achieving immunity in the absence of drug ILRI-made mAbs in Europe and the USA, and intervention? These were and are reasonable the American Type Culture Collection holds a questions that inform strategies to better control few hybridomas that were generated at ILRAD theileriosis in cattle, but were they the right available for purchase. Currently under way is questions? Would it perhaps have been better to 184 S.J. Black and C.L. Baldwin have invested a larger amount of effort in typing transmission. Stuart Shapiro and others at ILRAD T. parva strain diversity and improving the ITM, obtained some evidence using immunized rab- which has obvious immediate translational value, bits to support this approach (Shapiro et  al., or in focusing on prophylactic immunity against 1989), but this was not pursued in depth tick infestation? because elucidating and testing possible protect- The T. parva sporozoite ITM developed at ive antigens in tick saliva and their polymor- Muguga was addressed at ILRAD under the lead- phisms was considered a less-assured invest- ership of Tony Irvin and continues today under ment than identifying protective T. parva the guidance of Philip Toye. To effectively under- antigens. However, interest in an anti-vector ap- stand the efficacy of ITM as a potential vaccine, proach, while not a focus of ILRAD’s attention, T. parva strain diversity needed to be defined and has received renewed attention in Europe, where was addressed at ILRAD by many investigators investigators from several countries are using using mAbs against schizont antigens, bovine proteomic and transcriptomic approaches to T-cell clones specific for T. parva-infected cells and identify candidate vaccine antigens in the saliva molecular genetic approaches. Immunization of Ixodes ricinus (Sprong et al., 2014). It can be against theileriosis by ITM has some notable hoped that these studies will be successful and successes, e.g. protection of 97.6% of immun- will accelerate development of an effective vac- ized cattle in two districts of northern Tanzania cine against Rhipicephalus appendiculatus, the vec- against field challenge (Lynen et al., 2012). This tor of T. parva, which is also an ixodid tick. How- ‘live’ vaccine continues to be produced at ILRI, ever, even if candidate anti-tick vaccine antigens with nearly two million cattle immunized with are identified, these may be polymorphic and the ILRI-produced vaccine. With assistance from protective immune responses difficult to sustain GALVmed, a not-for-profit global alliance that through natural challenge alone, making the provides livestock treatments in developing coun- success of this approach not assured. Before tries, ILRI has registered the vaccine in Kenya, leaving the discussion of anti-vector immunity, it Malawi and Tanzania. is noteworthy that ILRI staff have recently iden- While the live vaccine approach serves as an tified and cloned tick gut proteins that, when re- interim disease control measure, it remains less peatedly administered to cattle, significantly de- desirable than a killed or subunit vaccine because crease the moulting success of R. appendiculatus of the requirement for infective particles (sporo- nymphal ticks to adults (Olds et al., 2012). While zoites) with the concomitant problems associated this approach has relevance to vaccine-based with the need for a reliable cold chain, carefully vector control, the need for repeated immuniza- titrated infecting particles and renewable propa- tion and the absence of a natural boost impose gation of stabilates to sustain the infection regime, constraints to its efficacy. The scientific consen- which is not straightforward given their genetic sus in ILRAD’s early years was that the focus on im- diversity (Patel et  al., 2011). Indeed, cocktail mune responses against T. parva and not on the combinations appropriate to a given treatment tick vector was the most promising approach. area must be developed. Because of live infection, Thus, towards the question of whether the the T. parva sporozoite ITM also carries the dan- ILRAD team indeed asked the right questions in ger of establishing carrier status in the treated the right way, including being creative and with animals. In addition, it is relevant to note that a focus on translational outcomes and taking ILRAD was mandated to perform intensive im- advantage of advances in understanding the bo- munological research on trypanosomiasis and vine immune systems made at ILRAD and else- theileriosis leading to improved control of dis- where, and whether they further implemented ease, a call to action that required more than an new technologies in a timely manner to achieve intensified focus on the ITM. its goals of generating a subunit T. parva vaccine, With regard to the tick vaccine approach, it we turn to the quantitative impact analysis pre- can reasonably be argued that the induction of sented above. It makes a case for solid and even an antibody response against a conserved and prolific productivity of ILRAD investigators, evi- physiologically relevant tick salivary protein denced through the body of their published work might be sustained by natural boosting and and, moreover, through continued focus on the hence prevent both tick infestation and T. parva fields of veterinary immunology and immuno- Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 185 parasitology in post-ILRAD careers, including cells in a parasite strain- and host MHC class on theileriosis vaccine development. Hence, we I-restricted manner (Goddeeris et  al., 1986c), will highlight what we consider substantive find- that CD4+ T-cells expedite priming of the T. parva- ings in theileriosis research and their temporal specific CD8+ T-cells (Taracha et al., 1997) and relationship to each other and to discoveries in that adoptive transfer of CD8+ T-cells from an ruminant immunology made at ILRAD. immune to a non-immune chimeric twin conveys Rapid progress was made in identifying protective immunity to homologous parasite candidate mechanisms of protective immunity. challenge (McKeever et al., 1994b). In addition, Terry Pearson and colleagues in 1979 showed T. parva sequences encoding protective CD8+ T-cell that T. parva-infected leukocytes stimulate au- epitopes have been identified and shown to induce tologous peripheral blood lymphocytes (PBLs) T. parva-specific cytotoxic T-cells (Graham et al., from naïve and primed cattle to proliferate in 2006, 2008). Work from Kariuki et  al. (1990) vitro and, in the case of PBLs from primed ani- showed that immunization of full-sibling cattle mals only, induce the generation of cytotoxic by infection with Cape buffalo-derived T. parva cells that lyse T. parva-infected cells but not unin- lawrencei and subsequent pharmacotherapy gen- fected lymphoblasts (Pearson et al., 1979a). This erated cytotoxic T-cells that recognized common work established as research priorities the iden- T-cell epitopes of T. parva-infected lymphocytes, tification of cytotoxic cells and elucidation of which is good news for T. parva vaccine develop- both their role in mediating protective immun- ment, since inclusion of these common epitopes ity in vivo and the identity of their target antigens. in the vaccine are likely to increase its efficacy. A few years later, Tony Musoke and colleagues However, work by MacHugh et al. (2011) showed showed that antibodies of the IgG2 class in that immunodominant T-cell epitopes on T. parva- serum from cattle that had recovered from ECF, infected lymphocytes showed genetic diversity and in serum from rabbits immunized with T. parva consistent with antigenic variation, which raises sporozoites, prevented sporozoites from infecting issues with respect to the selection of candidate and transforming bovine PBLs in vitro (Musoke vaccine epitopes. et al., 1982). This finding identified the sporozo- With respect to a sporozoite-neutralizing vac- ite coat protein as a candidate vaccine antigen cine, the gene encoding the T. parva sporozoite sur- and heralded an additional set of research prior- face coat, p67, has been cloned, and immunization ities, namely, characterizing the protein, its with baculovirus expressing the p67 recombinant neutralizing epitopes and their genetic stability, protein has been shown to induce production of and ways of immunizing to induce sustained sporozoite-neutralizing antibodies (Kaba et  al., production of sporozoite-neutralizing antibodies. 2003) and to protect immunized cattle against syr- Table 4.7 lists findings that led to the identi- inge challenge with T. parva sporozoites (Kaba et al., fication of T. parva-specific T-cell epitopes that 2005). However, immunity against tick-transmit- stimulate protective immunity, while Table 4.8 ted sporozoites appears to be considerably lower lists the progress towards identifying B-cell epi- (Vish Nene, ILRI, 2015, personal communication), topes on the T. parva sporozoite coat that stimu- raising the possibility that components of tick saliva late neutralizing antibodies. These tables illus- might antagonize interactions between neutralizing trate the steady and sustained progress towards antibodies and sporozoites. development of a subunit vaccine that stimulates While the T. parva vaccine programme has production of antibodies that neutralize sporozo- not yet resulted in full disease control, contributing ites and induces primed MHC and T.  parva-re- scientists have identified, cloned and expressed stricted CD8+ T-cells that mount recall responses recombinant T. parva proteins that stimulate to and lyse T. parva-infected cells, and CD4+ T-cells protective T-cells and antibody responses (Nene that expedite development of the cytotoxic T-cell et al., 1992, 1995; Graham et al., 2006). They response. Many of the aims of the T. parva im- have further provided evidence that a subunit munology programme have been achieved, but vaccine, perhaps comprising a sporozoite p67 development of an effective subunit T. parva vac- subunit that elicits sporozoite-neutralizing anti- cine remains a challenge, even in 2020. bodies together with schizont epitopes that stimu- With respect to T-cell immunity, it was ele- late appropriate CD4+ helper T-cells and CD8+ gantly shown that CD8+ T-cells kill T. parva- infected cytotoxic T-cells in cattle, is a realistic possibility. 186 S.J. Black and C.L. Baldwin Table 4.7. Immunity to T. parva-infected lymphocytes and identification of candidate vaccine antigens: timeline of substantial findings. Year Finding Reference 1979 PBLs from T. parva-immune animals are shown to mount T. parva- Pearson et al. (1979a) infected cell-specific proliferative and cytotoxic responses upon stimulation with T. parva-infected lymphocytes in vitro 1981 T. parva-infected cell-specific cytotoxic cells are shown to arise Eugui and Emery (1981) in the blood and lymph of T. parva-immune cattle following reinfection and are restricted by host polymorphic antigens 1981 Resistance to lethal challenge with T. parva (Muguga) is shown Emery (1981) to be transferred from immune to non-immune chimeric twins by cells in thoracic-duct lymph 1986 Cytotoxic lymphocytes specific for T. parva-infected cells are Goddeeris et al. (1986b,c); shown to be restricted by target-cell class I MHC antigens Morrison et al. (1987) 1987 Proliferation and production of T-cell growth factor(s) by cloned Baldwin et al. (1987) bovine CD4+ T-lymphocytes, putatively helper T-cells, that are specific for T. parva-infected cells are shown to be both T. parva strain and host-cell MHC class II restricted 1989 T. parva-specific helper and cytotoxic bovine CD4+ T-cells are Brown et al. (1989b); shown to proliferate in response to macroschizont membrane Baldwin et al. (1992); and infected-cell extract high-speed supernatant processed Brown et al. (1989a) by antigen-presenting cells 1994 Transfer of bovine CD8+ T-cells from efferent lymph of immune McKeever et al. (1994b) T. parva-infected cattle into their non-immune monozygotic twins is shown to convey immunity to lethal T. parva sporozoite challenge 1997 T. parva-immune bovine CD4+ T-cells are shown to expedite Taracha et al. (1997) development of T. parva-restricted bovine CD8+ cytotoxic T-cells in vitro 1999 Bovine γδ T-cells proliferate in response to and lyse T. parva- Daubenberger et al. (1999) infected lymphocytes and are not restricted by class I or class II MHC antigens 2006 Candidate T. parva vaccine antigens that stimulate bovine CD8+ Graham et al. (2006) cytotoxic T-cells are identified by screening antigen-presenting cells transiently transfected with cDNAs of selected T. parva schizont genes encoding secretory proteins and schizont membrane antigens 2008 Cytotoxic lymphocyte epitopes on candidate T. parva vaccine Graham et al. (2008) antigens are identified by immunoscreening on peptide- pulsed fibroblasts and transfected COS-7 cells and confirmed as relevant using cytotoxic lymphocytes from cattle subjected to T. parva sporozoite ITM 2009 The MHC haplotype of responder cattle determines the T. parva MacHugh et al. (2009) epitopic bias of bovine CD8+ cytotoxic cells and hence specifies immunodominance 2011 An immunodominant T. parva epitope is shown to undergo MacHugh et al. (2011) antigenic variation From this promising work, the T. parva vaccine berger et al., 1999), indicating the presence on project has been revitalized by support from the T. parva-infected cells of T. parva-induced or Bill & Melinda Gates Foundation (2005) awarded -d erived antigens that are not MHC restricted to a consortium comprising mostly ILRAD and which should be identified and considered alumni and current ILRI faculty. Work at ILRI for inclusion in a subunit vaccine. has shown that T. parva-infected cells stimulate While not of direct importance to the devel- the development of cytotoxic γδ T-cells (Dauben- opment of a T. parva vaccine, several grace notes Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 187 Table 4.8. Timeline of substantial findings in immunology of theileriosis: humoral immunity. Year Finding Reference 1982 ECF immune serum and serum of rabbits immunized with T. parva Musoke et al. (1982) sporozoites contain neutralizing IgG2 anti-sporozoite antibodies 1985 Infection of bovine lymphocytes by T. parva sporozoites is inhibited Dobbelaere et al. (1985) by a mAb that recognizes a 68 kDa circumsporozoite protein 1992 The gene encoding a 67 kDa T. parva circumsporozoite protein is Nene et al. (1992) cloned and antibodies against recombinant proteins containing residues 9–316 and 397–709 neutralize sporozoite infectivity in vitro 1992 A recombinant fusion protein of T. parva sporozoite protein p67 and Musoke et al. (1992) a non-structural gene of influenza A virus induces sporozoite- neutralizing antibodies in cattle 1998 The antibody response of cattle to the p67 sporozoite antigen is Gentschev et al. (1998) enhanced by fusion to the C-terminal secretion signal of Escherichia coli haemolysin and expression in Salmonella enterica Dublin 1998 Immunization with live attenuated S. enterica Dublin expressing Heussler et al. (1998) sporozoite p67 confers partial protection against syringe injection of T. parva sporozoites 2003 Expression of sporozoite p67 N- and C-terminal domains fused to Kaba et al. (2003) the baculovirus envelope glycoprotein GP64 results in retention of native antigen conformation 2005 Immunization with a baculovirus-derived sporozoite p67 subunit Kaba et al. (2005) induces sporozoite-neutralizing antibodies and provides up to 85% protection against syringe challenge with T. parva sporozoites enrich the legacy of ILRAD’s contributions to bers of non-specific cytotoxic cells in the lymph theileriosis research. With respect to the biology node draining the site of T. parva infection in of theileriosis, bovine CD4+ T-cells, CD8+ T-cells, naïve cattle (Houston et al., 2008). This response null cells (now known to be γδ T-cells) and B-cells does not arise in T. parva-immune cattle, which were all shown to be targets for infection and instead generate large numbers of T. parva- and transformation by T. parva sporozoites (Baldwin host MHC-restricted cytotoxic T-cells. It is possible et  al., 1988a), and infected cloned T-cells were that this poorly understood non-specific cyto- shown to retain function for some months after toxic response contributes to the inability of the infection (Baldwin and Teale, 1987), although infected cattle to control T. parva-infected cells, this was not the case with infected B-cells. Host- for example by killing responding T. parva-specif- cell invasion by T. parva sporozoites was shown ic T-cells. If these non-specific cytotoxic cells are to result from an initial energy-independent activated by conserved T. parva antigen, such zippering of the sporozoite coat (later to be iden- antigens might provide additional targets for an tified as p67) and host-cell plasma membrane anti-pathology vaccine. receptors, and subsequent energy-dependent removal of the host endocytic membrane that Immunoparasitology of trypanosomiasis prevents lysosomal fusion and confounds intra- cellular defences (Fawcett et al., 1984). Work by As with T. parva, a good deal of information re- ILRAD alumna Dirk Dobbleaere and colleagues garding African trypanosomes was available to uncloaked much of the mystery surrounding researchers seeking to develop a vaccine against how T. parva transforms host lymphocytes and AAT. It was known at the inception of ILRAD that usurps the host-cell mitotic apparatus to ensure there are several species of mammal-infective partition of schizonts into daughter cells (von African trypanosomes, all of which can be trans- Schubert et al., 2010). In addition, relatively re- mitted in the saliva of tsetse flies, and that there cent work at ILRI has rediscovered an old unsolved are several species of tsetse that can serve as mystery involving the development of large num- vectors and are adapted to different habitats. In 188 S.J. Black and C.L. Baldwin addition, methods to cryopreserve bloodstream- stage in our research’. Murithi suggested that if stage African trypanosomes and to use these to the answer turns out to be ‘no’, it might be better infect other animals had been described (Herbert to invest research effort into other areas of tryp- et al., 1968), as had methods to purify the para- anosomiasis research, such as chemotherapy sites from host blood (Lanham, 1968). It was also and vector control. At the same time as the known that bloodstream-stage trypomastigotes i nauguration symposium, A.J.S. Davies, of the are killed by antibodies specific for their immu- Chester Beatty Research Institute, raised an nodominant variable surface glycoprotein (VSG), additional issue. After commenting on the cost- and that the parasites undergo antigenic variation effective development of earlier vaccines, which by differential expression of VSGs. Furthermore, resulted from empirical testing of attenuated the VSG from a Trypanosoma brucei variant had organisms or organisms that induce cross- been isolated and characterized (Cross, 1975, protective immunity, he summarized the im- 1977). While research towards a Theileria parva munological focus of current disease research as vaccine at ILRAD/ILRI continues to progress, follows, ‘Rather to the chagrin of older parasit- this is not the case with research to develop a ologists... the immunologists have crept into the vaccine effective against AAT. In the case of AAT, [vaccine development] act with little but their high variation in protective antigens blocked the enthusiasms and wish to do something useful development of an effective vaccine from the out- to support this’. He rightly pointed out that im- set and led to the programme’s closure in 2001. mune responses are complicated and that the When ILRAD work began in 1974, there was study of immune responses per se might not evidence that metacyclic trypanosomes, which reveal ways to manipulate them to ensure con- are the mammal-infective forms derived from trol of highly antigenically variable pathogens. tsetse, revert to one or a few basic antigen types Notwithstanding this oblique cautionary note when transmitted to cattle. However, by the time and bolstered by the institution’s mandate to ad- ILRAD’s campus at Kabete was inaugurated in dress the immunology of parasitic infections, 1978, this view had changed, and it was no the ILRAD faculty working on the immunology longer considered likely that trypanosome infec- of AAT pressed ahead with a ‘let’s have at it tions could be prevented by prior immunization anyway’ attitude. of hosts with a few basic variable surface anti- Thus, through the 1980s, investigators at gens (reviewed by Vickerman, 1978). Indeed, ILRAD continued to study trypanosome VSG only 11 years later, ILRAD’s Peter Gardiner antigenic variation and genetic mechanisms wrote ‘the feeling has arisen that antigenic vari- underlying this process, under the leadership of ation, as demonstrated by the Trypanozoon and Jack Doyle and Dick Williams, and to examine Nannomonas subgenera of trypanosomes, is too the extent to which cattle could be immunized extensive, the number of serodemes too large, against AAT by injection of dead or dying tryp- and the coexistence of different species in many anosomes, or components of trypanosomes, or areas too complicated, to allow any immuno- by ITM under the leadership of Max Murray. The prophylaxis based on antibodies to variable anti- latter investigations showed that immunization gens’ (Gardiner, 1989). Gardiner held out hope of cattle with lethally irradiated trypanosomes that the case might be simpler with Trypanosoma with or without their VSG surface coat, with vivax, which belongs to the Dutonella subgenera. so-called ‘uncoated’ insect (procyclic) forms of However, there is still no proof that the antigenic the parasites, or with trypanosome membrane, repertoire of T. vivax is particularly limited, either or by infection with an assembly of metacyclic in metacyclic or bloodstream-stage VSGs. parasites from various isolates followed by tryp- Concern about the development of a tryp- anosome pharmacotherapy, did not prevent or anosomiasis vaccine was implied in a presen- reduce the severity of disease upon challenge with tation made by I.E. Murithi, then Director of heterologous trypanosomes. Instead, immuniza- Veterinary Services in the Kenya Ministry of tion with VSG-coated trypanosomes induced im- Agriculture, at the 1978 ILRAD inauguration munity to homologous challenge (Emery et  al., symposium. Murithi stated with respect to the 1982) – as in some circumstances did infection trypanosomiasis vaccine that ‘our research ef- with tsetse-transmitted Trypanosoma congolense fort must be planned in such a manner that an metacyclic forms – followed at an appropriate authentic answer is obtained at a relatively early time by trypanocidal chemotherapy. Resistance Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 189 of animals made immune by ITM to the tset- et al., 2007). These molecules were predominantly se-transmitted metacyclic parasites was coinci- poly-N-acetyllactosamine-bearing endosomal dent with the presence in serum of antibodies proteins that contain the receptor for transfer- that neutralized the metacyclic parasites (Akol and rin, a putative receptor for serum low-density Murray, 1985; Dwinger et al., 1987b), suggest- lipoprotein and most likely other receptors ing that protection against homologous tsetse (Nolan et al., 1999). Selective immunizations of challenge was, at least in part, based on the mice with the putative receptor for low-density development of antibodies specific for expressed lipoproteins (Bastin et  al., 1996) and with a VSGs. While these studies, which are summar- receptor for transferrin (Steverding, 2006), ized below, contributed valuable information on performed by scientists unconnected to ILRAD, the ruminant host immune response to trypano- also failed to elicit non-variant immunoprotec- somes, the findings were not viewed at the time tion under nutrient-sufficient conditions. Des- as supporting the possibility of vaccine develop- pite the lack of success of the nutrient receptor ment because of the absence of immunity to vaccine approach so far, the choice to take this heterologous infection. approach still seems reasonable to the chapter With the loss in enthusiasm for effective im- authors. munization against AAT using a composite The remaining approach taken by scientists VSG-based vaccine or by tsetse-transmitted working on the immunoparasitology of trypano- ITM, researchers at ILRAD chose two main ap- somiasis at ILRAD was to identify mechanisms proaches to try to affect immunological control of trypanotolerance in Cape buffalo, N’Dama of the disease. The first was a search for immu- cattle and certain inbred strains of mice, with a noprotective conserved antigens expressed by focus on control of parasitaemia and anaemia. trypanosomes, an approach based on the know- Understanding cell and molecular mechanisms ledge that bloodstream-stage African trypano- of trypanotolerance was viewed as a route to somes require lipids and iron and hence may identify ways to decrease the trypanosomiasis have antibody-accessible receptors for their susceptibility of livestock, such as through the macromolecular carriers; the second was a development of an anti-pathology vaccine, if key search for mechanisms of trypanosomiasis trypanosome pathology-inducing ligands could tolerance expressed in animals with genetically be identified, or by selective breeding, or, as only acquired resistance to the disease, which is dis- dreamed then, by genetic engineering if key cussed below. r esistance genes could be identified. Trypanotol- Optimism that non-variant immunoprotec- erance in both livestock and Cape buffalo was tion could be induced by immunization with shown to be a genetically acquired trait in two conserved nutrient receptors has proven un- magnificent experiments. The first was the im- founded, at least so far, although the idea that portation of N’Dama embryos from a herd in trypanosomes require macromolecular nutrients West Africa and their successful implantation was confirmed using axenic cultures of blood- into surrogate mothers at the ILRAD farm ad- stream-stage T. brucei developed at ILRAD. The joining the ILRAD campus (Jordt et  al., 1986), parasites were shown to require serum lipopro- thus realizing the vision of ILRAD scientists Max teins (low, intermediate or high density) from any Murray and Jack Doyle, and ILRAD’s then Dir- of several mammal species to replicate in vitro, ector General, Ross Gray, to establish a research and their sustained replication also required the herd of N’Dama at ILRAD. These animals, which iron-carrying molecule transferrin (Black and were born from mothers that had never experi- Vandeweerd, 1989). Reasoning that receptors for enced trypanosomiasis and that were raised these macromolecular nutrients would concen- and bred on the ILRAD campus (i.e. in a tsetse- trate in endocytic vesicles, Stuart Shapiro assessed and trypanosomiasis-free area) expressed the immunization with purified clathrin-coated ve- trypanotolerance traits of their breed, as shown sicles from T. brucei. However, this immunization by their superior control of parasitaemia and regime did not induce non-variant immunopro- anaemia after tsetse or syringe infection with tection in a mouse model (Shapiro, 1994), a dis- African trypanosomes when compared with appointing result that was later confirmed by similarly infected trypanosomiasis-susceptible immunization of rabbits and mice with tomato Boran cattle (several references are provided in lectin-purified parasite components (Guirnalda Table 4.9). The second approach was taken by 190 S.J. Black and C.L. Baldwin Table 4.9. Findings about the immunology of AAT in ruminants. Year Finding Reference 1981 PBLs from Boran cattle infected with T. congolense have decreased Masake et al. responses to mitogens (lipopolysaccharide, phytohaemagglutinin) (1981) compared with PBLs from uninfected cattle 1982 Development of a chancre at the site of infected tsetse bites is due to the Akol and Murray metacyclic trypanosome and is not a response to the fly bite (1982) 1982 Trypano-destructive antibodies bind to a subset of VSG epitopes that are Emery et al. (1982) exposed on the surface of the parasites 1982 Cattle, particularly N’Dama, that survive trypanosome infection make Shapiro and Murray antibodies against a broader repertoire of trypanosome antigens than (1982) those that die 1983 Infection of cattle with T. congolense suppresses their antibody response Rurangirwa et al. to vaccination with Brucella abortus (1983) 1985 Immunity to metacyclic T. congolense is induced by infection and Berenil Akol and Murray treatment. A chancre does not develop upon second exposure to (1985) homologous reinfection 1985 Trypanotolerant breeds of cattle are better able to control trypanosomes Akol and Murray in the skin and to generate protective antibody responses than (1985) trypanosusceptible cattle 1987 Goats infected with T. congolense resist secondary infection by tsetse bite Dwinger et al. but do not develop immunity to the secondary infection. Infected (1987b) animals may develop non-specific immunity to skin trypanosomes 1987 T. congolense-infected N’Dama cattle are better able to control anaemia Ellis et al. (1987b) and parasitaemia than similarly infected Boran cattle and have higher numbers of B-cells in peripheral blood before and during infection than Boran cattle 1989 Ayrshire cattle infected with haemorrhage-causing T. vivax develop Assoku and auto-antibodies to normal red blood cells and platelets Gardiner (1989) 1991 Macrophages in lymph nodes of T. congolense-infected cattle suppress Flynn and Sileghem mitogenic responses of T-cells from infected and naïve cattle (1991) 1992 The inability of lymph-node T-cells from T. congolense-infected cattle to Sileghem and Flynn respond to mitogen is associated with impaired expression of the (1992) α-chain of the IL-2 receptor 1992 Splenic CD5+ B-cells, putatively bovine B1 B-cells, increase in the Naessens and spleens of T. congolense-infected cattle Williams (1992) 1993 T. congolense-infected N’Dama but not Boran cattle generate IgG Authié et al. (1993) antibodies to a 33-kDa parasite parasite cysteine protease, congopain 1993 N’Dama cattle primed to T. congolense antigens by ITM mount superior Authié et al. (1993) IgG recall responses to the antigens upon reinfection than similarly treated Boran cattle 1994 γδ T-cells are selectively depleted from the afferent lymph of responding Flynn and Sileghem lymph nodes of Boran cattle exposed to T. congolense-infected tsetse (1994) 1994 CD8+ T-cells and γδ T-cells, including γδ T-cell clones, of infected N’Dama Flynn and Sileghem but not Boran cattle proliferate in vitro in response to a 100-kDa (1994) complex trypanosome invariant antigen 1994 VSG-primed CD4+ T-cells from cattle immunized with purified VSG and McKeever et al. adjuvant recognize non-conserved areas of the molecule (1994a) 1994 Serum from trypanosome-infected Cape buffalo contains a 133-kDa Reduth et al. (1994) trypanocidal protein 1994 The severity of anaemia in T. vivax-infected cattle correlates with ex vivo Sileghem et al. production of TNF by monocytes (1994) 1995 N’Dama cattle primed by T. congolense ITM retain memory T- and B-cells Lutje et al. (1995) specific for trypanosome antigens in lymph nodes for at least 3 years Continued Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 191 Table 4.9. Continued. Year Finding Reference 1995 Depletion of CD8+ T-cells from T. congolense-infected cattle does not Sileghem and affect parasitaemia or anaemia Naessens (1995) 1996 T. congolense-infected N’Dama cattle have significantly more VSG- Taylor et al. (1996) specific IgG in blood than Boran cattle during infection 1996 Antibody responses against surface-exposed VSG epitopes are similar in Williams et al. T. congolense-infected N’Dama and Boran cattle but N’Dama mount (1996) higher-titre IgG1 responses against cryptic VSG epitopes 1997 CD5+ B-cells are the main source of antibodies reactive with non- Buza et al. (1997) trypanosome antigens in T. congolense-infected cattle 1998 Monocytes from T. congolense-infected cattle produce IL-10 but not nitric Taylor et al. (1998) oxide when stimulated with interferon-γ in vitro 1999 Serum of T. congolense-infected cattle contains polyreactive IgM Buza and Naessens antibodies (1999) (2002) Depletion of CD8+ T-cells and WC1+ γδ T-cells from T. congolense- Referred to in infected cattle does not affect parasitaemia or anaemia Naessens et al. (2002) (2002) Depletion of CD4+ T-cells from T. congolense-infected Boran cattle Referred to in reduces VSG-specific antibody responses and results in increased Naessens et al. parasitaemia but does not affect antibody responses or parasitaemia in (2002) infected N’Dama cattle 2003 Depletion of CD4+ T-cells from T. congolense-infected cattle results in Naessens et al. significantly reduced size of chancre (2003b) 2003 Studies in T. congolense-infected chimeric Boran/N’Dama twin calves and Naessens et al. singletons indicate that the severity of anaemia is a function of (2003a) haematopoietic tissue whereas the level of parasitaemia is regulated by non-haematopoietic tissue 2006 T. congolense-infected N’Dama cattle are more capable of early T-helper O’Gorman et al. 1 T-cell proinflammatory responses than similarly infected Boran (2006) 2009 Gene expression profiling of PBLs from T. congolense-infected N’Dama and O’Gorman et al. Boran cattle show a rapid tenfold higher level of expression in the former (2009) Jan Grootenhuis who, with the support of the 133 kDa that lysed all bloodstream-stage African Dutch government and some support from Tony trypanosomes, irrespective of trypanosome spe- Alison, then Director General of ILRAD, estab- cies, serodeme or variant antigen type (Reduth lished a captive herd of Cape buffalo on the cam- et al., 1994). This molecule was later identified pus of the veterinary school at Kabete, neigh- as xanthine oxidase (Muranjan et al., 1997). In bouring ILRAD. These animals, which were also fact, trypanosomes were killed by hydrogen per- bred in captivity in a tsetse- and trypanosomia- oxide, to which they are highly sensitive, gener- sis-free area, were shown by the ILRAD faculty to ated during catabolism of hypoxanthine and be highly tolerant to trypanosomiasis induced by xanthine by xanthine oxidase and facilitated by tsetse or syringe challenge. an infection-induced decline in Cape buffalo Trypanosome-infected Cape buffalo con- blood catalase, which decreased catabolism of trolled trypanosome parasitaemia after one or a the hydrogen peroxide (Wang et al., 1999). The few parasitaemic waves, whether they were in- processes controlling expression of xanthine oxi- fected by tsetse or by syringe. Thereafter, they dase and catalase in blood plasma proved com- suppressed parasitaemia to a cryptic level with plex, with no obvious translational component few or no signs of disease but carried enough (Wang et  al., 2002), and this line of research parasites in their blood for several months to in- was discontinued. However, subsequent work fect feeding tsetse. Post-infection, Cape buffalo by Sam Black and colleagues at the University serum was shown to contain protein of about of Massachusetts and ILRI showed that the 192 S.J. Black and C.L. Baldwin infection- induced decline in Cape buffalo plasma titres of serum IgG1 antibodies specific for cryp- catalase and non-specific killing of trypanosomes tic VSG epitopes, suggesting differential regula- was short lived, whereas parasitaemia in the in- tion of these responses. In addition, the infected fected Cape buffalo remained cryptic for months, N’Dama cattle were found to develop lower levels which correlated with the accumulation in blood of IgM antibodies reactive with non-trypanosome of IgG antibodies against VSGs of successive antigens than the infected Boran. These IgM antigenic variants (Guirnalda et al., 2007). While antibodies, which include polyreactive IgMs, are correlation does not denote causation, it was most likely the products of B1 B-cells (Buza et al., observed that trypanosomes harvested from the 1997), a B-cell type that does not require help blood of infected Cape buffalo during cryptic from T-cells to mount an antibody response. The parasitaemia and cultured in the absence of skewing of antibody responses in infected N’Da- serum antibodies multiplied every 6 h, while ma cattle towards IgG isotypes suggests robust those cultured in the presence of the serum anti- T-cell-dependent immune responses because bodies died or barely multiplied. Mechanisms production of IgG antibodies in response to anti- leading to the highly efficient development of gen challenge or infection is, for the most part, protective VSG-specific IgG antibody responses dependent on help from CD4+ T-cells and the in the infected Cape buffalo were not resolved. cytokines they produce. Consequently, the Comparative studies in trypanosome- elevated levels of trypanosome antigen-specific infected N’Dama and Boran cattle showed that IgG antibodies in trypanosome-infected N’Dama N’Dama cattle developed smaller chancres than versus Boran cattle are consistent with the strong- those arising in similarly infected Boran cattle er trypanosome antigen-specific T-cell responses (Akol et al., 1986). Chancres are trypanosome- that arise in T. congolense-infected N’Dama induced inflammatory sites that arise at the site compared with Boran cattle (Flynn et al., 1992). of infected tsetse bites and are populated by T. congolense-derived cysteine protease, called metacyclic-derived trypanosomes, neutrophils congopain, is among the trypanosome proteins and subsequently lymphocytes. Their size is a recognized by IgG antibodies produced by infected function of metacyclic dose (Dwinger et al., 1987a); N’Dama, and there is some evidence that specific hence, their smaller size in tsetse-infected N’Da- immunization with these proteins modulates in- ma compared with Boran cattle is consistent fection-induced pathology (Authié et al., 2001). with the possibility that N’Dama better control T-cell-specific antibodies developed at ILRAD the growth of metacyclic-derived trypanosomes were used by Jan Naessens at ILRAD/ILRI to in the skin. In addition, levels of parasitaemia determine the impact of T-cell depletion on im- were significantly lower in infected N’Dama mune responses to African trypanosomes. Using than in infected Boran cattle, suggesting that the this procedure, Naessens and colleagues showed N’Dama are better able to control systemic infec- that CD8+ T-cells play little or no role in defence tion. In this regard, infected N’Dama were found against the parasites, but CD4+ T-cells are re- to respond to a wider range of trypanosome quired to control (reduce) the size of chancres antigens (Shapiro and Murray, 1982), and to (Naessens et  al., 2003b) consistent with inhib- mount high-titre IgG responses against those ition of trypanosome growth in the skin, (e.g. by antigens, than similarly infected trypanosomiasis- T-cell amplified innate immune responses). In susceptible Boran cattle (Taylor et al., 1996, and addition, in the absence of CD4+ T-cells, antibody other references listed in Table 4.9). These stud- responses against trypanosome antigens were ies suggest that the capacity to generate IgG decreased and levels of parasitaemia increased antibodies specific for trypanosome antigens compared with those arising in intact cattle (dis- during infection correlates with, and may con- cussed by Naessens et al., 2002). These observa- tribute to, trypanotolerance. Subsequent studies tions indicate that CD4+ T-cells play a central by Williams and colleagues at ILRI (Williams role in controlling protective immune responses et  al., 1996) showed that T. congolense-infected in trypanosome-infected cattle and hence the N’Dama and Boran cattle develop similar titres severity of trypanosomiasis, a notion that has and isotypes of antibodies against exposed (i.e. support from studies conducted in the mouse protective) VSG epitopes on intact trypanosomes model of human African trypanosomiasis by in blood serum, but the N’Dama develop higher John Mansfield and colleagues at the University Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 193 of Wisconsin at Madison (Hertz et  al., 1998). hence is a candidate pathology-inducing antigen. The specificity of the putative protective CD4+ The GPI glycan is poorly immunogenic, and it T-cells needs to be established. In this regard, remains an intriguing question as to whether Mansfield and colleagues have shown that CD4+ GPI glycan-specific antibodies could be induced T-cells in infected mice and in mice immunized by appropriate immunization, and if so, whether with intact VSG respond to the variable but not the antibodies would block infection-associated the more conserved epitopes of trypanosome VSGs pathology. (Dagenais et al., 2009). This was also found to be Attempts to dissect mechanisms of trypa- the case in cattle by scientists at ILRI (McKeever notolerance in ruminants and mice at ILRAD/ et al., 1994a). However, although T. b. rhodesiense- ILRI were paralleled by analysis of quantitative infected mice do not develop CD4+ T-cells that trait loci (QTLs) responsible for the trait. Studies react with conserved VSG epitopes, Mansfield in mouse models at ILRI and the University of and colleagues have now shown that they can be Liverpool, UK, identified three QTLs called ‘tryp- induced to do so by immunization with con- anosome immune response’ (Tir) loci that con- served VSG C-terminal peptide and, indeed, trol survival after infection with T. congolense. T-cells in the immunized animals mount recall Tir1 was identified as the major trypanotoler- responses against conserved T-cell epitopes of ance QTL with Pram1 being the most plausible the peptide upon trypanosome infection. In add- candidate Tir1 QTL gene (Goodhead et al., 2010). ition, this procedure results in effective control Pram1 encodes an adaptor protein that regulates of trypanosomes in tissues and greatly reduces adhesion-dependent oxygen intermediate pro- infection-associated pathology. These important duction and degranulation in mature neutro- findings bring into play the possibility of devel- phils (Clemens et al., 2004) and thus aspects of oping a VSG-based anti-pathology vaccine, based the innate immune response. Three subregions on assemblies of conserved VSG T-cell epitopes. were identified within Tir3, with 2B4 being the It is an appealing notion that such immuniza- most plausible Tir3c QTL gene. The 2B4 molecule tion would increase control of tsetse-transmitted is a receptor expressed by natural killer (NK) trypanosomes in the skin as well as in other cells. It recognizes CD48, which is expressed on tissues, which has been proposed as a goal of and regulates the activation and differentiation trypanosome vaccination by Henry Tabel and of lymphocytes and dendritic and endothelial colleagues at the University of Saskatchewan cells. Ongoing work in the Black laboratory, in (Tabel et al., 2013). Amherst, Massachusetts, shows that splenic NK Work at ILRAD/ILRI showed that the effi- cells in T. brucei-infected mice suppress protect- cacy of T-cell responses in trypanosome-infected ive antibody responses by depleting splenic B2 cattle, which might underpin mechanisms of B-cells by cell-mediated cytotoxicity (Frenkel et al., trypanotolerance, may be regulated by immuno- 2016). These NK cells express 2B4, which is en- suppressive monocytes/macrophages (Flynn et al., coded by the Tir3c candidate QTL gene. NK cells 1994), which prevent their expression of the are also implicated in trypanosome pathology in α-chain of the IL-2 (T-cell growth factor) recep- infected cattle, with elevated recruitment/acti- tor and thus T-cell activation (Sileghem and vation of liver NK cells being a common feature Flynn, 1992). Immunosuppressive monocytes/ of trypanosomiasis in T. congolense-infected macrophages have also been identified in trypa- N’Dama and Boran cattle (Noyes et al., 2011). nosome-infected mice, although their mechanism Studies of the mechanisms of trypanosom- of action, which is dependent on prostaglandins iasis susceptibility and tolerance at ILRAD/ILRI and nitric oxide (Mabbott et al., 1995), is distinct covered a wide field and left a legacy of infection- from that arising in infected cattle. Species-specific induced immune response data in cattle but did differences in suppressor effector mechanisms not identify candidate vaccine antigens or de- do not necessarily rule out similarities in the finitively identify resistance genes for introgres- mechanism of induction of the suppressor cells, sion into the gene pool of desirable Boran live- but this has not been resolved in either species. stock. Nevertheless, this work made some However, the VSG glycosylphosphatidylinositol notable contributions to our understanding of (GPI) has been shown to be a potent activator of trypanosome biology and immunology. It is par- monocytes in mice (Magez et  al., 1998) and ticularly noteworthy that CD4+ T-cells were found 194 S.J. Black and C.L. Baldwin to play a key role in directing host-protective im- immunology, including basic immunology, im- mune responses in the skin and blood of infected munoparasitology and vaccinology, from the cattle. The chapter authors consider that further following categories: study of their specificity and mechanism of action is warranted. It is also noteworthy that strong • Employed at ILRAD/ILRI as a scientist, IgG antibody responses directed against numer- post-doctoral fellow, graduate student or ous trypanosome antigens is a serological signa- technician who subsequently went on for a ture of infected trypanotolerant compared with graduate degree trypanosomiasis-susceptible livestock, as it is in • Collaborators on the ILRAD/ILRI campus mice. It seems likely that this phenotype could be (short term or sabbaticals) in those fields used together with body condition to identify po- • Administrators or advisory board members tentially trypanotolerant Boran cattle for selective of ILRAD/ILRI breeding in herds subjected to ongoing trypano- • Opinion leaders in these fields not associ- some and tsetse challenge in trypanosomia- ated directly with ILRAD/ILRI sis-endemic areas. With a new focus on priming CD4+ T-cells against conserved VSG epitopes and Analysis of work environment possibly a straightforward way to select trypa- and subsequent influence notolerant livestock under natural challenge on the horizon, fundamental studies on AAT, to The results of the survey are shown in Fig. 4.3. which ILRAD/ILRI has made many contribu- The first four questions give a reflection of the tions, might yet yield important translational work environment and how being involved in outcomes. important work impacted the growth of the sci- entists (from technicians to senior scientists): 1. I had a productive period working at ILRAD/ Beyond Nairobi and Kabete: outreach ILRI. to other diseases and places 2. While working at ILRAD/ILRI, I was exposed to areas of science/techniques that I was previ- ously unfamiliar with. The information and analyses found in this 3. Working as part of a team was important to section are based on survey results. They include me at ILRAD/ILRI. an analysis of the work environment at ILRAD/ 4. Working at ILRAD/ILRI helped me to grow as ILRI and its influence on subsequent work as a scientist. well as a survey of the tools and techniques applied outside the ILRAD/ILRI campus and the The next question showed whether the person additional diseases to which they were applied. voluntarily engaged in capacity building, as A survey was sent to scientists in the fields of training students was not part of the mandate: 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10 11 Survey question number Fig. 4.3. Graph showing the survey responses. A rating of 1 is ‘agree completely,’ 2 is ‘strongly agree,’ 3 is ‘moderate agreement,’ 4 is ‘weak agreement’ and 5 is ‘not at all.’ This reflects answers from 35 survey respondents who were employed at ILRAD/ILRI. Mean response Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 195 5. I trained other scientists/graduate students at the geographical placement of the research and/ ILRAD/ILRI. or disease. All are important in world food production. The last questions reflect the longevity of the in- fluence the environment had on future scientists: 6. Working at ILRAD/ILRI helped me to obtain Conclusions my current position. 7. My current research is related to my previous This chapter has assessed the relevance and studies at ILRAD/ILRI. impact of ILRAD and ILRI’s basic bovine im- 8. I train graduate students/post-doctoral fel- munology and immunoparasitology as con- lows in veterinary immunology/immunopara- ducted in their trypanosomiasis and theileriosis sitology. programmes. Questions were posed and more 9. My graduate students/post-doctoral fellows quantitative measurements were applied, as out- plan to work in veterinary immunology/immu- lined in the main chapter and in the summary noparasitology. (Table 4.11). The two livestock diseases that were 10. I have maintained contact with colleagues the focus of research at ILRAD and to some ex- from ILRAD/ILRI. tent ILRI still appear, some 40 years later, at or 11. Contacts that I made at ILRAD/ILRI have towards the top of priority lists regarding the im- helped me succeed in my career. pediments they raise to raising cattle in sub- There was uniform strong agreement for Saharan Africa. The investment made at ILRAD/ the first four questions, indicating that the work- ILRI to develop vaccines for these diseases is a tiny ing environment at ILRAD/ILRI was conducive percentage of that invested in research to combat to productive scientific research and scientific HIV and malaria over the same time period, for growth. The results for question 5 showed that which vaccines also do not exist. According to most individuals voluntarily trained other more one early ILRAD research participant, Bruno junior scientists. The answers to questions 6–9 Goddeeris, the tools developed and research ranged between strong and moderate agree- undertaken ‘gave worldwide a tremendous push ment, indicating that the longevity of the influ- in veterinary immunology research and made ence was more moderate, suggesting that the bias veterinary immunoparasitology known to the is still towards continuing in the field engaged in medical world’. This is validated by Fidel Zavala at ILRAD/ILRI (although not irrevocably so) and (2015, personal communication), a leading mal- towards training the future generation of re- aria researcher at John Hopkins University, who searchers in the field. Answers to questions 10 indicated that ILRAD research on theileriosis in and 11 suggested that contacts have been main- cattle showed the way for studies on human mal- tained among the majority of ILRAD/ILRI arial vaccine research that continue to this day. alumni and that these have proven fruitful, even Immunological research at ILRAD/ILRI has after leaving ILRAD/ILRI. had a substantial impact worldwide. The right approach was taken and appropriate tools were developed to allow further understanding of the Applications of tools and techniques ruminant immune system in support of vaccine Application or transfer of tools such as mAbs development and efficacy testing according to and BoLA MHC typing and techniques such as the mandate of ILRAD. Moreover, the research T-cell cloning were used outside of ILRAD/ILRI. team that undertook these studies was reflective and creative in generating a vast number of Additional disease research that benefitted tools to study the bovine immune system using the most modern technologies and following The tools, protocols (techniques) and approaches closely the discoveries informing the fundamen- generated at ILRAD/ILRI were applied to other tals of the murine and human immune systems diseases, and thus research on these diseases (Table 4.12). In addition, in some cases know- benefitted from or was impacted by the work per- ledge about the components and function of the formed at ILRAD/ILRI. These external benefi- bovine system preceded that in either mice or ciaries are summarized in Table 4.10 along with humans. 196 S.J. Black and C.L. Baldwin Table 4.10. Diseases not targeted at ILRAD/ILRI but benefiting from tools and techniques and approaches developed there; unless stated, they are diseases of cattle. Country or region Disease/causative organism (of research or disease location) Scientist involved African swine fever (pigs) USA Mwangi Anaplasmosis USA McGuire, Brown Avian pathogenic Escherichia coli Belgium Goddeeris infection (poultry) Bluetongue virus (sheep) Canada Ellis Bovine babesiosis USA McGuire, Brown Bovine herpes virus (sheep) USA De Martini Bovine herpes virus-1 (cattle) Canada Ellis Bovine respiratory syncytial virus Canada Ellis Bovine viral diarrhoea virus UK, USA Graham, Mwangi Brucella abortus and B. melitensis USA, India, Israel, Brazil Baldwin, Splitter (cattle and goats) Middle East respiratory syndrome East Africa Jores coronavirus (camel) Caseous lymphadenitis (sheep) Canada Ellis Chlamydia abortus (sheep) Scotland Rocchi Contagious bovine pleuropneumonia Kenya Jores, Naessens Cowdriosis (heartwater) USA, Zimbabwe, South Africa Barbett, Mahan, McGuire Fasciola hepatica UK Williams Foot-and-mouth disease virus Scotland Morrison Haemonchus contortus USA, Scotland McGuire, Ballingal Histomonas spp. Belgium Goddeeris Leishmania spp. (humans) Canada, USA Pearson Leptospira interrogans serovar Hardjo USA Baldwin Malignant catarrhal fever USA, Europe, South America, De Martini Africa Mycobacterium avium UK, USA Glass, Shields, Davis, Ballingal paratuberculosis (cattle and sheep) Mycobacterium bovis UK, Italy, USA Glass, Shields, Rocchi, Baldwin Nematodes Australia, UK Emery Neospora caninum UK, Italy Williams, Rocchi Newcastle disease virus (diagnostic Austria Dwinger test on same principle as the trypanosome test) (poultry) Ostertagia spp. Belgium Goddeeris Porcine diarrhoea and oedema Belgium, Cuba, Vietnam Goddeeris disease (pigs) Pulmonary adenocarcinoma Scotland Ballingal Retroviruses (pulmonary carcinoma) USA De Martini (sheep) Small ruminant lentiviruses USA, Europe, South America, De Martini Africa Streptococcus agalactiae (camels) East Africa Jores Theileria annulata UK, India, Turkey, China Glass Theileria orientalis (buffalo) Belgium, Vietnam, Indonesia Goddeeris Trypanosoma equiperdum (dourine) Belgium, Ethiopia, Switzerland Brun, Goddeeris (horses) Trypanosoma evansi (horses) Switzerland Brun Trypanosomiasis (trypanosome Austria Dwinger diagnostic tests) Trypanosomiasis (camels) East Africa Masake Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 197 Table 4.11. Answers to questions about impact of immunology research. Basic bovine Trypanosomes Trypanosomes Quality of research Overall immunology T. parva in cattle in mice Did what was found by Yes Yes Yes Yes Yes the research and assumed to be ‘true’ at the time of formal publication remain true for years to come? Was the research built Yes Yes Yes Field analysis of Yes upon or taken up by parasitaemia and other scientists (i.e. anaemia continue, influenced others) but little additional immunological work has been done Were the right things Yes Yes Yes Yes, although research Additional effort being done regarding on CD4+ T-cell should have been the research direction responses to the directed at the according to the parasites may have regulation and mandate of ILRAD/ been stopped specificity of CD4+ ILRI? prematurely T-cell responses; however, this was effectively followed up by the Mansfield laboratory (USA) Was the team Yes Yes Yes In hindsight, it can be In hindsight, it can successful, i.e. was seen that some be seen that some the research reflective avenues of research avenues of and creative? remain open research remain open Table 4.12. Assessment of immunology and immunoparasitology. (Data compiled from surveys.) Manuscript questions Assessment Was there a substantive body of work that showed Yes, ~20 peer-reviewed manuscripts per year the step-by-step progression of knowledge in addition to key papers in top-tier journals? Were the results published in a timely manner? Yes Citation indices More than 50% had >25 citations i10 index (ten or more citations) 77% of peer-reviewed manuscripts met this criterion Journal impact factor of top-cited manuscripts Moderate-level journals for >95% of manuscripts; 2% in top-tier journals Quality of staff: As a group, were the scientists who conducted the Yes – equal to their peers relative to career research at ILRAD/ILRI highly respected and length accomplished (h-index)? Did the experience at ILRAD/ILRI influence the field >30% continued to work directly on the topics of research that these scientists subsequently researched at ILRAD; <5% do something engaged in? totally unrelated Continued 198 S.J. Black and C.L. Baldwin Table 4.12. Continued. Manuscript questions Assessment Was the knowledge gained at ILRAD/ILRI and the Yes tools generated expanded or continued to be used to evaluate other important infectious diseases of ruminants that affect food safety? Impact: Commercialization of products mAbs in Europe and North America, ITM vaccine in Africa Other disseminated tools not commercialized BoLA typing Simple point-of-care technique for detecting African trypanosomes – the most sensitive parasitological technique available for animals, widely used throughout Africa Did their availability change the landscape? Yes, for both research and the vaccine The Future Future research priorities should be as fol- lows: The immunology and immunoparasitology re- • In addition to current research aimed at de- search at ILRAD and later ILRI had high scientific veloping a subunit vaccine for ECF based on and technical impacts on a very limited budget an assembly of immunodominant T. parva compared with spending on human diseases of T-cell epitopes, identify antigens that are similar immunological complexity. Output from shared among T. parva and T. parva lawrencei this research was primarily publications and mAbs strains and that stimulate αβ and γδ T-cells specific for bovine leukocytes and immunoglobu- for inclusion in the subunit vaccine. lins, which are still being used in many laborator- • Identify trypanosomiasis-tolerant individ- ies throughout the world. Despite its academic/ uals in herds of Boran cattle subjected to scientific successes, this immunology/immuno- natural trypanosome challenge, based on parasitology research did not affect the incidence the serum signature of IgG antibodies re- or economic impact of theileriosis or trypanosom- active with trypanosome polypeptides, iasis in any substantial way, other than the benefits body condition and blood packed cell vol- that have accrued from the ITM method of im- ume, and determine whether these traits munizing cattle against ECF. However, scientific can be improved by selective breeding. opinion believes that the fundamental research • Determine whether immunization of cattle on bovine immunology and the immunoparasi- with conserved VSG components will en- tology of theileriosis and trypanosomiasis carried hance control of parasites in the skin and out at ILRAD/ILRI will ultimately have signifi- tissues and reduce infection-associated im- cant development outcomes beyond those it has munopathology, as discussed by Black and already had on scientific capacity development. Mansfield (2016). References Akol, G.W. and Murray, M. (1982) Early events following challenge of cattle with tsetse infected with Trypanosoma congolense: development of the local skin reaction. Veterinary Record 110, 295–302. Akol, G.W. and Murray, M. (1985) Induction of protective immunity in cattle by tsetse-transmitted cloned isolates of Trypanosoma congolense. Annals of Tropical Medicine and Parasitology 79, 617–627. Akol, G.W., Authié, E., Pinder, M., Moloo, S.K., Roelants, G.E. and Murray, M. (1986) Susceptibility and immune responses of zebu and taurine cattle of West Africa to infection with Trypanosoma congolense transmitted by Glossina morsitans centralis. Veterinary Immunology and Immunopathology 11, 361–373. Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 199 Assoku, R.K. and Gardiner, P.R. (1989) Detection of antibodies to platelets and erythrocytes during infection with haemorrhage-causing Trypanosoma vivax in Ayrshire cattle. Veterinary Parasitology 31, 199–216. Authié, E., Muteti, D.K. and Williams, D.J. (1993) Antibody responses to invariant antigens of Trypa- nosoma congolense in cattle of differing susceptibility to trypanosomiasis. Parasite Immunology 15, 101–111. Authié, E., Boulange, A., Muteti, D., Lalmanach, G., Gauthier, F. and Musoke, A.J. (2001) Immunisation of cattle with cysteine proteinases of Trypanosoma congolense: targetting the disease rather than the parasite. International Journal for Parasitology 31, 1429–1433. Baldwin, C.L. and Teale, A.J. (1987) Alloreactive T cell clones transformed by Theileria parva retain cyto- lytic activity and antigen specificity. European Journal of Immunology 17, 1859–1862. Baldwin, C.L., Teale, A.J., Naessens, J.G., Goddeeris, B.M., MacHugh, N.D. and Morrison, W.I. (1986) Characterization of a subset of bovine T lymphocytes that express BoT4 by monoclonal antibodies and function: similarity to lymphocytes defined by human T4 and murine L3T4. Journal of Immunology 136, 4385–4391. Baldwin, C.L., Goddeeris, B.M. and Morrison, W.I. (1987) Bovine helper T-cell clones specific for lympho- cytes infected with Theileria parva (Muguga). Parasite Immunology 9, 499–513. Baldwin, C.L., Black, S.J., Brown, W.C., Conrad, P.A., Goddeeris, B.M., et al. (1988a) Bovine T cells, B cells, and null cells are transformed by the protozoan parasite Theileria parva. Infection and Immunity 56, 462–467. Baldwin, C.L., Machugh, N.D., Ellis, J.A., Naessens, J., Newson, J. and Morrison, W.I. (1988b) Monoclonal antibodies which react with bovine T-lymphocyte antigens and induce blastogenesis: tissue distribu- tion and functional characteristics of the target antigens. Immunology 63, 439–446. Baldwin, C.L., Iams, K.P., Brown, W.C. and Grab, D.J. (1992) Theileria parva: CD4+ helper and cytotoxic T-cell clones react with a schizont-derived antigen associated with the surface of Theileria parva- infected lymphocytes. Experimental Parasitology 75, 19–30. Baldwin, C.L., Sathiyaseelan, T., Rocchi, M. and McKeever, D. (2000) Rapid changes occur in the percent- age of circulating bovine WC1+γδ Th1 cells. Research in Veterinary Science 69, 175–180. Barbet, A.F. and McGuire, T.C. (1978) Crossreacting determinants in variant-specific surface antigens of African trypanosomes. Proceedings of the National Academy of Sciences USA 75, 1989–1993. Bastin, P., Stephan, A., Raper, J., Saint-Remy, J.M., Opperdoes, F.R. and Courtoy, P.J. (1996) An Mr 145,000 low-density lipoprotein (LDL)-binding protein is conserved throughout the Kinetoplastida order. Molecular and Biochemical Parasitology 76, 43–56. Bensaid, A., Naessens, J., Kemp, S.J., Black, S.J., Shapiro, S.Z. and Teale, A.J. (1988) An immuno- chemical analysis of class I (BoLA) molecules on the surface of bovine cells. Immunogenetics 27, 139–144. Bensaid, A., Kaushal, A., Baldwin, C.L., Clevers, H., Young, J.R., et al. (1991) Identification of expressed bovine class I MHC genes at two loci and demonstration of physical linkage. Immunogenetics 33, 247–254. Bill and Melinda Gates Foundation (2005) Gates Foundation commits $258.3 million for malaria research and development. Available at: www.gatesfoundation.org/Media-Center/Press-Releases/2005/10/ Gates-Foundation-Commits-2583-Million-for-Malaria-Research (accessed 17 January 2019). Black, S.J. and Mansfield, J.M. (2016) Prospects for vaccinating against pathogenic African trypanosomes. Parasite Immunology 38, 735–743. Black, S.J. and Vandeweerd, V. (1989) Serum lipoproteins are required for multiplication of Trypanosoma brucei brucei under axenic culture conditions. Molecular and Biochemical Parasitology, 37, 65–72. Black, S.J., Hewett, R.S. and Sendashonga, C.N. (1982) Trypanosoma brucei variable surface antigen is re- leased by degenerating parasites but not actively dividing parasites. Parasite Immunology 4, 233–244. Black, S.J., Sendashonga, C.N., O’Brien, C., Borowy, N.K., Naessens, M., et al. (1985) Regulation of para- sitaemia in mice infected with Trypanosoma brucei. Current Topics in Microbiology and Immunology 117, 93–118. Brown, C.G., Stagg, D.A., Purnell, R.E., Kanhai, G.K. and Payne, R.C. (1973) Letter: Infection and transform- ation of bovine lymphoid cells in vitro by infective particles of Theileria parva. Nature 245, 101–103. Brown, W.C. and Grab, D.J. (1985) Biological and biochemical characterization of bovine interleukin 2. Studies with cloned bovine T cells. Journal of Immunology 135, 3184–3190. Brown, W.C., Sugimoto, C., Conrad, P.A. and Grab, D.J. (1989a) Differential response of bovine T-cell lines to membrane and soluble antigens of Theileria parva schizont-infected cells. Parasite Immun- ology 11, 567–583. 200 S.J. Black and C.L. Baldwin Brown, W.C., Sugimoto, C. and Grab, D.J. (1989b) Theileria parva: bovine helper T cell clones specific for both infected lymphocytes and schizont membrane antigens. Experimental Parasitology 69, 234–248. Buza, J. and Naessens, J. (1999) Trypanosome non-specific IgM antibodies detected in serum of Trypano- soma congolense-infected cattle are polyreactive. Veterinary Immunology and Immunopathology 69, 1–9. Buza, J., Sileghem, M., Gwakisa, P. and Naessens, J. (1997) CD5+ B lymphocytes are the main source of antibodies reactive with non-parasite antigens in Trypanosoma congolense-infected cattle. Immunology 92, 226–233. Chen, C., Herzig, C.T., Telfer, J.C. and Baldwin, C.L. (2009) Antigenic basis of diversity in the γδ T cell co-receptor WC1 family. Molecular Immunology 46, 2565–2575. Clemens, R.A., Newbrough, S.A., Chung, E.Y., Gheith, S., Singer, A.L., et al. (2004) PRAM-1 is required for optimal integrin-dependent neutrophil function. Molecular and Cellular Biology 24, 10923–10932. Clevers, H., MacHugh, N.D., Bensaid, A., Dunlap, S., Baldwin, C.L., et al. (1990) Identification of a bovine surface antigen uniquely expressed on CD4–CD8– T cell receptor γδ+ T lymphocytes. European Journal of Immunology 20, 809–817. Cooper, M.D., Raymond, D.A., Peterson, R.D., South, M.A. and Good, R.A. (1966) The functions of the thymus system and the bursa system in the chicken. Journal of Experimental Medicine 123, 75–102. Creemers, P. (1982) Lack of reactivity of sera from Theileria parva-infected and recovered cattle against cell membrane antigens of Theileria parva transformed cell lines. Veterinary Immunology and Immu- nopathology 3, 427–438. Cross, G.A. (1975) Identification, purification and properties of clone-specific glycoprotein antigens consti- tuting the surface coat of Trypanosoma brucei. Parasitology 71, 393–417. Cross, G.A. (1977) Isolation, structure and function of variant-specific surface antigens. Annales de la Société Belge de Médecine Tropicale 57, 389–402. Cunningham, M.P., Brown, C.G., Burridge, M.J. and Purnell, R.E. (1973) Cryopreservation of infective particles of Theileria parva. International Journal for Parasitology 3, 583–587. Dagenais, T.R., Demick, K.P., Bangs, J.D., Forest, K.T., Paulnock, D.M. and Mansfield, J.M. (2009) T-cell responses to the trypanosome variant surface glycoprotein are not limited to hypervariable subre- gions. Infection and Immunity 77, 141–151. Daubenberger, C.A., Taracha, E.L., Gaidulis, L., Davis, W.C. and McKeever, D.J. (1999) Bovine γδ T-cell responses to the intracellular protozoan parasite Theileria parva. Infection and Immunity 67, 2241–2249. Davis, W.C., Brown, W.C., Hamilton, M.J., Wyatt, C.R., Orden, J.A., et al. (1996) Analysis of monoclonal antibodies specific for the γδ TcR. Veterinary Immunology and Immunopathology 52, 275–283. Dobbelaere, D.A., Shapiro, S.Z. and Webster, P. (1985) Identification of a surface antigen on Theileria parva sporozoites by monoclonal antibody. Proceedings of the National Academy of Sciences USA 82, 1771–1775. Dobbelaere, D.A., Prospero, T.D., Roditi, I.J., Kelke, C., Baumann, I., et al. (1990) Expression of Tac anti- gen component of bovine interleukin-2 receptor in different leukocyte populations infected with Thei- leria parva or Theileria annulata. Infection and Immunity 58, 3847–3855. Donaldson, M.R. and Cooke, S.J. (2013) Scientific publications: moving beyond quality and quantity to- ward influence. Bioscience 64, 12–13. Dwinger, R.H., Lamb, G., Murray, M. and Hirumi, H. (1987a) Dose and stage dependency for the develop- ment of local skin reactions caused by Trypanosoma congolense in goats. Acta Tropica 44, 303–314. Dwinger, R.H., Murray, M. and Moloo, S.K. (1987b) Potential value of localized skin reactions (chancres) induced by Trypanosoma congolense transmitted by Glossina morsitans centralis for the analysis of metacyclic trypanosome populations. Parasite Immunology 9, 353–362. Ellis, J.A., Baldwin, C.L., MacHugh, N.D., Bensaid, A., Teale, A.J., et  al. (1986) Characterization by a monoclonal antibody and functional analysis of a subset of bovine T lymphocytes that express BoT8, a molecule analogous to human CD8. Immunology 58, 351–358. Ellis, J.A., Morrison, W.I., Goddeeris, B.M. and Emery, D.L. (1987a) Bovine mononuclear phagocytic cells: identification by monoclonal antibodies and analysis of functional properties. Veterinary Immunology and Immunopathology 17, 125–134. Ellis, J.A., Scott, J.R., Machugh, N.D., Gettinby, G. and Davis, W.C. (1987b) Peripheral blood leucocytes subpopulation dynamics during Trypanosoma congolense infection in Boran and N’Dama cattle: an analysis using monoclonal antibodies and flow cytometry. Parasite Immunology 9, 363–378. Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 201 Ellis, J.A., Davis, W.C., MacHugh, N.D., Emery, D.L., Kaushal, A. and Morrison, W.I. (1988) Differentiation antigens on bovine mononuclear phagocytes identified by monoclonal antibodies. Veterinary Immun- ology and Immunopathology 19, 325–340. Emery, D.L. (1981) Adoptive transfer of immunity to infection with Theileria parva (East Coast fever) between cattle twins. Research in Veterinary Science 30, 364–367. Emery, D.L. and Moloo, S.K. (1981) The dynamics of the cellular reactions elicited in the skin of goats by Glossina morsitans morsitans infected with Trypanosoma (Nannomonas) congolense or T. (Duttonel- la) vivax. Acta Tropica 38, 15–28. Emery, D.L., Morrison, W.I., Buscher, G. and Nelson, R.T. (1982) Generation of cell-mediated cytotoxicity to Theileria parva (East Coast fever) after inoculation of cattle with parasitized lymphoblasts. Journal of Immunology 128, 195–200. Emery, D.L., MacHugh, N.D. and Ellis, J.A. (1987) The properties and functional activity of non-lymphoid cells from bovine afferent (peripheral) lymph. Immunology 62, 177–183. Eugui, E.M. and Emery, D.L. (1981) Genetically restricted cell-mediated cytotoxicity in cattle immune to Theileria parva. Nature 290, 251–254. Fawcett, D., Musoke, A. and Voigt, W. (1984) Interaction of sporozoites of Theileria parva with bovine lymphocytes in vitro. I. Early events after invasion. Tissue Cell 16, 873–884. Flynn, J.N. and Sileghem, M. (1991) The role of the macrophage in induction of immunosuppression in Trypanosoma congolense-infected cattle. Immunology 74, 310–316. Flynn, J.N. and Sileghem, M. (1994) Involvement of γδ T cells in immunity to trypanosomiasis. Immunology 83, 86–92. Flynn, J.N., Sileghem, M. and Williams, D.J. (1992) Parasite-specific T-cell responses of trypanotoler- ant and trypanosusceptible cattle during infection with Trypanosoma congolense. Immunology 75, 639–645. Flynn, J.N., McKeever, D.J., Sileghem, M. and Naessens, J. (1994) Modulation of the phenotype and func- tion of bovine afferent lymph cells during infection with Trypanosoma congolense. Veterinary Immun- ology and Immunopathology 40, 17–29. Frenkel, D., Zhang, F., Guirnalda, P., Haynes, C., Bockstal, V., et al. (2016) Trypanosoma brucei co-opts NK cells to kill splenic B2 B cells. PLoS Pathogens 12, e1005733. Fritsch, G. and Nelson, R.T. (1990) Bovine erythroid (CFU-E, BFU-E) and granulocyte-macrophage colony formation in culture. Experimental Hematology 18, 195–200. Gardiner, P.R. (1989) Recent studies of the biology of Trypanosoma vivax. Advances in Parasitology 28, 229–317. Gentschev, I., Glaser, I., Goebel, W., McKeever, D.J., Musoke, A. and Heussler, V.T. (1998) Delivery of the p67 sporozoite antigen of Theileria parva by using recombinant Salmonella dublin: secretion of the product enhances specific antibody responses in cattle. Infection and Immunity 66, 2060–2064. Gillis, S. and Smith, K.A. (1977) Long term culture of tumour-specific cytotoxic T cells. Nature 268, 154–156. Goddeeris, B.M., Baldwin, C.L., ole-MoiYoi, O. and Morrison, W.I. (1986a) Improved methods for puri- fication and depletion of monocytes from bovine peripheral blood mononuclear cells. Functional evaluation of monocytes in responses to lectins. Journal of Immunological Methods 89, 165–173. Goddeeris, B.M., Morrison, W.I. and Teale, A.J. (1986b) Generation of bovine cytotoxic cell lines, specific for cells infected with the protozoan parasite Theileria parva and restricted by products of the major histocompatibility complex. European Journal of Immunology 16, 1243–1249. Goddeeris, B.M., Morrison, W.I., Teale, A.J., Bensaid, A. and Baldwin, C.L. (1986c) Bovine cytotoxic T-cell clones specific for cells infected with the protozoan parasite Theileria parva: parasite strain specificity and class I major histocompatibility complex restriction. Proceedings of the National Academy of Sci- ences USA 83, 5238–5242. Goddeeris, B.M., Morrison, W.I., Naessens, J. and Magondu, J.G. (1987) The bovine autologous mixed leukocyte reaction: a proliferative response of non-T cells under the control of monocytes. Immunob- iology 176, 47–62. Goddeeris, B.M., Dunlap, S., Bensaid, A., MacHugh, N.D. and Morrison, W.I. (1991) Cell surface pheno- type of two cloned populations of bovine lymphocytes displaying non-specific cytotoxic activity. Veter- inary Immunology and Immunopathology 27, 195–199. Goodhead, I., Archibald, A., Amwayi, P., Brass, A., Gibson, J., et al. (2010)A comprehensive genetic ana- lysis of candidate genes regulating response to Trypanosoma congolense infection in mice. PLoS Neglected Tropical Diseases 4, e880. 202 S.J. Black and C.L. Baldwin Graham, S.P., Pelle, R., Honda, Y., Mwangi, D.M., Tonukari, N.J., et al. (2006)Theileria parva candidate vaccine antigens recognized by immune bovine cytotoxic T lymphocytes. Proceedings of the National Academy of Sciences USA 103, 3286–3291. Graham, S.P., Saya, R., Awino, E., Ngugi, D., Nyanjui, J.K., et al. (2007) Immunostimulatory CpG oligode- oxynucleotides enhance the induction of bovine CD4+ cytotoxic T-lymphocyte responses against the polymorphic immunodominant molecule of the protozoan parasite Theileria parva. Veterinary Immun- ology and Immunopathology 115, 383–389. Graham, S.P., Pelle, R., Yamage, M., Mwangi, D.M., Honda, Y., et al. (2008)Characterization of the fine specificity of bovine CD8 T-cell responses to defined antigens from the protozoan parasite Theileria parva. Infection and Immunity 76, 685–694. Grewal, A.S., Rouse, B.T. and Babiuk, L.A. (1976) Erythrocyte rosettes – a marker for bovine T cells. Canad- ian Journal of Comparative Medicine 40, 298–305. Guirnalda, P., Murphy, N.B., Nolan, D. and Black, S.J. (2007)Anti-Trypanosoma brucei activity in Cape buffalo serum during the cryptic phase of parasitemia is mediated by antibodies. International Journal for Parasitology 37, 1391–1399. Herbert, W.J., Lumsden, W.H. and French, A.M. (1968) Survival of trypanosomes after rapid cooling and storage at –196°C. Transactions of the Royal Society of Tropical Medicine and Hygiene 62, 209–212. Hertz, C.J., Filutowicz, H. and Mansfield, J.M. (1998) Resistance to the African trypanosomes is IFN-γ dependent. Journal of Immunology 161, 6775–6783. Heussler, V.T., Taracha, E.L., Musoke, A., Duchateau, L. and McKeever, D.J. (1998) Immunisation with live attenuated Salmonella dublin expressing a sporozoite protein confers partial protection against Theileria parva. Vaccine 16, 834–841. Houston, E.F., Taracha, E.L., Brackenbury, L., MacHugh, N.D., McKeever, D.J., et al. (2008) Infection of cattle with Theileria parva induces an early CD8 T cell response lacking appropriate effector function. International Journal for Parasitology 38, 1693–1704. Howard, C.J. and Naessens, J. (1993) Summary of workshop findings for cattle (tables 1 and 2). Veterinary Immunology and Immunopathology 39, 25–47. Howard, C.J., Morrison, W.I., Brown, W.C., Naessens, J. and Sopp, P. (1989) Demonstration of two allelic forms of the bovine T cell antigen Bo5 (CD5) and studies of their inheritance. Animal Genetics 20, 351–359. Howard, C.J., Sopp, P., Parsons, K.R., McKeever, D.J., Taracha, E.L., et al. (1991) Distinction of naive and memory BoCD4 lymphocytes in calves with a monoclonal antibody, CC76, to a restricted determinant of the bovine leukocyte-common antigen, CD45. European Journal of Immunology 21, 2219–2226. Hozumi, N. and Tonegawa, S. (1976) Evidence for somatic rearrangement of immunoglobulin genes coding for variable and constant regions. Proceedings of the National Academy of Sciences USA 73, 3628–3632. Jordt, T., Mahon, G.D., Touray, B.N., Ngulo, W.K., Morrison, W.I., et  al. (1986) Successful transfer of N’Dama embryos into Boran recipients. Veterinary Record 119, 246–247. Kaba, S.A., Hemmes, J.C., van Lent, J.W., Vlak, J.M., Nene, V., et al. (2003) Baculovirus surface display of Theileria parva p67 antigen preserves the conformation of sporozoite-neutralizing epitopes. Protein Engineering 16, 73–78. Kaba, S.A., Musoke, A.J., Schaap, D., Schetters, T., Rowlands, J., et al. (2005) Novel baculovirus-derived p67 subunit vaccines efficacious against East Coast fever in cattle. Vaccine 23, 2791–2800. Kariuki, T.M., Grootenhuis, J.G., Dolan, T.T., Bishop, R.P. and Baldwin, C.L. (1990) Immunization with Thei- leria parva parasites from buffaloes results in generation of cytotoxic T cells which recognize antigens common among cells infected with stocks of T. parva parva, T. parva bovis, and T. parva lawrencei. Infection and Immunity 58, 3574–3581. Kemp, S.J., Tucker, E.M. and Teale, A.J. (1990) A bovine monoclonal antibody detecting a class I BoLA antigen. Animal Genetics 21, 153–160. Koch, H.T., Kambeva, L., Norval, R.A., Ocama, J.G., Masaka, S., et al. (1990) Age resistance to Theileria parva bovis infection in calves. Veterinary Parasitology 37, 197–206. Kohler, G. and Milstein, C. (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 495–497. Lalor, P.A., Morrison, W.I., Goddeeris, B.M., Jack, R.M. and Black, S.J. (1986) Monoclonal antibodies iden- tify phenotypically and functionally distinct cell types in the bovine lymphoid system. Veterinary Im- munology and Immunopathology 13, 121–140. Lanham, S.M. (1968) Separation of trypanosomes from the blood of infected rats and mice by anion- exchangers. Nature 218, 1273–1274. Lutje, V. and Black, S.J. (1991) Cellular interactions regulating the in vitro response of bovine lymphocytes to ovalbumin. Veterinary Immunology and Immunopathology 28, 275–288. Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 203 Lutje, V., Taylor, K.A., Boulange, A. and Authié, E. (1995) Trypanosoma congolense: tissue distribution of long-term T- and B-cell responses in cattle. Immunology Letters 48, 29–34. Lynen, G., Yrjo-Koskinen, A.E., Bakuname, C., Di Giulio, G., Mlinga, N., et al. (2012) East Coast fever immunisation field trial in crossbred dairy cattle in Hanang and Handeni districts in northern Tanzania. Tropical Animal Health and Production 44, 567–572. Mabbott, N.A., Sutherland, I.A. and Sternberg, J.M. (1995) Suppressor macrophages in Trypanosoma brucei infection: nitric oxide is related to both suppressive activity and lifespan in vivo. Parasite Immunology 17, 143–150. MacHugh, N.D., Bensaid, A., Davis, W.C., Howard, C.J., Parsons, K.R., et al. (1988) Characterization of a bovine thymic differentiation antigen analogous to CD1 in the human. Scandinavian Journal of Im- munology 27, 541–547. MacHugh, N.D., Wijngaard, P.L., Clevers, H.C. and Davis, W.C. (1993) Clustering of monoclonal antibodies recognizing different members of the WC1 gene family. Veterinary Immunology and Immunopathology 39, 155–160. MacHugh, N.D., Mburu, J.K., Carol, M.J., Wyatt, C.R., Orden, J.A. and Davis, W.C. (1997) Identification of two distinct subsets of bovine γδ T cells with unique cell surface phenotype and tissue distribution. Immunology 92, 340–345. MacHugh, N.D., Connelley, T., Graham, S.P., Pelle, R., Formisano, P., et al. (2009) CD8+ T-cell responses to Theileria parva are preferentially directed to a single dominant antigen: implications for parasite strain-specific immunity. European Journal of Immunology 39, 2459–2469. MacHugh, N.D., Weir, W., Burrells, A., Lizundia, R., Graham, S.P., et al. (2011) Extensive polymorphism and evidence of immune selection in a highly dominant antigen recognized by bovine CD8 T cells specific for Theileria annulata. Infection and Immunity 79, 2059–2069. Mackay, C.R. and Hein, W.R. (1989) A large proportion of bovine T cells express the γδ T cell recep- tor and show a distinct tissue distribution and surface phenotype. International Immunology 1, 540–545. Mackay, C.R., Beya, M.F. and Matzinger, P. (1989) γδ T cells express a unique surface molecule appearing late during thymic development. European Journal of Immunology 19, 1477–1483. Magez, S., Stijlemans, B., Radwanska, M., Pays, E., Ferguson, M.A. and de Baetselier, P. (1998) The gly- cosyl-inositol-phosphate and dimyristoylglycerol moieties of the glycosylphosphatidylinositol anchor of the trypanosome variant-specific surface glycoprotein are distinct macrophage-activating factors. Journal of Immunology 160, 1949–1956. Masake, R.A. and Morrison, W.I. (1981) Evaluation of the structural and functional changes in the lymphoid organs of Boran cattle infected with Trypanosoma vivax. American Journal of Veterinary Research 42, 1738–1746. Masake, R.A., Pearson, T.W., Wells, P. and Roelants, G.E. (1981) The in vitro response to mitogens of leucocytes from cattle infected with Trypanosoma congolense. Clinical and Experimental Immunology 43, 583–589. McDevitt, H.O., Deak, B.D., Shreffler, D.C., Klein, J., Stimpfling, J.H. and Snell, G.D. (1972) Genetic con- trol of the immune response. Mapping of the Ir-1 locus. Journal of Experimental Medicine 135, 1259–1278. McGuire, T.C. and Musoke, A.J. (1981) Biologic activities of bovine IgG subclasses. Advances in Experi- mental Medicine and Biology 137, 359–366. McGuire, T.C., Musoke, A.J. and Kurtti, T. (1979) Functional properties of bovine IgG1 and IgG2: interaction with complement, macrophages, neutrophils and skin. Immunology 38, 249–256. McHardy, N., Haigh, A.J. and Dolan, T.T. (1976) Chemotherapy of Theileria parva infection. Nature 261, 698–699. McKeever, D.J., MacHugh, N.D., Goddeeris, B.M., Awino, E. and Morrison, W.I. (1991) Bovine afferent lymph veiled cells differ from blood monocytes in phenotype and accessory function. Journal of Im- munology 147, 3703–3709. McKeever, D.J., Awino, E., Kairo, A., Gobright, E. and Nene, V. (1994a) Bovine T cells specific for Trypanoso- ma brucei brucei variant surface glycoprotein recognize nonconserved areas of the molecule. Infec- tion and Immunity 62, 3348–3353. McKeever, D.J., Taracha, E.L., Innes, E.L., MacHugh, N.D., Awino, E., et al. (1994b) Adoptive transfer of immunity to Theileria parva in the CD8+ fraction of responding efferent lymph. Proceedings of the Na- tional Academy of Sciences USA 91, 1959–1963. Morgan, D.A., Ruscetti, F.W. and Gallo, R. (1976) Selective in vitro growth of T lymphocytes from normal human bone marrows. Science 193, 1007–1008. 204 S.J. Black and C.L. Baldwin Morrison, W.I. and Davis, W.C. (1991) Individual antigens of cattle. Differentiation antigens expressed predominantly on CD4– CD8– T lymphocytes (WC1, WC2). Veterinary Immunology and Immunopathol- ogy 27, 71–76. Morrison, W.I. and Murray, M. (1979) Trypanosoma congolense: inheritance of susceptibility to infection in inbred strains of mice. Experimental Parasitology 48, 364–374. Morrison, W.I., Roelants, G.E., Mayor-Withey, K.S. and Murray, M. (1978) Susceptibility of inbred strains of mice to Trypanosoma congolense: correlation with changes in spleen lymphocyte populations. Clin- ical and Experimental Immunology 32, 25–40. Morrison, W.I., Goddeeris, B.M., Teale, A.J., Groocock, C.M., Kemp, S.J. and Stagg, D.A. (1987) Cytotoxic T-cells elicited in cattle challenged with Theileria parva (Muguga): evidence for restriction by class I MHC determinants and parasite strain specificity. Parasite Immunology 9, 563–578. Muranjan, M., Wang, Q., Li, Y.L., Hamilton, E., Otieno-Omondi, F.P., Wang, J., Van Praagh, A., Grooten- huis, J.G. and Black, S.J. (1997) The trypanocidal Cape buffalo serum protein is xanthine oxidase. Infection and Immunity 65, 3806–3814. Murray, M., Murray, P.K. and McIntyre, W.I. (1977) An improved parasitological technique for the diagnosis of African trypanosomiasis. Transactions of the Royal Society of Tropical Medicine and Hygiene 71, 325–326. Musoke, A.J., Nantulya, V.M., Buscher, G., Masake, R.A. and Otim, B. (1982) Bovine immune response to Theileria parva: neutralizing antibodies to sporozoites. Immunology 45, 663–668. Musoke, A., Morzaria, S., Nkonge, C., Jones, E. and Nene, V. (1992) A recombinant sporozoite surface antigen of Theileria parva induces protection in cattle. Proceedings of the National Academy of Sci- ences USA 89, 514–518. Mwangi, D.M., Honda, Y., Graham, S.P., Pelle, R., Taracha, E.L., et  al. (2011) Treatment of cattle with DNA-encoded Flt3L and GM-CSF prior to immunization with Theileria parva candidate vaccine anti- gens induces CD4 and CD8 T cell IFN-γ responses but not CTL responses. Veterinary Immunology and Immunopathology 140, 244–251. Naessens, J. (1993) Leukocyte antigens of cattle and sheep. Nomenclature. Veterinary Immunology and Immunopathology 39, 11–12. Naessens, J. and Howard, C.J. (1993) Leukocyte antigens of cattle and sheep. Monoclonal antibodies submitted to the Second Workshop. Veterinary Immunology and Immunopathology 39, 5–10. Naessens, J. and Williams, D.J. (1992) Characterization and measurement of CD5+ B cells in normal and Trypanosoma congolense-infected cattle. European Journal of Immunology 22, 1713–1718. Naessens, J., Newson, J., Bensaid, A., Teale, A.J., Magondu, J.G. and Black, S.J. (1985) De novo expres- sion of T cell markers on Theileria parva-transformed lymphoblasts in cattle. Journal of Immunology 135, 4183–4188. Naessens, J., Newson, J., Williams, D.J. and Lutje, V. (1988) Identification of isotypes and allotypes of bovine immunoglobulin M with monoclonal antibodies. Immunology 63, 569–574. Naessens, J., Newson, J., McHugh, N., Howard, C.J., Parsons, K. and Jones, B. (1990) Characterization of a bovine leucocyte differentiation antigen of 145,000 MW restricted to B lymphocytes. Immunology 69, 525–530. Naessens, J., Teale, A.J. and Sileghem, M. (2002) Identification of mechanisms of natural resistance to African trypanosomiasis in cattle. Veterinary Immunology and Immunopathology 87, 187–194. Naessens, J., Leak, S.G., Kennedy, D.J., Kemp, S.J. and Teale, A.J. (2003a) Responses of bovine chi- maeras combining trypanosomosis resistant and susceptible genotypes to experimental infection with Trypanosoma congolense. Veterinary Parasitology 111, 125–142. Naessens, J., Mwangi, D.M., Buza, J. and Moloo, S.K. (2003b) Local skin reaction (chancre) induced fol- lowing inoculation of metacyclic trypanosomes in cattle by tsetse flies is dependent on CD4 T lympho- cytes. Parasite Immunology 25, 413–419. Nantulya, V.M. and Lindqvist, K.J. (1989) Antigen-detection enzyme immunoassays for the diagnosis of Trypanosoma vivax, T. congolense and T. brucei infections in cattle. Annals of Tropical Medicine and Parasitology 40, 267–272. Nantulya, V.M., Musoke, A.J., Rurangirwa, F.R. and Moloo, S.K. (1984) Resistance of cattle to tset- se-transmitted challenge with Trypanosoma brucei or Trypanosoma congolense after spontaneous recovery from syringe-passaged infections. Infection and Immunity 43, 735–738. Nene, V., Iams, K.P., Gobright, E. and Musoke, A.J. (1992) Characterisation of the gene encoding a candidate vaccine antigen of Theileria parva sporozoites. Molecular and Biochemical Parasitology 51, 17–27. Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 205 Nene, V., Inumaru, S., McKeever, D., Morzaria, S., Shaw, M. and Musoke, A. (1995) Characterization of an insect cell-derived Theileria parva sporozoite vaccine antigen and immunogenicity in cattle. Infec- tion and Immunity 63, 503–508. Ngaira, J.M., Nantulya, V.M., Musoke, A.J. and Hirumi, K. (1983) Phagocytosis of antibody-sensitized Trypanosoma brucei in vitro by bovine peripheral blood monocytes. Immunology 49, 393–400. Nolan, D.P., Geuskens, M. and Pays, E. (1999) N-linked glycans containing linear poly-N-acetyllac- tosamine as sorting signals in endocytosis in Trypanosoma brucei. Current Biology 9, 1169–1172. Noyes, H., Brass, A., Obara, I., Anderson, S., Archibald, A.L., et al. (2011) Genetic and expression ana- lysis of cattle identifies candidate genes in pathways responding to Trypanosoma congolense infec- tion. Proceedings of the National Academy of Sciences USA 108, 9304–9309. O’Gorman, G.M., Park, S.D., Hill, E.W., Meade, K.G., Mitchell, L.C., et al. (2006) Cytokine mRNA profiling of peripheral blood mononuclear cells from trypanotolerant and trypanosusceptible cattle infected with Trypanosoma congolense. Physiological Genomics 28, 53–61. O’Gorman, G.M., Park, S.D., Hill, E.W., Meade, K.G., Coussens, P.M., et al. (2009) Transcriptional profil- ing of cattle infected with Trypanosoma congolense highlights gene expression signatures underlying trypanotolerance and trypanosusceptibility. BMC Genomics 10, 207. Olds, C., Mwaura, S., Crowder, D., Odongo, D., van Oers, M., et al. (2012) Immunization of cattle with Ra86 impedes Rhipicephalus appendiculatus nymphal-to-adult molting. Ticks and Tick-borne Dis- eases 3, 170–178. Olobo, J.O. and Black, S.J. (1989) In vitro secretion of bovine immunoglobulins during pokeweed mitogen or pokeweed mitogen and antigen activation of lymphocytes. Veterinary Research Communications 13, 193–197. Patel, E.H., Lubembe, D.M., Gachanja, J., Mwaura, S., Spooner, P. and Toye, P. (2011) Molecular char- acterization of live Theileria parva sporozoite vaccine stabilates reveals extensive genotypic diversity. Veterinary Parasitology 179, 62–68. Pearson, T.W., Lundin, L.B., Dolan, T.T. and Stagg, D.A. (1979a) Cell-mediated immunity to Thei- leria-transformed cell lines. Nature 281, 678–680. Pearson, T.W., Roelants, G.E., Lundin, L.B. and Mayor-Withey, K.S. (1978) Immune depression in trypano- someinfected mice. I. Depressed T-lymphocyte responses. European Journal of Immunology 8, 723–727. Pearson, T.W., Roelants, G.E., Lundin, L.B. and Mayor-Withey, K.S. (1979b) The bovine lymphoid system: binding and stimulation of peripheral blood lymphocytes by lectins. Journal of Immunological Methods 26, 271–282. Pearson, T.W., Pinder, M., Roelants, G.E., Kar, S.K., Lundin, L.B., et al. (1980) Methods for derivation and detection of anti-parasite monoclonal antibodies. Journal of Immunological Methods 34, 141–154. Pearson, T.W., Hewett, R.S., Roelants, G.E., Stagg, D.A. and Dolan, T.T. (1982) Studies on the induction and specificity of cytotoxicity to Theileria-transformed cell lines. Journal of Immunology 128, 2509–2513. Pigott, D.M., Atun, R., Moyes, C.L., Hay, S.I. and Gething, P.W. (2012) Funding for malaria control 2006– 2010: a comprehensive global assessment. Malaria Journal 11, 246. Pinder, M., Musoke, A.J., Morrison, W.I. and Roelants, G.E. (1980a) The bovine lymphoid system: III. A monoclonal antibody specific for bovine cell surface and serum IgM. Immunology 40, 359–365. Pinder, M., Pearson, T.W. and Roelants, G.E. (1980b) The bovine lymphoid system: II. Derivation and partial characterlzation of monoclonal antibodies against bovine peripheral blood lymphocytes. Veter- inary Immunology and Immunopathology 1, 303–316. Pinder, M., Kar, S., Withey, K.S., Lundin, L.B. and Roelants, G.E. (1981a) Proliferation and lymphocyte stimulatory capacity of Theileria-infected lymphoblastoid cells before and after the elimination of intra- cellular parasites. Immunology 44, 51–60. Pinder, M., Withey, K.S. and Roelants, G.E. (1981b) Theileria parva parasites transform a subpopulation of T lymphocytes. Journal of Immunology 127, 389–390. Radley, D.E., Brown, C.G., Cunnigham, M.P., Kimber, C.D., Musisi, F.L., et  al. (1975) East Coast fever: challenge of immunised cattle by prolonged exposure to infected ticks. Veterinary Record 96, 525–527. Randolph, T.F., Saatkamp, H. and Perry, B.D. (2003) A short history of animal health economics. In: Pro- ceedings of the 10th International Symposium on Veterinary Epidemiology and Economics, 2003, Vina del Mar, Chile, p. 827. Reduth, D., Grootenhuis, J.G., Olubayo, R.O., Muranjan, M., Otieno-Omondi, F.P., et al. (1994) African buffalo serum contains novel trypanocidal protein. Journal of Eukaryotic Microbiology 41, 95–103. Rouse, B.T. and Babiuk, L.A. (1974) Host responses to infectious bovine rhinotracheitis virus. III. Isolation and immunologic activities of bovine T lymphocytes. Journal of Immunology 113, 1391–1398. 206 S.J. Black and C.L. Baldwin Rowlands, G.J., Nagda, S., Rege, J.E.O., Mhlanga, F., Dzama, K., et al. (2003) A Report to FAO on ‘The design, execution, and analysis of livestock breed surveys – a case study in Zimbabwe.’ ILRI, Nairobi. Rurangirwa, F.R., Musoke, A.J., Nantulya, V.M. and Tabel, H. (1983) Immune depression in bovine trypanosomiasis: effects of acute and chronic Trypanosoma congolense and chronic Trypano- soma vivax infections on antibody response to Brucella abortus vaccine. Parasite Immunology 5, 267–276. Scollay, R., Hall, J. and Orlans, E. (1976) Studies on the lymphocytes of sheep. II. Some properties of cells in various compartments of the recirculating lymphocyte pool. European Journal of Immunology 6, 121–125. Sendashonga, C.N. and Black, S.J. (1982). Humoral immune responses against Trypanosoma brucei vari- able surface antigen are induced by degenerating parasites. Parasite Immunology 4, 245–257. Shapiro, S.Z. (1994) Failure of immunization with trypanosome endocytotic vesicle membrane proteins to provide nonvariant immunoprotection against Trypanosoma brucei. Parasitology Research 80, 240–244. Shapiro, S.Z. and Murray, M. (1982) African trypanosome antigens recognized during the course of infec- tion in N’dama and Zebu cattle. Infection and Immunity 35, 410–416. Shapiro, S.Z., Naessens, J., Liesegang, B., Moloo, S.K. and Magondu, J. (1984) Analysis by flow cytometry of DNA synthesis during the life cycle of African trypanosomes. Acta Tropica 41, 313–323. Shapiro, S.Z., Voigt, W.P. and Ellis, J.A. (1989) Acquired resistance to ixodid ticks induced by tick cement antigen. Experimental and Applied Acarology 7, 33–41. Sileghem, M. and Flynn, J.N. (1992) Suppression of interleukin 2 secretion and interleukin 2 receptor ex- pression during tsetse-transmitted trypanosomiasis in cattle. European Journal of Immunology 22, 767–773. Sileghem, M. and Naessens, J. (1995) Are CD8 T cells involved in control of African trypanosomiasis in a natural host environment? European Journal of Immunology 25, 1965–1971. Sileghem, M., Saya, R., Ellis, J.A., Flynn, J.N., Peel, J.E. and Williams, D.J. (1992) Detection and neutral- ization of bovine tumor necrosis factor by a monoclonal antibody. Hybridoma 11, 617–627. Sileghem, M., Flynn, J.N., Logan-Henfrey, L. and Ellis, J. (1994) Tumour necrosis factor production by monocytes from cattle infected with Trypanosoma (Duttonella) vivax and Trypanosoma (Nannomonas) congolense: possible association with severity of anaemia associated with the disease. Parasite I mmunology 16, 51–54. Smith, J.B., McIntosh, G.H. and Morris, B. (1970) The traffic of cells through tissues: a study of peripheral lymph in sheep. Journal of Anatomy 107, 87–100. Sopp, P., Howard, C.J. and Parsons, K.R. (1991) Investigating monoclonal antibodies to bovine ‘null’ cell anti- gens using two-colour immunofluorescence. Veterinary Immunology and Immunopathology 27, 163–168. Spits, H., de Vries, J.E. and Terhorst, C. (1981) A permanent human cytotoxic T-cell line with high killing capacity against a lymphoblastoid B-cell line shows preference for HLA A, B target antigens and lacks spontaneous cytotoxic activity. Cellular Immunology 59, 435–447. Sprong, H., Trentelman, J., Seemann, I., Grubhoffer, L., Rego, R.O., et al. (2014) ANTIDotE: anti-tick vac- cines to prevent tick-borne diseases in Europe. Parasites and Vectors 7, 77. Steverding, D. (2006)On the significance of host antibody response to the Trypanosoma brucei transferrin receptor during chronic infection. Microbes and Infection 8, 2777–2782. Tabel, H., Wei, G. and Bull, H.J. (2013) Immunosuppression: cause for failures of vaccines against African trypanosomiases. PLoS Neglected Tropical Diseases 7, e2090. Takashima, I., Olson, C., Driscoll, D.M. and Baumgartener, L.E. (1977) B-lymphocytes and T-lymphocytes in three types of bovine lymphosarcoma. Journal of the National Cancer Institute 59, 1205–1209. Taracha, E.L., Goddeeris, B.M., Scott, J.R. and Morrison, W.I. (1992) Standardization of a technique for analysing the frequency of parasite-specific cytotoxic T lymphocyte precursors in cattle immunized with Theileria parva. Parasite Immunology 14, 143–154. Taracha, E.L., Awino, E. and McKeever, D.J. (1997) Distinct CD4+ T cell helper requirements in Theileria parva-immune and -naive bovine CTL precursors. Journal of Immunology 159, 4539–4545. Taylor, K.A., Lutje, V., Kennedy, D., Authié, E., Boulange, A., et  al. (1996) Trypanosoma congolense: B-lymphocyte responses differ between trypanotolerant and trypanosusceptible cattle. Experimental Parasitology 83, 106–116. Taylor, K., Mertens, B., Lutje, V. and Saya, R. (1998) Trypanosoma congolense infection of trypanotolerant N’Dama (Bos taurus) cattle is associated with decreased secretion of nitric oxide by interferon-γ activated monocytes and increased transcription of interleukin-10. Parasite Immunology 20, 421–429. Impact Assessment of Immunology and Immunoparasitology at ILRAD and ILRI 207 Teale, A.J. and Kemp, S.J. (1987) A study of BoLA class II antigens with BoT4+ T lymphocyte clones. Animal Genetics 18, 17–28. Teale, A.J., Morrison, W.I., Goddeeris, B.M., Groocock, C.M., Stagg, D.A. and Spooner, R.L. (1985) Bovine alloreactive cytotoxic cells generated in vitro: target specificity in relation to BoLA phenotype. Immunology 55, 355–362. Teale, A.J., Baldwin, C.L., Ellis, J.A., Newson, J., Goddeeris, B.M. and Morrison, W.I. (1986) Alloreactive bovine T lymphocyte clones: an analysis of function, phenotype, and specificity. Journal of Immunology 136, 4392–4398. Teale, A.J., Baldwin, C.L., Morrison, W.I., Ellis, J. and MacHugh, N.D. (1987) Phenotypic and functional characteristics of bovine T lymphocytes. Veterinary Immunology and Immunopathology 17, 113–123. Toye, P.G., MacHugh, N.D., Bensaid, A.M., Alberti, S., Teale, A.J. and Morrison, W.I. (1990) Transfection into mouse L cells of genes encoding two serologically and functionally distinct bovine class I MHC molecules from a MHC-homozygous animal: evidence for a second class I locus in cattle. Immunology 70, 20–26. Tuo, W., Bazer, F.W., Davis, W.C., Zhu, D. and Brown, W.C. (1999) Differential effects of type I IFNs on the growth of WC1– CD8+ γδ T cells and WC1+ CD8– γδ T cells in vitro. Journal of Immunology 162, 245–253. Usinger, W.R. and Splitter, G.A. (1981) Two molecularly independent surface receptors identify bovine T lymphocytes. Journal of Immunological Methods 45, 209–219. Vickerman, K. (1978) Antigenic variation in trypanosomes. Nature 273, 613–617. von Schubert, C., Xue, G., Schmuckli-Maurer, J., Woods, K.L., Nigg, E.A. and Dobbelaere, D.A. (2010) The transforming parasite Theileria co-opts host cell mitotic and central spindles to persist in continuously dividing cells. PLoS BiolOGY 8, e1000499. Wang, J., van Praagh, A., Hamilton, E., Wang, Q., Zou, B., et al. (2002)Serum xanthine oxidase: origin, regulation, and contribution to control of trypanosome parasitemia. Antioxidants and Redox Signaling 4, 161–178. Wang, Q., Murphy, N. and Black, S.J. (1999) Infection-associated decline of cape buffalo blood catalase augments serum trypanocidal activity. Infection and Immunity 67, 2797–2803. Wells, P.W., Emery, D.L., Hinson, C.A., Morrison, W.I. and Murray, M. (1982) Immunization of cattle with a variant-specific surface antigen of Trypanosoma brucei: influence of different adjuvants. Infection and Immunity 36, 1–10. Williams, D.J., Newson, J. and Naessens, J. (1990) Quantitation of bovine immunoglobulin isotypes and allotypes using monoclonal antibodies. Veterinary Immunology and Immunopathology 24, 267–283. Williams, D.J., Taylor, K., Newson, J., Gichuki, B. and Naessens, J. (1996) The role of anti-variable surface glycoprotein antibody responses in bovine trypanotolerance. Parasite Immunology 18, 209–218. Zinkernagel, R.M. and Doherty, P.C. (1974) Immunological surveillance against altered self components by sensitised T lymphocytes in lymphocytic choriomeningitis. Nature 251, 547–548. 5 Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 Brian Perry1, Bernard Bett2, Eric Fèvre2, Delia Grace3 and Thomas Fitz Randolph2 1University of Oxford and University of Edinburgh, UK; 2International Livestock Research Institute, Nairobi, Kenya; 3International Livestock Research Institute, Nairobi, Kenya and University of Greenwich, UK Contents Executive Summary 209 The problem 209 ILRI’s role in the global context 209 Impact of ILRI’s research 210 Scientific impacts 210 Economic impact assessment 210 Developmental impacts 210 Capacity development 210 Partnerships 210 Impacts on human resources capacity in veterinary epidemiology 211 Impacts on national animal health departments and services 211 Impacts on animal health constraints in developing countries 211 Impacts on ILRI’s research and strategy 211 Introduction 211 The introduction of veterinary epidemiology and economics at ILRAD 212 Field studies in Kenya 212 Tick-borne disease dynamics in eastern and southern Africa 214 The heartwater studies in Zimbabwe 215 Economic impact assessments of tick-borne diseases 216 Tick and tick-borne disease distribution modelling 217 Modelling the infection dynamics of vector-borne diseases 218 Impacts of Trypanosomiasis and its Control 218 Economic impact of trypanosomiasis 218 The epidemiology of resistance to trypanocides 219 The development of a modelling technique for evaluating control options 219 Sustainable trypanosomiasis control in Uganda 220 Sustainable trypanosomiasis control in the Ghibe Valley of Ethiopia 220 Spatial modelling of tsetse distributions 220 Preventing and containing trypanocide resistance in the cotton zone of West Africa 220 Rabies Research: A Networking and Capacity-building Role in Africa 221 © International Livestock Research Institute 2020. The Impact of the International 208 Livestock Research Institute (eds J. McIntire and D. Grace) Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 209 The Economic Impacts of Rinderpest Control 222 Applying Economic Impact Assessment Tools to Foot-and-mouth disease Control 223 The southern Africa FMD economic impact study 224 Economic Impacts of FMD in Peru, Colombia and India 224 Economic impacts of FMD control in endemic settings in low- and middle-income countries 224 The Global Foot-and-Mouth Disease Research Alliance (GFRA) 224 Rift Valley Fever 225 Economic impact assessment of control options and calculation of disability-adjusted life years (DALYs) 226 RVF risk maps for eastern Africa 226 Land-use change and RVF infection and disease dynamics 226 Epidemiology of Gastrointestinal Parasites 227 Priorities in Animal Health Research for Poverty Reduction 227 The Wellcome Trust Epidemiology Initiatives 228 The Broader Economic Impact Contributions 228 The Responses to Highly Pathogenic Avian Influenza 229 The ISVEE Experience 230 The Role of Epidemiology in ILRAD and ILRI 230 The Impacts of ILRAD and ILRI’s Epidemiology 231 Capacity development in veterinary epidemiology and impact assessment 231 Impacts on national animal health departments and services 231 Impacts on animal health constraints in developing countries 231 Impacts on ILRI’s research and strategy 231 References 232 Executive Summary to national to global. Furthermore, veterinary epidemiological and economic impact sciences The problem are key components in a number of the global grand challenges relating to disease control, The effectiveness of detecting and controlling climate change and food security. animal diseases is dependent on a solid under- standing of their dynamics and impacts through scientifically sound qualitative and quantita- ILRI’s role in the global context tive methods by trained personnel. Veterinary epidemiology is the systematic characterization The International Livestock Research Institute and explanation of patterns of animal diseases (ILRI) and its predecessor, the International and the use of this information in the reso- Laboratory for Research on Animal Diseases lution of animal and human health problems. (ILRAD), have played an important international This discipline exploits an increasing inventory role in identifying and developing epidemio- of tools for effective data gathering, assembly logical tools for the investigation and resolution and analysis, modelling and reporting, all of animal health constraints to livestock pro- targeted at decision making by producers, duction. When the group was formed in 1987, it governments and international development was charged with addressing the two diseases agencies. Furthermore, the integration of considered the priority in Africa, namely East epidemiology with agricultural economics Coast fever (ECF) and trypanosomiasis, to sup- and other social sciences provides a uniquely port the development of vaccines against these effective tool for evaluating disease as a con- two diseases, which was the mandate of ILRAD. straint to broader development agendas, for Following the transition from ILRAD to ILRI, assessing the absolute and relative economic the institution has been a leader in exploring importance of diseases, and for evaluating the new epidemiological approaches, and in widen- costs and benefits of alternative intervention ing the disciplinary spectrum of epidemiological options, at different levels ranging from farm investigations. However, arguably most important 210 B. Perry et al. of all, ILRI has played a facilitating role in collab- around the world and has contributed substan- orating with countries, institutions and organ- tially to research and development networking izations in Africa, Asia and Latin America to re- in many areas of animal health, including tick- spond to requests for both short- and long-t erm borne diseases, rabies, and the fundamental tools partnerships and support at international, re- of epidemiology and socio-economic impacts. gional, national and local levels, and in exten- sive capacity building in epidemiological tools, Scientific impacts techniques and approaches. Initially, ILRAD and ILRI focused almost The group brought the science of structured exclusively on the dynamics and impact of tick- epidemiological analysis to the institute in meth- borne pathogens of livestock in Africa, but during odological terms (particularly in observational the 1990s, the geographical focus, disciplinary field studies), and also expanded the disciplinary make-up and range of tools used broadened sub- contributions to include social, economic and stantially, tackling multiple diseases in Africa, environmental considerations in the analysis Asia and Latin America and building capacity in of disease impacts, and the impacts of disease epidemiological and economic impact assess- control interventions. ment techniques. For a period of 15 years (1987–2002) ILRAD/ Economic impact assessment ILRI’s epidemiology and socio-economic impact assessment capacity was assembled in one team Most of the earlier economic impact assessments serving a range of institutional and externally carried out by ILRAD and ILRI were disease commissioned needs and was recognized inter- specific, such as on ECF, rinderpest and FMD, nationally for its focus of health issues affecting and built on evidence derived from underlying development and poverty reduction. Through a epidemiological data. Economic impact assess- major study of animal health research priorities ments gained increasing momentum as ILRI’s commissioned by UK Department for International mandate broadened. The epidemiology group Development (DFID), the team made a substantial concluded that no longer should studies of the contribution to the design of ILRI’s new strategy, economics of diseases of production animals be which emerged in 2002, and to the poverty- limited to animal scientists seeking the collabor- focused agendas of other organizations such as ation of agricultural economists to affix prices the Food and Agriculture Organization of the to estimated productivity losses, and the new United Nations (FAO) and the Wellcome Trust. discipline of animal health economics emerged in which the quality of economic evaluations de- pended on integrating the products of good epi- Impact of ILRI’s research demiological studies into economic frameworks. In 1998, the World Organisation for Ani- Veterinary epidemiological and economic impact mal Health (Office International des Epizooties, sciences at ILRAD and ILRI have left a valuable or OIE) approached ILRI to compile and edit a legacy of publications in peer-reviewed journals, special edition of the OIE Scientific and Technical strategic reports and policy documents, as well Review on Animal Health Economics, which com- as methodologies and approaches that have prised nine chapters on various topics such as the been applied in virtually all corners of the world, demands of economic impact knowledge, and and many trained epidemiologists now serving seven case study chapters addressing specific different institutional needs in Africa, Asia, Latin diseases and different scenarios affecting the America, Europe and Australia. validity of the economic studies. The range of diseases subject to epidemio- Developmental impacts logical and economics research has been wide, and included the viral infections rinderpest, foot- While we cannot ascribe higher productivity to and-mouth disease (FMD), Rift Valley fever (RVF) ILRI’s veterinary epidemiology research per se, and highly pathogenic avian influenza (HPAI), the research has increased our understanding and the context of impact assessments has been of infection dynamics, which has influenced applied to trade, livelihoods and poverty reduction. disease control and the role of interventions in ILRI has worked with a multitude of partners different settings. Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 211 Capacity development the preparedness and responses to RVF in east- ern Africa, to a greater understanding of the Over the years, epidemiology has gone through economic impact of rinderpest in Africa and of both administrative and locality changes, seen FMD in Africa, Asia and Latin America, and to as a unified entity for 15 years, and later under regional understanding of the drivers of rabies the new ILRI strategy it was fragmented and control. More recently, epidemiology research at diminished. However, during the 15-year period ILRI has contributed substantially to our under- from 1987 to 2002, the institution had sup- standing of food safety risks in formal and ported approximately 15 MSc students and 37 informal markets and to the dynamics and risks PhD students with approximately 50 scientists of zoonotic diseases. The research has also con- predominantly from African countries. In add- tributed to the global understanding of the im- ition, there have been several postdoctoral fel- portance of these and other diseases to African lows who have been trained. As most of the epi- livestock systems and to the particular animal demiology is conducted under a specific disease, health constraints facing the poorer sectors of it is difficult to determine how many more have Africa’s livestock-engaged communities. been trained to date. Partnerships Impacts on ILRI’s research and strategy During the days of ILRAD, the epidemiology ILRI has worked with a multitude of partners and socio-economic programme had little or around the world and contributed substan- no impact on ILRAD’s research and strategy; tially to research and development networking rather, it was seen as providing evidence justi- in many areas of animal health, including fying the existence of the laboratory-based tick-borne diseases, rabies and the fundamen- vaccine research for the two target haemopara- tal tools of epidemiology and socio-economic sitic diseases. Nevertheless, after ILRI’s birth in impacts. In 2007, the Participatory Epidemi- 1995, the programme played an important role ology Network for Animal and Public Health in providing impact assessment services, which (PENAPH) was developed to connect groups progressively enhanced the engagement of the and individuals who apply participatory epi- institution with different national, regional and demiology in controlling emerging and exist- donor clients. This situation changed dramat- ing diseases. ically in 2002 following the publication of the Impacts on human resources capacity in DFID-commissioned study on animal research veterinary epidemiology priorities for poverty reduction. The matrix of three ‘pathways out of poverty’ provided the During the 15-year period from 1987 to 2002 a framework of the new institutional thematic substantial number of MSc and PhD students structure, not just for animal health research were trained through incorporation into ILRI’s but also for ILRI’s entire programme. research activities; these were predominantly from African countries. Introduction Impacts on national animal health departments and services Veterinary epidemiology is the systematic char- The epidemiology group provided a role model acterization and explanation of patterns of animal of investigative problem solving, which was diseases and, importantly, the use of this infor- picked up, copied and adopted by some institu- mation in the resolution of animal and human tions in African countries. health problems. It is a subject that exploits an increasing inventory of tools for effective data Impacts on animal health constraints in gathering, assembly and analysis, targeted at de- developing countries cision making in the field of animal disease con- trol and sustainable livestock enterprise develop- ILRI’s epidemiology research has made sub- ment. The integration of epidemiology with stantial contributions to our understanding and agricultural economics and other social sciences control of ECF and trypanosomiasis in Africa, to provides a uniquely effective tool for evaluating 212 B. Perry et al. disease as a constraint to broader development quently extended for a further 2 years) supported agendas, for assessing the absolute and relative what became the Epidemiology and Socioeco- economic importance of diseases, and for evalu- nomics Programme. The group was joined in 1992 ating the costs and benefits of alternative inter- by an ecologist to explore the environmental vention options, at different levels ranging from impacts of trypanosomiasis control (Reid et al., farm to national to global. ILRI and its predecessor, 1995). The team expanded in the late 1990s to ILRAD, have played an important international include another staff epidemiologist and two role in exploiting epidemiological tools for the postdoctoral epidemiologists. investigation and resolution of animal health The group rapidly laid out a work plan, constraints to livestock production and poverty beginning with the establishment of databases reduction in many regions of the developing on African production systems at risk from the world. Furthermore, ILRI has been a leader in two diseases and on methodologies for determining exploring new epidemiological approaches, their impact. This challenge led to the realization and in widening the disciplinary spectrum of that disease incidence and prevalence data in Af- epidemiological investigations. However, argu- rica were scarce and unreliable, notably on the ably most important of all, ILRI has played a fa- structure and ownership of the livestock popula- cilitating role in collaborating with countries, tions at risk. The need for structured quantita- institutions and organizations in Africa, Asia tive epidemiology capacity emerged, which led and Latin America in response to requests for both to a sustained programme of data assembly, digi- short- and long-term partnership and support at tal data documentation and assembly, the devel- international, regional, national and local levels, opment of modelling techniques and, of course, and in extensive capacity building in epidemio- the gathering of field data. logical tools, techniques and approaches. Field studies in Kenya The introduction of veterinary epidemiology and economics at ILRAD The first field site of diverse ecosystems and dis- ease impacts was in Kilifi District on the Kenyan The fundamental belief at the creation of ILRAD coast. A collaborative programme between the in 1974 was that vaccines against ECF and tryp- International Livestock Centre for Africa (ILCA) anosomiasis were the mandate of ILRAD. The and the Kenya Agricultural Research Institute evidence available to answer these questions at (KARI) was established in 1988 at KARI’s Mtwapa the time was derived almost entirely from African Regional Research Centre, near Mombasa. veterinary services and diagnostic laboratories, The smallholder dairy group was setting up which had, for the previous 60 or so years, been a broad study on the constraints to smallholder servicing livestock production enterprises of the milk production in the coastal lowlands of colonial powers. There were little if any economic Kenya and how extension services covering the data to quantify the impacts of these diseases, areas of feed and health could be improved. Spe- even in commercial systems, and the potential cifically, the study estimated the demand for milk returns from vaccines. and dairy products, identified technical and pol- ILRAD alone among the international agri- icy constraints on production in mixed small- cultural research centres of CGIAR had a unique holder farming systems, evaluated dairy cattle mandate to carry out basic research and as such breed resources, estimated disease risk to dairy undertook very little of the technology transfer cattle and tested disease control methods, and functions. Pressure progressively increased from developed feeding systems appropriate to small- partners to quantify the impacts of these two holder dairy production systems. diseases on African agriculture in order to better The ILRAD epidemiology team provided justify the scientific investment in the study of support to the studies on the epidemiology and these two diseases. impact of ECF in the form of design and analysis Thus, it was some 13 years after the estab- of studies led by KARI staff. The challenge was lishment of ILRAD that veterinary epidemiology a total lack of data on ECF occurrence, and so a was introduced into the institute. The Rockefeller series of cross-sectional studies was set up. Key was Foundation initially for 3  years (and subse- understanding the link between infection preva- Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 213 lence, as measured by antibodies to Theileria par- first to Kiambu District, with a 1-year study of va in an indirect fluorescent antibody test, and the dynamics of theileriosis (O’Callaghan et al., disease incidence. ILRAD’s entry into this re- 1998). This was followed in 1994 by investiga- search area was at a time when infection preva- tions in the neighbouring Muranga District, lence, as measured by antibody prevalence, had which hosted a range of livestock production not been correlated with disease incidence, and, systems in diverse AEZs. There were five distinct where prevalence studies had been undertaken, AEZs within Muranga District, giving the oppor- they had often been reported on the basis of ad- tunity to investigate the influence of a range of ministrative boundaries such as the FAO’s 1975 climate, livestock breeds and farming practice study in Kenya (Kariuki, 1988). variables on ECF dynamics. The work started In 1983, the Farm Management Handbook of with a cross-sectional serological study on 750 Kenya was published (Jaetzold and Schmidt, smallholder dairy farms in Muranga District, se- 1983), which provided a unique landscape syn- lected in a stratified random sample (Gitau et al., thesis of Kenya’s agricultural environment, ag- 1997), which showed the markedly different gregating a number of variables into a kaleido- prevalences of T. parva infection across AEZs. scope of colours representing the suitability for The area was typical of the highland areas of different crops and agricultural enterprises. With eastern Africa in which the process of smallholder the knowledge that the epidemiology and impacts dairy intensification was gaining momentum. of ECF were highly dependent on environmental The investigation continued with a study that suitability for the main vector tick, Rhipicephalus related prevalence with incidence, case morbid- appendiculatus (the brown ear tick), the zone ity and case mortality (Gitau et al., 1999), and boundaries provided a new and useful sampling with a study of how these infections affected frame that had previously been unexploited. weight gain in calves (Gitau et al., 2001). The A series of studies was set up in coastal Kenya work concluded with a synthesis of the implica- to determine the prevalence and incidence of tions of the research on disease risk and on the ECF and the other tick-borne infections such as potential role of vaccination against ECF (Gitau anaplasmosis and babesiosis, and to evaluate et al., 2000). the role of immunization against ECF using the The synthesis concluded that ECF risk is infection-and-treatment method (ITM). The low in predominately zero-grazing areas. Thus, studies provided an initial quantitative assess- tick control or future vaccination programmes ment of antibody prevalence to the spectrum of will probably only be used by very risk-averse tick-borne disease parasites in order to assess the farmers who wish to protect their highly valu- epidemiological status of these infections in both able cows from the low risk of ECF mortality. In indigenous Zebu cattle kept, and in improved dairy contrast, for open grazing systems, particularly cattle in three different agroecological zones (AEZs) in the lower- elevation upper-midland 4 (UM4) (Deem et al., 1993; Maloo et al., 2001a,b,c). zone, the risk of ECF is much greater and prob- The coast work provided an opportunity to ably much more variable. In this system, there engage at the front line with national partners will be much more substantial direct impact of and to explore impact study design; it also illus- ECF control programmes. In areas where ECF trated the need to disaggregate factors affecting control will be through vaccination, irrespect- ECF epidemiology and impact. However, as the ive of the grazing management system, there coastal systems were not fully representative of will be a greater likelihood of the development the intensifying livestock systems in the temper- of endemic stability. Increased vaccination ate highland areas of eastern Africa, additional coverage to enhance the development of herd studies were set up. The first was in Uasin Gishu, immunity, combined with modification of where larger-scale dairy and beef production acaricide control strategies to allow sufficient was rapidly being replaced by small-scale com- challenge, seemed to offer the best prospect for mercial dairy enterprises (Mukhebi et al., 1992a). establishing endemic stability. It was quite clear This work coincided with the introduction of the from these studies that attention needs to be discipline of human nutrition into the impact paid to the variation in ECF risk, both spatially equation (Curry et al., 1996). (as ECF risk changes over relatively short In 1992, the focus of the ECF epidemiology geographical distances) and temporally (sea- studies moved to the central highlands of Kenya, sonally), to develop optimal combinations of 214 B. Perry et al. control measures for ECF under different eco- which all exert their influence as a gradient (or logical and grazing situations. cline) of effects: It was in 1997 that the ECF epidemiology work expanded into other parts of Kenya, follow- • The ecological cline, in which the climatic ing the submission of a research proposal to the suitability for the tick vector varies with International Fund for Agricultural Develop- rainfall and altitude; the ecological cline ment (IFAD). The revised proposal strengthened gradient can be affected by differences in the epidemiology and impact assessment com- vegetation cover. ponents and placed them in an ex ante context. • The host genetic cline, in which purebred Vaccine efficacy trials were limited to two sites in taurine cattle bred under tick-free condi- Kenya, while the impact assessment broadened tions are highly susceptible to disease, and into new areas. The impact assessment studies taurine cattle bred in tick-borne infection included an evaluation of mechanisms for optimal endemic areas and some Zebu breeds (such delivery, adoption and impact of the p67 vaccine as Boran) bred in tick-free conditions are (p67 is the major surface protein of T. parva moderately susceptible, but Zebu cattle bred sporozoites), determining the impact of a recom- in tick-borne infection endemic areas are of binant vaccine on a series of productivity and low susceptibility to disease. economic indicators in smallholder dairy systems. • The feeding management cline, which con- The project also included key laboratory studies to trols the exposure of hosts to the ecological support the field studies and disease-modelling conditions; this can range from no influence, work (Ochanda et al., 1998). where cattle are herded on natural pasture, Table 5.1 outlines the various studies under- to complete influence, where cattle are kept taken in Kenya, illustrating the differences in on concrete and fed on cultivated forage impacts by region, AEZ and grazing management. grasses, as in the smallholder zero-grazing There were a wide variety of products units of eastern Africa. emerging from the IFAD study, ranging from • The tick control cline, where tick control Technical Advisory Notes such as ‘Assessing ranges from highly effective, regular appli- farmer preferences for the provision of livestock cation through to no tick control at all. health services’ to international presentations (Leneman et al., 2000; Kiara et al., 2000, 2003; Ndung’u et al., 2000, 2003; Wanyangu et al., 2000; Di Giulio et al., 2003, Karimi et al., 2003; Tick-borne disease dynamics in eastern O’Callaghan et al., 2003; Diaz et al., 2003) to and southern Africa peer-reviewed papers in scientific journals. In a review article of the ECF work (Perry and At the start of these intense epidemiological Young, 1995), it was postulated that the degree investigations in Kenya, a regional meeting of mortality and production losses from T. parva on tick-borne diseases was held in Lilongwe, infections were dependent on four key factors, Malawi, in 1988 (Dolan, 1989), which provided Table 5.1. ECF risks by region, AEZ and grazing management. Study O’Callaghan (1998) Gitau (1998) Maloo (1993) Calves (<1 year of age); Calves (≤6 months of age); Calves ( <1 year of age); Incidence rate type incidence density cumulative incidence incidence density Grazing management Zero grazing Pasture Zero grazing Pasture Zero grazing Pasture Number of animals (n) 93 108 134 91 38 50 ECF morbidity rate (%) 5.5 10.9 11.8 49 36.4 68.8 ECF mortality rate (%) 0 2.2 1.7 20.6 20.8 49.7 Case-fatality proportion (%) 0 25 28.6 38.1 57.1 72.2 Seroconversion rate (%) 41.4 56.5 58.4 74 0 0 Morbidity proportion (%) 13.3 7.7 12.2 36.1 0 0 Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 215 an opportunity for sharing of information and Development Community (SADC), the Univer- understanding on theileriosis throughout the sities of Florida (USA) and Warwick (UK), and eastern and southern African regions, and where ILRI determined and quantified the infection dy- it became apparent that there were significant namics of heartwater in the major production differences in the disease epidemiology between systems and AEZs of Zimbabwe, the economic the eastern and southern regions. The Lilongwe impact of the disease, and the technical and eco- meeting led to a textbook on theileriosis, The Epi- nomic viability of different control interven- demiology of Theileriosis in Africa (Norval et al., tions, with particular emphasis on the role of 1992b). The book was published by Academic inactivated vaccines. The project outputs were Press, with ILRAD supporting the time contribu- as follows: tions of the three authors and the preparation of camera-ready copy, and the book became the • Distributions of tick vectors in Zimbabwe reference point for all those working on control were defined and documented. A national of the disease. This book remains the only text- survey of 3000 collections determined that book on theileriosis, a product of ILRAD’s epi- Amblyomma hebraeum is the dominant tick demiology team of substantial impact. in the south and that it had spread into cen- The disease called ECF was said to have tral and eastern areas of the high veld. The been eradicated in southern Africa as a result survey also found that Amblyomma variega- of an intensive dipping programme, with the tum is present mainly in the north-west but last case occurring in Swaziland in 1960. that it is also found in central and eastern Nevertheless, T. parva infections persist, but with parts of the high veld, with some overlap the official eradication declared, the names of of the two species (Norval et al., 1994; Peter theileriosis and corridor disease have been used et al., 1998, 1999). to describe disease outbreaks. This curious con- • Spread of heartwater was documented and frontation of science and officialdom of disease quantified, and factors affecting the spread and parasite nomenclature was reviewed in an were determined. A. hebraeum had spread article entitled ‘The naming game: the changing far north due largely to movement of cattle fortunes of East Coast fever and Theileria parva’ (and some wildlife) to the high veld. A grad- (Perry and Young, 1993). It is encouraging ual reduction in acaricide use, particularly how these studies have led to our current in the communal lands, contributed to the understanding of theileriosis epidemiology, as expanding distribution of this tick (Norval illustrated by a recent review by Gachohi et al. et al., 1992a). (2012). • Infection dynamics in the tick vector and mammalian hosts were determined and quan- tified. Endemic stability, in which population The heartwater studies in Zimbabwe immunity develops, was found to be wide- spread but not present where acaricides were Through funding from the US Agency for used intensively to interrupt natural infec- International Development (USAID), a 5-year tion. These results suggest that use of inacti- project on the epidemiology and impact of an- vated vaccines in many circumstances will other important tick-borne disease of livestock allow a reduction in acaricide use with a in Africa, heartwater (also known as ehrlichio- transition to endemic stability and subse- sis; caused by Ehrlichia ruminantium infection, quent natural infection boosting the vac- formerly Cowdria ruminantium) was initiated in cinal immunity. 19941. This work subsequently received ILRI’s • The impact of endemic stability, and carrier award (and ILRI’s nomination for the CGIAR infections, on sheep productivity was deter- Chairman’s Award) for Scientific Partner- mined. Studies in sheep revealed that creating ship, 2000. endemic stability artificially with vaccines The collaborative research in epidemiology does not harm the health and reproductive and economics between the Veterinary Research performance of breeding ewes or the growth Laboratory (VRL) in Harare, the Heartwater and milk consumption of their lambs (Mar- Research Project of the Southern African tinez et al., 1999a,b). 216 B. Perry et al. • Infection dynamics models were devel- The successes of this collaborative project oped using data generated by the research. were considerable, with all objectives met and all A mathematical model of the infection dynam- findings published within a period of 5  years. ics of the heartwater pathogen, E. ruminan- The results of the project have given scientists a tium, showed that endemic stability is due sound understanding of the factors influencing principally to the protection of calves and the distribution of tick vectors, and the infection lambs against disease by innate or maternally dynamics and impacts of the disease in different derived factors (O’Callaghan et al., 1998). AEZs and production systems, and predictions • Economics of livestock production in heart- of the technical and economic impacts of control water areas was determined. Both large- with a new generation of inactivated vaccines now and small-scale livestock production could emerging. The results generated have strong im- be increased significantly with more, and plications for heartwater control in other coun- more cost-effective, heartwater control tries of Africa. Design and modelling features of methods (Perry et al., 1998; Chamboko the study have been used in studies of other tick- et al., 1999). borne livestock diseases. • The economic impact of heartwater, and of Apart from its technical achievements, the future vaccine use, was determined. The project boosted scientific capacity, particularly annual total direct losses in Zimbabwe (acari- in Zimbabwe, through postgraduate training for cide costs, milk losses, treatment costs) from national scientists. Project members produced heartwater were estimated to be US$5.6 mil- 23 papers in peer-reviewed journals, 15 of which lion. A new inactivated vaccine was predicted were authored by VRL scientists. Project staff to have a benefit:cost ratio of 2.4:1 in the produced another 36 publications, including 22 communal sectors and 7.6:1 in the com- presentations at scientific meetings and ten mercial sectors (Mukhebi et al., 1999). articles in the project newsletter. • The efficacy of future vaccine use was evaluated in epidemiological models. The timing of vaccination and frequency of Economic impact assessments revaccination were shown to have greater of tick-borne diseases effects on population protection than vac- cine efficacy. In the face of an epidemic, the The first opportunity to undertake an economic frequency of administration is critical to a impact assessment came in the late 1980s with vaccine’s success. Vaccines of relatively the ILRAD/KARI partnership on the Kenya low efficacy (about 50%) can significantly coast, where an immunization trial was being reduce livestock morbidity and mortality if carried out using ITM to control ECF. Adrian administered with appropriate frequency Mukhebi showed greater profitability in immun- (O’Callaghan et al., 1999). ized cattle compared with unimmunized, through • The economic impact of the disease and of lower mortality and higher weight gains (Muk- its control through vaccines was evaluated hebi et al., 1989). This was a solid first piece of and quantified in the countries of the SADC evidence, although it was compiled on a state-run region. In total, 31 million cattle and 28 Agricultural Development Corporation beef ranch. million small ruminants were found to be at The authors commented that ‘these results and risk of heartwater in the nine SADC coun- the recommendation apply to one immunization tries affected: Angola, Botswana, Malawi, trial on one farm which was under an atypical Mozambique, South Africa, Swaziland, Tan- management system for the region’. zania, Zambia and Zimbabwe. The total annual Building on the potential for ITM, Mukhebi losses were estimated at US$47.6 million, of et al. (1990) then dissected the complicated vac- which 61% were production losses and 39% cine preparation process and calculated the costs control costs. New inactivated heartwater of establishing a vaccine production facility (at a vaccines could yield benefit:cost ratios of up hypothetical site in Kenya but using a method- to 4.4:1, particularly in commercial and ology generic to other countries). emerging market-orientated systems of the These initial forays into the economics region (Minjauw et al., 1998, 2000). of ECF and its control led to an Africa-wide Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 217 assessment, presenting the estimated total costs assessments, running the software ARC/INFO of ECF in affected countries in 1989 as US$168 from the Environmental Systems Research Insti- million (Mukhebi et al., 1992b). The authors tute (ESRI). A memorandum of understanding calculated benefit:cost ratios for the control of was set up between ILRAD and UNEP, and two ECF through vaccination of between 8.9 and UNEP scientists contributed data assembly and 16.8, depending on the intensity of post- analysis time over the following 18  months, vaccination acaricide use. The authors warned before the epidemiology and socio-economics that ‘the input values and hence results pre- group at ILRAD obtained core funds to establish sented in this paper are dependent upon sparse its own GIS capacity, led by Russ Kruska. Research and inadequate data and are largely illustra- in the late 1980s and early 1990s centred on tive of the methodology and data needs. Never- vector-borne disease distribution and impact theless, they provide an estimated magnitude studies; the group also contributed to the estab- of the economic losses attributed to theilerio- lishment in 1992 of the UNEP/CGIAR partner- sis, and the economics of its control by the ship on the development of digital datasets for infection and treatment method in the infected research on a wide range of topics including region’. natural resources, ecology, environment and socio- A series of further economic studies of economic factors. ECF was undertaken by PhD student Hezron Efforts to model the potential distribution of Nyangito, who, in a partnership with the R. appendiculatus were assisted by two key inputs. D epartment of Agricultural Economics at The first was the database on tick field samplings Texas A&M University, used a whole-farm assembled by Jane Walker. The second was a cli- simulation model to estimate the financial mate matching model, Climex, developed by and economic pay-offs from the use of ITM Robert Sutherst, with parameters for the condi- vaccination, drawing on data collected from tions favoured by the brown ear tick (among Uasin Gishu, Kenya (Nyangito et al., 1994, others). Instead of the model being run on cli- 1995, 1996). mate data for any given location, it was run for The study by Mukhebi et al. (1999) pro- the whole of Africa, in each of the 25 km2 pixel vided one of the first attempts to truly inte- cells of an interpolated climate surface for the grate epidemiology and economics models to continent. The results plotted the potential dis- predict future economic impacts of different tribution of R. appendiculatus (Lessard et al., control scenarios under a set of epidemiological 1990; Perry et al., 1990, 1991), but when scenarios. compared with the database of Jane Walker and others of where the tick had been recorded, the potential distribution exceeded the historical records, suggesting that Climex did not tell the Tick and tick-borne disease distribution whole story. This finding stimulated interest modelling in other predictive modelling approaches to estimate tick distribution (Randolph, 1993), in The need to understand the geographical scale which much closer attention was paid to the of impact of both ECF and trypanosomiasis very climatic requirements of all three instars of the quickly led into the area of modelling, primarily tick, which eventually led to a more biologically of the vectors but also to a degree of the disease sound spatial prediction platform (Randolph itself. In the absence of high-quality field data on and Rogers, 1997). R. appendiculatus distribution, the group first In a follow-up case study of predicting sought data on the key drivers of climate and outbreaks of theileriosis in Zimbabwe using mul- vegetation, and in late 1987 and early 1988, a tiple climatic variables (Duchateau et al., 1997), contract was drawn up between ILRAD and the the methodology for assessing distribution Global Resource Information Database (GRID) drivers using climate databases was addressed. group led by Harvey Croze at the United Nations The database was considered to suffer from col- Environment Programme (UNEP) in Nairobi. linearity, because most climatic variables share This group was using geographical information qualities with (or are influenced by) other vari- systems (GIS) for various African continent-wide ables in the database. Fitting logistic regression 218 B. Perry et al. models to disease occurrence with highly correl- exploration of modelling in impact assessment, ated independent variables can lead to mislead- and ILRAD, in collaboration with FAO, organ- ing conclusions if the true biological meaning is ized a modelling workshop in Nairobi in No- not clearly understood. This case study used the vember 1992 to explore the approaches being analysis of principal components to reduce large made by different research groups (Perry and numbers of variables to smaller sets of variables Hansen, 1994). that more efficiently describe the important fea- The collaboration with Imperial College tures of the database. London, and subsequently the University of An important early stage in the impact Warwick (where Graham Medley moved), led to assessment process had been to determine more the first attempt to develop a quantitative frame- accurately the distribution of diseases and their work of the infection dynamics of theileriosis vectors. This had four major impacts: (Medley et al., 1993). It was able to demonstrate 1. It enhanced the understanding of disease how infection was maintained in cattle populations vector distributions, and factors affecting these, and quantified the important role of carrier ani- such as climate and vegetation. mals. With the progressive understanding emer- 2. It enhanced the understanding of the role of ging from field studies in different regions, the GIS in predicting disease and vector distributions, model was updated and reported by O’Callaghan and the need for appropriate high-resolution geo- et al. (2003). referenced databases, including the use of satellite- The principles behind this first model were derived imagery. then applied to heartwater, and a quantitative 3. It allowed exploration of new methods for framework was produced that demonstrated improving the predictive capacity of distribu- for the first time the concept of endemic stability tion models (illustrated by Duchateau et al., (O’Callaghan et al., 1998). The approach went 1997). on to explore the effects of vaccination against 4. It alerted Ethiopia, where R. appendiculatus heartwater (O’Callaghan et al., 1999). had never been recorded, to the climatic suit- The various observational field studies and ability of the survival of this vector in certain the supportive modelling initiatives raised many parts of the country (Norval et al., 1991). This issues, particularly the mechanism for establish- stimulated the Ethiopian government to revise ment of endemic stability to both theileriosis and its policy on importation of live cattle from heartwater, and the implication of different inter- Kenya. A new study confirming the susceptibil- ventions, particularly tick control, on the devel- ity of Ethiopia to ECF has since been published opment and maintenance of endemic stability. (Leta et al., 2013), repeating the warnings made This led to an extrapolation of the findings to by ILRAD in 1991. other diseases, including malaria, warning that certain interventions might interrupt endemic stability and lead to outbreaks of disease (Cole- man et al., 2001). This Lancet publication won Modelling the infection dynamics the ILRI’s award (and nomination for the CGIAR of vector-borne diseases Chairman’s Award) for the Outstanding Scientific Article of 2002. Soon after the observational studies on ECF on the Kenya coast had started, it was felt that there was a need to develop a mathematical model (Anderson and May, 1992) of T. parva in- fection dynamics that could contribute to our Impacts of Trypanosomiasis understanding of disease impacts and offer a and its Control framework to test the effect of interventions, such as vaccination. This started a 15-year col- Economic impact of trypanosomiasis laboration on infection dynamics of tick-borne infections with Graham Medley, which eventu- The focus of impact assessment remained largely ally moved from theileriosis to E. ruminantium on tick-borne pathogens until 1997. In the early infection (heartwater). This launched a wider 1990s, the epidemiology group adapted a model Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 219 that had been used for theileriosis impact assess- The epidemiology of resistance ment to quantify the impacts of trypanosomia- to trypanocides sis and its control in Cameroon, Gambia, Côte d’Ivoire and Zimbabwe. The model estimated In the late 1990s, ILRI began research on the the proportion of the national herd at risk of the epidemiology and impact of trypanosomiasis in disease and the annual economic cost of the dis- several African countries. Field work was under- ease and produced a breakdown of the costs into taken in Uganda, Kenya, Tanzania and Zambia production losses and input costs. Work based (in collaboration with the national agricultural on this model was reported in Shaw (1992), in research system and the University of Glasgow) Mukhebi et al. (1993) and in the later review of and Burkina Faso (in collaboration with the Shaw (2009). Centre International de Recherche-Développement Subsequently an ex ante model of a poten- sur l’Elevage en zone Subhumide (CIRDES) and tial trypanosomiasis vaccine (Kristjanson et al., the Free University of Berlin (FUB). Highlights 1999) indicated that the potential benefits of this work included the following: of improved trypanosomiasis control, in terms of meat and milk productivity alone, were • Support from the Bundesministerium für US$700 million per year in Africa. The disease wirtschaftliche Zusammenarbeit (BMZ) for cost to livestock producers and consumers was several phases of an ILRI/CIRDES/FUB an estimated US$1340 million annually, with- project on the epidemiology of trypanocide out including indirect livestock benefits such as resistance in West Africa (see Chapter 3, manure and traction. Given an adoption period this volume); of 12 years, a maximum adoption rate of 30%, • Collaboration on trypanocide resistance a discount rate of 5%, and a 30% probability of in  eastern Africa (Kenya, Tanzania and the research being successful within 10 years, Zambia); the net present value of the vaccine research is • Epidemiology studies of drug resistance in estimated to be at least US$288 million, with Mukono County, Uganda. an internal rate of return of 33%, and a bene- • Support for drug resistance studies under- fit:cost ratio of 34:1. taken by the Kenya Trypanosomiasis Research While praising the estimated returns to tryp- Institute (KETRI) in collaboration with anosomiasis control, the critics were uncomfort- Glasgow University. able with these returns being attributed exclusively These studies resulted in a series of multi- to the effects of a vaccine. The predicted reduc- author and multi-institutional publications (e.g. tion of trypanosomiasis could in fact be achieved Gall et al., 2004; Knoppe et al., 2006). by several different interventions, including some for which technologies were already avail- able. It could also be an evaluation of more The development of a modelling effective deployment of tsetse traps, or of genetic- technique for evaluating control options ally engineered livestock resistance to the effects of trypanosomiasis or of effective chemother- An area of emphasis was the development of apy; the productivity impacts may be similar. models to understand factors influencing the What will differ will be the probability of suc- transmission dynamics of trypanosomiasis and cess, the cost of the research/implementation, assessing and predicting the impact of control the time to achieving that success, and the adop- strategies. The objective of this research was to tion rates. This is important, because there are determine whether transition models, as pro- those who believe that a vaccine will be out of posed by Diggle et al. (2002), could be applied. our reach for a long time to come, and while the This modelling approach offered two major ad- evaluation demonstrated probable benefits from vantages over standard methods. The first is that trypanosomiasis control, it was not specific to a risk factor associations can be assessed simul- vaccine as the way to achieve this. Notably, taneously for both new (incident) infections and 15 years on from this study, there is still no vac- recurrent infections after chemotherapy. The cine on the horizon, whereas the results were second is that such statistical methods can allow based on one being available 9 years ago. monthly rather than weekly or fortnightly 220 B. Perry et al. sampling intervals in the field. This latter feature Sustainable trypanosomiasis control in is crucial logistically and would allow the ana- the Ghibe Valley of Ethiopia lysis of data from a much wider variety of sam- pling sites, as monthly sampling is commonly The initial modelling work of ILCA’s research in employed. The use of transition models allowed the Ghibe Valley of Ethiopia moved on into an the distinction to be made between key factors environmental impact study, supported by the influencing both the incidence and persistence International Atomic Energy Agency (IAEA), of trypanosome infections in cattle in the Ghibe who were at the time exploring the potential role Valley, Ethiopia, over a 12-year period from 1986 of the sterile insect technique to eradicate tryp- to 1998 (Schukken et al., 2004). With an ob- anosomiasis. While there was general scepticism served average prevalence, based on microscopic over the widespread use of this technique, there examination, of approximately 50%, Ghibe ranked was at the time substantial political support for as an area of severe trypanosomiasis impact rela- wider tsetse eradication under the Programme tive to other tsetse-infested areas in Africa (Snow against African Trypanosomiasis (PAAT) pro- and Rawlings, 1999). The real benefit of using a gramme. This work also built on previous collab- transition model to investigate infection dynam- orative ILCA/ILRAD studies of environmental ics of trypanosomes in cattle is its ability to assess impact of long-term trypanosomiasis control in both the incidence and persistence of infections. the Ghibe Valley, including impacts on bird spe- This was particularly useful because the main cies richness (Wilson et al., 1997). factor influencing changes in incidence, namely tsetse control, differed from the factor most likely to Spatial modelling of tsetse distributions be responsible for increased duration of infection, namely resistance to commonly used trypano- Underlying several studies on trypanosomiasis cidal drugs. Age and the number of previous in- were studies on predicting the distribution of fections also influenced incidence and duration tsetse species. The GIS capacity set up in the late of infection, raising interesting hypotheses for fur- 1980s was subsequently applied to support ther investigation (e.g. potential of selection of studies on the impact of trypanosomiasis con- trypanotolerance in local Ethiopian breeds). trol (Perry et al., 1994) and later exploited by Robin Reid, who went on to explore various as- Sustainable trypanosomiasis pects of the environmental impacts of tsetse control in Uganda control (Reid et al., 2000) before moving into broader ecosystems research at ILRI. The ILRI epidemiology group began work The leaders in the use of statistical methods e xploring the historical resurgence of human and spatial climate and vegetation databases African trypanosomiasis (HAT; also known as were David Rogers and colleagues at the Univer- sleeping sickness) in Uganda (Fèvre et al., 2001; sity of Oxford (e.g. Rogers et al., 1996; Wint and Welburn et al., 2001) and went on to explore Rogers, 2000). However, with the greater en- different potential control options using modelling gagement of ILRI in predicting the effects of cli- techniques (McDermott and Coleman, 2001). mate change on the length of the growing period HAT remains an important disease in and the implications this had on livestock pro- Uganda, and cattle are its main reservoir. This duction systems, an assessment of the potential project assessed the role of cattle in human dis- for changing tsetse distributions was considered ease and how control of cattle trypanosomiasis (McDermott et al., 2001). Subsequent research can be used to reduce the public health burden has included work on the economic impact of of T. brucei rhodesiense HAT. Activities in- trypanosomiasis (Robinson et al., 2014a). cluded: (i) development of tests to differentiate human-infective and -non-infective T. brucei Preventing and containing trypanocide spp.; (ii) studies into cattle movement in new resistance in the cotton zone of West Africa outbreaks of HAT; and (iii) studies to evaluate factors that influence HAT risk, burden and In April 2012, a final report on ‘Preventing and under-reporting. containing trypanocide resistance in the cotton Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 221 zone of West Africa’ was issued, which was one required to implement tsetse control interven- of a series of projects exploring drug resistance tions and strategic treatment of cattle to suppress in trypanosomiasis control. Previous work had trypanosome populations. Actions improving focused on methods to evaluate resistance to cattle health status, such as helminth control, may trypanocides in north-east Guinea, southern also help further suppress surviving trypano- Mali (Affognon et al., 2009; Talaki et al., 2009) somes. Longer-term research is needed to confirm and south-west Burkina Faso (Der et al., 2011), the subsequent dynamics of resistant trypano- and on testing integrated control strategies to some populations if trypanosomes re-e stablish. reduce the risk of new drug resistance. In the Finally, a preliminary assessment of the final phase of the project, the project continued potential impact of the investments made to date to evaluate resistance and raise awareness and in research on trypanocide resistance indicates capacity to address the problem across much of that adequate returns will be achieved to justify the rest of the zone and to scale up the prevention the investment (Affognon et al., 2010). strategies developed earlier. Appropriate strat- Research findings and outputs from the egies were also being developed for containing – project have been taken up by continuing efforts and, if possible, reversing – resistance in the in the region to improve control of trypanosom- pockets previously characterized, and a specific iasis, notably by a network to monitor drug resist- study was undertaken to assess the impact of the ance (RESCAO: Réseau d’épidémiosurveillance trypanocide resistance research efforts to date. de la résistance aux trypanocides et aux acari- This series of projects has generated an cides en Afrique de l’Ouest), the FAO PAAT and a important body of evidence for improving the 5-year, €3.1 million project involving the princi- sustainability of trypanosomiasis control in West pal German partners and CIRDES to extend the Africa and elsewhere in sub-Saharan Africa project approach and findings to new countries (McDermott et al., 2003; Clausen et al., 2010). (Togo, Ethiopia and Mozambique). The final project focused on four main outputs, with capacity strengthening in the region as a cross-cutting objective. First, national research Rabies Research: A Networking and teams generated evidence that trypanocide re- Capacity-building Role in Africa sistance occurs in several locations across the cotton zone of West Africa, and the partnership ILRAD had no mandate in rabies research, but, has provided national services improved tools for beginning in the early 1990s, joined the South- detecting and monitoring it. Through collabor- ern and Eastern Rabies Group (SEARG). Its first ation with the Institute of Tropical Medicine contribution was an overview paper on the (Antwerp), progress was made in developing epidemiology of rabies in Africa (Perry, 1992). markers for identifying resistance in trypanosomes As part of the capacity-building function of the (Delespaux et al., 2010); this is expected to pro- epidemiology and socio-economics programme, vide even more rapid and increasingly accurate partnership with the rabies control groups in diagnostics for detecting and monitoring drug Kenya and neighbouring countries was estab- resistance. lished, including the supervision of Philip Kitala Second, informational aids, decision tools in his PhD research on rabies in the Machakos and media messages targeting farmers and ani- District of Kenya, and a series of papers emerged mal health service providers to reduce the risk of (Kitala et al., 2000, 2001, 2002). Other stra- resistance were further developed (Grace et al., tegic contributions emerged from ILRAD and 2008, 2009). There is a better understanding of ILRI (Bingham et al., 1993, 1995; Perry, 1993, how farmers access information about animal 1995; Perry and Wandeler, 1993). In addition, health care and of the most important actors in the epidemiology team was called in to evaluate national-level information networks that com- the controversy surrounding the role of rabies and municate such information. A third set of activ- rabies vaccination in the demise of the African ities demonstrated the effectiveness of integrated wild dog packs in the Aitong region of Kenya’s control strategies for containing trypanocide Maasai Mara (Macdonald et al., 1992). resistance in a location once it has established. The impacts of the engagement with rabies In such situations, the public sector is probably were substantial, and mostly centred on the 222 B. Perry et al. building of a rabies epidemiology and control ILRAD in 1995, the EU, which at the time was network through SEARG, in capacity building making large investments in rinderpest eradica- on rabies epidemiology, diagnosis and control tion through the AU-IBAR, approached both throughout the eastern and southern African institutions. The PARC programme had been region, and in highlighting priority research planning for some time to undertake an economic needs. Furthermore, ILRI’s work contributed to evaluation of the rinderpest control programme. some of the principles of dog rabies control, such However, the funds made available by the EU were as the need to understand the vaccination cover- not considered sufficient to employ an independ- age required to prevent rabies (Coleman and Dye, ent economist. As a result, the task was offered 1996), the need for a sound understanding of in 1994 to both ILRAD and ILCA, on the suppos- population ecology (in this case, dog ecology) in ition that they could supplement the limited order to target vaccination initiatives (Perry, 1993) funds with their institutional capacities in agri- and the need to exploit community engagement in cultural economics. ILRAD, with its historical rabies vaccination campaigns (Perry et al., 1995). focus on vector-borne haemoparasitic diseases, turned the offer down, while ILCA accepted it. The two institutions were then amalgamated, and ILRI was born, and the newly created Systems The Economic Impacts Analysis and Impact Assessment Group (SA/IA, of Rinderpest Control the successor to the Epidemiology and Socioeco- nomics Programme) ‘inherited’ the project. This Rinderpest has had a devastating effect on the provided a new opportunity for ILRI to work livestock industries of Africa since its introduc- with AU-IBAR, and two agricultural economists tion to the continent in the late 19th century. In were recruited under the leadership and super- its classical form, it was responsible for high vision of the SA/IA group. levels of mortality and its mere presence con- The study indicated that, for a sample of ten strained trade in livestock. During the 1960s, sub-Saharan African countries, the rinderpest the first coordinated international control pro- campaign was implemented in a cost-effective gramme was put in place, known as the Joint manner, with average per livestock unit costs Project (JP) 15. Although largely successful, rin- appearing within the narrow range of US$0.30– derpest returned in a major epidemic through- 0.66 (Tambi et al., 1999). Benefit–cost analysis out much of the continent after JP15 concluded revealed that the benefits of the campaign in in the late 1970s. As a result, the Pan-African each of the ten countries covered the value of Rinderpest Control (PARC) programme was ini- the investment. The estimated average return tiated under the auspices of the African Union– over the ten countries of US$1.98 for each US Interafrican Bureau for Animal Resources dollar invested in the campaign indicated that (AU-IBAR) funded by the European Union (EU) rinderpest control in Africa has been economic- and national governments to control and ultim- ally profitable. The net present value of US$32 ately eradicate rinderpest from Africa. A decade million indicates that the rinderpest campaign after the campaign started in 1986, increasing has been a wise public investment decision. donor concern about its impact, coupled with an The work by ILRI scientists demonstrated increasing public and private demand for infor- further that rinderpest control has also im- mation on the benefits and costs of rinderpest con- proved the well-being of livestock farmers in trol, prompted the call for an economic impact sub- Saharan Africa, as well as that of consumers assessment of the campaign. of livestock products. Analysis of the distribu- Despite Africa being the last bastion of rin- tion of welfare gains from rinderpest control be- derpest before its global eradication in 2011, tween producers and consumers revealed that ILRAD did not become involved in research into producers derived the greater share (80%) of the its control, although following its eradication, US$64 million in net value of production losses ILRI did exploit the participatory epidemiology avoided due to rinderpest control in the ten and surveillance tools used in the final phases of countries, while consumers derived approximately the campaign during its research into HPAI in 20% in net benefits from increased supplies lead- Indonesia. Just before the merger of ILCA and ing to lower prices (Roeder and Rich, 2009). Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 223 Applying Economic Impact A regional FMD coordination unit was set up, Assessment Tools to Foot-and-mouth based in Bangkok, Thailand, led by Laurie Gleason, Disease Control and an advisory committee was established, on which Brian Perry of ILRI was invited to sit. The With the formation of ILRI, the new institution first meeting was held in Bangkok on 1 March embarked on the development of an appropriate 1998. This set the scene for the first of the FMD research role in what was a new sphere of influ- economic impact studies, which focused on Thai- ence for the organization, and an Asia Action land within a South-east Asia regional context. Group was established to plan strategic engage- Thai epidemiologist Wantanee Kalpravidh was ment. For the epidemiology group, the first assigned to the study, and agricultural economics contact with the new region was a strategic support was provided by Suzan Horst of Wagen- attendance at the Federation of Asian Veterin- ingen University, the Netherlands. The World ary Associations (FAVA) Conference in Cairns, Reference Laboratory for FMD (WRL-FMD) in Australia, 24–28 August 1997, where a series Pirbright, UK, provided the FMD-specific tech- of meetings was held with those attending. The nical support. Later in 1998, after he was recruited representatives of the Australian Centre for to ILRI’s epidemiology and impact assessment International Agricultural Research (ACIAR; group, Tom Randolph took over the economic John Copland), and the OIE (Yoshihiro Ozawa) impact analysis components of the initiative. facilitated a discussion with multiple partners The first product of this work was presented on how ILRI could provide added value to on- at the annual meeting of the South-east Asia going initiatives within the Asian region. Foot-and-Mouth Disease (SEAFMD) group, under For ILRI to make an effective move into Asia the auspices of OIE, in Phnom Penh, Cambodia, in in animal health research, it was considered that February 1999, and emerged as a peer-reviewed it should exploit the generic research capacity it publication later the same year (Perry and Ran- had developed in epidemiology and economic dolph, 1999). This group provided a cost–benefit impact, rather than its traditional disease focus analysis of different FMD control scenarios and of vector-borne haemoparasites of ruminants, different emerging trading opportunities that as these were not considered to be a high priority would result from greater FMD control. in the region. In addition, it was considered that ILRI work on FMD in the Philippines had a the first phase of ILRI’s involvement in the re- significant impact. Randolph et al. (2002) devel- gion should be to better define constraints to oped scenarios, based on the plans and timetable livestock production and trade, and ways of alle- of the Government of the Philippines, and more viating these. To this end, it was suggested that optimistic and pessimistic assumptions, each dis- enhancing the regional animal disease surveil- cussed in detail with national stakeholders. It also lance and monitoring programme, the Animal had an additional component, which was an ana- and Plant Health Information System for Asia lysis of the distribution of the benefits. It illustrated (APHISA), in the field of epidemiological and that, while the FMD control programme was funded economic impact assessment would be the most entirely from public sector government coffers, appropriate entry point. ILRI was invited to in eradication scenarios the major beneficiaries undertake an initial case study on the impact of would be the private sector pig producers, traders FMD in the region, and the impact of alternative and marketers; the commercial swine sector was FMD control strategies. This provided an oppor- estimated to capture 84% of the benefits gener- tunity to enhance impact assessment capacity in ated by the public investment in eradication, ver- the region, initiate longer-term disease control sus 4% by backyard swine producers. priority evaluations and provide immediate sup- ILRI later undertook a collaborative study port to the newly created OIE-coordinated FMD on FMD on smallholder agricultural enter- control and eradication programme that had prises in southern Laos (Perry et al., 2002a). been set up. The OIE operation was funded by This demonstrated the widespread impacts the the Australian, Swiss and Japanese governments, disease had on multiple species and enterprises with the Swiss Government offering to provide in the smallholder systems of southern Laos and the ILRI portion of the funding. has been often cited as evidence of the disrup- tion to the livelihoods of smallholders globally. 224 B. Perry et al. The southern Africa FMD economic 16% of the increased value of economic activity impact study resulting from trade is eventually transferred as income to low-income households in both rural The results of a benefit–cost analysis showed and urban areas. that FMD control would benefit the economy of The direct impacts on the poor of FMD, and Zimbabwe (Perry et al., 2003). First, in a com- of measures established to control it, are very parison between the Baseline Scenario and the limited. FMD has not been a problem in commu- pessimistic FMD Control Scenario 3 (in which nal areas where the majority of the poor live, disinvestments in FMD control by 50% and re- and its effects on indigenous cattle are consider- sultant loss of beef export markets was pre- ably less than on commercially orientated herds. dicted), it was shown that for every US$1 that Furthermore, despite the fact that about 75% of Zimbabwe disinvests in the FMD control pro- poor households own or have access to cattle, gramme, a further US$5 would be lost by the over 60% of these households own fewer than country. No transboundary effects were taken five animals. Most of these households use cattle into account, and the losses calculated were in- for wealth storing and other livelihood functions curred by Zimbabwe alone. However, the associ- such as traction, and do not have the herd size ation of the outbreak of FMD in south-eastern capacity to engage actively in commercial cattle Botswana in March 2002 (after over 30 years of marketing. As such, only about 2% of house- freedom from the disease) with the outbreaks in holds are engaged in regular marketing of cattle. western Zimbabwe suggested that the costs to This study provided one of the most extensive the region as a whole of Zimbabwe’s disinvest- analyses of FMD impacts carried out, applying ments could be much greater. new methods such as the SAM/CGE modelling to Second, the results showed that if Zim- a most complex subject. Randolph et al. (2005) babwe were to invest further in the fences and explored further the highly skewed equity impacts the veterinary service infrastructures required emerging from this study. to create a much larger and much more secure export zone that was internationally recognized as FMD free by the OIE, there would be returns of Economic impacts of FMD in Peru, approximately US$1.5 for every US$1 invested. Colombia and India As in the disinvestment scenario, this does not incorporate benefits to the region as a whole In 1995, the Joint FAO/IAEA Division of the through greater disease security for FMD control, IAEA requested ILRI support for an economic nor does it include the other benefits that would assessment of FMD control. An economic impact result from an enhanced national veterinary assessment plan emerged (Romero et al., 2001). service. This analysis did not consider whether Unlike the South-east Asia partnerships, these Zimbabwe would be able to maintain the cap- Andean initiatives did not result in completed acity, in terms of quantity and quality of beef, to benefit–cost analyses; rather, the impacts of supply the export market on a sustainable basis. these studies in the region were in the field of Importantly, the distributions of the costs networking, training, capacity building, aware- and benefits turned out to be highly skewed. Ex- ness raising and methodology development. penditures from FMD control are borne almost entirely by the public sector, but when losses from trade bans resulting from FMD outbreaks are included, private sector costs are dominant. Economic impacts of FMD control in The majority of impacts of FMD and the benefits endemic settings in low- and from its control are related to the ability to trade middle-income countries internationally, and so most of the benefits ac- crue to the commercial sector, comprising cattle In April 2006, Brian Perry approached the Well- production, beef processing, and related input come Trust and the EU for support for an inter- industries and services. The Social Accounting national workshop on the research needs for Matrix/Computable General Equilibrium (SAM/ better FMD control in endemic settings of many CGE) modelling indicates that approximately low- and middle-income countries. This was Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 225 approved, and the Global Roadmap for Improv- FMD-free countries, and their trading opportun- ing the Tools to Control Foot-and-Mouth Disease ities, while Programme 2 focused on the needs in Endemic Settings was duly held in Agra, of endemic settings, principally in developing India, in late November 2006 (Perry and Sones, countries. At this point, ILRI assumed the lead- 2007b) and the Global Roadmap was launched ership of Programme 2, and initiated contact with in April 2007. potential research sponsors and development agencies. Programme 1 was assigned to the four participating research laboratories in the USA, the UK, Canada and Australia, funded through The Global Foot-and-Mouth Disease national bodies supporting each laboratory. The Research Alliance (GFRA) outcome would be a better set of tools to manage the risk to the four countries currently posed by ILRI developed a proposal for a GFRA with five FMD in endemic areas. This would focus on im- research pillars: proved vaccines and diagnostics and the further development of antivirals. 1. A detailed understanding of the host immune Program 2 was designed to focus on devel- responses to FMD virus. oping better tools for use in endemic areas with 2. Development of a new generation of inexpen- the overall aim of a gradual reduction of the dis- sive and thermostable vaccines that meet the ease in endemic areas. It was recognized that requirements of both endemic and epidemic FMD targeted research would be needed to develop control and management. specific tools for Programme 2 and that this work 3. A full understanding of the factors that per- could occur both inside and outside of the cur- mit the development of virus carrier animals, rent GFRA partner institutes. It was foreseen the risk that they pose and options for managing that, for Programme 2, much still needed to be them. done in the area of epidemiology and impact as- 4. The identification of antiviral compounds to sessment, and that ILRI would take the lead in inhibit virus replication and rapidly reduce virus this, and that it would be a core component of release. funding under Programme 2. 5. Quantitative predictions of the performance However, ILRI management was not at the of the new technologies developed in different time supportive of ILRI’s engagement with GFRA, settings through the use of epidemiological and in part because it was considered that FMD did economics models. not rank highly enough in the health constraints The leadership of each pillar was assigned to dif- facing smallholder producers. ferent institutions, with ILRI taking on this latter pillar. The US$70 million proposal was launched as a Strategic Global Research Partnership for the Control of FMD in April 2004. Rift Valley Fever The proposal was received with enthusiasm on the scientific side, but participants urged the Research into RVF at ILRI commenced with an development of a business plan. This was duly evaluation of the impacts of the 2006/2007 commissioned and in June 2005 representatives outbreak that occurred in eastern Africa. This of the partnership set out to visit key donors (DFID work was commissioned by USAID and FAO, and and the Department for Environment, Food & focused primarily on the north-eastern region of Rural Affairs (DEFRA) in the UK; the EU; the Kenya, thought to be the epicentre of the epi- Canadian International Development Agency demic in Kenya. The outbreak occurred between (CIDA) in Canada; and various partners in December 2006 and March 2007 and affected the USA). more than 700 people; approximately 150 human Considering the different contexts of FMD fatalities occurred throughout the country. It control in the developed and developing world, was believed that people suffering severe clinical GFRA revised its approach to adopt two comple- disease had close contact with infected livestock. mentary programmes. Programme 1 was targeted The impact assessment was implemented at the FMD vaccine and diagnostic needs of by a team of epidemiologists, economists and 226 B. Perry et al. social scientists. A memorandum of understand- suggest that the 2006/2007 outbreak caused a ing was established with Kenya’s Department of total of 3974 DALYs, or 1.5 DALYs per 1000 Veterinary Services (DVS), enabling the DVS to population. Provisional results further show participate in the project as a key partner. This that strategies to enhance mass vaccination of work was later extended to the Arusha region of cattle and camels over a sustained 2-year period Tanzania with additional support from FAO. would greatly reduce DALYs. It also showed that Surveys conducted in both sites utilized partici- integrating vector control measures, for instance patory epidemiological tools and the data col- through the application of larvicides, would yield lected were synthesized and published (Jost et al., even better results, although the practicability of 2010). The key observations made were that there implementing such interventions through insti- were major weaknesses in preparedness and the tutional collaboration has not been fully resolved. response to the outbreak, and that pastoralists noticed RVF-compatible events long before offi- cial notifications were made by the government. RVF risk maps for eastern Africa At the same time, economic impact assessments were conducted by Karl Rich and Francis ILRI epidemiologists have developed risk maps Wanyoike (Rich and Wanyoike, 2010). This dem- for the eastern Africa region that can be used to- onstrated that the disease induced substantial gether with the decision support tool to enhance production losses, employment losses and reduc- targeting and evaluation of RVF interventions. tions in operating capital among various value This work builds on previous studies done by the chain actors including producers, livestock traders, National Aeronautics and Space Administration animal transporters, and slaughterhouse and (NASA) and other research institutions such as butchery operators. It was estimated that the out- the Centers for Disease Control and Prevention break cost the Kenyan economy US$32 million. (CDC). Two methods that have been applied for These findings fuelled discussions on the this analysis are: (i) ecological niche modelling need for improved warning systems and a based on the Genetic Algorithm for Rule-set Pre- structured contingency plan for managing the diction (GARP); and (ii) a logistic regression model, disease. More importantly, timelines developed followed by mapping predicted probabilities on a with local communities showing events that spatial landscape. Both models use historical preceded the outbreak were transformed into a data on RVF outbreaks from the 2006/2007 decision support tool (Consultative Group for outbreak. Statistical analyses demonstrate that RVF Decision Support, 2010). This is considered RVF risk is significantly associated with excep- a major contribution by ILRI and FAO to RVF tionally high rainfall, low altitude, clay soils and contingency planning given that this tool has high normalized difference vegetation indices. now been incorporated into the Ministry of Live- Such maps could also be used to enhance our stock’s Contingency Plan for RVF. More work still understanding of ecological niches for the virus, needs to be done, however, to develop a harmon- particularly if the existing hotspots can be classi- ized contingency plan that unites the public fied based on their abilities to support disease health and veterinary sectors in line with the persistence. More importantly, these maps are One Health paradigm. being integrated with socio-economic variables to determine areas that are most vulnerable to the Economic impact assessment of disease given their livelihood patterns, capacities to control options and calculation of access public health services and literacy levels. disability-adjusted life years (DALYs) The outputs of the impact study supported the Land-use change and RVF infection formulation of a new study to assess the cost- and disease dynamics effectiveness of RVF control options from a multidisciplinary perspective. This work also With increasing awareness of the impacts of aimed to estimate economic costs of RVF in RVF epidemics, there is a growing interest in de- people using DALYs. The estimates generated termining processes that cause RVF occurrence Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 227 and transmission, as well as those that promote of a process to bring greater coordination among its persistence during inter-epidemic periods. its membership in the funding of livestock re- ILRI is currently leading a project in Kenya that search. As part of this process, it sought to better seeks to understand RVF drivers from a multidis- define the research options and priorities, and ciplinary perspective. The project is founded on DFID proposed that these be placed in the con- the premise that intact ecosystems can regulate text of poverty reduction. This required first disease epidemics, and that factors that disrupt defining poverty and the association with live- ecosystem structure and function, such as cli- stock, and then quantifying the association, a mate, land use and demographic changes, con- process that continues (Robinson et al., 2014b). tribute to disease emergence and spillovers. The There were seven major components to the project involves local partners such as the Kenya study (Perry et al., 2002b). The first was to de- Medical Research Institute (KEMRI), DVS, Uni- scribe and quantify the distribution and extent versity of Nairobi and Ministry of Health. of poverty in South-east Asia, South Asia and Preliminary observations indicate that there sub-Saharan Africa, and to determine the asso- is a great potential for endemic transmission of ciation of poverty with different agricultural RVF in irrigated areas established in arid and production systems that involve livestock. These semi-arid zones, as poorly managed drainage two tasks were accomplished in a companion systems and watersheds provide ideal conditions study commissioned by DFID (Thornton et al., for the development of primary and secondary 2002), which developed maps to quantify popu- vectors of RVF. Observations also show that lations of poor livestock keepers and to predict areas that are RVF endemic tend to be vulnerable how they would change over the next five dec- to other infectious/zoonotic diseases, malnutri- ades. The results provided data on the number of tion or insecurity, presenting multiple chal- poor (people on less than US$1 per day) in each lenges to the implementation of sustainable RVF of the major livestock production systems of the control strategies. world. These figures served as a weighting factor in determining the importance of different live- stock diseases to the poor. Epidemiology of Gastrointestinal The second component was to determine Parasites the priority species to the poor in each region and production system. This was undertaken by a literature review and through stakeholder work- ILRI published a field and laboratory handbook shops in West Africa, eastern, central and south- on the epidemiology and diagnosis of gastro- ern Africa, South Asia and South-east Asia. intestinal parasites entitled The Epidemiology, The third component was to quantify the Diagnosis and Control of Gastro-Intestinal Parasites disease constraints by species. Diseases and of Ruminants in Africa (Hansen and Perry, 1990). syndromes considered to negatively affect the A second edition, The Epidemiology, Diagnosis and livelihoods, productivity outputs and marketing of Control of Helminth Parasites of Ruminants, was livestock products by the poor were identified in published several years later (Hansen and Perry, a set of stakeholder workshops. The socio- 1994). economic (primarily production losses and con- trol costs incurred by the poor), zoonotic (for those diseases transmissible from animals to Priorities in Animal Health Research humans) and national impacts (a combination for Poverty Reduction of marketing impacts on the poor with public sector expenditures on disease control) were In 2000, ILRI began work on animal health and identified and scored. poverty reduction involving scientists and opin- Published literature on the impact of live- ion leaders in Africa, Asia, Europe and North stock diseases and of their control in the target America. This eventually delivered one of the regions was scrutinized and synthesized by com- highest-impact products of ILRI’s epidemiology missioned reviews. Research opportunities to al- group (Perry et al., 2002b). The Inter-Agency leviate these constraints were then identified. Donor Group (IADG) had just been born, as part First, research needs were identified from the end 228 B. Perry et al. users’ perspectives by participants in several was again approached by the Trust, and follow- regional workshops. Second, research oppor- ing discussions in London, the group submitted tunities were identified from the upstream per- a pre-proposal for a Wellcome Centre for Stra- spective by international experts specializing in tegic Veterinary Epidemiology based at ILRI in different diseases. In addition to identifying rele- Nairobi. After deliberations by the Trust, the vant research opportunities, the experts were proposal was not accepted for funding, but it did asked to estimate the cost, time frame, probability reopen the door to dialogue. Strongly influenced of success and available capacity to undertake by the report by Perry et al. (2002b), in July such research. To ensure that issues other than 2002 the Trust announced a new funding pro- technology generation were addressed, additional gramme entitled ‘Animal Health in the Develop- reviews of research opportunities for the better ing World’, under which it set aside £25 million delivery of animal health services were commis- over a period of 5  years to fund researchers to sioned. A specific review of the role of research develop methods of predicting and controlling into the genetics of resistance to disease was also outbreaks of animal diseases. commissioned. A large number of research proposals were The next step was to score the disease im- developed by ILRI in partnership with institutions pacts (Shaw et al., 2003), synthesize the disease in the UK and USA, including gastrointestinal impacts on the poor with the research needed parasitism, FMD epidemiology and dynamics, to reduce them and identify priority research anti-tick vaccines, livestock/disease information opportunities. A conceptual framework matrix platforms for East Africa and African swine fever, was developed to classify different types of disease- among many others. This later led to the Live- specific research: (i) transferring knowledge and stock for Life Programme, launched in December available tools; (ii) developing improved tools 2005. and strategies that were better delivered; and (iii) In January 2007, the Wellcome Trust held developing new tools and approaches by the a meeting to give scientists funded under earlier contribution the research product will make to grant programmes the opportunity to present poverty reduction (by securing the assets of the research findings, and to consider future needs poor; reducing the constraints to intensification in the field (the meeting was entitled ‘Animal or improving marketing opportunities). Health Research: Recent Developments and This study has had a lasting impact on re- Future Directions’). To coincide with the meet- search priorities for development and is still the ing, a policy review paper was commissioned most cited reference in this context. In addition, (Perry and Sones, 2007a), and ILRI presented it set the stage for measuring the association be- an invited talk on the challenges of research tween poverty and livestock, and for applying outputs influencing policy (Perry and Hooton, greater emphasis to the impacts that research in 2007). animal health have on the processes of poverty In summary, the epidemiology group at reduction, rather than simply on national agri- ILRI undoubtedly had a substantial impact on cultural development. The methodology has the shaping and development of the Wellcome been further developed (Perry and Grace, 2009). Trust’s programmes of funding for livestock dis- ease control in the developing world, and the impacts were spread to many different research The Wellcome Trust Epidemiology institutions and countries. Initiatives In January 2002, the DFID-commissioned study entitled ‘Investing in Animal Health Research to The Broader Economic Impact Alleviate Poverty’ was published (Perry et al., Contributions 2002b). As described above, this had been com- missioned by IADG on pro-poor livestock research Economic impact assessments gained increasing and development, of which the Wellcome Trust momentum as ILRI’s mandate broadened. In a was a member, and had been represented by paper prepared as an invited plenary presenta- Catherine Davies. The ILRI epidemiology group tion to the 17th International Conference of the Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 229 World Association for the Advancement of Vet- high-value livestock product markets in Europe erinary Parasitology, Perry and Randolph (1999) and elsewhere. The studies in Ethiopia showed wrote: ‘The traditional veterinarian views dis- that investments in the quarantine and testing ease as evil, and often embarks on a career with required to ensure that beef feedlots were free of a “Superman” like determination to destroy it, FMD were not the limiting factors affecting the regardless of how important it is. To the classical economic viability of beef exports to Middle East healer, economic considerations are secondary. markets; rather, it was the high cost of feeding The economist on the other hand sees animal animals to ensure that the product arriving in disease as just one, and often an insignificant the market was competitive with others coming one, of a great spectrum of constraints to human from Australia, Brazil and other sources. and societal wellbeing that needs to be put in ILRI later studied the potential role of com- context.’ They concluded that no longer should modity-based trade on international market studies of the economics of diseases of produc- access by developing countries (Rich and Perry, tion animals be limited to animal scientists seek- 2009, 2011). This work concluded that, on a ing the collaboration of agricultural economists geographical basis, the benefits of commodity- to affix prices to estimated productivity losses. based trade are much more likely to be felt in Rather, a new discipline of animal health eco- countries like Argentina, Brazil and India than in nomics emerged in which the quality of economic African countries. Opportunities exist for south- evaluations depended on integrating the prod- ern Africa but are predicated largely on continued ucts of good epidemiological studies into economic preferential access that may or may not be frameworks. sustainable in the long term. While there are In addition, during 1998, the ILRI Epidemi- numerous opportunities for some African coun- ology and Disease Control group leader was tries in niche markets, it is also important to bal- approached by the director general of the OIE ance this potential with the sound exploitation to coordinate the design, compilation and edit- of livestock resources and a pragmatic under- ing of a special edition of the OIE Scientific and standing of the challenges in marketing and Technical Review on Animal Health Economics. competitiveness. The constraints that compli- This peer-reviewed edition comprised nine chap- cate market access for Africa are much more ters on different demands for economic impact those related to infrastructure, productivity and knowledge, and seven case study chapters ad- efficiency throughout the livestock supply chain, dressing specific diseases and different circum- and it is in these areas that policy attention is stances affecting the validity of economics studies. urgently required. This book, in combination with the study by Perry and Randolph (1999) mentioned above, served as important milestones for the inte- grated science of epidemiology and economics. The Responses to Highly Pathogenic Rushton (2009) described as it as ‘the first book Avian Influenza to bring together a number of important themes in animal health economics: farm-level economic The emergence of HPAI, initially in East and assessments; trade implications of sanitary re- South-east Asia with subsequent spread to Africa, quirements; and veterinary service delivery’. caused disquiet in all animal health research ILRI later did an economic analysis of the communities and institutions, and ILRI, in col- potential costs, benefits and competitiveness of laboration with the International Food Policy trade in meat from Ethiopia to the Middle East Research Institute (IFPRI), initiated a wide- (Rich et al., 2008). This report was further devel- ranging consultation to discuss where research oped for a peer-reviewed publication, presented could contribute (ILRI/IFPRI, 2006). It was not by Karl Rich at the International Food and Agri- a straightforward process, providing the chal- business Management Association (IAMA) in lenge of developing a framework and methods June 2009 (Rich and Perry, 2009), where it won for designing appropriate control strategy the Best Paper Award. At the time, it was widely interventions, and generating evidence of the believed that poor countries with abundant live- potential trade-offs with poverty reduction stock were well placed to develop exports to objectives. 230 B. Perry et al. ILRI was active in providing several back- first for ISVEE, in a special edition of the Kenya ground guidance and methodological frame- Veterinarian. The full meeting proceedings are works for the global response to HPAI. In 2009, now also available online (ISVEE, 1994). ILRI was a partner in the production of a man- The next ISVEE was held in Paris in 1998 ual entitled Introduction to Participatory Epidemi- with a strong representation from ILRI and its ology and its Application to Highly Pathogenic partners. But from here the level of participation Avian Influenza Participatory Disease Surveillance. grew substantially, and the 9th and 10th ISVEEs A Manual for Participatory Disease Surveillance (in Colorado and Viña del Mar, Chile, respectively) Practitioners (Ameri et al., 2009). ILRI also devel- brought the research of ILRI’s epidemiology oped a user guide for initial bird flu risk maps as group to new levels of recognition; in the Chile a contribution to improving the surveillance for meeting of 2003, the group had 29 papers and bird flu (Stevens et al., 2009). These have recently posters accepted. At this meeting, under joint been built on and updated in a new risk mapping sponsorship with the International Association report (Gilbert et al., 2014). of Agricultural Economists, Tom Randolph or- A Nigerian Avian Influenza Control and ganized a mini-symposium on Animal Health Human Pandemic Preparedness and Response Economics, which comprised plenary papers, in- Project (NAICP) began in July 2006 and invited dependent papers and a discussion forum. In ILRI to do an independent impact assessment this, he concluded that animal health economics of the project. The evaluation developed ten has established a solid, although remarkably ‘outcome pillars’ to depict the benchmark ‘gold narrow, foundation in the literature, but that it standard’ of best practices against which to had not exploited its potential (Randolph et al., evaluate NAICP (as discussed in Perry et al., 2003). While ILRI continues to be represented at 2010). The evaluation report was presented subsequent meetings, the commitment to and to the Government of Nigeria (Perry et al., impact of ILRI seen during the period 1994–2006 2011) and there were also two independent peer- has waned. reviewed publications (Henning et al., 2013; Bett et al., 2014). The Role of Epidemiology in ILRAD and ILRI The ISVEE Experience Veterinary epidemiology and socio-economic The International Symposium on Veterinary impact research has gone through both admin- Epidemiology and Economics, known commonly istrative and locality changes during its existence by its acronym ISVEE, has been held every over the last 27 years. It was a unified entity for 3 years since the inaugural meeting in Reading, 15 years, but in 2002, when the new ILRI strat- UK, in 1976. It brings together directors of vet- egy was developed, epidemiology and impact erinary services, disease control planners, quan- assessment became both fragmented and dimin- titative epidemiologists, agricultural economists, ished in human resource capacity. modellers and statisticians to present and dis- From 1987 to 1994 under ILRAD, it was the cuss on a wide range of diseases and issues. Epidemiology and Socioeconomics Programme, Kenya was proposed by ILRAD’s epidemiology also varyingly referred to as the Epidemiology group and later confirmed as the venue for 1994 and Socioeconomics Unit and Socioeconomics at the 6th ISVEE in Ottawa in 1991, with Brian Programme in emerging documentation. In the Perry as secretary of the organization (ISVEE, early days of ILRI from 1995 to 1997, epidemi- 1991). This provided the first opportunity to ology was accommodated under the newly cre- bring ISVEE to Africa, and to engage national ated Systems Analysis and Impact Assessment and regional programmes in presenting their Group, placing it under the Production Systems work and participating in the meeting. Through Programme, but this did not last for long, and the hard work and commitment of John Row- from 1997 to 2002 it became the Epidemi- lands, a statistician at ILCA, the full proceedings ology and Disease Control group under the Ani- were handed to participants as they registered, a mal Health Programme. Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 231 The Impacts of ILRAD and ILRI’s Africa, to a greater understanding of the eco- Epidemiology nomic impact of rinderpest in Africa and of FMD in Africa, Asia and Latin America, and to Capacity development in veterinary regional understanding of the drivers of rabies epidemiology and impact control. More recently, epidemiology research assessment at ILRI has contributed substantially to our understanding of food safety risks in formal and informal markets, and to the dynamics During the 15-year period from 1987 to 2002, and risks of zoonotic diseases. The research has a substantial number of MSc and PhD stu- also contributed to the global understanding of dents were trained through incorporation into the importance of these and other diseases research activities; these were predominantly to African livestock systems, and to the par- from African countries. In addition, as men- ticular animal health constraints facing the tioned in the section on ISVEE, the group poorer sectors of Africa’s livestock-engaged and its associated students presented at many communities. international meetings and in most cases pub- lished their research findings in peer-reviewed journals. Impacts on ILRI’s research and strategy Impacts on national animal health departments and services During the days of ILRAD, the epidemiology and socio-economics programme had little or no im- The epidemiology group provided a role model pact on ILRAD’s research and strategy; rather, it of investigative problem solving, which was was seen as providing evidence justifying the ex- picked up, copied and adopted by other institu- istence of the laboratory-based vaccine research tions. However, this mostly occurred where for the two target haemoparasitic diseases. there was a specific donor-funded project to sup- Nevertheless, after ILRI’s birth in 1995, the port the establishment of an epidemiology group group did play an important role in providing and was more common in academic than in impact assessment services, which progressively public service bodies such as veterinary depart- enhanced the engagement of the institution ments. Many newly trained graduates return to with different national, regional and donor cli- their institutions with a sound training but do ents. This situation changed dramatically in not have the opportunity to build on that, often 2002 following the publication of the DFID- because of institutional weaknesses, with inad- commissioned study on animal research prior- equate financial resources for research and for ities for poverty reduction (Perry et al., 2002b). staff development. Veterinary epidemiologists rely The matrix of three ‘pathways out of poverty’ on collaboration with colleagues at the bench, in (see Perry et al., 2002b, Table ES1: securing the field and in the planning arena, and par- assets, reducing constraints to intensifica- ticularly with agricultural economists and other tion and improving market opportunities) social scientists, so may find substantial diffi- and three research and development opportun- culty functioning in a ‘conservative’ public sector ities (transferring knowledge and available environment. tools; improved tools, better strategies better delivered; and new tools and approaches) pro- vided the framework of the new institutional Impacts on animal health constraints in thematic structure, not just for animal health developing countries research but also for ILRI’s entire programme. Ironically, while a key product of the epidemi- ILRI’s epidemiology research has made substan- ology and disease control group provided the tial contributions to the understanding and con- framework for ILRI’s new strategy, by the trol of ECF and trypanosomiasis in Africa, to the same token it also triggered the decline of epi- preparedness and responses to RVF in eastern demiology as an institutional entity in ILRI. 232 B. Perry et al. Acknowledgements contributions of ILRI’s products in the fields of epidemiology and economics. This chapter is an We acknowledge the support of John McIntire updated version of Perry, B.D. (2015) Towards a and Delia Grace for commissioning this review. Healthier Planet: Veterinary Epidemiology Research We thank Bill Thorpe for his insightful com- at the International Laboratory for Research on Animal ments on an earlier draft. We also thank Carolyn Diseases (ILRAD) and the International Livestock Benigno, Martyn Jeggo and Roger Morris for pro- Research Institute (ILRI), 1987–2014. ILRI viding independent impact statements on the Research Report 38. ILRI, Nairobi. Notes 1 It had been conceived at the University of Florida by former ILRAD tick ecologist Andy Norval. On the tragic death of Norval in April 1994, ILRI assumed project leadership. 2 Currently, the OIE Sub-Regional Representation for South-East Asia (SRR-SEA) is engaged in FMD control in the region. The SRR-SEA evolved from the South-east Asia Foot and Mouth Disease Regional Coordination Unit (SEAFMD RCU), which was created in 1997 for the control of FMD in South-east Asia, coordinating various pre- vention and control initiatives in countries of the region, in particular Cambodia, Lao People’s Democratic Republic, Malaysia, Myanmar, the Philippines, Thailand and Vietnam. In 2010, the OIE and Association of Southeast Asian Nations (ASEAN) supported the membership of the remaining ASEAN countries (Brunei Darussalam and Singa- pore) and China, which has resulted in a vastly expanded programme, now renamed the South-East Asia and China Foot and Mouth Disease campaign (SEACFMD). References Affognon, H., Coulibaly, M., Diall, O., Grace, D., Randolph, T. and Waibel, H. (2009) Étude des politiques relatives aux stratégies de gestion de la chimiorésistance dans le cadre de la lutte contre la trypano- somose en Afrique de l’Ouest: cas du Mali. ILRI Research Report 17. ILRI, Nairobi. Affognon, H.D., Randolph, T.F. and Waibel, H. (2010) Economic analysis of animal disease control inputs at farm level: the case of trypanocide use in villages under risk of drug resistance in West Africa Live- stock. Research for Rural Development 22, 224. Ameri, A.A., Hendrickx, S., Jones, B., Mariner, J., Mehta, P. and Pissang, C. (2009) Introduction to Partici- patory Epidemiology and its Application to Highly Pathogenic Avian Influenza Participatory Disease Surveillance. A Manual for Participatory Disease Surveillance Practitioners. ILRI, Nairobi. Available at: https://cgspace.cgiar.org/bitstream/handle/10568/367/BirdFlu-Manual_final.pdf?sequence=2 (ac- cessed 13 February 2020). Anderson, R.M. and May, R.M. (1992) Infectious Diseases of Humans: Dynamics and Control. Oxford University Press, New York. Bett, B., Henning, J., Abdu, P., Okike, I., Poole, J., et al. (2014) Transmission rate and reproductive number of the H5N1 highly pathogenic avian influenza virus during the December 2005–July 2008 epidemic in Nigeria. Transboundary and Emerging Diseases 61, 60–68. Bingham, J., Perry, B.D., King, A.A., Schumacher, C.L., Aubert, A., et al. (1993) Oral vaccination of jackals against rabies: recent progress in Zimbabwe. Onderstepoort Journal of Veterinary Research 60, 1156–1157. Bingham, J., Kappeler, A., Hill, F., King, A., Perry, B.D. and Foggin, C. (1995) Efficacy of SAD (Berne) rabies vaccine given by the oral route in two species of jackal (Canis mesomelas and Canis adustus). Journal of Wildlife Diseases 31, 416–419. Chamboko, T., Mukhebi, A.W., Callaghan, C.J.O., Peter, T.F., Kruska, R.L., et al. (1999) The control of heart- water on large scale commercial and smallholder farms in Zimbabwe. Preventive Veterinary Medicine 39, 191–210. Clausen, P.H., Bauer, B., Zessin, K.H., Diall, O., Bocoum, Z., et al. (2010) Preventing and containing trypanocide resistance in the cotton zone of West Africa. Transboundary and Emerging Diseases 57, 28–32. Consultative Group for RVF Decision Support (2010) Decision-support tool for prevention and control of Rift Valley fever epizootics in the Greater Horn of Africa. American Journal of Tropical Medicine and Hy- giene 83, 75–85. Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 233 Coleman, P.G. and Dye, C. (1996) Immunization coverage required to prevent outbreaks of dog rabies. Vac- cine 14, 185–186. Coleman, P.G., Perry, B.D. and Woolhouse, M.E.G. (2001) Endemic stability: a veterinary idea applied to human public health. Lancet 357, 1284–1286. Curry, J.J., Huss-Ashmore, R.A., Perry, B.D. and Mukhebi, A.W. (1996) A framework for the analysis of gen- der, intra-household dynamics and livestock disease control, with examples from Uasin Gishu District, Kenya. Journal of Human Ecology 24, 161–189. Deem, S.L., Perry, B.D., Katende, J.M., McDermott, J.J., Mahan, S.M., et al. (1993) Variations in prevalence rates of tick-borne diseases in Zebu cattle by agro-ecological zone: implications for East Coast fever immunisation. Preventive Veterinary Medicine 16, 171–187. Delespaux, V., Vitouley, H.S., Marcotty, T., Speybroeck, N., Berkvens, D., et al. (2010) Chemosensitization of Trypanosoma congolense strains resistant to isometamidium chloride by tetracyclines and enroflox- acin. PLoS Neglected Tropical Diseases 4, e828. Der, D., Affognon, H. and Sidibe, I. (2011) Réseaux sociaux et échange d’informations vétérinaires: cas de la trypanosomose animale chez les éleveurs de Solenzo dans la province des Banwa, Burkina Faso. Sciences and Medicines in Africa 3, 402–411. Di Giulio, G., Lynen, G., Medley, G.F., O’Callaghan, C.J., Kiara, H.K. and McDermott, J.J. (2003) Character- ization of theileriosis in pastoral production systems in Northern Tanzania. In: ISVEE 10: Proceedings of the 10th International Symposium for Veterinary Epidemiology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Colorado, p. 482. Diaz, C.M., Massawe, S.C., Clemence, A., Gitau, G.K., Kiara, H.K., et al. (2003) Risk mapping of East Coast fever in coastal and highland regions of Kenya based on predicted mortality and morbidity in- cidences. In: ISVEE 10: Proceedings of the 10th International Symposium for Veterinary Epidemiology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Colorado, p. 426. Diggle,P.J., Heagerty, P.J., Liang, K.-Y. and Zeger, S. (2002) Analysis of Longitudinal Data. Oxford Univer- sity Press, Oxford. Dolan, T.T. (1989) Theileriosis in Eastern, Central and Southern Africa: Proceedings of a Workshop on East Coast Fever Immunization Held in Lilongwe, Malawi, 20–22 September 1988. ILRI, Nairobi. Duchateau, L., Kruska, R. and Perry, B.D. (1997) Reducing a spatial database to its effective dimension- ality for logistic-regression analysis of incidence of livestock disease. Preventive Veterinary Medicine 32, 207–218. Fèvre, E.M., Coleman, P.G., Odiit, M., Magona, J.W., Welburn, S.C. and Woolhouse, M.E. (2001) The origins of a new Trypanosoma brucei rhodesiense sleeping sickness outbreak in eastern Uganda. Lancet 358, 625–628. Gachohi, J., Skilton, R., Hansen, F., Ngumi, P. and Kitala, P. (2012) Epidemiology of East Coast fever (Theileria parva infection) in Kenya: past, present and the future. Parasites and Vectors 5, 194–194. Gall, Y., Woitag, T., Bauer, B., Sidibé, I., McDermott, J., et al. (2004) Trypanocidal failure suggested by PCR results in cattle field samples. Acta Tropica 92, 85–95. Gilbert, M., Golding, N., Zhou, H., Wint, G.W., Robinson, T.P., et al. (2014) Predicting the risk of avian influenza A H7N9 infection in live-poultry markets across Asia. Nature Communications 5, 4116. Gitau, G.K., Perry, B.D., Katende, J.M., McDermott, J.J., Morzaria, S.P. and Young, A.S. (1997) The preva- lence of tick-borne infections in smallholder dairy farms in Muranga District, Kenya: a cross-sectional study. Preventive Veterinary Medicine 30, 95–107. Gitau, G.K. (1998) A study of factors influencing endemic stability and instability to Theileriosis and Babesiosis on dairy production in Murang’a District, Kenya. PhD thesis, University of Nairobi, Kenya. Gitau, G.K., Perry, B.D. and McDermott, J.J. (1999) The incidence, calf mortality and calf morbidity due to Theileria parva infections in smallholder dairy farms in Murangéa District, Kenya. Preventive Veterin- ary Medicine 39, 65–79. Gitau, G.K., McDermott, J.J., Katende, J.M., Brown, R.N. and Perry, B.D. (2000) Differences in the epidemi- ology of theileriosis on smallholder dairy farms in contrasting agro-ecological and grazing strata of highland Kenya. Epidemiology and Infection 124, 3–8. Gitau, G., McDermott, J.J., McDermott, B. and Perry, B.D. (2001) The impact of Theileria parva infections and other factors on calf mean daily weight gains in smallholder dairy farms in Murang’a District, Kenya. Preventive Veterinary Medicine 51, 149–160. Grace, D., Randolph, T., Diall, O. and Clausen, P.-H. (2008) Training farmers in rational drug-use improves their management of cattle trypanosomosis: a cluster-randomised trial in south Mali. Preventive Vet- erinary Medicine 83, 83–97. 234 B. Perry et al. Grace, D., Randolph, T., Affognon, H., Dramane, D., Diall, O. and Clausen, P.-H. (2009) Characterisation and validation of farmers’ knowledge and practice of cattle trypanosomosis management in the cotton zone of West Africa. Acta Tropica 111, 137–143. Hansen, J. and Perry, B. (1990) The Epidemiology, Diagnosis and Control of Gastro-Intestinal Parasites of Ruminants in Africa. ILRAD, Nairobi. Hansen, J. and Perry, B. (1994) The Epidemiology, Diagnosis and Control of Helminth Parasites of Rumin- ants. A Handbook. ILRAD, Nairobi. Henning, J., Bett, B., Okike, I., Abdu, P. and Perry, B. (2013) Incidence of highly pathogenic avian influenza H5N1 in Nigeria, 2005–2008. Transboundary and Emerging Diseases 60, 222–230. ILRI/IFPRI (2006) How research can support efforts to control avian flu in developing countries: first steps towards a research action plan. ILRI/-IFPRI Working Group on Avian Influenza. IFPRI, Washington, DC. ISVEE (1991) ISVEE 6: Proceedings of the 6th International Symposium on Veterinary Epidemiology and Economics, 12–16 August, Ottawa, Canada. ISVEE, Guelph, Canada. Available at: www.sciquest. org.nz/elibrary/edition/4891%20 (accessed 13 February 2020). ISVEE (1994) ISVEE 7: Proceedings of the 7th International Symposium on Veterinary Epidemiology and Economics, 15–19 August, Nairobi, Kenya. Published as The Kenya Veterinarian, Vol. 18(2). Available at: www.sciquest.org.nz/elibrary/edition/4896%20 (accessed 13 February 2020). Jaetzold, R. and Schmidt, H. (1983) Farm Management Handbook of Kenya, Vol. II, Part C: Natural Conditions and Farm Management Information. Ministry of Agriculture, Kenya. Jost, C.C., Nzietchueng, S., Kihu, S., Bett, B., Njogu, G., et al. (2010) Epidemiological assessment of the Rift Valley fever outbreak in Kenya and Tanzania in 2006 and 2007. American Journal of Tropical Medicine and Hygiene 83, 65–72. Karimi, S.K., McDermott, J.J., McDermott, B., Gitau, G.K., Gathuma, J.M. and Kinuthia, R.N. (2003). Risk factors for sero-prevalence of tick-borne diseases of calves in Maasai pastoral herds in Kajiado Dis- trict, Kenya. In: ISVEE 10: Proceedings of the 10th Symposium of the International Society for Veterin- ary Epidemiology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Color- ado, p. 507. Kariuki, D.P. (1988) Theileriosis in Kenya. In: Dolan, T.T. (ed) Theileriosis in Eastern, Central and Southern Africa. Proceedings of a Workshop on East Coast Fever Immunization Held in Lilongwe, Malawi, 20–22 September 1988. ILRI, Nairobi. Kiara, H., O’Callaghan, C.J., Randolph, T.F., McDermott, J.J. and Perry, B.D. (2000) Targeting East Coast fever control strategies based on the assessment of biological risk. In: ISVEE 9: Proceedings of the 9th Symposium of the International Society for Veterinary Epidemiology and Economics, 6–11 August, Breckinridge, Colorado. ISVEE, Fort Collins, Colorado, p. 292. Kiara, H.K., McDermott, J.J., O’Callaghan, C.J., Randolph, T.F. and Perry, B.D. (2003) Estimating bio- logical parameters for modelling the transmission dynamics of Theileria parva infection. In ISVEE 10: Proceedings of the 10th International Symposium for Veterinary Epidemiology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Colorado, p. 471. Kitala, P.M., McDermott, J.J., Kyule, M.N. and Gathuma, J.M. (2000) Community-based active surveil- lance for rabies in Machakos District, Kenya. Preventive Veterinary Medicine 44, 73–85. Kitala, P., McDermott, J.J., Kyule, M., Gathuma, J., Perry, B.D. and Wandeler, A. (2001) Dog ecology and demography information to support the planning of rabies control in Machakos District, Kenya. Acta Tropica 78, 217–230. Kitala, P., McDermott, J., Coleman, P. and Dye, C. (2002) Comparison of vaccination strategies for the con- trol of dog rabies in Machakos District, Kenya. Epidemiology and Infection 129, 215–222. Knoppe, T., Bauer, B., McDermott, J.J., Peregrine, A.S., Mehlitz, D. and Clausen, P.H. (2006) Isometamid- ium sensitivity of Trypanosoma congolense stocks from cattle in West Africa tested in mice and the drug incubation infectivity test. Acta Tropica 97, 108–116. Kristjanson, P.M., Swallow, B.M., Rowlands, G.J., Kruska, R.L. and de Leeuw, P.N. (1999) Measuring the costs of African animal trypanosomiasis, the potential benefits of control and returns to research. Agricultural Systems 59, 79–98. Leneman, J.M., McDermott, J.J., Okuthe, O.S. and Randolph, T.F. (2000) Farmers’ perceptions of East Coast fever risk and adoption of control strategies in Kenya. In: ISVEE 9: Proceedings of the 9th Symposium of the International Society for Veterinary Epidemiology and Economics, 6–11 August, Breckinridge, Color- ado. ISVEE, Fort Collins, Colorado, p. 308. Lessard, P., l’Eplattenier, R., Norval, R., Kundert, K., Dolan, T., et al. (1990) Geographical information sys- tems for studying the epidemiology of cattle diseases caused by Theileria parva. Veterinary Record 126, 255–262. Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 235 Leta, S., de Clercq, E. and Madder, M. (2013) High-resolution predictive mapping for Rhipicephalus appendic- ulatus (Acari: Ixodidae) in the Horn of Africa. Experimental and Applied Acarology 60, 531–542. Macdonald, D., Artois, M., Aubert, M., Bishop, D.H.L., Ginsberg, J.R., et al. (1992) Cause of wild dog deaths. Nature 360, 633–634. Maloo, S.H. (1993) Vector-borne diseases and preventive medicine practises in smallholder dairy cattle in coastal Kenya. PhD thesis Veterinary School, University of Glasgow, Glasgow, UK. Maloo, S., Ngumi, P., Mbogo, S., Williamson, S., Thorpe, W., et al. (2001a) Identification of a target population for immunisation against East Coast fever in coastal Kenya. Preventive Veterinary Medicine 52, 31–41. Maloo, S., Rowlands, G., Thorpe, W., Gettinby, G. and Perry, B.D. (2001b) A longitudinal study of disease incidence and case-fatality risks on small-holder dairy farms in coastal Kenya. Preventive Veterinary Medicine 52, 17–29. Maloo, S., Thorpe, W., Kioo, G., Ngumi, P., Rowlands, G. and Perry, B.D. (2001c) Seroprevalences of vector-transmitted infections of small-holder dairy cattle in coastal Kenya. Preventive Veterinary Medicine 52, 1–16. Martinez, T., Meltzer, M., Perry, B.D., Burridge, M. and Mahan, S. (1999a) The effect of subclinical experi- mental Cowdria ruminantium infection in ewes on the growth and milk consumption of pre-weaning lambs. Preventive Veterinary Medicine 41, 105–118. Martinez, T., Meltzer, M., Perry, B.D., Burridge, M. and Mahan, S. (1999b) The effect of subclinical experi- mental Cowdria ruminantium infection on the health and reproductive performance of breeding ewes. Preventive Veterinary Medicine 41, 89–103. McDermott, J. and Coleman, P.G. (2001) Comparing apples and oranges – model-based assessment of different tsetse-transmitted trypanosomosis control strategies. International Journal for Parasitology 31, 603–609. McDermott, J.J., Kristjanson, P.M., Kruska, R.L., Reid, R.S., Robinson, T.P., et al. (2001) Effects of climate, human population and socio-economic changes on tsetse-transmitted trypanosomiasis to 2050. In: Seed, R. and Black, S. (eds) World Class Parasites, Vol. 1. The African Trypanosomes. Kluwer, Boston. McDermott, J., Woitag, T., Sidibé, I., Bauer, B., Diarra, B., et al. (2003) Field studies of drug-resistant cattle trypanosomes in Kenedougou Province, Burkina Faso. Acta Tropica 86, 93–103. Medley, G.F., Perry, B.D. and Young, A.S. (1993) Preliminary analysis of the transmission dynamics of Theileria parva in eastern Africa. Parasitology 106, 251–264. Minjauw, B., Perry, B.D., Kruska, R., Peter, T., Mahan, S., et al. (1998) Economic impact assessment of heartwater in southern Africa. In: Economic Impact Assessment of Heartwater in Southern Africa. Ninth International Conference of Association of Institutions of Tropical Veterinary Medicine (AITVM), 14–18 September, Harare, Zimbabwe. Minjauw, B., Kruska, R., Odero, A., Randolph, T.F., McDermott, J.J., et al. (2000) Economic impact assess- ment of Cowdria ruminantium infection and its control in southern Africa. In: ISVEE 9: Proceedings of the 9th Symposium of the International Society for Veterinary Epidemiology and Economics, 6–11 August, Breckinridge, Colorado. ISVEE, Fort Collins, Colorado, p. 645. Mukhebi, A., Wathanga, J., Perry, B.D., Irvin, A. and Morzaria, S. (1989) Financial analysis of East Coast fever control strategies on beef production under farm conditions. Veterinary Record 125, 456–459. Mukhebi, A.W., Morzaria, S.P., Perry, B.D., Dolan, T.T. and Norval, R.A.I. (1990) Immunization against East Coast fever: production and delivery costs of the infection and treatment method. Preventive Veterinary Medicine 9, 207–219. Mukhebi, A.W., Curry, J.J., Perry, B., Mining, S.K., Delehanty, J.M., et al. (1992a) Targeting smallholder livestock against East Coast fever in Uasin Gishu District, Kenya. Poster presented to the Inter- national Seminar on Livestock Services for Smallholders, Yogyakarta, Indonesia. Mukhebi, A., Perry, B.D. and Kruska, R. (1992b) Estimated economics of theileriosis control in Africa. Preventive Veterinary Medicine 12, 73–85. Mukhebi, A.W., Curry, J., Perry, B.D., Reid, R., Ellis, J., et al. (1993) Comparative case studies of the economic costs of trypanosomiasis. In: ILRAD 1993 Annual Scientific Report. ILRAD, Nairobi, p. 105–106. Mukhebi, A., Chamboko, T., O’Callaghan, C., Peter, T., Kruska, R., et al. (1999) An assessment of the eco- nomic impact of heartwater (Cowdria ruminantium infection) and its control in Zimbabwe. Preventive Veterinary Medicine 39, 173–189. Ndung’u, L.W., Randolph, T.F., Coetzee, G., Krecek, R.C. and Perry, B.D. (2000) An economic assess- ment of current delivery pathways for the control of tick-borne diseases in Kenya. In: ISVEE 9: Pro- ceedings of the 9th Symposium of the International Society for Veterinary Epidemiology and Eco- nomics, 6–11 August, Breckinridge, Colorado. ISVEE, Fort Collins, Colorado, p. 333. 236 B. Perry et al. Ndung’u, L.W., Randolph, T.F., McDermott, J.J., Kiara, H.K. and Perry, B.D. (2003). Best-bet pathways for the delivery of East Coast fever vaccines to smallholder dairy systems in Kenya. In: ISVEE 10: Pro- ceedings of the 10th International Symposium for Veterinary Epidemiology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Colorado. p. 432. Norval, R.A.I., Perry, B.D., Gebreab, F. and Lessard, P. (1991) East Coast fever: a problem of the future for the horn of Africa? Preventive Veterinary Medicine 10, 163–172. Norval, R.A.I., Perry, B.D. and Hargreaves, S.K. (1992a) Tick and tick-borne disease control in Zimbabwe: What might the future hold. Zimbabwe Veterinary Journal 23, 1–15. Norval, R.A.I., Perry, B.D. and Young, A.S. (1992b) The Epidemiology of Theileriosis in Africa. Academic Press, Orlando, Florida. Norval, R.A., Perry, B.D., Meltzer, M.I., Kruska, R.L. and Booth, T.H. (1994) Factors affecting the distribu- tions of the ticks Amblyomma hebraeum and A. variegatum in Zimbabwe: implications of reduced acaricide usage. Experimental and Applied Acarology 18, 383–407. Nyangito, H.O., Richardson, J.W., Mukhebi, A.W., Mundy, D.S., Zimmel, P., et al. (1994) Whole farm eco- nomic analysis of East Coast fever immunisation strategies in Kilifi District, Kenya. Preventive Veter- inary Medicine 21, 215–235. Nyangito, H.O., Richardson, J.W., Mukhebi, A.W., Mundy, D.S., Zimmel, P. and Namken, J. (1995) Whole farm economic evaluation of East Coast fever immunization strategies on farms in Uasin Gishu Dis- trict of Kenya. Computers and Electronics in Agriculture 15, 19–33. Nyangito, H.O., Richardson, J.W., Mukhebi, A.W., Zimmel, P., Namken, J. and Berry, B.P. (1996) Whole farm simulation analysis of economic impacts of East Coast fever immunization strategies on mixed crop-livestock farms in Kenya. Agricultural Systems 51, 1–27. O’Callaghan, C.J. (1998) A study of the epidemiology of theileriosis of smallholder dairy farmers in Kiambu district, Kenya. PhD thesis, Faculty of Graduate Studies, University of Guelph, Guelph, Canada O’Callaghan, C., Medley, G., Peter, T. and Perry, B. (1998) Investigating the epidemiology of heartwater (Cow- dria ruminantium infection) by means of a transmission dynamics model. Parasitology 117, 49–61. O’Callaghan, C., Medley, G., Peter, T., Mahan, S. and Perry, B.D. (1999) Predicting the effect of vaccin- ation on the transmission dynamics of heartwater (Cowdria ruminantium infection). Preventive Veter- inary Medicine 42, 17–38. O’Callaghan, C., Medley, G., Kiara, H.K., McDermott, J.J., Musoke, A., et al. (2003) Quantitative analysis of dose–response effects in the transmission dynamics of Theileria parva and the impact of sporozoite- neutralising vaccines. In: ISVEE 10: Proceedings of the 10th International Symposium for Veterinary Epidemi- ology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Colorado, p. 872. Ochanda, H., Young, A.S., Medley, G.F. and Perry, B.D. (1998) Vector competence of 7 rhipicephalid tick stocks in transmitting 2 Theileria parva parasite stocks from Kenya and Zimbabwe. Parasitology 116, 539–545. Perry, B.D. (1992) The epidemiology of dog rabies and its control in eastern and southern Africa. In: King, A. (ed.) Proceedings of an International Meeting on Rabies in Eastern and Southern Africa, June, Lusaka, Zambia. Editions Fondation Marcel Merieux, Lyon, France, pp. 107–121. Perry, B.D. (1993) Dog ecology in eastern and southern Africa: implications for rabies control. Onderste- poort Journal of Veterinary Research 60, 429–436. Perry, B.D. (1995) Rabies control in the developing world: can further research help? Veterinary Record 137, 521–522. Perry, B. and Grace, D. (2009) The impacts of livestock diseases and their control on growth and development processes that are pro-poor. Philosophical Transactions of the Royal Society B: Biological Sciences 364, 2643–2655. Perry, B.D. and Hansen, J.W. (eds) (1994) Modelling Vector-borne and Other Parasitic Diseases. Proceed- ings of a Workshop organised by ILRAD in collaboration with FAO, 23–27 November 1992, Nairobi, Kenya. ILRAD, Nairobi, p. 369. Perry, B.D. and Hooton, N. (2007) Translating research outputs into policy and intervention outcomes: the art of knowing and influencing one’s ultimate clients. Animal health research: recent developments and future directions. Wellcome Trust Scientific Meeting, Wellcome Trust, London. Perry, B.D. and Randolph, T.F. (1999) Improving the assessment of the economic impact of parasitic diseases and of their control in production animals. Veterinary Parasitology 84, 145–168. Perry, B. and Sones, K. (2007a) Poverty reduction through animal health. Science 315, 333–334. Perry, B.D. and Sones, K.R. (2007b) Global roadmap for improving the tools to control foot-and-mouth dis- ease in endemic settings. Report of a workshop held at Agra, India, 29 November–1 December 2006, and subsequent roadmap outputs. ILRI, Nairobi. Veterinary Epidemiology at ILRAD and ILRI, 1987–2018 237 Perry, B.D. and Wandeler, A.I. (1993) The delivery of oral rabies vaccines to dogs: an African perspective. Onderstepoort Journal of Veterinary Research 60, 451–457. Perry, B.D. and Young, A.S. (1993) The naming game: the changing fortunes of East Coast fever and Theileria parva. Veterinary Record 133, 613–616. Perry, B.D. and Young, A.S. (1995) The past and future roles of epidemiology and economics in the control of tick-borne diseases of livestock in Africa: the case of theileriosis. Preventive Veterinary Medicine 25, 107–120. Perry, B.D., Lessard, P., Norval, R.A.I., Kundert, K. and Kruska, R. (1990) Climate, vegetation and the distribution of Rhipicephalus appendiculatus in Africa. Parasitology Today 6, 100–104. Perry, B.D., Kruska, R., Lessard, P., Norval, R. and Kundert, K. (1991) Estimating the distribution and abun- dance of Rhipicephalus appendiculatus in Africa. Preventive Veterinary Medicine 11, 261–268. Perry, B.D., Kruska, R. and Reid, R.S. (1994) The development and use of geographic information sys- tems to assist trypanosomiasis control. FAO Animal Production and Health Papers. FAO, Rome. Perry, B., Kyendo, T., Mbugua, S., Price, I. and Varma, S. (1995) Increasing rabies vaccination coverage in urban dog populations of high human population density suburbs: a case study in Nairobi, Kenya. Preventive Veterinary Medicine 22, 137–142. Perry, B.D., Chamboko, T., Mahan, S.M., Medley, G.F., Minjauw, B., et al. (1998) The economics of inte- grated tick and tick-borne disease control on commercial farms in Zimbabwe. Zimbabwe Veterinary Journal 29, 21–29. Perry, B.D., Gleeson, L., Khounsey, S., Bounma, P. and Blacksell, S. (2002a) The dynamics and impact of foot and mouth disease in smallholder farming systems in South-East Asia: a case study in Laos. Revue Scientifique et Technique 21, 663–670. Perry, B.D., Randolph, T.F., McDermott, J.J., Sones, K.R. and Thornton, P.K. (2002b) Investing in animal health research to alleviate poverty. ILRI, Nairobi. Perry, B.D., Randolph, T.F., Ashley, S., Chimedza, R., Forman, T., et al. (2003) The impact and pov- erty reduction implications of foot and mouth disease control in southern Africa with special refer- ence to Zimbabwe. ILRI, Nairobi. Perry, B.D., Camus, E., Ellis, T., Mbugua, H., Moore, R., et al. (2010) Second real time global evaluation of FAO’s work on highly pathogenic avian influenza. FAO, Rome. Perry, B.D., Randolph, T.F., Bett, B.K., Grace, D., Globig, A., et al. (2011) Independent impact assessment of the World Bank-funded Nigeria Avian Influenza Control and Human Pandemic Preparedness and Response Project (NAICP). ILRI, Nairobi. Peter, T., Perry, B.D., O’Callaghan, C., Medley, G., Mlambo, G., et al. (1999) Prevalence of Cowdria ruminantium infection in Amblyomma hebraeum ticks from heartwater-endemic areas of Zimbabwe. Epidemiology and Infection 123, 309–316. Peter, T., Perry, B.D., O’Callaghan, C., Medley, G., Shumba, W., et al. (1998) Distributions of the vectors of heartwater, Amblyomma hebraeum and Amblyomma variegatum (Acari: Ixodidae), in Zimbabwe. Experimental and Applied Acarology 22, 725–740. Randolph, S.E. (1993) Climate, satellite imagery and the seasonal abundance of the tick Rhipicepha- lus appendiculatus in southern Africa: a new perspective. Medical and Veterinary Entomology 7, 243–258. Randolph, S.E. and Rogers, D.J. (1997) A generic population model for the African tick Rhipicephalus ap- pendiculatus. Parasitology 115, 265–279. Randolph, T.F., Perry, B.D., Benigno, C.C., Santos, I.J., Agbayani, A.L., et al. (2002) The economic impact of foot and mouth disease control and eradication in the Philippines. Revue Scientifique et Technique 21, 645–659. Randolph, T.F., Saatkamp, H. and Perry, B.D. (2003) A short history of animal health economics. In: ISVEE 10: Proceedings of the 10th International Symposium for Veterinary Epidemiology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Colorado, p. 827. Randolph, T.F., Morrison, J.A. and Poulton, C. (2005) Evaluating equity impacts of animal disease con- trol: the case of foot and mouth disease in Zimbabwe. Review of Agricultural Economics 27, 465–472. Reid, R., Kruska, R., Perry, B., Ellis, J. and Wilson, C. (1995) Environmental impacts of trypanosomiasis control through land-use change: conceptual model, approach and preliminary results. In: Proceed- ings of the 22nd Meeting of the International Scientific Council for Trypanosomiasis Research and Control, 25–29 October 1993, Kampala, Uganda. OAU/STRC, Nairobi, pp. 235–243. Reid, R.S., Kruska, R.L., Deichmann, U., Thornton, P.K. and Leak, S.G.A. (2000) Human population growth and the extinction of the tsetse fly. Agriculture, Ecosystems and Environment 77, 227–236. 238 B. Perry et al. Rich, K. and Perry, B.D. (2009) The impact of changing global animal health trading standards on market access for livestock products by developing countries; assessment of commodity-based trade. Commissioned report to the Department for International Development (DFID) of the Government of the UK. Rich, K.M. and Perry, B.D. (2011) Whither commodity‐based trade? Development Policy Review 29, 331–357. Rich, K.M. and Wanyoike, F. (2010) An assessment of the regional and national socio-economic impacts of the 2007 Rift Valley fever outbreak in Kenya. American Journal of Tropical Medicine and Hygiene 83, 52–57. Rich, K.M., Perry, B.D., Kaitibie, S., Gobana, M. and Tewolde, N. (2008) Enabling livestock product ex- ports from Ethiopia: understanding the costs, sustainability and poverty reduction implications of sani- tary and phytosanitary compliance. Final report for the Texas A&M University Sanitary and Phytosani- tary Livestock and Meat Marketing Program, Texas Agricultural Experiment Station, Texas. Robinson, T.P., Cecchi, G., Wint, G., Mattioli, R. and Shaw, A. (2014a) Benefits of trypanosomosis control in the Horn of Africa. In: Sebastian, K. (ed.) Atlas of African Agriculture Research and Development: Revealing Agriculture’s place in Africa. IFPRI, Washington, DC, pp. 62–63. Robinson, T.P., Wint, W.R.G., Conchedda, G., van Boeckel, T.P., Ercoli, V., et al. (2014b) Mapping the global distribution of livestock. PLoS One 9, 1371–1371. Roeder, P. and Rich, K.M. (2009) Conquering the cattle plague: the global effort to eradicate rinderpest. In: Spielman, D.J. and Pandya-Lorch, R. (eds) Millions Fed: Proven Successes in Agricultural Develop- ment. IFPRI, Washington, DC, pp. 109–116. Rogers, D.J., Hay, S.I. and Packer, M.J. (1996) Predicting the distribution of tsetse flies in West Africa using temporal fourier processed meteorological satellite data. Annals of Tropical Medicine and Para- sitology 90, 225–241. Romero, J.R., Villamil, L.C., Perry, B.D., Randolph, T.F., Vera, V.J. and Ramírez, G.C. (2001) Valoración del impacto económico del programa de control y erradicación de la fiebre aftosa en Colombia. Revista Colombiana de Ciencias Pecuarias Supplement 14. Rushton, J. (2009) The Economics of Animal Health and Production. CAB International, Wallingford, UK. Schukken, Y., van Schaik, G., McDermott, J.J., Rowlands, G., Nagda, S., et al. (2004) Transition models to assess risk factors for new and persistent trypanosome infections in cattle – analysis of longitudinal data from the Ghibe Valley, southwest Ethiopia. Journal of Parasitology 90, 1279–1288. Shaw, A.P. (1992) Development of a methodology for assessing the losses due to trypanosomiasis in Africa. Consultancy report. ILRAD, Andover, UK. Shaw, A.P. (2009) Assessing the economics of animal trypanosomosis in Africa – history and current perspectives. Onderstepoort Journal of Veterinary Research 76, 27–32. Shaw, A.P., Randolph, T.F., Heffernan, C., Rushton, J. and Perry, B.D. (2003) A framework for prioritizing animal health constraints in developing countries according to their impact on the poor. In: ISVEE 10: Proceedings of the 10th International Symposium for Veterinary Epidemiology and Economics, 17–21 November, Vina del Mar, Chile. ISVEE, Fort Collins, Colorado, p. 431. Snow, W. and Rawlings, P. (1999) Methods for the rapid appraisal of African animal trypanosomosis in the Gambia. Preventive Veterinary Medicine 42, 67–86. Stevens, K., de Glanville, W., Costard, S., Métras, R., Theuri, W., et al. (2009) Risk mapping for HPAI H5N1 in Africa-Improving surveillance for bird flu: initial bird flu risk map reports. ILRI, Nairobi. Talaki, E., Sidibé, I., Diall, O., Affognon, H., Grace, D., et al. (2009) Variations saisonnières et facteurs de risques des trypanosomoses animales dans un contexte de chimiorésistance dans la zone de Sikas- so au Mali. Bulletin of Animal Health and Production in Africa 57, 149–160. Tambi, E., Maina, O.W., Mukhebi, A. and Randolph, T.F. (1999) Economic impact assessment of rinder- pest control in Africa. Revue Scientifique et Technique 18, 458–477. Thornton, P.K., Kruska, R.L., Henninger, N., Kristjanson, P.M., Reid, R.S., et al. (2002) Mapping poverty and livestock in the developing world. ILRI, Nairobi. Wanyangu, S.W., Kiara, H., Randolph, T.F., Leneman, J.M., Okuthe, O.S., et al. (2000) The infection and treatment method for East Coast fever immunization: assessing its impact in Kenya. In: ISVEE 9: Proceedings of the 9th Symposium of the International Society for Veterinary Epidemiology and Economics, 6–11 August, Breckinridge, Colorado. ISVEE, Fort Collins, Colorado, p. 339. Welburn, S.C., Fèvre, E.M., Coleman, P.G., Odiit, M. and Maudlin, I. (2001) Sleeping sickness: a tale of two diseases. Trends in Parasitology 17, 19–24. Wilson, C.J., Reid, R.S., Stanton, N.L. and Perry, B.D. (1997) Effects of land use and tsetse fly control on bird species richness in southwestern Ethiopia. Conservation Biology 11, 435–447. Wint, W. and Rogers, D. (2000) Consultants’ report: predicted distribution of tsetse in Africa. FAO, Rome. 6 The Management and Economics of East Coast Fever Philip Toye1, Henry Kiara1, Onesmo ole-MoiYoi2, Dolapo Enahoro3 and Karl M. Rich4 1International Livestock Research Institute, Nairobi, Kenya; 2Senior Visiting Scientist, International Centre of Insect Physiology and Ecology, Nairobi, Kenya; 3International Livestock Research Institute, Accra, Ghana; 4International Livestock Research Institute, Dakar, Senegal Contents Executive Summary 240 The problem 240 ILRI’s contribution in the global context 240 Impacts of ILRI research 241 Scientific impacts 241 Development impacts 241 Economic impacts 242 Policy impacts 242 Capacity building 242 Partnerships 242 Introduction 242 The Parasite 243 Control methods 244 Impacts of ECF 244 The Infection-and-Treatment Method (ITM) of Vaccination 245 Development of the ITM vaccine 246 ILRI’s contribution to the Muguga cocktail 246 Development and application of molecular tools 247 Monoclonal antibodies (mAbs) 248 DNA-based strain identification 249 Research to Develop a Subunit Vaccine 250 Protection of exposed animals 250 The anti-schizont vaccine 250 Lack of protection with serum 250 The central role of CTLs 250 Identification of CTL antigens 251 Antigenic diversity 251 Immunodominance 252 The anti-sporozoite vaccine 252 ITM: The Future 252 The buffalo problem 252 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 239 240 P. Toye et al. Molecular tools 253 Cold chain 253 Vaccine production 253 Performance monitoring 253 Irradiation of sporozoites: potential lessons from candidate malaria vaccine 253 The Proliferative Response in T. parva-infected Lymphocytes 254 Casein kinase 2 254 Background 254 Experimental approach 254 Current implications of CK2 overproduction in human medicine 255 The Economic Impacts of ECF Research 255 Modelling the economic impacts of ECF 256 Methods 257 ITM price 259 Grazing type 259 Returns to ECF Investment 262 Model results 263 Model gaps 267 Conclusions and the Future 267 References 267 Executive Summary It is widely accepted that vaccination is the most attractive control option, and the development The problem of a vaccine to protect cattle against ECF was one of the founding aims of the International La- East Coast fever (ECF) is a fatal bovine disease boratory for Research on Animal Diseases (IL- caused by the protozoan parasite Theileria parva. RAD). The disease occurs in 16 countries in eastern, At about the time of ILRAD’s establishment central and southern Africa where the vector, in 1973, a vaccination procedure was being de- the brown ear tick (Rhipicephalus appendiculatus), veloped at the East African Veterinary Research is found. ECF causes major economic losses by Organization (EAVRO) at Muguga, Kenya. The affecting both dairy cows and young Zebu cattle infection- and- treatment method (ITM) is an im- in pastoralist systems and ranches. It is among munization procedure against ECF. It involves the most serious constraints to cattle productiv- inoculation of live sporozoites of T. parva, usu- ity in the countries in which it is found. ally in the form of a semi-purified homogenate The costs of ECF include both direct and in- of T. parva-infected ticks, combined with simul- direct losses. Direct losses are due to cattle taneous treatment with a dose of a long-acting deaths, the stunting of calves, reduced milk pro- formulation of the antibiotic oxytetracycline. duction in survivors, and the costs of preventing Whilst safe and very effective when adminis- and controlling the disease. Indirect losses in- tered correctly, production and delivery of this clude the lack of adoption of more productive live ECF vaccine is complicated, expensive and breeds of cattle and the avoidance of areas of time consuming, and at the time of ILRAD’s high infection risks. ECF affects households by founding, there were doubts as to whether such reducing milk supplies, depleting assets and re- a procedure was commercially viable. ducing incomes, all of which harm household food and nutritional security. ECF has been controlled predominantly ILRI’s contribution in the global context through acaricide application, but this treat- ment is expensive and not always successful. An The International Livestock Research Institute alternative option is for farmers to keep local (ILRI) has played a pivotal role in overseeing ECF breeds of cattle, which tend to be more disease management in the affected regions of Africa. resistant but less productive than exotic breeds. From the beginning, ECF and trypanosomiasis The Management and Economics of East Coast Fever 241 were central to the work of ILRAD. The institute has sequencing of the t. parva genome. Sequen- led activities in various vaccine development strat- cing of the T. parva genome was carried out by egies, including commercial production of the ITM Gardner et al. (2005). This was the second api- vaccine. ILRI has also overseen the development complexan to be sequenced and was essential in and application of molecular tools to c haracterize screening for CTL antigens. A related paper (Pain the vaccine and to address concerns by veterinary et al., 2005) compared the genome of Theileria authorities about the risks of using ITM in the annulata with that of T. parva. field. ILRI has also undertaken major research efforts to develop an alternative (subunit) vaccine strain characterization. Scientists at ILRI in and has furthered our understanding of the bo- collaboration with many other researchers de- vine immune response in support of these efforts. veloped DNA-based methods to characterize Theileria spp. parasites, including restriction fragment length polymorphisms of repetitive re- Impacts of ILRI research gions of the T. parva genome, analysis of poly- morphisms in ribosomal RNA genes, in telomere Scientific impacts regions of the parasite chromosomes and in genes encoding T. parva antigens and the use of ILRI has generated important research findings microsatellites and minisatellites. in several aspects of ECF research. These are out- lined as follows. sporozoite subunit work. Scientists at ILRI demonstrated that immunity to T. parva could itm. Scientific contributions to the development, also be induced through vaccination with the production and use of ITM immunization have p67 protein, which is present on the surface of been significant; together, these contributions the infective sporozoite stage of the parasite. The have enabled the immunization of hundreds of protection is believed to be mediated primarily by thousands of cattle in both the pastoralist and antibodies. dairy sectors and have assisted the commercial production, distribution and use of the vaccine cellular proliferation. A spillover impact of in eastern Africa. ILRI’s ECF research, which turned out to be important for the medical research commu- immunology of t. parva infection. In terms of nity, was the discovery that upregulated casein scientific impacts, ILRI scientists provided convin- kinase 2 is the cause of uncontrolled cell prolif- cing evidence that a response by cytotoxic T eration in cattle suffering from theileriosis. lymphocytes (CTLs) was the main effector mech- This discovery in this ‘bovine cancer’ research anism deployed by immune cattle against T. parva advanced our understanding of the role of this infection; this work was important in the broader enzyme in the development of certain human context of vaccine development as it was one of cancers, and thus of potential targets for treat- the first demonstrations that CD8+ CTLs could me- ment regimes. diate protection against intracellular protozoan Epidemiological scientific impacts are de- parasites. ILRI scientists made the first successful scribed in Chapters 5 and 8 (this volume).    A identification of T. parva proteins recognized by major ILRAD product was the book The Epidemi- CTLs from immune cattle; the identification of ology of Theileriosis in Africa (Norval et al., 1992). these CTL antigens was a major achievement in This not only focused on epidemiology but also vaccine research, and further examination of the covered all aspects of the parasite and infection antigens and the CTLs directed against them has life cycles. yielded some very valuable insights into the im- munobiology of the host–parasite relationship. Development impacts advances in bovine immunology. As part of the In 1996 at the request of the FAO, ILRI produced work investigating the immunology of T. parva 600,000 doses of the ‘Muguga cocktail’, a ver- infection, significant advances were made in our sion of the ITM vaccine, which was subsequently understanding of the bovine immune system. These distributed commercially. In 2007, when stocks are discussed here and in Chapter 4 (this volume). of this batch were depleted, ILRI was asked by 242 P. Toye et al. AU-IBAR to produce a second batch at its own prices paid by consumers but also on food im- cost. With no other institutions in the region ports, livestock feed demand, and land use. with the facilities and expertise to oversee this task, ILRI again complied, producing 1.2 million doses of the ‘Muguga cocktail’ vaccine (‘ILRI08’ Policy impacts batch), almost all of which has been commer- ILRI scientists helped obtain official registration cially distributed. In 2008, ILRI addressed con- of the ECF ITM Muguga cocktail vaccine in cerns of smallholder dairy farmers regarding the Kenya, Malawi and Tanzania and approval for large number of doses that were included in use in Uganda pending registration. Until the each vaccine straw. To make relatively few doses vaccine was registered in each country, it had available to those keeping just a few cows, ILRI only been used with special permission given by produced vaccine straws with just five to eight the national veterinary authorities. doses. The doses in these straws proved to be as safe and effective as those in the larger-dose Capacity building straws. Use of the ECF ITM vaccine has had major development impacts. It has protected ap- ILRI supported at least 34 graduate fellows in proximately 1.6 million cattle against the disease ECF studies, including 20 PhD degrees, 13 MSc from 1997 to 2014, preventing the untimely degrees and one post-doctoral scientist. ILRI also deaths of some 400,000 animals over that period, contributed significantly to vaccine production assuming a typical annual calf mortality of 40% capacity CTTBD, Malawi. (Di Giulio et al., 2009). About 80% of the ECF ITM vaccine has been sold in the pastoral pro- Partnerships duction systems of northern Tanzania, about 10% in the pastoral systems of Kenya and the re- Development of the ECF ITM vaccine and overall mainder in the smallholder dairy systems of management of ECF in the region has led to sev- Kenya. ILRI also facilitated the transfer of pro- eral ILRI partnerships with other institutions, in- duction of the vaccine to the Centre for Ticks cluding: the Africa Union–Interafrican Bureau for and Tick-Borne Diseases (CTTBD) in Malawi. Animal Resources (AU-IBAR), Centre for Ticks and Tick-Borne Diseases (CTTBD, Malawi), Food Economic impacts and Agriculture Organization of the United Na- tions (FAO), GALVmed (UK), Kenya Agricultural The production and distribution of the ECF ITM and Livestock Research Organization (KALRO), vaccine has protected millions of cattle, prevent- United Nations Development Programme (UNDP), ing the deaths of thousands of valuable animals. Vet Agro Limited (Tanzania), VETAID (Uganda) Use of the vaccine also improved milk production and Vétérinaire Sans Frontière-Germany (VSFG). and reduced stunting in calves and the costs of preventing and controlling the disease. Earlier studies estimated that adoption of multi-compo- Introduction nent ECF vaccines in affected countries would reduce the value of calf mortality annually (by ECF is a devastating tick-borne disease of cattle US$10.1 million) and increase the value of milk caused by a protozoan parasite, Theileria parva production (by US$1.7 million), with estimated (Norval et al., 1992, pp. 64–97, on the classifica- economic returns of US$9–17 for every dollar in- tion of Theileria spp.; Coetzer et al., 1994). The vested. Other economic modelling that considers majority of the discussion in this chapter and in vaccine adoption processes over time and a wide Chapter 5 (this volume) is of T. parva. The form range of economic impacts is presented in this of theileriosis known as ECF is present in 12 chapter. In Kenya, production of beef and dairy countries in eastern, central and southern Af- in 2030 is shown to increase by up to 40% and rica where the vector, the brown ear tick (Rh- 56%, respectively, compared with baseline condi- ipicephalus appendiculatus), is found (CABI, tions of no new investments in ECF manage- 2020). ECF causes major economic losses by af- ment. Changes such as these have potential fecting both dairy cows and young Zebu cattle i mplications for incomes obtained by farmers and in pastoralist systems and ranches. A clinically The Management and Economics of East Coast Fever 243 similar disease, Corridor disease, is found in cat- hyalommid ticks, it occurs in North Africa, tle infected with T. parva transmitted by ticks southern Europe, the Near and Middle East, which have fed on buffalo. The chief distinguish- India, China and Central Asia. It causes both ing feature of Corridor disease is the low number mortality and reduced production, and has sig- of the piroplasm (blood) stage of the parasite. A nificant economic impacts as a result. milder form of ECF in Southern Africa with strong seasonal occurrence is referred to as Jan- uary disease in Zimbabwe. It is among the most The Parasite serious constraints to cattle productivity in the countries where it is found. Development of a vac- The T. parva parasite has a complex life cycle cine to protect cattle against ECF was one of the involving bovine hosts (African buffalo and founding aims of ILRAD (see Introduction chap- domestic cattle) and the tick vector (Fig. 6.1). ter, this volume). T. parva also causes corridor dis- The tick feeds on the host three times, as a ease if buffalo-adapted parasites are transmitted larva, nymph and adult. Theileria sporozoites de- to cattle, and January disease in Zimbabwe. These velop in the salivary glands of infected ticks and have similar clinical signs to ECF but last for only are passed to cattle along with tick saliva when a few days, and emaciation and diarrhoea are not the ticks feed. In cattle, these sporozoite forms of seen. Turning sickness is an aberrant infection the parasite attach themselves to the animal’s characterized by neurological signs caused by white blood cells (lymphocytes): some sporozo- parasites in cerebral blood vessels. ites enter lymphocytes and develop into multi- T. annulata causes tropical theileriosis or nucleate parasite forms called macroschizonts. Mediterranean Coast fever. Transmitted by The infected bovine lymphocytes become enlarged LYMPHOCYTE Sporozoites LYMPHOBLAST SALIVARY GLAND Macroschizont ACINUS Synchronous division of host cell and parasite Sporoblast Microschizont Kinete Zygote ERYTHROCYTES Merozoites GUT Piroplasms Fig. 6.1. Life cycle of T. parva (anon, ILRI). 244 P. Toye et al. cells (lymphoblasts) that multiply in synchrony rate of development of resistance has ranged with the parasites, resulting in a rapidly expand- from 2 years (synthetic pyrethroids) to 40 years ing population of infected bovine cells. Some of (arsenic) influenced by, among other factors, the the macroschizonts differentiate into micro- class of acaricide and the frequency of applica- schizonts and then merozoites during infection. tion. Acaricides also expose users to potential These merozoites are released into the blood- health risks, which are exacerbated by lack of stream and invade red blood cells, where they protective clothing and can also cause environ- further develop into forms called piroplasms. mental contamination. Frequent use of acari- Ticks feeding on cattle become infected when cides is expensive – in high ECF-challenge areas, they ingest red blood cells containing piro- application might be needed every 5 days. plasms. In the tick gut, the parasites differentiate A second strategy to reduce the losses asso- into male and female gametes, which fuse to ciated with ECF is to use less susceptible, but less form zygotes. These develop further and eventu- productive, indigenous cattle rather than im- ally migrate to the tick’s salivary glands. Here, proved cattle breeds, which are more susceptible stimulated by tick feeding, 30,000–50,000 sporo- to the disease (Stobbs, 1966; Ndungu et al., 2005). zoites develop in an average tick. These are intro- As a third strategy, drugs are available for the duced with tick saliva into a new bovine host, treatment of clinical ECF, but to be effective, initiating a new cycle of parasite development. these must be used at an early stage of the disease The initial clinical signs of infection include and thus require constant monitoring of cattle, pyrexia and enlargement of the superficial which presents difficulties, especially in pastoral lymph nodes, most notably the parotid and pres- systems. In addition, the cost of the drug treat- capular nodes. As the infection progresses, the ment is high (US$40 per animal), which can be animals become listless and anorexic and exhibit prohibitive, especially for less valuable Zebu severe respiratory distress. There is often a pro- animals. More recently, a vaccination protocol gressive loss of white blood cells. On post-mortem known as the infection-and-treatment method examination, infected lymphocytes are found in (ITM) has become available commercially and is several organs including the lymph nodes, lungs, increasingly being used, as discussed below. liver, kidneys, gastrointestinal tract and sometimes the brain. A particularly prominent post-mortem finding is frothy exudate in the trachea and se- Impacts of ECF vere congestion of the lungs (Irvin et al., 1983). In contrast, T. parva-infected buffalo show few if The impacts of the disease include direct losses any clinical signs of disease. due to cattle deaths, the stunting of calves, reduced It is believed that T. parva co-evolved with the milk production in survivors, and the cost of African Cape buffalo and has undergone a ‘host measures to prevent and control ECF. Indirect jump’ to cattle, where it causes the disease re- losses include the lack of adoption of more pro- ferred to as ECF (reviewed by Norval et al., 1992). ductive breeds of cattle and the avoidance of areas of high infection risk. ECF affects house- holds by reducing milk supply, depleting assets and Control methods reducing incomes, all of which harm household food and nutritional security. Moreover, for small- The dominant method for controlling ECF has holder dairy farmers with just one or two animals, been the application of acaricides to cattle to the loss of a valuable cow can be a devastating limit their tick infestations. The acaricides are blow. In one pastoralist area in northern Tanza- applied as sprays, in dips and more recently as nia, overall calf mortality was shown to be 40–80%, oil-based pour-ons. This control method has with 75% of all deaths due to ECF (Di Giulio several disadvantages. After prolonged use, the et al., 2009). Chapter 5 (this volume) reports ticks can develop resistance to the active ingredi- more detailed epidemiological and economic ent in the acaricides (Abbas et al., 2014) and few studies of the impact of ECF. if any novel acaricides are expected on the mar- It is estimated that 40 million of the 75 mil- ket in the immediate future. Ticks have developed lion cattle in the eastern, central and southern resistance to all known classes of acaricides. The Africa region are at risk of contracting ECF The Management and Economics of East Coast Fever 245 (Morzaria and Williamson, 1999; McLeod and homogenate of T. parva-infected ticks, combined Randolph, 2000; FAO, 2014, 2017). with simultaneous treatment with a dose of a In 1992, an ILRI-led study estimated that long-acting formulation of the antibiotic oxy- ECF caused annual economic losses totalling tetracycline. Without the antibiotic treatment, US$170 million, including more than 1 million the sporozoite inoculation would be lethal. cattle deaths a year (Mukhebi et al., 1992). By The oxytetracycline suppresses the infection by 2005, cattle deaths alone, at 1.1 million head, inhibiting the development of sporozoites to accounted for more than US$300 million, based schizonts in lymphocytes (Spooner, 1990). The on the unit price of cattle. ECF-related cattle outcome is an asymptomatic infection or a mild deaths in 2005 thus represented around 44% ECF episode followed by the development of a of the combined value of beef production in protective immune response. The immunity has B urundi, Kenya and Rwanda in that year (FAO/ been shown to last for at least 43 months in the AU-IBAR/ILRI, 2017). absence of challenge (Burridge et al., 1972) and Various studies have estimated the impact it is accepted that a single inoculation confers of the disease at national or local levels. In 1999, life-long immunity to the disease under field an ILRI-led study estimated that total losses due conditions. Cattle that have been immunized by to ECF in Kenya were more than US$95 million a ITM, or that have survived a natural challenge, year, with mortality accounting for almost may become ‘carriers’ of a persistent, potentially three-quarters of this amount; loss of milk pro- tick-transmissible, infection. duction due to morbidity and the cost of acari- Adoption of the ITM vaccine has been re- cides made up most of the balance. For Tanzania, ported to be associated with a range of benefits. the estimated total annual loss was US$43.8 mil- Calf mortality rates have been dramatically re- lion, most of which was due to mortality. These duced, such as in pastoralist areas of northern estimates can be of significant economic import- Tanzania where calf mortality rates dropped from ance. For example, the annual ECF-related losses 80% to as low as 2% in a study of 2178 crossbred in Kenya estimated in 1999 would have been calves (1434 immunized and 744 controls) in equivalent to 11% of the total value of output from 167 smallholder households in two districts (Ly- the livestock subsector, while the annual losses in nen et al., 2012). Pastoralists also report that cat- Tanzania translated to up to 46% of the same. tle wearing the distinctive round ECF ear tag at- tract higher prices than non-v accinated animals of equal size. In another study in Tanzania, it was The Infection-and-Treatment found that households that had used the ITM vac- Method of Vaccination cine sold twice as many animals as non-adopting households (Homewood et al., 2006). ITM is an immunization procedure against ECF. Although acaricides may still be required for The basis for the development of ITM lay in the other tick-borne diseases, vaccination by ITM re- observations, made shortly after the disease was duces the frequency of acaricide application. In characterized, that cattle that survived an episode both smallholder dairy and pastoralist areas of of ECF were unlikely to experience a second clin- northern Tanzania, acaricide application is now ical episode. Early experiments of Arnold Theiler often done once a month rather than once a showed that cattle infected ‘artificially’ with para- week. Pastoralists also report that when their ani- site preparations were subsequently resistant to mals are vaccinated, they no longer have to avoid field challenge. Proof of concept of the ITM ap- areas where they know cattle are at high risk of proach as a practical means of vaccination was ECF. A study by Lynen et al. (2012) reported that achieved during the 1970s by the group at EAV- 80% of the farmers reduced the frequency of RO led by Matt Cunningham, as has recently acaricide treatment after immunization, while been documented in detail by Perry (2016), and 38% reduced this treatment by more than 75%. much of the credit for current field application Other methods of controlling ECF are of ITM against theileriosis belongs to these sci- available. They include treating animals with entists, who long preceded ILRAD and ILRI. anti-theilerial drugs when they fall sick. The ITM involves inoculation of live T. parva therapeutic drugs are quite effective but require sporozoites, usually in the form of a semi-purified early diagnosis as the disease progresses very 246 P. Toye et al. quickly. The drugs also tend to be expensive, (DANIDA), FAO, International Fund for Agricul- costing US$40–60, and are beyond the reach of tural Development (IFAD), Netherlands govern- many smallholder farmers. This drug treatment ment, Overseas Development Administration/ also has the disadvantage that once an animal Department for International Development suffers ECF, production losses, including stunt- (ODA/DFID) and Wellcome Trust, among others. ing in calves, follow. The ITM method was built on three key In smallholder dairy systems where cattle practical developments: are kept confined rather than allowed to graze, exposure to ticks, and therefore diseases, is min- • The demonstration that animals could be imized. Where farmers can grow their own reproducibly infected with a ground-up fodder, this is quite an effective disease control suspension of whole infected ticks. strategy. Quite often, however, farmers must get • The observation that such tick prepar- feed from outside the farms and this can intro- ations could be stored for extended periods duce ticks, leading to disease outbreaks. As men- in liquid nitrogen and were infectious tioned, another approach has been to maintain when thawed. indigenous cattle breeds that are relatively toler- • The use of antibiotics, particularly long- ant to tick-borne diseases, including ECF. How- acting tetracycline, as a reliable means ever, the local breeds are less productive than to prevent the clinical effects of a dose of improved breeds, so this strategy is unsuitable the tick suspension without impairing the for intensive smallholder dairy farming. development of immunity. One challenge that remained was that of parasite strain specificity, or the frequent inabil- Development of the ITM vaccine ity of a single parasite strain when used in ITM to protect against all other strains of the parasite. Work to develop an ECF vaccine began in the Researchers at EAVRO assessed the protective cap- 1960s at EAVRO, located in Muguga, Kenya, acity of several parasite isolates and showed that under the auspices of the East African Commu- a combination of parasites from three isolates – nity. Following the collapse of the East African Muguga, Serengeti-transformed and Kiambu 5 – Community, it continued at the National Veter- offered the most complete protection to heterol- inary Research Centre, KARI (now the Veterin- ogous challenge (Radley et al., 1975). The ary Research Centre, KALRO). This culminated combination was named the ‘Muguga cocktail’. in the introduction of the ITM Muguga cocktail Other forms of ITM have been produced and tested vaccine on a commercial basis in various east- over the years using different T. parva isolates. The ern African countries for various classes of cat- veterinary authorities in Kenya, Zambia and Zim- tle between 1998 (Tanzania pastoral sector) and babwe favoured the locally derived, single iso- 2012 (Kenya dairy sector). lates Marikebuni, Boleni and Katete/Chitongo, For close to 50 years, many partners have respectively, while the rest of the eastern African been involved in the testing, refinement, re-testing, region chose the Muguga cocktail. Kenya later production, registration and commercialization also adopted the Muguga cocktail. of the vaccine. Key among this team has been (in alphabetical order): the Africa Union–Intera- frican Bureau for Animal Resources (AU-IBAR), ILRI’s contribution to the Muguga cocktail CTTBD (Malawi), Food and Agriculture Organ- ization of the United Nations (FAO), Global Alli- Production of the ITM vaccine is a long and ance for Livestock and Veterinary Medicines complicated process, involving artificial infec- (GALVmed, UK), KARI, UNDP, Vet Agro Limited tion of ‘production’ cattle, application of several (Tanzania), VETAID (Uganda) and VSFG. Finan- hundred thousand ticks to the cattle and hom- cial support for more than four decades has been ogenization of the subsequently infected ticks provided by many donors and investors, includ- (Patel et al., 2016). The clean ticks, which come ing the Bill & Melinda Gates Foundation (BMGF), from a selected population of highly susceptible Biotechnology and Biological Sciences Research ticks maintained by ILRI’s tick unit, are fed on Council (BBSRC), CGIAR Funders through ILRI, production cattle infected with one of the three Danish International Development Agency component stabilates. Infected ticks are analysed The Management and Economics of East Coast Fever 247 for infection rates to ensure that equal numbers Further production of the vaccine will be of sporozoites from each isolate are included undertaken by CTTBD, in Malawi. In conjunc- when the ticks are combined before homogen- tion with GALVmed, ILRI facilitated the ization. The Muguga cocktail stabilate is evalu- e stablishment of the process in CTTBD and ated in extensive in vivo trials to ensure that the s upplied the vaccine seed stabilates, which vaccine is both safe and effective and to deter- were made at the same time as the ‘ILRI08’ mine the appropriate dilution with respect to a vaccine batch, and well-characterized patho- fixed dose of oxytetracycline. gen-free Rhipicephalus appendiculatus ticks. Given the complicated production process, ILRI scientists continue to provide technical one of the perceived obstacles to the widespread advice to M alawi. use of the ITM Muguga cocktail was the pro- One of the commonly voiced issues sur- spect of producing standardized, large-scale rounding the ITM vaccine, especially in the small- batches of several hundred thousand doses of holder dairy systems, is the number of doses in the vaccine. In 1996, ILRI was asked by FAO to the vaccine straws (ILRI/GALVmed, 2015). The undertake this challenge, which resulted in the standard presentation of the ITM vaccine pro- production of over 600,000 doses of a safe and duced at ILRI is plastic straws of 32–40 doses, effective vaccine. At the time, there was no other cryopreserved in liquid nitrogen. Once thawed institution in the region with the facilities and and prepared for vaccination with diluent, the expertise to undertake this task. The vaccine was vaccine has a working life of only a few hours provided to CTTBD, in Malawi, for subsequent (Mbao et al., 2006); any vaccine unused by that sales and distribution. All the available vaccine time must be discarded. Whereas this is not a was sold unsubsidized on a commercial basis, major problem when vaccinating large herds in which provided strong evidence that a commer- pastoralist settings, use of the vaccine in small- cially viable demand for the vaccine existed. holder settings, where each smallholder may In 2007, when the stocks of the FAO- have only two or three cattle, can lead to substan- requested ILRI batch were about to be depleted, tial wastage of the vaccine. To overcome this, ILRI ILRI was asked to produce a second commer- demonstrated that the vaccine straws can be cial-scale batch. At its own expense, the institute thawed, diluted, refrozen and repackaged with produced a batch of about 1.2 million doses minimal loss of vaccine viability. An initial experi- (named ‘ILRI08’). These were sold to distributors mental production resulted in straws containing authorized by the respective national veterinary five to eight doses (Patel et al., 2019). In collabor- authorities. As in 1998, no other institution ation with Vet Agro Limited, in Tanzania, ILRI in the region could mass produce a quality- showed that the vaccine was safe and effective assured ITM Muguga cocktail vaccine, so ILRI’s under both experimental and field conditions. Im- production again ensured continued vaccine portantly, cattle vaccinated with the diluted vac- supply at a time when demand for the vaccine cine are protected against challenge with the par- was gradually increasing across the region, es- ent stabilate, suggesting that any important pecially in Tanzania. In association with GAL- antigenic types are not lost during the process. Vmed, ILRI also compiled the documentation required to support the registration of the ITM Muguga cocktail, which was subsequently offi- Development and application cially registered in Kenya, Malawi and Tanza- of molecular tools nia, and progress was made towards registra- tion of the vaccine in Uganda. Until the Over the past 35 years or so, ILRI has developed vaccine’s registration in each country, the vac- and utilized a series of increasingly sophisticated cine had been used through special permission molecular tools to investigate various aspects of by the national veterinary authorities, which T. parva infection. These have taken advantage discouraged private distributors from taking up of, and kept pace with, advances in the fields of the vaccine. ILRI scientists also published de- immunology, molecular biology, genomics and tails of the production process in a peer- biotechnology. Some of these tools, which were reviewed journal to ensure future production developed primarily to study the biology and epi- of standardized batches of the Muguga cocktail demiology of the parasite or to facilitate the pur- stabilate (Patel et al., 2016). suit of a subunit vaccine against T. parva, have 248 P. Toye et al. found subsequent use in the analysis of the ITM with three of the mAbs. The original 16 mAbs vaccine. These tools have improved the quality described above, plus an additional four gener- control of the immunizing stabilate and have ated against buffalo-derived parasites, were used provided greater insights into the true genetic to characterize infected lymphocytes from buf- complexity of the Muguga cocktail and the effi- falo (Conrad et al., 1987b; Baldwin et al., 1988). cacy and biological impact of ITM in the field. Among other things, the results showed that They have also helped to allay concerns of the parasites with different mAb reactivity patterns veterinary authorities about the risks of using were present in a single isolate and that different ITM in the field. parasite types were isolated from the same a nimal when sampled at different times but that cloned parasite lines did not alter their mAb re- activity profile. Although the possibility of anti- Monoclonal antibodies (mAbs) genic variation (the ability of a single organism to express different antigens or forms of the mAbs were raised against the schizont stage same antigen at different times) in T. parva had of  the parasite specifically to try to define the been proposed earlier (Young et al., 1988), the immunological heterogeneity that had been ob- last observation suggested that this was not the served during the early stages of the develop- case. Thus, it became clear that the same animal ment of the ITM vaccine. The first set of seven can be infected with several parasite types con- mAbs did indeed show differential reactivity currently, and that while an initial infection can in  immunofluorescent assays with a set of result in protection against subsequent disease, schizont-infected cells, especially those derived it may not necessarily protect against subsequent from buffalo (Pinder and Hewitt, 1980). These infections. This observation has important observations were confirmed with a further nine implications in the ability of vaccinated animals mAbs tested on an extended array of parasite to transmit parasites to other, non-vaccinated, isolates (Minami et al., 1983). An additional animals. observation from this work was that immunized The specificity of several of the mAbs has cattle were susceptible to heterologous challenge been determined and all were shown to react from a parasite of different mAbs reactivity pro- with a single antigen known as the polymorphic files, and that the ‘breakthrough’ parasites had immunodominant molecule, or PIM (Shapiro the same profile as those comprising the heterol- et al., 1987; Toye et al., 1991). The differential re- ogous challenge. In a companion publication, it activity with strains of T. parva is believed to be was reported that cattle were protected from due to sequence variation of the epitopes located heterologous challenge from parasites having a within the PIM antigen (Toye et al., 1995a). The similar mAb profile but not those with a different size variation of the antigen among parasites profile (Irvin et al., 1983). from different isolates allowed an additional One of the practical aims of developing the layer of discrimination. It  was noted, for ex- mAbs was to use them to characterize antigeni- ample, that analysis of cloned cell lines, all de- cally distinct strains for inclusion in an ITM sta- rived from the Marikebuni isolate, revealed four bilate for use in field immunization. This has not different parasite types (Goddeeris et al., 1990; been possible, primarily because most of the an- Toye et al., 1991). ti-schizont mAbs recognize the same antigen PIM was also shown to be the major anti- and it is now known that several other antigens gen recognized by sera from infected cattle, are the targets of the protective immune re- which has led to its use in an enzyme-linked im- sponse induced by ITM (see below). A panel of munosorbent assay (ELISA) developed by ILRI nine of the mAbs was used as part of a study in- scientists to detect antibodies to T. parva (Katende volving several approaches to characterize the et al., 1998). This ELISA is used extensively dur- component stocks of the Muguga cocktail ing the production of the ITM vaccine to screen (Bishop et al., 2001). The results showed that cattle to be used for production of the sporozo- two of the stocks (Muguga- and Serengeti-trans- ites for the stabilate and for assessing the safety formed) displayed identical patterns of reactiv- and efficacy of the final product (Patel et al., ity, whereas the Kiambu 5 stock did not react 2016). In the field, the assay is relied upon to The Management and Economics of East Coast Fever 249 assess seroconversion levels following vaccin- parasite strains were also detected in animals ation, and it has been used widely in epidemi- after vaccination, indicating that the vaccinated ology studies (Gitau et al., 1997; Okuthe and animals can become re-infected but not show Buyu, 2006; Swai et al., 2009; Kivaria et al., clinical disease. The studies also showed that 2012; Malak et al., 2012; Toye et al., 2013; Kiara non-vaccinated animals co-grazing with vaccin- et al., 2014). ated animals could be infected with the vaccine Similar ELISAs were developed at ILRI for parasites. This phenomenon has been put forward other tick-borne diseases caused by Theileria mu- as a case against the use of live parasite vac- tans (Katende et al., 1990), Anaplasma marginale cines, particularly those comprising parasites (Morzaria et al., 1999) and Babesia bigemina originating from areas outside those in which (Tebele et al., 2000) and were transferred for cattle are being vaccinated (McKeever, 2007). commercial distribution to Svanova Biotech AB, These concerns are allayed by observations that now part of Boehringer-Ingelheim Animal no deleterious effects from the use of ITM have Health. The Theileria assays were subsequently yet to be reported, that such mixing of parasite withdrawn from sale due to insufficient global strains has been occurring for millennia, par- demand. T. parva is limited to parts of eastern, ticularly through the presumably unrestricted central and southern Africa, and T. mutans is movement of infected buffalo, and that the para- generally considered to be non-pathogenic. site strains used in the vaccines originate from Nevertheless, the assays are still offered as kits natural infections and are not ‘artificial’ or the and as a diagnostic service by ILRI. result of genetic manipulation in the laboratory. ILRI scientists were also part of a large international team that, in 2005, reported the sequencing of the T. parva genome. In the con- DNA-based strain identification clusion to the paper reporting this in Science (Gardner et al., 2005), the team described the Scientists at ILRI in collaboration with other re- genome data as ‘a critical knowledge base for searchers have developed a series of DNA-based a pathogen of significance to agriculture in methods to characterize Theileria parasites, in- Africa’. One immediate use of the genome se- cluding restriction fragment length polymor- quence was the development of a panel of ge- phisms of repetitive regions of the T. parva genome nome-wide mini- and microsatellite markers, (Conrad et al., 1987a; Allsopp et al., 1989), ana- which greatly enhanced the power and utility of lysis of polymorphisms in ribosomal RNA genes molecular analyses, as described above. The ap- (Bishop et al., 1992), telomere regions (Morzaria plication of the mini- and microsatellite markers et al., 1990) and genes encoding T. parva anti- coupled with high-t hroughput electrophoresis gens (Geysen et al., 1999), and the use of micro- showed that the M uguga cocktail is more com- satellites and minisatellites (Oura et al., 2003). An plex than previously thought, containing at early application of these techniques to the ITM least 14 different genotypes of T. parva, although vaccine revealed that two of the component they express a limited number of alleles (Patel s tabilates (Muguga and Serengeti-transformed) et al., 2011). The study also showed, by compar- were remarkably similar and quite distinct from ing the genotypic composition between different the Kiambu stabilate (Bishop et al., 2001). In a col- batches of the v accine, that the batches were laborative study with scientists from the Univer- very similar. High- throughput electrophoresis sity of Glasgow Veterinary School and the UK’s with the satellite markers has been shown to be Institute of Animal Health at Pirbright, these a useful and reproducible approach that can be markers were applied to analyse samples from cat- used to monitor the genetic composition of fu- tle vaccinated in the field with the Muguga cock- ture ITM vaccine batches. Such a tool is essential tail (Oura et al., 2004, 2007). to facilitate standardized, consistent, quality- With the caution that these markers repre- assured vaccine production. sent a very small but none the less highly poly- More recently, high-throughput sequen- morphic segment of the parasite genome, the cing has been used by ILRI scientists and collab- work showed that the vaccine strains can persist orators to characterize the parasite more deeply. in vaccinated animals for several years. Local Norling et al. (2015) analysed the whole-genome 250 P. Toye et al. sequences derived from the three component required for the rational design of an ECF vac- stocks of the Muguga cocktail and showed that cine was an understanding of the mediators of two of the stocks (Muguga- and Serengeti-trans- this immunity and the corresponding parasite formed) were remarkably similar. Somewhat components that induced these mediators. surprisingly, the total diversity residing in the three component stocks together represents only a small fraction of the T. parva genetic diversity The anti-schizont vaccine observed in field isolates. This result was sup- ported by satellite genotyping and high-through- put multilocus genotyping of genes encoding Lack of protection with serum antigens recognized by cytotoxic T lymphocytes The initial focus on the mediators of the protec- (CTLs) and believed to be important in the pro- tion seen in immune animals was on the role of tective immunity seen in cattle (Hemmink et al., serum. This is not surprising given the state 2016). There was very limited diversity in many of knowledge of, and tools available to study, of the antigen gene or satellite loci and certainly mammalian immune responses at the time. The far less than is seen in field populations of T. par- expectation was that infection with the parasite va. The results raise the intriguing question of induced the production of protective serum how the Muguga cocktail can protect cattle ex- antibodies. However, it was shown early on that posed to field challenge from T. parva. This is one cattle with high levels of anti-T. parva antibodies of many questions yet to be addressed concern- were nevertheless fully susceptible to infection ing this method of vaccination, as discussed (Wagner et al., 1974). below. The central role of CTLs Research to Develop a Subunit Attention at ILRAD then turned to investigating Vaccine a direct role for white blood cells as the primary mediators of immunity. The first evidence was obtained by Terry Pearson and co-workers, who A subunit vaccine is one that avoids the use of a showed that infected lymphocytes are specific- whole organism and instead relies on inocula- ally recognized and killed by effector cells from tion of those components that can stimulate, immune animals (Pearson et al., 1979). Shortly and are the target of, a protective immune re- thereafter, Eugui and Emery (1981) showed that sponse. The advantages of subunit vaccines are the cytotoxic (killing) activity was genetically that they avoid the risk of infection with virulent restricted. In other words, the killing was only organisms and, if they are manufactured by observed when the infected cells and the cyto- ‘synthetic’ means such as recombinant DNA toxic effector cells came from the same animal or technologies, are easier and cheaper to produce. closely related animals. This would be expected with killing mediated by classical CD8+ CTLs, where recognition of the infected cell is depend- Protection of exposed animals ent on presentation of the parasite components by molecules of the class I major histocom- A fundamental observation that serves as a basis patibility complex (MHC), which varies among for vaccine development is that animals that animals. This was subsequently confirmed by survive an infection are immune to the clinical Goddeeris et al. (1986a,b) who showed that the effects of a subsequent exposure. For T. parva, killing was mediated by cells of the CD8+ lineage this observation was made shortly after ECF was and, by using a panel of mismatched and semi- recognized (reviewed by Lawrence, 1992). The matched target cells, that it was restricted by a key knowledge gained from these and many class I MHC molecule (KN104). Importantly, it subsequent experiments is that cattle have the was demonstrated that the killing was specific capacity to prevent the clinical effects of T. parva for strains of the parasite – the immune cells infection, thus instilling confidence that an ef- were derived from the Muguga isolate of T. parva fective vaccine could be developed. What was and did not recognize cells infected with the The Management and Economics of East Coast Fever 251 Marikebuni isolate. As this reflected earlier in vivo characterize lymphocyte subsets (Naessens et al., challenge experiments, such as the work under- 1992; MacHugh et al., 1993). taken in the development of the Muguga cocktail The feasibility of using transient transfec- (discussed earlier), it provided convincing evi- tion technology to screen libraries of genes for dence that the CTL response was the main effector candidate CTL antigens was also demonstrated mechanism deployed by immune cattle. Further (Toye et al., 1995c). Eventually the use of a ran- confirmation of this was provided by cell transfer dom immunoscreen coupled with targeted gene experiments where a purified population of CD8+ analysis yielded the first successful identification lymphocytes from immune animals was shown to of T. parva proteins recognized by CTLs from im- provide protection when transferred into a naïve mune cattle (Graham et al., 2006). In further twin recipient (McKeever et al., 1994). work, the short (9–11 amino acids) peptide re- This work was important in the broader gions precisely recognized by the CTLs were context of vaccine development as it was one of identified, together with the class I MHC mol- the first demonstrations that CD8+ CTLs could ecule that presented the peptides on the cell sur- mediate protection against intracellular proto- face (Graham et al., 2008). Unfortunately, the zoan parasites. The challenge was then to repli- goal of inducing protective immune responses by cate the induction of the protective CTL response using these antigens as vaccines was only par- by using isolated parasite components (anti- tially achieved (Graham et al., 2006), and no gens) rather than the whole organism – in other method for consistently inducing CTLs in large words, a subunit vaccine. mammals with isolated proteins is currently available. Nevertheless, the identification of these Identification of CTL antigens CTL antigens was a major achievement in vac- cine research, and further examination of the By the early 1990s, the techniques to identify, antigens and the CTLs directed against them has characterize, isolate and administer parasite yielded some very valuable insights into the im- components that are the targets of antibody re- munobiology of the host–parasite relationship. sponses were well advanced and commonplace. The same situation did not apply to those recog- Antigenic diversity nized by CTLs. In general, CTLs are induced by and recognize antigens when they are presented Antigenic diversity is the phenomenon through on the surface of infected cells in the context of which parasites and other infectious organisms the class I MHC molecule. The most efficient way evade a protective immune response by chan- to evaluate individual components from infec- ging the sequence of the epitope recognized by tious organisms for CTL recognition is to intro- an existing immune response such that the mu- duce the gene that encodes the antigen into a tated organism can infect an already exposed cell expressing the appropriate MHC class I mol- host. Evidence for antigenic diversity in T. parva ecule, and one of the technologies that was was first shown for two of the CTL antigens, Tp1 emerging to do this was transfection technology. and Tp2 (MacHugh et al., 2009; Pellé et al., Among the first uses of transfection tech- 2011). The diversity was particularly striking in nology at ILRI was the production of cells suit- the Tp2 gene, with single nucleotide polymor- able for expressing candidate antigen genes. phisms identified at 61% of positions in the gene. Thus, a widely used mouse cell line was trans- Equally striking was the finding that there was fected with two bovine class I MHC genes, which, much greater diversity in the T. parva parasites at the same time, provided the first conclusive originating from buffalo than those from cattle, evidence that a second class I MHC locus existed suggesting that the T. parva population transmit- in cattle (Toye et al., 1990; Bensaid et al., 1991). ted among cattle may represent only a subset of Transfection technology also proved useful in the entire T. parva population, presumably that isolating the gene encoding an early candidate which is most fully adapted for transmission CTL antigen (PIM), which had proved recalci- among cattle. The studies did not detect evidence trant to isolation by the traditional bacterial among the currently identified epitopes that the expression systems (Toye et al., 1995b), and in mutations were driven by positive immune selec- establishing the specificity of mAbs used to tion (Pellé et al., 2011), although there was a 252 P. Toye et al. lack of recognition by naturally derived CTLs of being taken within ILRI, which was aimed at mutated epitope sequences in both Tp1 and Tp2 developing a vaccine to prevent the entry of (MacHugh et al., 2009; Connelley et al., 2011). sporozoites into infected cells. This was based on evidence that sera of animals from endemic Immunodominance areas or animals that had been hyperimmun- ized against T. parva were able to neutralize the CD8+ CTLs recognize antigens presented by class entry of sporozoites into lymphocytes in vitro I MHC molecules. The genes encoding the MHC (Musoke et al., 1982). The availability of neu- molecules are carried on two genetic regions, or tralizing mAbs led to the identification of the haplotypes, with each parent contributing one p67 surface molecule as the target of the protect- of the haplotypes. For CTLs specific for T. parva, it ive antibodies (Dobbelaere et al., 1985). These was noted very early on that the CTLs were re- observations culminated in the remarkable stricted predominantly or even completely by demonstration that cattle vaccinated with a re- only one of the MHC haplotypes (Morrison et al., combinant version of the p67 antigen were im- 1987). More recent work led by Ivan Morrison mune to subsequent challenge with T. parva, at the University of Edinburgh, UK, and in collab- which raised the prospects of a commercial vac- oration with ILRI scientists has shown that the cine against ECF (Musoke et al., 1992). How- great majority of the CTLs in any given respond- ever, subsequent field trials have returned mixed ing animal recognize the same peptide presented results in terms of the protective ability of the by the same MHC molecule (MacHugh et al., vaccine (reviewed by Nene et al., 2016) and the 2009). The precise peptide that is recognized is goal of a vaccine based on the p67 antigen is yet governed by the MHC type of the animal. This is to be realized. a quite remarkable phenomenon, given that T. parva is predicted to encode over 4035 proteins (Gardner et al., 2005), each of which could, the- ITM: The Future oretically at least, contain hundreds of peptide epitopes. The focus of the CTL response on one or two epitopes, termed immunodominance, had ILRI’s scientific contributions to the develop- been described previously in viral infections, ment, production and use of the ITM vaccine but this was the first description of the phenom- have been significant. Together, these contribu- enon in a complex organism such as T. parva. tions have enabled the ITM Muguga cocktail to Immunodominance has significant impli- be used to immunize hundreds of thousands of cations on the performance of a potential sub- cattle in both extensive (pastoralist and ranch) unit vaccine. The focus of the response on one or and dairy sectors. They have also helped pave two epitopes leaves the animal vulnerable to in- the way for commercial production, distribu- fection with a second parasite that carries muta- tion and use of the vaccine in much of eastern tions in those epitopes. Given the large number Africa. of MHC haplotypes that are likely to be present New research and development problems in outbred populations of cattle, there may be a must be addressed, however, as discussed below. similarly large number of epitopes presented by Many of these challenges and opportunities those MHC molecules. For there to be a subunit are also described by Perry (2016), based on a vaccine of practical use, it may rely on the exist- review commissioned by CGIAR. ence of a few antigens of limited diversity that can induce CTLs specific for epitopes present on those antigens when given in isolation and that The buffalo problem can still recognize and kill infected cells. Although the Muguga cocktail appears to pro- vide good protection against cattle-derived para- sites in both laboratory and field conditions, the The anti-sporozoite vaccine group at EAVRO observed that this protection did not extend to cattle exposed to buffalo- While the CTL response was being examined, derived parasites (Radley et al., 1979). Break- another approach, led by Tony Musoke, was through infections were recently observed in The Management and Economics of East Coast Fever 253 Kenya, where immunized cattle graze with or near Vaccine production buffalo (Sitt et al., 2015), although the vaccine appeared to perform well in northern Tanzania, The production and testing of batches of stabi- where buffalo and cattle graze together (Home- late is expensive, demanding and time consum- wood et al., 2006). To combat this, an obvious ing, and raises animal welfare issues: production solution is to produce a vaccine stabilate con- of 1 million doses requires 18 months, 130 cattle, taining buffalo-derived parasites, although this 500 rabbits and at least 600,000 nymphal ticks. presents two additional challenges. First, given A further problem in the production of the ITM the extensive heterogeneity in the parasite popu- vaccine is that the process includes the sexual lation found in buffalo, it may be difficult if not stage of the T. parva life cycle. Because of this, re- impossible to select a ‘buffalo’ parasite stabilate combination can occur, causing stabilates to that will protect against all buffalo challenges. vary both qualitatively and quantitatively, com- Second, the very low parasitaemia found in cat- plicating efficacy and safety testing. Studies have tle infected with buffalo-derived parasites will in- begun at ILRI to develop an in vitro correlate of crease considerably the number of cattle re- potency, which may reduce or eliminate the quired in the production process, which may need for the extensive in vivo testing that is now render this approach economically unsustain- required. In addition, in vitro production (grow- able. The current alternative is to provide clear ing of parasites in vessels under laboratory guidelines as to where the ITM Muguga cocktail c onditions) could minimize the need for cattle, should and should not be used. rabbits and ticks, thereby simplifying production and reducing the opportunity for interbatch Molecular tools variation to occur. The powerful molecular tools developed by ILRI and other scientists can now be used to track Performance monitoring quantitative variation in the parasite composition of stabilates. The availability of these tools means As use of the ITM vaccine increases, it will be im- that formal procedures must be developed and portant to continue to monitor the dynamics of documented as standards to evaluate production local T. parva populations and to investigate ap- batches of stabilate to ensure the production of parent vaccine failures and breakthroughs. a consistent, standardized, safe and effective Standard protocols should be developed to guide product. More sophisticated and powerful mo- such studies and the molecular and other tools lecular tools are likely to be needed in the future, as well as capacity building and technical sup- for example to detect parasite components that port to national laboratories charged with this might be important for immunity or transmis- responsibility. sion but that are present at low, currently un- detectable, levels. Also, more studies are needed to identify which antigens are responsible for the protection conferred by the Muguga cock- Irradiation of sporozoites: tail vaccine. potential lessons from candidate malaria vaccine Cold chain Encouraging results have recently been an- The need for a liquid nitrogen cold chain is in- nounced for a malaria vaccine based on malaria convenient and logistically challenging. Attempts sporozoites attenuated by exposure to radiation. to date to lyophilize (or freeze-dry) T. parva sporo- The resulting experimental vaccine is similar to zoites to make them stable at or near room tem- the ITM vaccine, although it does not require perature have resulted in low and variable rates simultaneous treatment with drugs. of recovery of viable parasites. Lyophilization of In August 2013, an article in Science found organisms as large and complex as sporozoites is that six adult volunteers who received the high- likely to present a significant challenge and may est dose of an experimental malaria vaccine, not be possible with the technology available P fSPZ, were protected from subsequent chal- today. lenge with malaria parasite-infected mosquitoes 254 P. Toye et al. (Seder et al., 2013). This is the first time that The Proliferative Response 100% protection has ever been achieved for a in T. parva-infected Lymphocytes malaria vaccine, albeit in a very small-scale phase I safety trial. This outperforms the results Casein kinase 2 reported in 2012 for the other leading malaria vaccine candidate, RTS,S/AS01, a subunit vac- ILRI’s molecular biology laboratory sought to cine, which protected just 31% of young infants determine the underlying molecular mechan- and 56% of older babies and toddlers. isms that drove bovine lymphocytes to divide The promising result for PfSPZ was achieved uncontrollably in theileriosis. The goal of under- using an approach that is similar to ITM: Plas- standing the mechanisms was to enhance the modium falciparum sporozoites are attenuated by ability to create a vaccine against the disease. radiation and the weakened but live parasites The results of this investigation, however, had a are injected intravenously. Antibodies to other much broader significance, which became evident stages in the parasite life cycle were undetect- only several years later (ole-MoiYoi, 1989, 1995; able, indicating that the irradiated sporozoites ole-MoiYoi et al., 1988, 1989, 1992, 1993). were effectively attenuated and did not develop beyond the early liver stage. The malaria team believes that the result has established proof of Background concept for the vaccine and has demonstrated that The generation time of T. parva-infected lympho- PfSPZ is safe and meets regulatory standards. cytes in vitro varies from 16 to 27 h. The intra- There would be significant advantages in lymphocytic form of the parasite, the macro- terms of cost and safety if an ECF vaccine com- schizont, is considered cancerous because it prising irradiated parasites were to be developed. induces uncontrolled growth and clonal expan- The possibility of using irradiated sporozoites to sion of the infected lymphocytes, which are plei- immunize cattle was explored by Cunningham’s omorphic and show alterations in surface group at EAVRO (Cunningham et al., 1973). The phenotype. The mode of death of the infected experiments were unsuccessful and the EAVRO animals is very much like that of acute leukae- scientists concluded that vaccination against mia in people, i.e. massive tissue and organ infil- ECF was unlikely to be achieved using irradiated tration in the lung, kidneys and other organs. parasites. Nevertheless, researchers at ILRAD T. parva, together with its Mediterranean coun- pursued the use of irradiated T. parva sporozoites terpart T. annulata, is unique among intracellu- in the late 1980s (ILRAD, 1989). Cattle inocu- lar protozoan parasites in causing this bovine lated with irradiated sporozoites showed weak leukaemia. The disease is unusual because if in- antibody and cytotoxic responses and were not fected cattle are treated in a timely manner with protected on challenge. Further advances in this anti-theilerial drugs, the uncontrolled cell division area will require methods to quantitate sporozo- stops with the death of the parasite. This is the ites with much greater precision and to expose reason why this disease is often described as revers- them more uniformly to the irradiation dose. ible lymphocyte transformation (leukaemia). The issues discussed above raise new ques- tions for vaccination against ECF: Experimental approach • Is a subunit vaccine attainable or could a better use of available resources be achieved Initial experiments to determine the driving if these were redirected to making the ITM force for lymphocyte division in theileriosis vaccine easier and cheaper to produce and focused on small segments of DNA known as deliver? t umour viruses or oncogenes. Many of these • Why are reactors and vaccine failures seen were originally identified in diseases caused by in Kenya in areas where buffalo-derived viruses, such as src of Epstein–Barr virus asso- parasites are present but not in northern ciated with Burkett’s lymphoma or v-myc of Tanzania? bird myelocytomatosis. None of the oncogenes • How does the Muguga cocktail provide that were available at the time showed any sig- such broad protection in the field? nificant hybridization signals with materials The Management and Economics of East Coast Fever 255 from T. parva-infected lymphocytes separated being tested for the treatment of other white according to their size in gels. blood cell cancers as well as cancers of solid tis- The next set of experiments employed con- sues such as those of the breast, lung, kidney ventional gel electrophoresis, which separates and prostate, and metastatic tumours. CK2 has proteins according to their size. Using various been shown to be a ‘pro-life’ enzyme. It protects inhibitors and activators of enzymes called pro- cells from natural death (apoptosis). As such, it tein kinases, which attach phosphate groups to minimizes the effectiveness of chemotherapy in certain amino acids in proteins, the researchers many human cancers. Surprisingly, CK2 inhib- could determine the class of protein kinase that ition also shows great promise in slowing was predominant in the T. parva-infected cells. down degenerative neurological diseases such There are two classes of protein kinase called as Alzheimer’s disease and Parkinson’s disease. tyrosine kinase oncogenes and serine/threonine Scientists at ILRAD/ILRI were vaguely aware of oncogenes. There was little or no information this potential while doing their experiments about serine/threonine oncogenes in the late because CK2 is the kinase that phosphorylates 1980s when these experiments were carried proteins in the brain that somehow become de- out. It was therefore surprising to find casein natured and precipitate out as fibrillary tangles, kinase 2 (CK2), which preferentially phosphor- compromising brain function. ylated only serine and threonine amino acid res- idues in the T. parva-infected cells. There were no detectable tyrosine oncogene signals in the T. parva-infected materials. CK2 was not known The Economic Impacts to function as an oncogene at that time. That it of ECF Research was discovered to be the cause of uncontrolled cell proliferation in theileriosis prompted med- Earlier studies have estimated various aspects of ical studies that confirmed it was also involved in the economic benefits of ECF disease control. some human cancers, which advanced under- According to McLeod and Randolph (2000), standing of their potential treatment. Follow-up adoption of a multi-component ECF vaccine in experiments could have included the generation small dairy and large commercial livestock sys- of lysates from T. parva-infected lymphocytes or tems across endemic countries in southern, cen- such lysates from purified macroschizonts to tral and eastern Africa would reduce the value determine what the parasite may secrete into of cattle mortality by US$10.1 million, while in- the cytoplasm of the infected lymphocytes that creasing value of milk production by US$1.7 activates CK2. It would have been interesting to million annually. Mukhebi et al. (1992) looked at have treated infected animals with a CK2 inhibi- the economics of immunization using ITM in tor, such as trace doses of heparin, to test ECF-affected countries in Africa. This analysis whether such charged molecules could enter the showed high potential economic returns, with a cells. However, at this time, ILRAD merged with benefit:cost ratio in the range of 9–17 under the International Livestock Centre for Africa various assumptions. Minjauw (1999) demon- (ILCA) to become ILRI and there was a change in strated the cost-effectiveness of ITM as an ECF the institute’s mandate. control strategy in traditionally managed cross- bred cattle (Bos indicus × B. taurus) in Zambia. Current implications of CK2 overproduction Other work has focused on the economic burden in human medicine of ECF in the livestock sector and general econ- omy. For example, ECF-related spending in Although the implications of CK2 as an onco- Kenya was estimated at US$10 million in 1987 gene were not known, and this work did not lead (Young et al., 1978), while Zimbabwe expended to a vaccine against T. parva, CK2 has recently an estimated US$9 million on ECF-related bills become a major target for the treatment of in the 1988/89 fiscal year (Perry et al., 1990). acute, chronic leukaemia in humans using com- Although these costs included control of other petitive inhibitors of the adenosine triphosphate tick-borne diseases, ECF is arguably the major (ATP) molecule. These studies are in phase II disease requiring acaricide application in the trials and show promise. In addition, CK2 is also region (Cunningham, 1977). 256 P. Toye et al. Modelling the economic markets can also be affected by ECF, the former impacts of ECF through the changing use of pasture land and/ or reduced land for crops based on the unavail- ECF causes a range of economic impacts associ- ability of draught labour, and the latter from re- ated with the morbidity and mortality of animals. duced throughput in local abattoirs and milk Following the classifications of disease impact processors that diminishes the demand for la- developed by Rich et al. (2014) and expanded by bour (see Rich and Wanyoike, 2010, for a simi- Rich and Niemi (2017), we can categorize the lar discussion of Rift Valley fever in Kenya). distinct levels of impacts – and associated mod- While macroeconomic impacts outside the agri- elling needs – of ECF to inform our modelling cultural and livestock sectors associated with approach. Rich et al. (2014) identified impacts ECF outbreaks are likely to be small, important taking place from the micro-level (i.e. house- environmental spillovers such as water and land hold or farm level) to different types of me- contamination exist with the use of acaricides so-level impacts (species, sector and value for dipping (Mukhebi and Perry, 1992). chain) and to aggregate macro-impacts on the Most economic analyses of ECF are derived local, regional or global economy. Their frame- from farm budget data. Various authors esti- work also considered externalities that could mated the regional impacts of ECF on meat and come from control strategies themselves (e.g. milk based on a combination of available sec- the effects of acaricides on the environment) as ondary data on livestock production and pri- well as spillovers from regional and inter- mary survey data (Mukhebi et al., 1992, 1995; national trade. Martins et al., 2010; Kivaria et al., 2012). An ex- The impacts of ECF are primarily through ception to these partial budget approaches is high mortality affecting the assets and incomes that of Nyangito et al. (1996), who developed a of farmers. In smallholder settings that are less whole-farm simulation model with linkages commercially oriented, this can severely restrict o utside the livestock sector, including crops and the ability of farmers to cope with market shocks products derived and used by livestock (e.g. feed, or to meet irregular livelihood needs such as draught power, manure). Financial and eco- school fees and family emergencies. As with other nomic impacts derived from these simulations diseases of livestock, the risk of disease itself can were computed over a 10-year horizon. The lead to stocking patterns that are suboptimal model allowed the simulation of alternative from the standpoint of market productivity, such investments and technologies and was used to as maintaining older, less productive (from a develop ECF control scenarios based on combin- commercial standpoint) herds that are less dis- ations of ITM and acaricides. Based on small- ease prone. For more commercially oriented holder farm data from Kilifi in Kenya, they found farmers, ECF reduces sales of meat and milk and that a control strategy that used ITM alongside a thwarts investment into scale economies, such 75% reduction in acaricide use generated an in- as fencing for beef and buildings, processing for ternal rate of return of over 34% and a benefit:- dairy or expanding herd sizes to meet growing cost ratio of 5.18. demand. Substitution effects with other sources A drawback of this approach (and of the of protein could also arise, causing changes in partial budget models cited above) is that prices supply and demand in other livestock markets. and other market parameters are exogenous to ECF can induce important spillover effects the system and thus fail to capture market dy- outside the immediate livestock sector. For in- namics related to disease control. Moreover, stance, where animals are used as draught la- most of the available economic analyses do not bour, ECF can have strong, negative impacts on directly consider the adoption process over the the productivity of staple crops such as maize. time involved in the implementation of ECF con- Mukhebi et al. (1992) estimated that the nega- trol strategies. Partial or slow uptake of ITM tive impacts of ECF on animal traction com- could reduce its impact both on the disease and prised 13% (US$21 million) of the total annual on markets over time. This suggests the need to loss associated with ECF in 11 African countries use more robust modelling frameworks that cap- in 1989. This in turn has indirect effects on ture a wider range of economic phenomena over prices of other food crops. Land and labour longer periods. The Management and Economics of East Coast Fever 257 Methods The Muguga cocktail ITM vaccine has been commercially available for the past 15  years. To assess the economic impacts of ECF, Rich and Over that period, close to 1.8 million doses of the Perry (2011) investigated the impacts of ECF on vaccine have been distributed. About 80% of production, prices, trade and livelihoods. First, these have been sold in the pastoral production they used the DynMod model (http://livtools. systems of northern Tanzania and about 10% in cirad.fr/dynmod; accessed 13 February 2020) the pastoral systems in Kenya. The rest have developed by Lesnoff (2007) to assess the impact been sold in smallholder dairy systems across of ECF on herd demography. DynMod traces Kenya. Although considerable effort has been livestock herd growth, based on exogenously put into the intensive dairy systems, adoption of defined births, fecundity and mortality; exogen- ITM in dairying has been low. ously defined purchases and offtake rates; and Many reasons have been advanced for this the initial structure of herd populations by age apparent contradiction, because when the and gender. These parameters calibrate the evo- vaccine was first developed the main target lution of herd growth based on a state-transition was smallholder dairy farmers who, it was then (age-cohort) model. Animals move between age believed, had the incentive to protect their cohorts based on pre-defined parameters associ- high-value dairy animals and were more com- ated with their time (in months) spent in a par- mercially oriented than other livestock keepers ticular age class (juveniles, subadults or adults). (Perry, 2016). Rich et al. (2014) used DynMod to develop Some of the reasons suggested include scenarios of alternative rinderpest control regimes alternative disease control methods in dairy sys- in which varying levels of rinderpest-associated tems. Acaricides are very effective in controlling mortality were generated to define herd popula- ECF when correctly used. They are also relatively tions under different control regimes. Rich et al. economical because one can buy enough for just (2014) adopted a similar approach for comput- one spray per week, and they also protect against ing the ex ante benefits of ECF control. Based on other tick-borne diseases. Curative drugs are also a review of the literature (Gachohi et al., 2012) available because most private animal health and expert consultations, the authors first de- service providers are in the high agricultural rived an average of annual incidence and fatality potential areas where most smallholder dairying associated with ECF in four different production is located. systems (intensive dairy (ID), open-grazing dairy The reluctance of the smallholder dairy (OD), agropastoral (AP) and pastoral (P)), as farmers to experiment with new untested methods shown in Table 6.1. The product of incidence on their valuable animals has also played a role. (I, with units of new annual cases of ECF/total As much as acaricides have their shortcomings, population) and case fatality (CF, with units of most farmers have used them and know they annual deaths from ECF/new annual cases of work. A novel approach would have to prove ECF) rates gives the percentage of the population itself before most farmers want to risk using it. dying annually from ECF. This product was Strong marketing of acaricides and anti- weighted by the population share (w) in each theilerial drugs by pharmaceutical companies in system to give a number for the national ECF the dairy areas has also contributed. Unfortu- mortality (MECF; Equation 6.1). A triangular dis- nately, they may sometimes give incomplete tribution comprising a minimum, most likely information, and the current distributors of the and maximum level of ECF mortality was gener- vaccine are small local companies whose staff ated for sensitivity analysis. may lack capacity to present scientifically cor- MECF å rect information.= wiIiCFi (6.1) { } The packaging of the last several batches of iÎ ID,OD,AP,P the vaccine in 30–40-dose packages is also a dis- Next, we developed alternative scenarios of incentive to smallholder dairy farmers. A single adoption patterns of ECF control through ITM. straw of the current batch of the vaccine covers 40 It is instructive to first review some of the issues calves, and to collect this number from small- governing adoption of ITM to inform our choice holder farms that on average keep one or two ani- of parameters. mals is problematic. Had small-dose packages 258 P. Toye et al. Table 6.1. Parameters used for deriving population trajectories in DynMod. (Calculations from DynMod.) ECF incidence (%) Case-fatality rate (%) Adoption rate (%) Predicted adoption by 2026 (%) Cattle population System share (%) Base Low High Base Low High 2007 2016 Base Low High Intensive dairy 8.7 20 10 30 10 8 15 0 3 20 15 25 Open-grazing dairy 10.5 20 10 30 15 10 20 0 5 40 30 50 Agropastoral 27.3 20 10 30 4 3 5 0 0 20 15 25 Pastoral 22.5 20 10 30 30 20 40 1 15 33 20 40 Not affected by ECF 31.5 0 0 0 0 0 0 0 0 0 0 0 Population-weighted 13.7 6.9 20.6 incidence Population-weighted case 10.1 7.0 13.6 fatality Population (number of cattle 17.5 in millions) The Management and Economics of East Coast Fever 259 been available for smallholder dairy systems, for up to 3  months without using acaricides. adoption of the vaccine might have been greater. Only when farmers get external hay do they The reason for the dose packaging is re- experience increased tick-borne disease risk. lated to vaccine production. Because of the Many factors that block tick-borne disease need to maximize on the number of sporozoites in dairy systems are the opposite in pastoral sys- and reduce the amount of tick material in the tems. First, pastoralists do not have alternative stabilate, ticks with high infection rates are nor- disease control options. Second, they keep large mally selected. The estimation of the sporozo- herds, and any one herd can easily include the ites is also a crude estimate based on the esti- 40 calves required for immunization with one mated number of infected ticks in a batch and straw of the vaccine. Third, whereas the vaccine the number of infected acini in an infected tick is relatively expensive, it is easier for pastoralists based on a random sample of ticks assessed. The to sell one animal to raise enough money to final dose was therefore only determined at the vaccinate their calves. Fourth, in many pas- end of the production run and there was no toral areas, the risk of ECF is increasing as a re- easy way to dilute the stabilate once it was pro- sult of upgrading indigenous stock with breeds duced. Recently, Patel et al. (2019) demon- that are more productive but also more suscep- strated that, once made, the stabilate can be tible to ECF. thawed, diluted and repackaged without loss of Few studies have investigated adoption efficacy. Another attempt to reduce the number trends for ITM, partly because the vaccine has of doses in a straw has been to predilute the sta- not been used extensively other than in the pas- bilate based on the expected concentration toral systems. Attempts at predicting future (based on the number of sporozoites/ml) and adoption trends are therefore speculative, based then determine the final dose afterwards. New on projected developments in the three produc- techniques to more accurately count the num- tion systems. ber of sporozoites in the stabilate are being de- Intensification of smallholder dairy farm- veloped. If these studies are successful, it may be ing is likely to increase, driven by population possible in the future to produce vaccines of a increases and land pressure. As pointed out, desired number of doses. the risk of disease declines with intensification. Provision of animal health services in these systems is also likely to improve. Although ITM price more acaricide resistance is projected to de- The price of ITM relative to other veterinary vac- velop generally, because acaricides are not in- cines is quite high. Although smallholder dairy tensively applied in intensive smallholder dairy farmers make money from the sale of milk, rais- systems, acaricide resistance in these dairy sys- ing at once US$8–12 for each animal to be im- tems is unlikely to be a problem in the immedi- munized is not easy. Unlike pastoralists who can ate future. Demand for the vaccine in the inten- sell one animal to raise enough money to vaccin- sive systems will arguably remain low in the ate all their calves, the smallholder dairy farmer near future. does not have a surplus animal to sell. This can In the more open-grazing dairy systems, de- work only in areas where there are strong co- mand for ITM is likely to rise due to increasing operatives that can advance farmers money for acaricide resistance. However, this will be con- vaccination. trasted by the trend towards intensification and zero grazing in some of the currently medium- Grazing type sized farms. Overall, there is likely to be a moder- ate increase in demand for the ITM vaccine in Zero grazing dominates the intensive dairy sys- open-grazing dairy systems. tems. This involves keeping cattle in stalls and The greatest potential of ITM is expected bringing fodder to them. This method reduces from agropastoral systems, which have the most the risk of tick-borne diseases, including ECF, cattle and which are spread across areas most because animals move less in tick-infested pas- suitable for ticks. Currently, the disease risk is tures. In some highland areas where animals are low due to an endemically stable situation. The constantly kept in stalls, many farmers can go indigenous breeds kept and a scarcity of ticks 260 P. Toye et al. due to overgrazing leads to low incidence and The new population mortality rate per age rare fatalities. However, many farmers are upgrad- class and time step ( MC1,t ; Equation 6.3) is thus ing their cattle by crossing their local animals the difference between the original age-cohort with exotic breeds to produce greater amounts mortality rate ( MC0 ) and the avoided mortality of milk that they can sell for high prices. This (AMt) achieved from the adoption of ITM: trend is likely to continue, and as more and more C C ECF farmers begin keeping ECF-susceptible breeds, M1,t = M0 -AMt (6.3) the disease risk might rise, and with it, demand for the ITM vaccine. This new path of mortality rates was subse- Demand for the vaccine in pastoral systems quently put into DynMod for simulation and is unlikely to change significantly. As more farm- compared against a counterfactual (baseline) ers become aware of the vaccine, more are likely scenario of control measures. to adopt its use until a peak is reached. After that, The International Model for Policy Analysis only newly born calves that have not been vac- of Agricultural Commodities and Trade (IM- cinated in the past and very few unvaccinated PACT) was then employed to compute impacts adults will be vaccinated. The trend towards up- on the agricultural economy and livelihoods grading animals is also likely to continue, which (the model is documented in Robinson et al., will lead to more disease-susceptible animals and 2015). IMPACT is a system of simulation models a greater demand for the vaccine. It is unlikely incorporating economic, crop, livestock, hydrol- that animal health services in the pastoral sys- ogy and climate change components. IMPACT tems will improve significantly in the near future uses a partial equilibrium economic model that and that there may not be alternative methods represents global supply, demand and trade of for controlling ECF other than vaccination. agricultural commodities among open econ- Given these dynamics in the different pro- omies. The demand for agricultural commod- duction systems, the authors developed a variety ities is simulated for 159 countries. Agricultural of different paths of adoption that are summar- production is modelled at a subnational level ized in Table 6.1 and in Fig. 6.2 in scenarios of known as food production units (FPUs). There low, most likely (shown as ‘base’) and high levels are 320 FPUs in IMPACT that represent intersec- of adoption. For dairy systems during 2007– tions among 159 nations and 154 water basins. 2016, it was assumed that adoption took place Solving the system of country demand and FPU in the last 3  years only (starting from 1% in supply equations in IMPACT produces measures 2014), while for pastoral systems, a gradual lin- of each country’s crop and livestock production ear increase in the adoption range was assumed. and land use, as well as measures of prices and For 2017–2026, roughly linear rates of adop- trade of agricultural commodities consistent with tion were assumed, albeit with slightly slower trade balances in global agricultural markets. rates of increase in the first part of the period, Model results are then translated into indicators rising steadily in the latter part and levelling off of agricultural incomes, food security, nutrition by 2026. Each of these adoption levels was and environmental health, among others. weighted in its respective system to obtain a na- The IMPACT model has been used to ana- tional weighted average. lyse global food security, rural development and The estimated impact on ECF mortality of natural resource management to 2050, includ- adoption of ITM was then computed. Adoption ing different scenarios of policy, technological, of ITM in each period (t) was assumed to result economic and climatic change. Some of the mod- in perfect control of ECF, consequently reducing el’s applications are relevant to assessment of population-weighted ECF mortality by the rate transformations in global livestock. These appli- of incremental adoption relative to the first cations highlight the drastically changing roles of period (2007) of the simulation. This percentage developing countries in the demand and produc- is defined as avoided mortality (AM; Equation tion of livestock products globally (Delgado et al., 6.2) from ECF for each period: 1999); a more recent study assessing the poten- tial impacts on food prices and the management AMECFt = å wiARi,tIiCFi (6.2) of natural resources of changing demand and iÎ{ID,OD,AP,P} supply of meat and milk in fast-growing regions The Management and Economics of East Coast Fever 261 Low adoption Most likely adoption High adoption 50 40 30 20 10 0 2007 2012 2017 2022 2026 2012 2017 2022 2026 2012 2017 2022 2026 Year System Intensive dairy Open dairy Agropastoral Pastoral Weighted mean Fig. 6.2. Projected ECF vaccine adoption rates by adoption scenario and production system, 2007– 2026. Constructed from DynMod model for ECF in Kenya. Intensive dairy cattle were about 8.7% of the cattle population in 2007, open dairy were 10.5%, agropastoral were 27.3%, pastoral were 22% and cattle unaffected by ECF approximately 31.5% of the national cattle herd of 17.5 million head. (Rosegrant et al., 2013); and an assessment of offsetting shortfalls in the supply of animal- trajectories of livestock demand and supply in sourced foods. Furthermore, the impact effects Africa and South Asia that explores the potential are captured over the long run, providing a tool impacts of improving livestock productivity in useful for more long-term planning. IMPACT key livestock-producing countries (McDermott may thus offer a much more comprehensive ap- et al., 2013). proach to measuring impacts of ECF control in The IMPACT model allows for international the affected regions than has been offered by trade, providing a framework to assess the sub- previous approaches (e.g. Mukhebi et al., 1992; national factors of ECF and its management that Minjauw, 1999). In addition, by introducing are driving dynamics of the demand and produc- reliably informed estimates of the parameters tion of livestock products and the global context depicting incidence and control of a specific dis- of food systems. As such, the model is useful in ease, this work builds directly on the earlier ana- assessing the effects of ECF control on livestock lysis of livestock futures by McDermott et al. (2013). production and food security, and on the poten- The current study applies the IMPACT tial (including through higher imports) for model to medium- to long-run impacts of the % of system farmers adopting 262 P. Toye et al. ITM vaccine for ECF control. Only the main as- Returns to ECF Investment pects of the model structure that are most rele- vant to the current context are presented below. Data from DynMod and IMPACT were used to As mentioned, FPU is the unit of analysis in derive the benefits associated with ECF control IMPACT, with FPU production of livestock calcu- under different scenarios. To measure the re- lated as the product of an assumed ‘average’ turns to investment for the ITM vaccine, we yield (AY) per head and the number of animals compared these benefits with the costs of achiev- in the FPU. Animals are distinguished by species ing them to derive benefit:cost ratios associated and include beef cattle, dairy cattle, sheep and with investments in ECF control. goats, poultry and pigs, while livestock product Two sets of costs were generated based on types (j) accounted for in the model are beef, two different sources. First, data on expenditures milk, lamb, poultry meat, eggs and pork. Live- on research by ILRAD/ILRI from 1975 to 2015 stock yield or production per animal (AY; Equa- were obtained to quantify the sunk costs associ- tion 6.4) is driven by factors of improved animal ated with research expenditures on vaccine de- and animal management practices that are ex- velopment, etc. It is worth highlighting that ogenous (Int) to the system of demand and sup- these costs are global costs that are wholly at- ply equations in IMPACT: tributed to the case study of Kenya. As such, = ´ they overestimate the costs incurred in Kenya. AYj,fpu AYIntj,fpu AYInt2j,fpu (6.4) Second, for each scenario, costs were generated on vaccine delivery based on expert opinions. Animal numbers (AN; Equation 6.5) are These costs included the number of doses de- functions of endogenous and exogenous fac- ployed per scenario, distributor costs and train- tors. These are species and system specific and ing costs for vaccinators. As delivery costs differ are a function of endogenously determined by production system, a weighted-average vac- input (c) and output (j) price indices (PNET) and cine cost was derived based on the share of ani- feed costs (PC), country-specific (cty) parameters mals in different systems. The weighted-average of feed efficiency (Feeds). The exogenous current cost of vaccine dose plus delivery was component of the equation adjusts year-to-year estimated at US$6.70. It was assumed that the changes in herd populations through a herd ex- costs of a new distributor of vaccines was pansion rate (ANInt2) that denotes system-level US$100,000 per 250,000 doses administered in differences. the first 5 years of the scenario, and US$100,000 AN = ANInt ´ANInt2 per 500,000 doses from year 6. We assumed j,fpu j,fpu j,fpu ANe Feede these costs were recurrent costs based on the æ PNETj,cty ö æ PC ö´ç ÷ ´Õç c,cty ÷ number of vaccines delivered. Finally, we as-ç PNET0 ÷ ç PC0 ÷è j,cty ø feedsè c,cty ø sumed that for every 10,000 doses of vaccine delivered, a training cost of US$300 was in- (6.5) curred. This was assumed as a one-time cost in Equation 6.5 provides a convenient entry the year when this occurred. All scenario costs point for modelling production system growth in were assumed to increase by a 5% inflation rate the context of ECF, as cattle mortality, a major annually. disease effect, can be factored into the animal ITM was assumed to confer life-long immun- population growth rates. The effects of ITM vac- ity to ECF. Therefore, the number of doses admin- cine use were thus simulated in IMPACT using istered had to account for the natural survival adjustments to the historical and projected and presence of animals that had been vaccinated growth rates (ANInt2) of beef and dairy cattle in previous years. A simple Markov chain was herds. These adjustments are based on the mor- constructed over the 20-year scenario period to tality and population growth rates generated in adjust the number of administered vaccinations DynMod for the different scenarios of ECF con- for animals that had already been vaccinated and trol, as described above. Animal mortality rates either survived or had not exited the system in turn reflect incidence, fatality and adoption through offtakes. The Markov chain was age- rates related to ECF disease and vaccine use, as cohort weighted to account for different mortality presented in Table 6.1. and sales transitions by age and sex of animal. The Management and Economics of East Coast Fever 263 Sunk research costs and scenario costs adoption from 2007, thus providing us with an were added to generate a stream of current costs ex post impact of adoption of ITM to date (2007– from 1975 to 2026. These were discounted at a 2016) and ex ante projections against such a 5% discount rate to generate net present values baseline from 2017 to 2026. As summarized in for each scenario. Benefit:cost ratios compare Fig. 6.3, there are higher impacts on herd growth additional agricultural production value derived from ITM adoption than in our most likely and from IMPACT/DynMod with the added costs low mortality scenarios. In these scenarios, we from the different scenarios, with the counter- see an increase in cattle stocks of 60% over factual assuming no expenditure on ECF at all. 2007–2026 versus a no-control baseline growth of just over 20% over the same period. Taking our most likely scenarios into account, Model results cumulative herd numbers remain about 10% higher relative to the baseline. First, results are presented of the projected paths Different rates of animal mortality simu- of cattle production from DynMod under nine lated using DynMod were translated into cattle scenarios of ECF control. These are contrasted populations in Kenya for scenarios of ECF con- with a baseline scenario that assumes no ITM trol using IMPACT. These scenarios include a Low adoption Most likely adoption High adoption 160 140 120 100 2007 2012 2017 2022 2027 2012 2017 2022 2027 2012 2017 2022 2027 Year Avoided mortality Low mortality Most likely mortality High mortality Baseline projection without vaccination Fig. 6.3. Projected cattle herds by vaccine adoption and mortality avoided, 2007–2026. (Constructed from DynMod model for ECF in Kenya.) Cattle herd index = 100 264 P. Toye et al. baseline that assumes that current levels of ECF scenario represent increases of 3–40% over year control are maintained over the simulation period 2026 production under baseline conditions. (2005–2026) and nine plausible situations of ECF Dairy production similarly increases 6% to 56% management. The baseline condition accounts in 2026 under alternative ECF management. for the current management of ECF in Kenya, in- Looking at the baseline trend for lamb, eggs, cluding acaricides and low rates of adoption of poultry meat and pork, it is observed that na- the ITM vaccine. Because future adoption is un- tional production increases by 80%, 19%, 113% certain, the ECF control scenarios necessarily and 44%, respectively, for these product types, cover a range of adoption rates. Because there is from 2005 to 2026. Furthermore, relative to the insufficient information on current rates of cat- baseline in 2026, production may decline very tle mortality due to ECF (particularly as distinct slightly under the ECF control scenarios for from other disease-related animal deaths), the lamb, for poultry and for eggs, with no change scenarios also use a range of ECF mortality rates. observed in pork production to 2026. The de- In applying the data on animal growth rates as cline in production of lamb, eggs and poultry input into the IMPACT model, impacts were meat is probably explained by market substitu- simulated of interactions among ECF, herd dy- tion effects. As cattle meat and dairy supplies in- namics, markets and socio-economic variables. crease under ECF management, the prices of Table 6.2 presents IMPACT projections of these products decline, causing a weakening in the supply of animal-source foods associated the demand (and subsequent production) of with the baseline and ECF scenarios in Kenya. comparable products. The market shifts are, Compared with the status quo in which beef pro- however, small. Related to the expansion in cat- duction increases from 392,000  t in 2005 to tle production, the aggregate supply of meat, 706,000 t in 2026, beef output in 2026 is pro- milk and eggs increases by 48% from 2005 to jected to increase to between 726,000 and 2026. Compared with the baseline in 2026, sup- 990,000  t under various assumptions of ECF ply increases under the ECF control scenarios in mortalities. These scenarios correspond to low to a range of 0.5–5%. high levels of vaccine adoption countrywide. The Figure 6.4 presents an index of the value estimates of beef production under the ECF control of national agricultural production associated Table 6.2. Production and net imports of selected commodities in Kenya, 2005–2026. (Calculations from DynMod and from IFPRI IMPACT; Robinson et al., 2015). 2026: 2026: low mortality, 2026: medium mortality, 2026: high mortality, 2005 status quo medium adoption medium adoption medium adoption Production in 1000 t Beef 392 706 732 768 942 Dairy 2178 5018 5200 5254 6694 Lamb 77 140 140 140 140 Eggs 61 73 73 73 73 Poultry meat 19 43 43 43 42 Pork 14 21 21 21 21 Feedsa 286 461 475 494 590 Net imports in 1000 t Beef 0.83 36.68 9.26 0.00 0.00 Dairy 3.30 0.00 0.00 0.00 0.00 Lamb 0.10 0.00 0.00 0.00 0.00 Eggs 0.11 52.33 52.54 52.55 52.56 Poultry meat 0.00 10.76 10.77 10.77 10.79 Pork 0.00 0.00 0.00 0.00 0.00 aFeed grain for livestock production, not quantity of feeds produced. The Management and Economics of East Coast Fever 265 Low adoption Most likely adoption High adoption 200 150 100 2007 2012 2017 2022 2026 2012 2017 2022 2026 2012 2017 2022 2026 Year Avoided mortality Low mortality Most likely mortality High mortality Baseline projection without vaccination Fig. 6.4. Index of agricultural production in Kenya by vaccine adoption and mortality avoided, 2026–2027. (Constructed from DynMod model for ECF in Kenya.) with the vaccination scenarios. Relative to the animal mortality, the net import of beef is be- reference case, agricultural incomes improve by tween 5000 and 15,000 t in 2026, or roughly between 0.8% and 12% in 2027. On a commodity- 1–2% of the demand in that year. Net dairy by-commodity basis, revenue from beef and imports at 3300  t, are only 0.1% of national dairy production increases by 3% to 40% over demand in 2005. ECF control in cattle does the baseline. Benefits also accrue to the con- influence the demand and supply of other live- sumers, but these are more modest. Aggregate stock commodities, probably through competi- household expenditures on food are lowered by tion for feeds, but these effects are also small. 0.01% to 0.07% in 2027 under the ECF scen- The model projections on crop use for arios compared with the reference case. Food ex- feed are substantial. Baseline feed demand in- penditure as a share of national income remains creases 61% from 2005 (285,000  t) to 2026 the same or declines (by 0.01% at most) relative (461,000 t). Demand for cereals is 29% of this to the base case in 2027. demand in 2005 and 35% of the same in 2016. The vaccination scenarios have faint effects Under the ECF control scenarios, demand for on international trade in livestock commodities. livestock feeds in 2026 is 2.3–34% higher than The scenarios lead to reduced imports in both the baseline volume. However, there is little or volume and value. Under assumptions of low no increase in cultivated area. This suggests that Agricultural production index = 100 in 2007 266 P. Toye et al. cropland is being reallocated (e.g. from use as unlike the projections of Kristjanson et al. (1999) food) to the production of livestock feeds. As cat- of the returns to a trypanosomiasis vaccine, tle production expands and meat and milk prices which does not exist. fall, households are probably able to replace Model results indicate modest impacts of staples in their diets with the higher nutrient ECF control through ITM. There was a steady in- animal-source foods, so that no additional land crease in domestic supply of cattle, meat and is needed to support the growth in feed demand. milk and an associated reduction in net imports. Figure 6.5 illustrates benefit:cost ratios The value of production in the agricultural sec- associated with the use of the ITM vaccine in tor rises by about 12% relative to the baseline by Kenya. The ITM vaccine, under a wide range of 2026. Agricultural revenue is increased, mostly hypotheses about vaccine adoption rates across through expanded livestock production, while livestock systems and about avoided mortality, food expenditures are lower in 2026 under the would produce a high return to ILRI’s historical ECF scenarios than in the baseline case. The im- investment in all types of ECF research. This plication is that producer welfare is improved in example of a proven technology against ECF is the near to medium term but not at the expense 15 10 5 0 Low adoption Most likely adoption High adoption Vaccine adoption levels Avoided mortality Low mortality Most likely mortality High mortality Fig. 6.5. Benefit:cost ratios for ECF vaccination research and development programmes by vaccine adoption levels and mortality avoided. (Constructed from DynMod model for ECF in Kenya.) Benefit:cost ratio The Management and Economics of East Coast Fever 267 of consumer welfare. Impacts on nutrition and younger, more productive animals, and might food security are relatively small, suggesting allow farmers greater access to commercial that increased production probably reduces reli- markets that demand younger stock. ance on food imports and improves the supply of key macro- and micronutrients in human diets but does not create new food consumption. Conclusions and the Future ILRI and its historical partners have had signifi- Model gaps cant scientific and development impact on our understanding of theileriosis and on manage- An important research gap is ECF’s impact on ment of ECF. The demonstration that commer- morbidity and productivity. Our analysis focuses cial-scale batches of the ITM vaccine could be on mortality avoided by ECF vaccination and produced was a major achievement and under- does not consider disease effects on meat and pinned the sale of the vaccine in Tanzania. In a milk productivity or possible rising costs of treat- more direct sense, it also resulted in the immun- ment. We may have therefore understated the ization of more than 1.5 million cattle. The many potential benefits from ITM in Kenya. years of basic research in the search for a subunit Another model gap is that IMPACT does not vaccine have not only resulted in a greater consider the power output of cattle. We expect that understanding of the biology and immunology improved control of ECF would have a positive of T. parva infection but have also had a much i mpact on the use of cattle as an input to crop pro- broader scientific impact on our general know- duction, which would raise producer incomes. ledge of protozoan–host interactions and of the However, as Kenyan agriculture mechanizes, bovine immune system, especially the role and draught power will become less important and po- function of the MHC and CTLs. The scientific out- tential benefits from increasing draught power will comes unexpectedly found application in human become smaller, so it is not possible to project a net tumour biology. The numerous serological and effect of omitting power benefits from the present nucleic acid-based tools that were generated dur- model. ing this time later proved immensely valuable in A second indirect effect of ITM adoption the characterization and quality control of the that we did not capture is the potential benefit of ITM vaccine, and in dissecting the functioning of controlling high-mortality diseases such as rin- the bovine immune system. derpest and ECF. This would, in theory, allow The future for ILRI in the pathology and im- farmers to stock herds in a more efficient man- munoparasitology of theileriosis will be guided ner and one less dictated by risk preferences (e.g. by the degree of success in the uptake of the ITM stocking older animals instead of more product- vaccine, balanced against the evolving prospects ive younger ones, or keeping less productive but for a subunit vaccine. The future in the epidemi- disease-resistant breeds; Rich et al., 2014). These ology and economics of ECF management will shifts could have significant effects by providing be developing and evaluating current or novel farmers with higher prices for the sales of control methods. References Abbas, R.Z., Zaman, M.A., Colwell, D.D., Gilleard, J. and Iqbal, Z. (2014) Acaricide resistance in cattle ticks and approaches to its management: the state of play. Veterinary Parasitology 203, 6–20. Allsopp, B.A., Carrington, M., Baylis, H., Sohal, S., Dolan, T. and Iams, K. (1989) Improved characterization of Theileria parva isolates using the polymerase chain reaction and oligonucleotide probes. Molecular and Biochemical Parasitology 35, 137–147. Baldwin, C.L., Malu, M.N. and Grootenhuis, J.G. (1988) Evaluation of cytotoxic lymphocytes and their parasite strain specificity from African buffalo infected with Theileria parva. Parasite Immunology 10, 393–403. 268 P. Toye et al. Bensaid, A., Kaushal, A., Baldwin, C.L., Clevers, H., Young, J.R., et al. (1991) Identification of expressed bovine class I MHC genes at two loci and demonstration of physical linkage. Immunogenetics 33, 247–254. Bishop, R., Baylis, H., Allsopp, B., Toye, P., Nene, V., et al. (1992) Genomic polymorphisms in Theileria parva. In: Morzaria, S. (ed.) Genome Analysis of Protozoan Parasites. Proceedings of a Workshop held at ILRAD Nairobi, Kenya, 11–13 November 1992. ILRAD, Nairobi, pp. 70–76. Bishop, R., Geysen, D., Spooner, P., Skilton, R., Nene, V., et  al. (2001) Molecular and immunological characterisation of Theileria parva stocks which are components of the ’Muguga cocktail’ used for vaccination against East Coast fever in cattle. Veterinary Parasitology 94, 227–237. Burridge, M.J., Morzaria, S.P., Cunningham, M.P. and Brown, C.G.D. (1972) Duration of immunity to East Coast fever (Theileria parva infection of cattle). Parasitology 64, 511–515. CABI (2020). Invasive Species Compendium: distribution table. www.cabi.org/isc/datasheet/62109#todistri- butionDatabaseTable (accessed 30 April 2020). Coetzer, J.A.W., Thomson, G.R. and Tustin, R.C. (eds) (1994) Infectious Diseases of Livestock with Spe- cial Reference to Southern Africa, Vols I–III. Oxford University Press, Cape Town. Connelley, T.K., MacHugh, N.D., Pellé, R.P., Weir, W. and Morrison, W.I. (2011) Escape from CD8+ T cell response by natural variants of an immunodominant epitope from Theileria parva is predominantly due to loss of TCR recognition. Journal of Immunology 187, 5910–5920. Conrad, P.A., Iams, K., Brown, W.C. and Sohanpal, B. (1987a) DNA probes detect genomic diversity in Theileria parva stocks. Molecular and Biochemical Parasitology 25, 213–226. Conrad, P.A., Stagg, D.A., Grootenhuis, J.G., Irvin, A.D., Newson, J., et al. (1987b) Isolation of Theileria parasites from African buffalo (Syncerus caffer) and characterization with anti-schizont monoclonal antibodies. Parasitology 94, 413–423. Cunningham, M.P. (1977) Immunization of cattle against Theileria parva. In: Henson, H.B., and Campbell, M (eds) Theileriosis. Report of a Workshop, 7–9 September, Nairobi, Kenya. IDRC, Ottawa, pp. 66–75. Cunningham, M.P., Brown, C.G.D., Burridge, M.J., Musoke, A.J., Purnell, R.E. and Dargie, J.D. (1973) East Coast fever of cattle: 60Coirradiation of infective particles of Theileria parva. Journal of Protozoology 20, 298–300. Delgado, C.L., Rosegrant, M.W., Steinfeld, H, Ehui, S. and Courbois, C. (1999) Livestock to 2020: the next food revolution. Food, Agriculture, and Environment Discussion Paper 28. IFPRI, Washington, DC. Di Giulio, G., Lynen, G., Morzaria, S., Oura, C. and Bishop R. (2009) Live immunization against East Coast fever – current status. Trends in Parasitology 25, 85–92. Dobbelaere, D.A., Shapiro, S.Z. and Webster, P. (1985) Identification of a surface antigen on Theileria parva sporozoites by monoclonal antibody. Proceedings of the National Academy of Sciences USA 82, 1771–1775. Eugui, E.M. and Emery, D.L. (1981) Genetically restricted cell-mediated cytotoxicity in cattle immune to Theileria parva. Nature 290, 251–254. FAO (2014) Global Livestock Production and Health Atlas (GLiPHA). FAO, Rome. FAO (2017) FAOSTAT Statistics Database of the Food and Agricultural Organization of the United Nations. FAO, Rome. FAO/AU-IBAR/ILRI (2017) Regional strategy for the control of African swine fever in Africa. FAO, Rome. Gachohi, J., Skilton, R., Hansen, F., Ngumi, P. and Kitala, P. (2012) Epidemiology of East Coast fever (Thei- leria parva infection) in Kenya: past, present and the future. Parasites & Vectors 5, 194. Gardner, M.J., Bishop, R., Shah, T., de Villiers, E.P., Carlton, J.M., et al. (2005) Genome sequence of Thei- leria parva, a bovine pathogen that transforms lymphocytes. Science 309, 134–137. Geysen, D., Bishop, R., Skilton, R., Dolan, T.T. and Morzaria, S. (1999) Molecular epidemiology of Theileria parva in the field. Tropical Medicine & International Health 4, A21–A27. Gitau, G.K., Perry, B.D., Katende, J.M., McDermott, J.J., Morzaria, S.P. and Young, A.S. (1997) The preva- lence of tick-borne infections in smallholder dairy farms in Murang’a district; a cross-sectional study. Preventive Veterinary Medicine 30, 95–107. Goddeeris, B.M., Morrison, W.I., Teale, A.J., Bensaid A. and Baldwin, C.L. (1986a) Bovine cytotoxic T-cell clones specific for cells infected with the protozoan parasite Theileria parva: parasite strain specificity and class I major histocompatibility complex restriction. Proceedings of the National Academy of Sci- ences USA 83, 5238–5242. Goddeeris, B.M., Morrison, W.I. and Teale, A.J. (1986b) Generation of bovine cytotoxic cell lines, specific for cells infected with the protozoan parasite Theileria parva and restricted by products of the major histo- compatibility complex. European Journal of Immunology 16, 1243–1249. The Management and Economics of East Coast Fever 269 Goddeeris, B.M., Morrison, W.I., Toye, P.G. and Bishop, R. (1990) Strain specificity of bovine Theileria parva- specific cytotoxic T cells is determined by the phenotype of the restricting class I MHC. Immunology 69, 38–44. Graham, S.P. Pellé, R., Yamage, M., Mwangi, D.M., Honda, Y., et al. (2006) Theileria parva candidate vac- cine antigens recognized by immune bovine cytotoxic T lymphocytes. Proceedings of the National Academy of Sciences USA 103, 3286–3291. Graham, S.P., Pellé, R., Yamage, M., Mwangi, D.M., Honda, Y., et al. (2008) Characterization of the fine specificity of bovine CD8 T cell responses to defined antigens from the protozoan parasite Theileria parva. Infection and Immunity 76, 685–694. Hemmink, J.D., Weir, W., MacHugh, N.D., Graham, S.P., Patel, E., et al. (2016) Limited genetic and anti- genic diversity within parasite isolates used in a live vaccine against Theileria parva. International Journal for Parasitology 46, 495–506. Homewood, K., Trench, P., Randall, S., Lynen, G. and Bishop, B. (2006) Livestock health and socioeconomic impacts of veterinary intervention in Masailand: infection-and-treatment vaccine against East Coast fever. Agricultural Systems 89, 248–271. ILRAD (1989) Bovine immune responses to irradiated Theileria parva sporozoites. In: ILRAD Annual Scien- tific Report. ILRAD, Nairobi, pp. 11–13. ILRI/GALVmed (2015) Distribution, delivery and improvement of the infection and treatment method vac- cine for East Coast fever. Report of a Workshop, Nairobi, Kenya, 19–20 August 2014. ILRI, Nairobi. Irvin, A.D., Dobbelaere, D.A.E., Mwamachi, D.M., Minami, T., Spooner, P.R. and Ocama, J.G.R. (1983) Immunisation against East Coast fever. Correlation between monoclonal antibody profiles of Theileria parva stocks and cross immunity in vivo. Research in Veterinary Science 35, 341–346. Katende, J., Morzaria, S., Toye, P., Skilton, R., Nene, V., et al. (1998) An enzyme-linked immunosorbent assay for the detection of Theileria parva antibodies in cattle using a recombinant polymorphic immu- nodominant molecule. Parasitology Research 84, 408–416. Katende, J.M., Goddeeris, B.M., Morzaria, S.P., Nkonge, C.G. and Musoke, A.J. (1990) Identification of a Theileria mutans-specific antigen for use in an antibody and antigen detection ELISA. Parasite Immun- ology 12, 419–433. Kiara, H., Jennings, A., Bronsvoort, B.M. de C., Handel, I.G., Mwangi, S.T., et al. (2014) A longitudinal as- sessment of the serological response to Theileria parva and other tick-borne parasites from birth to one year in a cohort of indigenous calves in western Kenya. Parasitology 141, 1289–1298. Kivaria, F.M., Kapaga, A.M., Mbassa, G.K., Mtui,P.F and Wani, R.J. (2012) Epidemiological perspectives of ticks and tick-borne diseases in South Sudan: cross-sectional survey results. Onderstepoort Journal of Veterinary Research 79, E1–E10. Kristjanson, P.M., Swallow, B.M., Rowlands, G.J., Kruska, R.L. and de Leeuw, P. (1999) Measuring the costs of African animal trypanosomosis, the potential benefits of control and returns to research. Agri- cultural Systems 59, 79–98. Lawrence, J.A. (1992) History of bovine theileriosis in southern Africa. In: Norval, R.A.I., Perry, B.D. and Young, A.S. (eds) The Epidemiology of Theileriosis in Africa. Academic Press, London, pp. 27–29. Lesnoff, M. (2007) DynMod – A Tool for Demographic Projections of Ruminants Under Tropical Conditions. User’s Manual. ILRI, Nairobi. Lynen, G., Yrjo-Koskenen, S.E., Bakuname, C., Di Giulio, G., Mlinga, N., et al. (2012) East Coast fever im- munisation field trial in crossbred dairy cattle in Hanang and Handeni districts in Northern Tanzania. Tropical Animal Health and Production 44, 567–572. MacHugh, N.D., Connelly, T., Graham, S.P., Pellé, R., Formisano, P., et al. (2009) CD8+ T cell responses to Theileria parva are preferentially directed to a single dominant antigen: implications for parasite strain-specific immunity. European Journal of Immunology 39, 2459–2469. MacHugh, N.D., Taracha, E.L. and Toye, P.G. (1993) Reactivity of workshop antibodies on L cell and COS cell transfectants expressing bovine CD antigens. Veterinary Immunology and Immunopathology 39, 61–67. Malak, A.K., Mpoke, L., Banak,J., Muriuki, S., Skilton, R.A., et al. (2012) Prevalence of livestock diseases and their impact on livelihoods in Central Equatoria State, southern Sudan. Preventive Veterinary Medicine 104, 216–223. Martins, S.B., Di Giulio, G., Lynen, G., Peters, A. and Rushton, J. (2010) Assessing the impact of East Coast Fever immunisation by the infection and treatment method in Tanzanian pastoralist systems. Preventive Veterinary Medicine 97, 175–182. 270 P. Toye et al. Mbao, V., Berkvens, D., Dolan, T., Speybroeck, N., Brandt, J., et al. (2006) Infectivity of Theileria parva sporozoites following cryopreservation in four suspension media and multiple refreezing: evaluation by in vitro titration. Onderstepoort Journal of Veterinary Research 73, 207–213. McDermott, J., Enahoro, D. and Herrero, M. (2013) Livestock futures to 2020. In: Barrett, C.B. (ed.) How Will They Shape Food, Environmental, Health, and Global Security? Food Security and Sociopoliti- cal Stability. Oxford University Press, Oxford. McKeever, D.J. (2007) Live immunisation against Theileria parva: containing or spreading the disease? Trends in Parasitology 12, 565–568. McKeever, D.J., Taracha, E.L.N., Innes, E.L., MacHugh, N.D., Awino, E., et al. (1994) Adoptive transfer of immunity to Theileria parva in the CD8+ fraction of responding efferent lymph. Proceedings of the Na- tional Academy of Sciences USA 91, 1959–1963. McLeod, R. and Randolph, T. (2000) Product development plan: East Coast fever vaccine for Africa. ILRI, Nairobi (unpublished report). Minami, T., Spooner, P.R., Irvin, A.D., Ocama, J.G.R., Dobbelaere, D.A.E. and Fujinaga, T. (1983) Charac- terisation of stocks of Theileria parva by monoclonal antibody profiles. Research in Veterinary Science 35, 334–340. Minjauw, B. (1999) The economic impact of heartwater in Tanzania, Zambia, Mozambique, and of its con- trol through the use of new inactivated vaccines. ILRI, Nairobi (unpublished report). Morrison, W.I., Goddeeris, B.M., Teale, A.J., Groocock, C.M., Kemp, S.J. and Stagg, D.A. (1987) Cytotoxic T-cells elicited in cattle challenged with Theileria parva (Muguga): evidence for re- striction by class I MHC determinants and parasite strain specificity. Parasite Immunology 9, 563–578. Morzaria, S.P., Katende, J., Musoke, A., Nene, V., Skilton, R. and Bishop, R. (1999) Development of sero- diagnostic and molecular tools for the control of important tick-borne pathogens of cattle in Africa. Parassitologia 41, 73–80. Morzaria, S.P., Spooner, P.R., Bishop, R.P., Musoke, A.J. and Young, J.R. (1990) SfiI and NotI polymor- phisms in Theileria stocks detected by pulsed field gel electrophoresis. Molecular and Biochemical Parasitology 40, 203–211. Morzaria, S. and Williamson, S. (1999) Live Vaccines for Theileria parva: Deployment in Eastern, Central and Southern Africa. Proceedings of an FAO/OAU-IBAR/ILRI workshop, 10–12 March 1997. Nairobi. ILRI, Nairobi. Mukhebi, A.W. and Perry, B.D. (1992) Economic implications of the control of East Coast fever in Eastern, Central, and Southern Africa. In: Kategile, J.A. and Mubi, S. (eds) Future of Livestock Industries in Eastern and Southern Africa: Proceedings of the workshop held at Kadoma Ranch Hotel, Zim- babwe, 20–23 July 1992. ILCA, Addis Ababa, pp. 107–112. Mukhebi, A.W., Perry, B.D. and Kruska, R. (1992) Estimated economics of theileriosis control in Africa. Pre- ventive Veterinary Medicine 12, 73–78. Mukhebi, A.W., Kariuki, D.P., Mussukuya, E., Mullins, G., Ngumi, P.N., et al. (1995) Assessing the eco- nomic impact of immunisation against East Coast fever: a case study in coast province, Kenya. Vet- erinary Record 137, 17–22. Musoke, A.J., Nantulya, V.M., Buscher, G., Masake, R.A. and Otim, B. (1982) Bovine immune response to Theileria parva: neutralizing antibodies to sporozoites. Immunology 45, 663–668. Musoke, A., Morzaria, S., Nkonge, C., Jones, E. and Nene, V. (1992) A recombinant sporozoite surface antigen of Theileria parva induces protection in cattle. Proceedings of the National Academy of Sci- ences USA 89, 514–518. Naessens, J., Sileghem, M., MacHugh, N., Park, Y.H., Davis, W.C. and Toye, P. (1992) Selection of BoCD25 monoclonal antibodies by screening mouse L cells transfected with the bovine p55-interleukin-2 (IL-2) receptor gene. Immunology 76, 305–309. Ndungu, S.G., Brown, C.G.D. and Dolan, T.T. (2005) In vivo comparison of susceptibility between Bos indicus and Bos taurus cattle types to Theileria parva infection. Onderstepoort Journal of Veterinary Science 72, 13–22. Nene, V., Kiara, H., Lacasta, A., Pellé, R., Svitek, N. and Steinaa, L. (2016) The biology of Theileria parva and control of East Coast fever – current status and future trends. Ticks and Tick-borne Diseases 7, 549–564. Norling, M., Bishop, R.P., Pellé, R., Qi, W., Henson, S., et al. (2015) The genomes of three stocks compris- ing the most widely utilized live sporozoite Theileria parva vaccine exhibit very different degrees and patterns of sequence divergence. BMC Genomics 16, 729. The Management and Economics of East Coast Fever 271 Norval, R.A.I., Perry, B.D. and Young, A.S. (eds) (1992) The Epidemiology of Theileriosis in Africa. Aca- demic Press, London. Nyangito, H.O., Richardson, J.W., Mundy, D.S., Mukhebi, A.W., Zimmel, P. and Namken, J. (1996) Eco- nomic impacts of East Coast Fever immunization on smallholder farms, Kenya: a simulation analysis. Agricultural Economics 13, 163–177. Okuthe, O.S. and Buyu, G.E. (2006) Prevalence and incidence of tick-borne diseases in smallholder farm- ing systems in western Kenya highlands. Veterinary Parasitology 141, 307–312. ole-MoiYoi, O.K. (1989) Theileria parva: An intracellular protozoan parasite that induces reversible lympho- cyte transformation. Experimental Parasitology 69, 204–210. ole-MoiYoi, O.K. (1995) Casein kinase II in theileriosis. Science 267, 834–837. ole-MoiYoi, O.K., Iams, K.P., Nayar, A., Brown, W.C., Sugimoto, C., et al. (1988) Protein kinase activation in Theileria-infected cells. In: Lonsdale-Eccles, J.D. and Lenahan, J.K. (eds) Protein Traffic in Parasite and Mammalian Cells. Proceedings of an International Workshop. ILRAD, Nairobi, pp. 127–129. ole-MoiYoi, O.K., Nayar, A., Iams, K.P., Musoke, A.J. and Yilma, T. (1989) Molecular aspects of Theileria parva and approaches to vaccine development for animals. Annals of the New York Academy of Sci- ences 569, 174–182. ole-MoiYoi, O.K., Brown, W.C., Iams, K.P., Nayar, A. and Macklin, M.D. (1993) Evidence for the induction of casein kinase II in bovine lymphocytes transformed by the intracellular protozoan parasite Theileria parva. EMBO Journal 12, 1621–1631. ole-MoiYoi, O.K., Sugimoto, C., Conrad, P.A. and Macklin, M.D. (1992) Cloning and characterization of the casein kinase II alpha subunit gene from the lymphocyte-transforming intracellular protozoan parasite Theileria parva. Biochemistry 31, 6193–6202. Oura, C.A.L., Bishop, R., Asiimwe, B.B., Spooner, P., Lubega, G.W. and Tait, A. (2007) Theileria parva live vaccination: parasite transmission, persistence and heterologous challenge in the field. Parasitology 134, 1205–1213. Oura, C.A.L., Bishop, R., Wampande, E.M., Lubega, G.W. and Tait, A. (2004) The persistence of component Theileria parva stocks in cattle immunized with the ‘Muguga cocktail’ live vaccine against East Coast fever in Uganda. Parasitology 129, 27–42. Oura, C.A.L., Odongo, D.O., Lubega, G.W., Spooner, P.R., Tait, A. and Bishop, R.P. (2003) A panel of microsatellite and minisatellite markers for the characterisation of field isolates of Theileria parva. International Journal for Parasitology 33, 1641–1653. Pain, A., Renauld, H., Berriman, M., Murphy, L., Yeats, C.A., et al. (2005) Genome of the host-cell trans- forming parasite Theileria annulata compared with T. parva. Science 309, 131–133. Patel, E., Mwaura, S., Kiara, H., Morzaria, S., Peters, A. and Toye, P. (2016) Production and dose deter- mination of the Infection and Treatment Method (ITM) Muguga cocktail vaccine used to control East Coast fever in cattle. Ticks and Tick-Borne Diseases 15, 306–314. Patel, E.H., Lubembe, D.M., Gachanja, J., Mwaura, S., Spooner, P. and Toye, P. (2011) Molecular char- acterization of live Theileria parva sporozoite vaccine stabilates reveals extensive genotypic diversity. Veterinary Parasitology 179, 62–68. Patel, E.H., Mwaura, S., Di Giulio, G., Cook, E.A.J., Lynen, G. and Toye, P.G. (2019) Infection and treatment method (ITM) vaccine against East Coast fever: reducing the number of doses per straw for use in smallholder dairy herds by thawing, diluting and refreezing already packaged vaccine. BMC Veterinary Research 15, 46. Pearson, T.W., Lundin, L.B., Dolan, T.T. and Stagg, D.A. (1979) Cell-mediated immunity to Theileria- transformed cell lines. Nature 281, 678–680. Pellé, R., Graham, S.P., Njahira, M.N., Osaso, J., Saya, R.M., et al. (2011) Two Theileria parva CD8 T cell antigen genes are more variable in buffalo than cattle parasites, but differ in pattern of sequence diversity. PLoS One 6, e19015. Perry, B.D. (2016) The control of East Coast fever of cattle by live parasite vaccination: a science-to-impact narrative. One Health 2, 103–114. Perry, B.D., Mukhebi, A.W., Norval, R.A.I. and Barrett, J.C. (1990) A preliminary assessment of current and alternative tick and tick-borne disease control strategies in Zimbabwe. ILRAD, Nairobi. Pinder, M. and Hewitt, R.S. (1980) Monoclonal antibodies detect antigenic diversity in Theileria parva para- sites. Journal of Immunology 124, 1000–1001. Radley, D.E., Brown, C.G.D., Cunningham, M.P., Kimber, C.D., Musisi, F.L., et al. (1975) East Coast fever: 3. Chemoprophylactic immunization of cattle using oxytetracycline and a combination of Theilerial strains. Veterinary Parasitology 1, 51–60. 272 P. Toye et al. Radley, D.E., Young, A.S., Grootenhuis, J.G., Cunningham, M.P., Dolan, T.T. and Morzaria, S.P. (1979) Further studies on the immunization of cattle against Theileria lawrencei by infection and chemo- prophylaxis. Veterinary Parasitology 5, 117–128. Rich, K.M. and Niemi, J. (2017) Economic impact of a new disease: same impact in developed and developing coun- tries? Revue Scientifique et Technique de l’Office International des Epizooties 36, 115–124. Rich, K.M. and Perry, B.D. (2011) The economic and poverty impacts of animal diseases in developing countries: new roles, new demands for economics and epidemiology. Preventive Veterinary Medicine 101, 133–147. Rich, K.M. and Wanyoike, F. (2010) An assessment of the regional and national socio-economic impacts of the 2007 Rift Valley fever outbreak in Kenya. American Journal of Tropical Medicine and Hygiene 83, 52–57. Rich, K.M., Roland-Holst, D. and Otte, J. (2014) An assessment of the ex-post socio-economic impacts of global rinderpest eradication: methodological issues and applications to rinderpest control programs in Chad and India. Food Policy 44, 248–261. Robinson, S., Mason-d’Croz, D., Islam, S., Suler, T.B., Robertson, R., et al. (2015) The International Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT): model description for version 3. IFPRI Discussion Paper 01483. IFPRI, Washington, DC. Rosegrant, M.W., Tokgoz, S. and Bhandary, P. (2013) The new normal? A tighter global agricultural supply and demand relation and its implications for food security. American Journal of Agricultural Economics 95, 303–309. Seder, R.A., Chang, L.J., Enama, M.E., Zephir, K.L., Sarwar, U.N., et al. (2013) Protection against malaria by intravenous immunization with a nonreplicating sporozoite vaccine. Science 341, 1359–1365. Shapiro, S.Z., Fujisaki, K., Morzaria, S.P., Webster, P., Fujinaga, T., et al. (1987) A life-cycle stage-specific antigen of Theileria parva recognized by anti-macroschizont monoclonal antibodies. Parasitology 94, 29–37. Sitt, T., Poole, E.J., Ndambuki, G., Mwaura, S., Njoroge, T., et al. (2015) Exposure of vaccinated and naive cattle to natural challenge from buffalo-derived Theileria parva. International Journal for Parasitology: Parasites and Wildlife 4, 244–251. Spooner, P.R. (1990) The effects of oxytetracycline on Theileria parva in vitro. Parasitology 100, 11–17. Stobbs, T.H. (1966) The introduction of Boran cattle into an E.C.F. endemic area. East African Agricultural and Forestry Journal 31, 298–304. Swai, E.S., Karimuribo, E.D., Kambarage, D.M. and Moshy, W.E. (2009) A longitudinal study on morbidity and mortality in young stock smallholder dairy cattle with special reference to tick borne infections in Tanga region, Tanzania. Veterinary Parasitology 160, 34–42. Tebele, N., Skilton, R.A., Katende, J., Wells, C.W., Nene, V., et al. (2000) Cloning, characterization, and expression of a 200-kilodalton diagnostic antigen of Babesia bigemina. Journal of Clinical Microbiol- ogy 38, 2240–2247. Toye, P.G., MacHugh, N.D., Bensaid, A.M., Alberti, S., Teale, A.M. and Morrison, W.I. (1990) Transfection into mouse L cells of genes encoding two serologically and functionally distinct bovine class I MHC molecules from a MHC-homozygous animal: evidence for a second class I locus in cattle. Immunology 70, 20–26. Toye, P.G., Goddeeris, B.M., Iams, K., Musoke, A.J. and Morrison, W.I. (1991) Characterization of a poly- morphic immunodominant molecule in sporozoites and schizonts of Theileria parva. Parasite Immun- ology 13, 49–62. Toye, P., Gobright, E., Nyanjui, J., Nene, V. and Bishop, R. (1995a) Structure and sequence variation of the genes encoding the polymorphic, immunodominant molecule (PIM), an antigen of Theileria parva recognized by inhibitory monoclonal antibodies. Molecular and Biochemical Parasitology 73, 165–177. Toye, P.G., Metzelaar, M.J., Wijngaard, P.L.J., Nene, V., Iams, K.P., et al. (1995b) Characterization of the gene encoding the polymorphic immunodominant molecule (PIM), a neutralizing antigen of Theileria parva. Journal of Immunology 155, 1370–1381. Toye, P., Wijngaard, P., MacHugh, N. and Clevers, H. (1995c) An assay for the identification of antigens recognized by cytotoxic T cells based on transient transfection of COS cells. Journal of Immuno- logical Methods 187, 95–101. Toye, P., Handel, I., Gray, J., Kiara, H., Thumbi, S., et al. (2013) Maternal antibody uptake, duration and influence on survival and growth rate in a cohort of indigenous calves in a smallholder farming system in western Kenya. Veterinary Immunology and Immunopathology 155, 129–134. The Management and Economics of East Coast Fever 273 Wagner, G.G., Duffus, W.P.H. and Burridge, M.J. (1974) The specific immunoglobulin response in cattle i mmunized with isolated Theileria parva antigens. Parasitology 69, 43–53. Young, A.S., Brown, C.G.D., Burridge, M.J., Grootenhuis, J.G., Kanhai, G.K., et al. (1978) The incidence of theilerial parasites in East African buffalo (Syncerus caffer). Tropenmedizin und Parasitologie 29, 281–289. Young, A.S., Groocock, C.M. and Kariuki, D.P. (1988) Integrated control of ticks and tick-borne diseases of cattle in Africa, Parasitology 96, 403–432. 7 Transboundary Animal Diseases Delia Grace1, Tadelle Dessie2, Michel Dione3, Henry Kiara4, Anne Liljander5, Jeff Mariner6,, Jan Naessens7, Edward Okoth4, Ekta Patel4, Lucilla Stienna4, Phil Toye4 and Barbara Wieland2 1International Livestock Research Institute, Nairobi, Kenya and University of Greenwich, UK; 2International Livestock Research Institute, Addis Ababa, Ethiopia; 3International Livestock Research Institute, Dakar, Senegal; 4International Livestock Research Institute, Nairobi, Kenya; 5EUROIMMUN s AG, Lübeck, Germany; 6Cummings School of Veterinary Medicine at Tufts University, Grafton, Massachusetts, USA. 7De Haan, Belgium Contents Executive Summary 275 ILRI’s contribution in the global context 275 Impacts of ILRI’s research 275 Scientific impacts 275 Development impacts 276 Policy impacts 276 Capacity building 276 African Swine Fever 276 Research 277 Research impact: technologies 278 Diagnostics 278 Vaccines 278 Research impact: molecular epidemiology 278 Research impact: field epidemiology and control 279 Transmission dynamics 279 Risk factors 279 Control 279 Research impact: socio-economic studies 280 Mycoplasma Disease 280 Research 281 Research impact: technologies 281 Diagnostics 281 Vaccines 281 Therapeutics 283 Research impact: molecular epidemiology 283 Host-mycoplasma interactions 283 Research impact: field epidemiology and control 283 Modelling control strategies 283 Epidemiological surveys 284 Vaccine trial 284 © International Livestock Research Institute 2020. The Impact of the International 274 Livestock Research Institute (eds J. McIntire and D. Grace) Transboundary Animal Diseases 275 Research impact: socio-economics 284 Demand for vaccines 284 Vaccine delivery systems 285 Policy analysis 285 Peste des Petits Ruminants 285 Research 285 Research impact: technologies 286 Research impact: field epidemiology and control 286 Vaccination strategies 286 Global eradication 287 Newcastle Disease 287 Research 287 Research impact: technologies 287 Research impact: field epidemiology and control 287 Research impact: socio-economic studies 288 Vaccine adoption 288 Other Livestock Transboundary Diseases 288 Rinderpest 288 Foot-and-mouth disease 289 Classical swine fever 289 Porcine reproductive and respiratory syndrome 290 Lumpy skin disease 290 Infectious bursal disease 290 Transboundary Animal Diseases in Systems 290 Conclusion and Future Directions 293 References 294 Executive Summary ILRI’s contribution in the global context Transboundary animal diseases (TADs) are highly contagious epidemics with the potential for very ILRI has been involved in ASF research since rapid spread, causing serious economic and some- 2003. Originally, the focus was on diagnostics times public health consequences while threat- and m olecular epidemiology. A large project ening farmers’ livelihoods. TADs often cause high in western Kenya broadened this by introducing morbidity and mortality in susceptible animal a multidisciplinary approach. With the start of populations. Some TADs are also emerging in- the CGIAR Research Programme (CRP) on Live- fectious diseases, food-borne diseases and/or stock and Fish in 2012, a major research-to- zoonoses: these are covered in other chapters. development initiative started in Uganda. From This chapter covers those high-impact, highly 2008, ILRI has been a key player in conducting contagious animal diseases, such as foot-and- research on the mycoplasma diseases: first, CBPP, mouth disease (FMD), that do not infect humans and, more recently, CCPP. This research agenda but do affect food and nutrition s ecurity and started in response to the African Union’s listing trade that the International Livestock Research of CBPP as the most economically important Institute (ILRI) has been working on since the transboundary livestock disease on the African 1990s. These are: African swine fever (ASF), continent in the post-rinderpest era. ILRI has mycoplasma disease (both contagious bovine been significantly involved in PPR research since pleuropneumonia (CBPP) and contagious 2010, with the primary aim of developing caprine pleuropneumonia (CCPP), peste des appropriate and proven PPR vaccination strat- petits ruminants (PPR) and Newcastle disease egies that can progressively control the disease (ND). Other TADs, which were to a lesser degree in developing countries. Finally, ILRI has made the focus of ILRI research, are briefly mentioned small but strategic inputs to the understanding (including FMD, classical swine fever (CSF) and and control of rinderpest, FMD, CSF, ND and rinderpest). some other TADs. 276 D. Grace et al. Impacts of ILRI’s research Development impacts Mycoplasma studies at ILRI have been under- Scientific impacts taken since the mid-2000s, but the initial focus Molecular epidemiology highlights include: the was on upstream research, which can be expected isolation and genetic characterization of CBPP to bring about development impact several dec- and ASF and other pathogens; a better under- ades later. ASF research became more develop- standing of disease dynamics; and improved ment oriented with the start of the CRP on Livestock understanding of immunity in local African and Fish, and end-user benefits are starting to be breeds and European breeds. documented but not yet on a large scale. Already, Technical research contributions to vaccine ILRI evidence has helped national partners to development for CBPP, CCPP, ASF and PPR have better target ASF surveillance and response in been developing a new challenge model for Uganda, safeguarding smallholders’ stock. A CCPP for vaccine development, improved diag- thermostable PPR vaccine developed by ILRI is nostic tests for CBPP and ASF, and construction under production in Mali and Kenya and will play of biological components using synthetic biol- a major role in global control efforts as the vaccine ogy to identify vaccine candidates. does not require a cold chain. ILRI made a small con- Field epidemiology research highlights tribution to generating the evidence that motiv- include: a better understanding of disease ated the global eradication of rinderpest, one of the dynamics by mathematical modelling of ASF, greatest successes of global animal disease control. CBPP and rinderpest and also using social networks to understand ASF transmission; Policy impacts investigating reservoirs of ASF and rinderpest ILRI, the Food and Agriculture Organization and the role of carriers in ASF; conducting of the United Nations (FAO) and the African large-scale studies on the prevalence and risk Union-Interafrican Bureau for Animal Resources factors for ASF, showing the importance of this (AU-IBAR) have been collaborating since 2013 disease; contributing to a major gap analysis of to develop an Africa-wide strategy for the pre- FMD research, shaping the path of FMD con- vention and control of ASF (FAO/AU-IBAR/ILRI, trol; and a large vaccine trial of an improved 2017). ILRI evidence has stimulated the local mycoplasma vaccine, which did not demon- government in Uganda to invest in ASF control. strate superiority but did show the feasibility of In 2015, FAO and the World Organization CBPP control. ILRI and partners pioneered for Animal Health (OIE) launched an international the first attempt to describe the entire disease initiative for the progressive control of PPR burden of any naturally occurring animal popu- and its global eradication by 2030. ILRI hosted lation in the world by a landmark study in the second meeting of the Global PPR Research which a cohort of several hundred calves were Alliance (GPRA), and ILRI scientists contributed followed for a year and assessed for more than to the development of the first pan-African 100 pathogens. strategy for the progressive control of PPR. Socio-economic research highlights include: ILRI developed a policy analysis for the im- extending knowledge, attitude and practice plementation of CBPP control strategies in pas- related to TADs; understanding the economic toral regions of sub-Saharan Africa. ILRI has factors in the adoption of the CBPP and the ND also contributed to a major gap analysis for the vaccine; the first ex ante assessment of the costs Global Foot-and-Mouth Disease Research Alli- and benefits of ASF control through improving ance (GFRA) and to the Global African Swine biosecurity; developing estimates for the eco- Fever Research Alliance (GASFRA) and along nomic benefits of rinderpest, PPR and CBPP with FAO and the African Union helped develop control that influenced investment; challenging a regional strategy for ASF control. conventional wisdom that training and aware- ness were enough to motivate disease control by Capacity building farmers and traders; and incorporating gender, as a measure of inequality, into the analysis Several dozen graduate fellows have partici- of the distributional effects of most TADs (see pated in this research. Farmers, butchers, gov- Chapter 18, this volume). ernment veterinarians and scientists in Uganda Transboundary Animal Diseases 277 have benefitted from training. One manual on (FLI) in Germany started, aiming to develop an ASF has been developed and is widely being used ASF vaccine, led by Richard Bishop. by the private sector in training farmers. The capacity for diagnosis was used between 2006 and 2008 to further train laboratory African Swine Fever technicians and field epidemiologists in Kenya, Uganda, Tanzania, Rwanda and Burundi. Sub- sequently, diagnostic tests have been validated African swine fever (ASF) is a haemorrhagic viral with national veterinary services in much of disease in pigs resulting in mortality rates approach- East Africa. ILRI and the other organizations ing 100% (Steinaa et al., 2016). ASF is native to made available both facilities and material from southern and eastern Africa, where it was histor- recent outbreaks of ASF in the eastern Africa ically maintained in a wildlife (sylvatic) cycle involv- region for this purpose. The work focused on ing warthogs, the natural reservoir, and soft ticks. extensive studies of ASF outbreaks and sylvatic It first came to attention when susceptible exotic (wild pig-based) cycles in Kenya and Uganda. pig breeds were introduced in the early 1900s to ILRI was also a partner in a Swedish-led eastern Africa. It spread within Africa and, in the consortium to assess the impact of ASF in Uganda, 1950s, to Europe and subsequently to South Amer- starting in 2010. This project conducted out- ica and the Caribbean (Costard et al., 2009). In break investigations and also investigated the the 1990s, the disease was eradicated from Europe, role of bush pigs in transmission. It was fol- except in Sardinia (Galindo and Alonso, 2017). lowed by another Swedish-led initiative, which However, in 2007, the disease expanded once focused more on interventions. more out of Africa into the Caucasus (Georgia). From 2012, the Biosciences eastern and ASF was recently confirmed in China (Wang et al., central Africa-ILRI Hub (BecA-ILRI Hub), in 2018) and in Vietnam and Cambodia. partnership with the Commonwealth Scientific Unlike many other viral pathogens, ASF is and Industrial Research Organisation (CSIRO), remarkably stable, surviving in pork, pig waste conducted studies in the border region of Kenya and the environment. No drugs or vaccines exist and Uganda. This project represented the most for treating or preventing ASF. Control relies comprehensive study of ASF disease yet attempted. on biosecurity and surveillance, diagnosis and Using the biosciences capacity for diagnosis and slaughter, all difficult to apply given the persist- epidemiological surveillance at the BecA-ILRI Hub ence of the virus. Limited diagnostic capacities and epidemiological research incorporating and poor knowledge about the epidemiology of participatory and quantitative approaches, state- the ASF virus have long hindered its control. of-the-art diagnostics, molecular biology, genetics and genomics, mathematical modelling and social network studies enabled the development of Research evidence-b ased recommendations for disease mitigation. There were important advances in im- ILRI research on ASF began in 2003 through munology of ASF, and two potential vaccine candi- collaboration with the European Union Refer- dates were produced by deleting a gene in two ence Laboratory for ASF diagnostics at the Centre different ASF viruses – the Kenyan 1033 strain for Animal Health Research (CISA), National In- (genotype IX) and the Sardinia strain (genotype I). stitute for Agricultural and Food Research and In 2011, the CRP on Livestock and Fish Technology, Spain. This worked towards the de- identified the smallholder pig value chain in velopment of improved diagnostic tests and also Uganda as a high-potential target to translate assessed viral prevalence and molecular diversity research into development interventions (Ouma in East and Central Africa using assays devel- et al., 2015). ASF was identified by the farmers oped at CISA, as reported by Bishop et al. (2015), and pig value chain stakeholders as the major Gallardo et  al. (2009, 2011, 2013) and Okoth constraint to smallholder pig farming (Dione et al. (2013). The CISA provided staff expertise, et  al., 2014). This background analysis gener- intellectual input, diagnostic reagents and use ated information on the epidemiology of ASF of their Biosafety Level 3 laboratory for viral cul- and led to testing of ‘best-bet’ interventions to ture and in vivo infections of swine. In 2004, a manage ASF and other pig diseases (Dione et al., collaboration with the Friedrich-Loeffler-Institut 2018a,b). 278 D. Grace et al. In collaboration with Chinese researchers, isolates. Immune responses were characterized ILRI has been researching genomic character- by very low antibody titres but solid cellular im- ization of domestic pigs and wild boars in Asia mune responses (Steinaa et al., 2016). and the genetic resistance of domestic pigs and Another advance was sequencing MHC wild boars to ASF. With the spread of ASF into molecules from Kenyan pigs to identify which Asia, ILRI has been providing support to coun- antigens can be used for a vaccine that induces tries in the region. cellular immunity. This revealed new sequences (Sørensen et al., 2017); ILRI has received sequences from approximately 34 other Kenyan pigs, which Research impact: technologies are currently being evaluated. Of particular im- portance was the achievement of generating two Diagnostics vaccine candidates by deleting a gene, which should attenuate the virus. Recently, cutting- Reliable diagnostics are key to the rapid contain- edge CRISPR/Cas (clustered regularly interspaced ment and management of disease outbreaks. As short palindromic repeats and CRISPR-associated such, ILRI contributed to validation of diagnos- protein 9) gene-editing technology and synthetic tic tests developed by their international partner, biology approaches for generating ASF vaccine CISA. However, samples taken from outbreak areas, candidates have been introduced to the ILRI ASF surprisingly, did not test positive for ASF anti- research programme. Using CRISPR/Cas9 is bodies using the OIE protocols (Gallardo et  al., faster, cheaper, more accurate and more efficient 2013; Okoth et al., 2013). Analysis of blood and than other existing genome-editing methods. serum samples using a polymerase chain reac- tion (PCR) assay found positivity to ASF virus (ASFV) of 28% in two independent samplings in Research impact: molecular epidemiology south-western Kenya and 0% PCR positivity in central Kenya, but no animals were seropositive ILRI used molecular epidemiology to track vir- in either study site using the OIE indirect ELISA, uses causing outbreaks, helping to understand and none of the animals sampled exhibited clin- the spatial and temporal relationships among ical symptoms of ASF. Failure of the OIE protocol them. Complete sequencing of the p54 gene might be related to the characteristics of African from ASFV isolates revealed regional differences pigs (Gallardo et al., 2013), opening opportunities and the value of p54 gene sequencing as an add- for further research on host-pathogen interactions. itional, intermediate-resolution, molecular epi- ILRI scientists were part of a Swedish-led demiological tool for typing of ASFV (Gallardo team that evaluated a rapid diagnostic approach et  al., 2009). Whole-genome phylogenetics, in- using a portable, commercial real-time PCR (Zsak cluding a newly sequenced virulent isolate from et  al., 2005). Trials suggested that it could be Spain, identified two clusters. One contained used effectively at the pen side or in a field labora- South African isolates from ticks and warthog, tory with performance at a level comparable suggesting derivation from a sylvatic (wildlife- to sophisticated molecular laboratories (Leblanc to-pig) transmission cycle. The second contained et  al., 2013). This was subsequently confirmed isolates from West Africa and the Iberian Penin- in Uganda (Liu et al., 2016). sula, suggesting a domestic (pig-to-pig) transmis- sion cycle. This provides valuable insights into Vaccines control, as disease maintained in a sylvatic cycle is harder to control than disease maintained in a In collaboration with FLI, ILRI studied the immune domestic cycle. Comparative genomics revealed responses to African ASFV strains and generated high diversity within a limited sample of the ASFV additional knowledge of the Kenyan pig major gene pool (de Villers et al., 2010) but revealed that histocompatibility complex (MHC)1. Using an genetically similar ASFVs may be circulating be- experimental animal model with an isolated tween Kenya and Uganda (Gallardo et al., 2011). virulent virus from Kenya, it was found that im- Scientists later enriched the publicly avail- munity could be obtained by increasing doses of able ASFV genome bank with sequences from East the virulent strain. This method is now being Africa. Genome sequencing and annotation of used for studying the immune response to such a recent pig-derived p72 genotype IX and a Transboundary Animal Diseases 279 tick-derived genotype X isolate from Kenya were Kenya. The data from this and earlier studies carried out using the Illumina platform and this suggest that there has been transfer of viruses of was compared with a Kenya 1950 isolate. The at least two different p72 genotypes from wild to three genomes constituted a cluster that was domestic pigs in East Africa (Gallardo et al., 2011). phylogenetically distinct from other ASFV genomes Although warthogs are considered the main but 98–99% conserved within the group. There were wild vertebrate host of the virus in the endemic multiple differences among East African genomes African setting, they are not the only wild African in the 360 and 110 multi-copy gene families pigs with a potential role in ASF epidemiology. The (Bishop et  al., 2015). This information is not bush pig, Potamochoerus larvatus, is an elusive, only important in tracking movement of the vir- nocturnal pig known to be susceptible to ASF, and uses but also helps with recognition of new vir- might be a link between the sylvatic and domestic uses being introduced into East Africa, indicating cycle. Studies from south-west Kenya showed a breach in transboundary disease control. A similarity in viruses in bush pigs and domestic pigs practical implication of the genetic similarity of (Okoth et al., 2013), indicating possible transmis- the Kenyan and Ugandan viral isolates is that sion between the two species or the presence of ASF control requires a regional approach to con- an intermediary vector. Initial results following trol. There was also a classification of the ASFV a bush pig with a radio collar revealed close genome series of multi-gene families, with the interactions between the species (Ståhl et  al., goal of providing standard comparisons and 2014). Other studies have reported an apparent naming schemes, thereby enhancing the cap- strong association of viral infections with pig acity to search for novel vaccine targets in ASFV. breeds, suggesting that some breeds may be resist- ant to infection and offering opportunities for genetic resistance as a disease mitigation meas- Research impact: field epidemiology ure as well as encouraging the conservation of and control tolerant breeds (Mujibi et al., 2018). Participatory epidemiology studies in three Transmission dynamics districts of central Uganda found that farmers considered ASF and parasites to be the major Understanding transmission dynamics is needed health constraint to pig production. They also before control is feasible. Recent findings that reported ASF as the primary cause of pig mor- high levels of detection of ASFV DNA in pigs tality, with epidemics occurring mainly during slaughtered in Kenya during a period with no the dry season (Dione et al., 2014). Other studies reported outbreaks provided support for the hy- have revealed widespread under-reporting of pothesis that subclinical, chronically infected or ASF by farmers, traders and animal healthcare recovered pigs may be responsible for persistence professionals and the motivations for this (Ath- of the virus in endemic areas (Thomas et  al., erstone et al., 2019). 2016). Other findings indicated that carrier pigs may play a role in ASFV maintenance and help Risk factors explain the disease outbreaks that have occurred without any evidence of any of the known Understanding risk factors can help target con- transmission sources (pigs clinically ill with ASF trol efforts to where they will have most effect. or adjacent populations of resistant African wild ILRI scientists have explored risk factors for ASF pigs) (Abworo et al., 2017). These findings have in Uganda and Kenya (Dione et al., 2015; Nanti- significantly increased scientific knowledge of ma et al., 2015). Key drivers of outbreaks were the epidemiology of ASF in the field in Africa, panic sales of pigs and inadequate disposal of which has contributed to the design of effective dead pigs. In an innovative study, ASF transmis- surveillance and control strategies. sion paths and nodes were described using social The role of the ancestral sylvatic cycle of network analysis (Lichoti et  al., 2016, 2017). ASFV was not well understood in the endemic This showed the importance of commerce in areas of eastern Africa. Scientists therefore ex- spreading ASFV between farms and across re- plored for the first time the coexistence of differ- gions, implying that greater emphasis should be ent ASFV genotypes in the soft ticks found in placed on post-farm nodes in the prevention and warthog burrows and adult wild warthogs in control of the disease (Dione et al., 2016a). 280 D. Grace et al. Control Research impact: socio-economic studies By combining knowledge and data from the vari- Scientists also explored the knowledge, attitudes, ous studies, it was possible to estimate disease practices, capacities and incentives of pig value transmission dynamics using geospatial mapping chain actors in relation to ASF and biosecurity and mathematical modelling. This was used for (Chenais, 2016). This showed that respondents ex ante assessment of the effectiveness of different were well aware of the clinical signs of ASF, the ASF control strategies and to identify the optimum routes for disease spread and the measures for time for deployment of interventions in the field disease control. However, awareness of the con- to minimize the losses following ASF outbreaks trol measures did not guarantee their implemen- (Barongo et  al., 2015, 2016). This provided evi- tation. A majority of middlemen and butchers dence to help national and international partners acknowledged having sold live pigs, carcasses or conduct targeted surveillance of pig diseases, in- pork that they believed was infected with ASF. cluding important zoonoses (Dione et al., 2017). Factors that limit the adoption of biosecurity The evidence generated has informed the devel- measures by farmers include cost and cultural opment of an ASF control strategy for Africa factors such as the stigma related to the use of (FAO/AU-IBAR/ILRI, 2017). footbaths and the restriction of farm visits to In the absence of a vaccine, the most com- neighbours or traders. mon recommendation for ASF control has been The results of an ex ante model projected that, to implement biosecurity measures such as foot- although biosecurity measures would reduce ASF baths, fencing and quarantine. Such measures are outbreaks, they would also lead to a 6.2% reduc- usually promoted through training and public tion in profit per year while giving a 7.8 % increase information campaigns. A randomized control in profits accruing to butchers, traders, collectors trial was carried out to evaluate the effects of and wholesalers (Ouma et al., 2018). This could training farmers in biosecurity on the Kenya- explain the low adoption of biosecurity practices Uganda border. The trial found that, although by farmers and a need for other incentives for farmer knowledge improved after the trial, farmer farmers (Nantima et al., 2016). practices did not (Nantima et al., 2016). Another Other studies assessed the gender dimensions attempt in Bangladesh (with poultry) to promote of pig management and disease control. These biosecurity through training and public informa- found that, during disease outbreaks, especially tion also indicated that information was insuffi- of ASF, both men and women provided animal cient to change behaviour and that additional health care. Therefore, control information should motivation was needed (Rimi et al., 2016). explicitly target both men and women within the ILRI was a member of GASFRA. This organ- same household. This broader outreach would ized four scientific conferences from 2013 to help spread knowledge of pig husbandry and 2016 to conduct gap analyses of current know- ensure that action during outbreaks does not rely ledge and the available countermeasures to on a few individuals (Dione et al., 2016b,c). effectively control and mitigate the impact of a disease outbreak of ASF. Based on these analyses, a report was generated setting out the priority Mycoplasma Disease research needs (Seixas et al., 2018). ILRI, in partnership with FAO and the Mycoplasmas are the smallest and simplest African Union, elaborated a strategy joined self-replicating bacteria. Several species of myco- with an action plan to allow a progressive and plasma infect humans, causing pneumonia, coordinated control of ASF at the regional urinary tract infections and sexually transmit- level. To achieve this objective, it prioritized ted disease. Hundreds more species infect ani- the strengthening of capacities of technical mals. CBPP is an infectious disease of cattle services and the improvement of current pro- caused by the small-colony type of Mycoplasma duction systems, creating optimal conditions mycoides; CCPP is a devastating disease of goats for the modernization and development of the caused by Mycoplasma capricolum. pig industry in a healthy context (FAO/AU- CBPP was introduced to Africa in the 1800s IBAR/ILRI, 2017). and is now one of the most important livestock Transboundary Animal Diseases 281 diseases in sub-Saharan Africa. It is now esti- official (OIE) prescribed assays (complement fix- mated to menace the livelihoods of nearly 25 ation test and competitive ELISA) make them million people in 19 African countries. suitable for use at the herd level only (Nkando While eradication is the surest way to con- et al., 2012). This was confirmed by a compari- trol CBPP, it is expensive and can only be tried son of four serological assays: one complement during major outbreaks and on major trade fixation assay developed in house and three routes. Eradication is generally infeasible in Africa OIE-recommended tests (Schubert et al., 2011). because of animal movements and the limited None of the four tests detected all infected ani- resources of national veterinary services. The mals. ILRI scientists later contributed to the de- current control strategy relies on using a live velopment of a novel, successfully validated, vaccine. Unfortunately, this has limited efficacy, real-time PCR assay (Schnee et al., 2011). a short duration of immunity and causes severe ILRI scientists next set out to develop an im- side effects (including the loss of the animal’s proved serological test, seeking to find novel myco- tail). The current diagnostic tests have limited plasma antigens recognized by sera from infected sensitivity and are only useful at the herd level cattle, thus having diagnostic potential to improve and are not practical at individual levels. test sensitivity (Jores et  al., 2009; Naseem et  al., 2010). A systematic comparison of 17 selected Mycoplasma mycoides subsp. mycoides (Mmm) Research immunogens commenced, a standardized ELISA protocol was developed, and well-defined serum An ILRI review of animal health and poverty samples were used to compare individual proteins issues identified CBPP and CCPP as major prob- and protein combinations with respect to sensitiv- lems (Perry et al., 2002), although CBPP research ity and specificity (Heller et al., 2016). The result- had not begun until the late 1990s. At this time, ing assay, comprising the two best- performing Ethiopia was experiencing major new outbreaks. immunogens, had an overall diagnostic accuracy ILRI initially supported a thesis on the cost of comparable to the OIE-prescribed tests, and work CBPP control in Ethiopia (Laval, 1999). Scien- to optimize the test further is under way. tists subsequently estimated the presence and The assay was further transferred to a lat- prevalence of CBPP in the Ethiopian highlands eral flow test format in collaboration with a com- (Bonnet et al., 2005), defined the basic epidemio- mercial company, enabling rapid diagnosis (less logical parameters of the disease in southern than 30 min) of CBPP (Heller et al., 2016). The test Sudan (Mariner et  al., 2006a) and promised a is currently undergoing final optimization and control model in East Africa (Mariner et al., 2006b). evaluation. A rapid and field-applicable recom- Careful analysis of the existing literature on binase polymerase amplification assay was also CBPP suggested the key information and technolo- developed for a related pathogenic mycoplasma, gies needed to develop better control measures M. capricolum subsp. capripneumoniae (Mccp) that in an African environment (Jores et al., 2013b). could be further developed to make it commer- Research then focused on host-mycoplasma cial (Liljander et al., 2015). interactions, epidemiological models, improved diagnostics, elucidation of protective responses Vaccines and identification of potential vaccine antigens. An early study suggested a minor role for CD4+ T-lymphocytes, which are involved in helping and Research impact: technologies regulating immune responses, in protection in a primary infection (Jores et al., 2008), despite some Diagnostics earlier publications suggesting such a link. This was confirmed in a later study in cattle depleted In the absence of good vaccines, recurrent testing for CD4+ T-lymphocytes (Sacchini et  al., 2011). using improved diagnostic assays in combination Differences in disease severity are probably largely with elimination of CBPP-positive animals or due to differences in cattle genotypes. Inflamma- herds is a control option (Ssematimba et al., 2015). tory cytokines, as expected, were found during ILRI research found that the inaccuracy of the infections in lung tissues (Sterner-Kock et  al., 282 D. Grace et al. 2016) and in plasma (Sacchini et al., 2012). De- Jores et al., 2018). The polysaccharide molecules pletion of CD4+ T-cells did not significantly influ- were purified, coupled to a protein and adminis- ence cytokine levels, again suggesting their tered twice to cattle (Mwirigi et al., 2016b). After minor role in control of a primary infection. challenge, the severity of disease in the immunized Several approaches were followed to identify cattle was significantly reduced. It is interesting to protective antigens. Although antibody titres in know that carbohydrate can protect, and this is a primary infection did not seem to correlate in agreement with the hypothesis that preven- with a positive outcome for the animal (Schieck tion of adhesion to lung cells is protective, but et al., 2014), it was hypothesized that antibodies the polysaccharide coating would not be suitable protect in a secondary infection or after vaccin- for vaccine production, as it would be too expen- ation. ILRI scientists and colleagues contributed sive to produce. towards the characterization of the in vitro core Additional studies tested other potential surface proteome of cultured mycoplasma, thus pathogenic molecules for their capacity to protect identifying candidate Mmm antigens for the against disease. Two promising protein candi- development of improved control measures dates (Mulongo et al., 2013, 2015) were investi- (Krasteva et  al., 2014). Another study charac- gated. Neither protein induced protection: in terized proteins expressed in vivo by mycoplasma contrast, immunization with LppQ-N exacerbated obtained from pleural effusion (Weldearegay pathogenesis as a result of the formation of im- et al., 2015). mune complexes (Mulongo et  al., 2015). Im- Evidence was obtained by an ILRI student munization with the other candidate did induce that whole, inactive mycoplasma were protective good antibodies in cattle, but these antibodies (Mwirigi et al., 2016a). Immunizations were car- did not inhibit the enzymatic function of the en- ried out with the live vaccine and with two for- zyme, which may have been necessary for the mulations of inactivated Mmm. Heat-inactivated prevention of its pathogenicity (Mulongo et al., mycoplasma were found to be as protective as the 2013). These studies emphasize that care has to live vaccine. This confirmed that live mycoplasma be taken when selecting candidate vaccine anti- are not required for induction of immunity and gens, and that it is important to avoid including suggested that protection can be induced by antigens that do not protect or that counterpro- purified molecules, which are much preferable tect in a subunit vaccine (Jores et al., 2013b). as vaccines. A successful approach in a comprehensive Virulence factors are molecules produced project was aimed at identifying candidate vaccine by pathogens that allow them to infect hosts and molecules using reverse vaccinology. As the gen- evade the immune system and are currently the ome of Mmm was available, a bioinformatics focus of intense research. If vaccines could be study selected 66 proteins that were likely to be developed against virulence factors they could present on the mycoplasma membrane or were be helpful in preventing disease or reducing secreted and were therefore accessible to the im- severity. For Mmm, no virulence factors have mune system. ILRI contributed synthetic genes been confirmed in vivo. However, for the closely for 38 of these Mmm proteins and the remaining related goat mycoplasma, Mycoplasma mycoides were designed, and recombinant proteins pro- subsp. capri (Mmc), biology tools such as genome duced. The potential antigens were ranked on transplantation can be employed to gain insight their ability to elicit antibody responses, as tested into the function of genes. The whole genome of in sera from immune animals (Perez-Casal et al., Mmc is transferred into yeast cells, where it can be 2015). After prioritization, the recombinant modified using molecular tools available for yeast, proteins were then pooled into groups of five and the modified genome is then transplanted antigens and administered to cattle. After chal- back into mycoplasma cells and the resulting lenge, three of the formulations induced protec- mutant can be characterized. This approach was tion (Nkando et  al., 2016). This suggests that used to generate a mutant mycoplasma lacking more than one protein is protective, and further the polysaccharide coating (an outer layer com- research suggested that a cocktail of four pro- posed of carbohydrates), and in vivo experiments teins mixed with an adjuvant formed a good vac- demonstrated that the polysaccharide coating is cine. The Kenya Veterinary Vaccines Production indeed a virulence factor (Schieck et  al., 2016; Institute (KEVEVAPI) is currently acquiring the Transboundary Animal Diseases 283 capacity to produce and quality control the vac- (Mmc) to jump to cattle and differentiate into a cine. Studies have been planned to determine its pathogenic species. efficacy in the field and to determine how long immunity lasts. Host-mycoplasma interactions Although in laboratory studies the efficacy of the subunit vaccine was the same as the live Because Mmm causes a severe infection in the vaccine, it will still have a number of vital ad- lungs, one step in the disease process must be the vantages: a lower price, no need for a cold chain interaction of mycoplasma with cells in the lungs. (important to reach remote areas with little in- A study demonstrated strong attachment of cul- frastructure) and the absence of serious side ef- tured mycoplasma to in vitro-cultured bovine fects. This last condition is important to convince lung epithelial cells. While all cell types bound herders to vaccinate their animals. Another ad- mycoplasma to some degree, very high binding vantage of the subunit vaccine is that it may was only observed with lung or bronchial epi- allow the development of a test to discriminate thelial cells from cattle (Aye et al., 2015). How- vaccinated from infected animals, which will ever, when the epithelial cells were derived from greatly facilitate disease control. fetal lungs, no such binding was observed. This would explain why newborn and very young calves do not develop lung disease, as they may Therapeutics lack a receptor for mycoplasma attachment. As A study, initiated by the University of Nairobi and expected if strong attachment is an essential finalized at the BecA-ILRI Hub identified plants step in the disease process, the mycoplasma also used by local herders to treat lung infections in showed only strong specificity for bovine cells, their animals and tested extracts from these plants and not for cells from other species including sheep for their capacity to block in vitro mycoplasma and goat. Preventing this binding step may thus growth (Kama-Kama et al., 2016). Several had prevent disease. Monoclonal antibodies were activity, and further research identified at least created against Mmm, and a number that in- one chemical compound with bacteriostatic hibited attachment were selected. In one further activity for Mmm (Kama-Kama et al., 2017). study on these monoclonal antibodies, a myco- plasma molecule was identified as being involved in the adhesion process (Aye et  al., 2018) and Research impact: molecular epidemiology could be a target for vaccination. Further testing was not carried out, as, in the meantime, the re- ILRI contributed to the isolation of new myco- verse vaccinology approach had successfully plasma strains, as well as obtaining genome identified protective antigens (described earlier). sequencing of several strains, including Mmm (Fischer et al., 2015), Mccp (Falquet et al., 2014) Research impact: field epidemiology and Mycoplasma feriruminatoris sp. nov. (Fischer and control et  al., 2013; Jores et  al., 2013a). Additional important work was done by Gourgues and Barr Modelling control strategies (2016). ILRI scientists improved our knowledge of Treatment of affected cattle with antimicrobials the phylogenetic relationships and demographic has been officially discouraged on the basis that history of mycoplasma of the ‘mycoides cluster’, a it may favour the creation of chronic carriers, group of related mycoplasma that infect rumin- which are believed to be responsible for disease ants (Fischer et al., 2012). They used multi-locus spread. However, a simulation model developed sequence typing on seven housekeeping genes by ILRI and collaborators based on field data from over 120 strains. Interestingly, the origin of from Ethiopia found that antibiotics were the Mmm, the cause of CBPP, dates back approxi- most efficient strategy, suggesting that the use of mately 10,000 years, coinciding with the do- antimicrobials by smallholder farmers should be mestication of livestock. It is believed that the reconsidered (Lesnoff et al., 2004). tradition of keeping goats and cattle in close Jeffrey Mariner and colleagues also devel- proximity must have allowed a goat mycoplasma oped mathematical models assessing the impact 284 D. Grace et al. of alternative control measures on the transmis- the survey were not indicative of disease preva- sion dynamics of CBPP, incorporating field lence (Bonnet et al., 2005). These studies brought parameters from pastoral systems. The results to attention some of the complications in con- indicated that a control strategy based on the trolling CBPP. currently available vaccines alone would not be sufficient to eradicate CBPP unless the efficacy, Vaccine trial safety and duration of immunity could be sub- stantially improved. According to the models, The current CBPP control strategy in Africa the farmers would benefit from reduced disease is based on immunization with a live vaccine prevalence and mortality through vaccination (mainly strain T1/44). The vaccine confers im- of healthy animals combined with antimicrobial munity only for 6  months to 1  year and has treatment of clinical cases (Mariner et al., 2006b). o ccasionally shown severe side effects at the in- The models also suggested that, under the pre- oculation site: 3.8% of animals showed side ef- vailing inter-herd contact patterns, CBPP can be fects including loss of their tail (Kairu- Wanyoike maintained indefinitely, even in moderate-sized et  al., 2014b). The vaccine needs a cold chain herds (Mariner et al., 2006a,b). A recent model and must be applied within 1  h of reconstitu- predicted that CBPP could be eliminated in tion. These characteristics lead to low accept- approximately 2 years provided that 75% of the ance and willingness by the animal owners to animals in an isolated herd are vaccinated an- present their animals for vaccination. In add- nually (minimum vaccine protection required is ition, the vaccine is not widely available as the 18  months) and recommended that recurrent veterinary services regulate its use. testing be done using an improved diagnostic test ILRI and partners conducted a 3-year and elimination of positive animals (Ssematimba immunization trial using a modified vaccine for- et al., 2015). mulation – the commercially available T1/44 vaccine – while concurrently evaluating vaccine delivery systems. The programme concluded that Epidemiological surveys the modified vaccine did not result in improved When CBPP increased in the Ethiopian high- protection when deployed under field conditions lands, ILRI and partners responded by re- (Nkando et al., 2012). While stating that prop- searching the problem and solutions. At that erly delivered vaccination significantly reduced time, there was little information reported in the impact of CBPP in affected areas, the study the literature on within-herd spread of CBPP also highlighted the need for annual boosting for during outbreaks. In collaboration with the sustained protection. Centre de Coopération Internationale en Re- cherche Agronomique pour le Développement Research impact: socio-economics (CIRAD) and the National Animal Health Re- search Centre, a research programme was set Demand for vaccines up in a CBPP-infected zone in the Ethiopian highlands (Boji district in West Wellega Zone) ILRI scientists and partners used structured ques- to estimate the epidemiological parameters of tionnaires to assess farmer practices and percep- the disease and to assess the effects of different tions of CBPP control, the willingness of farmers disease-management strategies naturally im- to pay, their preferences regarding vaccination and plemented by the local farmers. an estimation of the impact of CBPP (Kairu- A longitudinal study was carried out, which Wanyoike et al., 2013, 2014a,b, 2017). Pastoral found that 34% of cattle became seropositive populations consider CBPP a threat to their cattle, over a 16-month observation period and 39% of but awareness of prevention control methods these became clinically ill, while 13% died (Les- varied. Many farmers were reluctant to vaccin- noff et  al., 2004). There was no evidence that ate their cattle due in part to adverse reactions. CBPP control measures used locally by farmers The need for a better vaccine was reflected in the (including treatment) had any benefit. A subse- low willingness of farmers to pay for the current quent study found that herd prevalence increased vaccine and vaccination. The willingness to pay with herd size and that clinical signs observed in was higher for a hypothetical preferred vaccine Transboundary Animal Diseases 285 and vaccine attributes such as the inclusion of a condition of goats in Nigeria in 1930 and was pH indicator (Kairu-Wanyoike et al., 2014a). first characterized in Côte d’Ivoire in 1942. Until the 1980s, it was seen as a problem of Vaccine delivery systems West Africa: since then, it has spread rapidly to around 70 countries in Africa, the Middle East CBPP control measures and disease perceptions and Asia. The disease re-entered China in 2015 were further studied from a gendered socio- and spread to the borders with Korea and Viet- economic perspective in north-eastern Kenya nam. In 2016, Mongolia reported its first case (Muindi et al., 2015). It was concluded that poor in livestock and a subsequent mass mortality road infrastructure represented the main obstacle event in wildlife, especially the Saiga antelope. for vaccine delivery in the study area and that In 2018, the disease was reported in the Euro- women, being better at recognizing the early clin- pean Union (Bulgaria). Kenya saw a series of ical signs of CBPP, could play an important role in epidemics in 2006–2008, killing almost 1.2 preventing disease spread by alerting the commu- million small ruminants and causing a decline nity and thus averting herd and trade losses by in milk production of 2.1 million litres (FAO, swift implementation of quarantine and ring vac- 2016). The mean annual global loss from PPR cinations (Waithanji et  al., 2015). Furthermore, mortality has been estimated at US$1.5 billion, the studies concluded that there is a market for ranging from US$0.8 billion to US$2.7 billion the implementation and adoption of an improved (Mariner et al., 2016). CBPP vaccine by both men and women if it is Safe, effective and inexpensive live vaccines, thermostable, more efficacious and safer than the based on the Nigeria 75/1 or Sungri strains, are current vaccine. Women also found affordability available and widely used. The Nigeria 75/1 an important attribute for a new vaccine (Muindi strain has been in use for decades without any et al., 2015). These types of study are well known adverse effects and has been shown to generate to overestimate demand and willingness to pay life-long immunity and to protect against all lin- and should be regarded as promising but incon- eages of PPR. The availability of these vaccines clusive in this context. combined with lessons learnt from the global eradication of rinderpest have inspired the inter- Policy analysis national animal health community to target PPR ILRI contributed to a policy analysis for the for global eradication by 2030. ILRI aims to con- implementation of CBPP control strategies in tribute to the global eradication effort while at pastoral regions of sub-Saharan Africa (Onono the same time providing evidence to inform and et al., 2017). This found that current vaccination guide it. was low and suggested the adoption of signed contractual agreements between the public and Research private sectors to support the vaccination of sus- ceptible herds raised in endemic regions. ILRI has been involved in PPR research since 2010 with the aim of developing vaccination Peste des Petits Ruminants strategies for developing countries. ILRI hosted the second meeting of GPRA and its scientists PPR, or ‘goat plague’, is a highly contagious dis- contributed to the development of the first pan- ease of wild and domestic sheep and goats (Tay- African strategy for the progressive control of lor, 1984; Robinson et al., 2011). The disease is PPR (Elsawalhy et al., 2010). Subsequently, ILRI caused by peste des petits ruminants virus, scientists contributed to a business analysis for genus Morbillivirus, which is closely related to the global eradication of PPR (Jones et al., 2016; rinderpest virus. PPR is a devastating disease in Mariner et  al., 2016), and remain closely en- naïve small ruminants. Outbreaks have occa- gaged with the FAO/OIE PPR Secretariat. ILRI sionally been observed in camels and some spe- scientists have been advocates for aggressive cies of wild Asian ungulates, but no clinical eradication of PPR through programmes that disease has been reported in African wildlife. emphasize surveillance, epidemiological analysis The disease was recognized as a ‘rinderpest-like’ and targeted vaccination. 286 D. Grace et al. Research impact: technologies scale lyophilizers are purpose-built machines, and each has a different configuration of compo- In southern Nigeria, ILCA developed and tested a nents that behaves uniquely. The lyophilization package consisting of a tissue-culture rinderpest process has to be optimized for each machine. vaccine and a dipping programme to control The result was a PPR vaccine that maintained the mange. They reported that, as a result, kid survival minimum required dose as a 25-dose presen- had greatly increased, reproductive efficiency of tation for over 5  months at 37°C and for over does improved, and monthly mortality declined by 10 days at 56°C. This level of thermostability is 87% in goats and by 79% in sheep (Adeoye, 1985). comparable to that achieved for the thermo- Research on PPR restarted in the 2000s stable rinderpest vaccine used in the global (Balamurugan et  al., 2014). ILRI scientists eradication of rinderpest (Mariner et al., 2017). demonstrated the feasibility of production of a Current efforts focus on producing sufficient thermostable vaccine for PPR, showing that ef- vaccine to enable pilot vaccination programmes. fective PPR vaccines based on the attenuated Ni- To this end, the Central Veterinary Laboratory in geria 75/1 strain could be thermostabilized using Mali and KEVEVAPI have started using the the protocol for the production of the thermo- protocol with technical support from ILRI and stable rinderpest vaccine developed by Tufts Uni- Tufts University scientists. The performance of versity and the US Department of Agriculture resulting batches will be assessed through the (USDA) (Mariner et al., 1990, 1991) and used in Pan African Veterinary Vaccine Centre (PANVAC) the Global Rinderpest Eradication Programme and its performance in the field and efficiency (GREP) (Mariner et  al., 2012). As the thermo- will be compared with conventional vaccine. stable rinderpest manufacturing process is unen- Thus, this vaccine will be an important control cumbered by intellectual property constraints and tool for remote areas, providing essential support could be used with the existing PPR vaccine pro- for global eradication efforts. duced in accordance with OIE norms, the vaccine In addition, ILRI has supported the testing was available for implementation in the field. of novel DIVA (differentiation between infected To identify the optimal process to achieve and vaccinated animals) vaccines using African thermostability, two main approaches were com- breeds, as ILRI is one of the few institutions pared, one based on the ThermoVax rinderpest where such tests can actually be done (Holzer vaccine developed by Tufts University and the et al., 2016). DIVA vaccines differentiate between USDA, which is free from intellectual property infected and vaccinated animals because the vac- constraints, the other based on the patented cines lack at least one antigenic protein that is pre- Xerovac process. Thermostability was assessed sent on the field virus. These vaccines are important in accelerated stability tests at a range of tem- for control because vaccination in poultry would peratures and using in vitro titration assays. The have greater worldwide acceptance if naturally ThermoVax method used in rinderpest vaccine infected and vaccinated-only animals could be manufacture was found to provide the best distinguished, and if culling were necessary then long-term thermostability at 37°C. These batches naturally infected animals could be targeted. consisted of existing PPR vaccine lyophilized in the same 72 h cycle that had been utilized in the production of thermostable rinderpest vaccine Research impact: field epidemiology (Mariner et al., 1991). Under OIE norms, changes and control to lyophilization procedures do not require reval- idation of the immunogenicity of the vaccine. Vaccination strategies The ILRI team concluded that the efficiency of the lyophilization cycle, a key variable that ILRI has been active in advancing approaches impacts the efficacy of chemical stabilization, using targeted vaccination as a more effective al- was the source of the enhanced stability, as this ternative to mass vaccination because it has was the only variable changed in the batches higher effectiveness and lower cost. ILRI has also found to have the highest levels of thermostabil- piloted more effective vaccination delivery systems ity. Lyophilization is an activity that combines based on public–private–community partnerships biological, chemical and physical sciences. Large- and quantity-based approaches to remuneration Transboundary Animal Diseases 287 for work done. One of the lessons of rinderpest is considered the most important TAD in village eradication was that mass vaccination was not a poultry. Periodic outbreaks result in high mortality successful approach for completing the eradication among free- ranging flocks and serve as a disin- process. The success of eradication depended on the centive for poultry keepers to invest time or re- development of approaches that identified popu- sources in their birds. Because of the difficulty of lations responsible for disease maintenance and attaining biosecurity in backyard poultry, vaccin- then focused vaccination on those critical com- ation is the preferred control strategy. Live vaccines munities. This finding was applied to PPR as ILRI have been widely used since the 1950s and have piloted vaccination activities in Karamoja, Uganda made great progress in preventing and controlling and Sudan. In Uganda, vaccination was imple- ND. Live lentogenic B1 and LaSota vaccine strains mented through community animal health workers of low virulence are commonly used worldwide for with costs covered by the public sector. In Sudan, protection. more conventional veterinary personnel were used, but the livestock owners covered the costs. Research Global eradication Poultry disease was not a focus of research dur- ing ILRI’s first decades, and for a number of There are four components in the global eradica- reasons, stakeholders did not prioritize poultry tion framework of PPR: (i) promoting an enabling health. Moreover, in Africa, poultry did not make environment and reinforcing veterinary capaci- important contributions to the diet, although ties through public information, updating legal that has changed due to imports and the estab- framework and preparing national and regional lishment of intensive production. Poultry dis- PPR plans; (ii) supporting diagnostic and surveil- ease first rose to prominence at ILRI because of lance systems, which includes assessment of the the highly pathogenic avian influenza (HPAI) epidemiological situation and strengthening of pandemic (see Chapter 8, this volume). Subse- regional epidemiology and laboratory networks; quently, a project investigated ND vaccine im- (iii) supporting PPR eradications and, if approxi- pact in East Africa, and further work studied a mately 1.5 billion sheep and goats need to be vac- range of poultry diseases in Ethiopia. cinated, controlling other small ruminant diseases in support of PPR eradication; and (iv) coordin- Research impact: technologies ation at country, regional and global levels. ILRI facilitated scientists who had played leadership roles in the global eradication of rin- Live vaccines can show low efficacy due to ad- derpest to assess key lessons from rinderpest ministration challenges or failure of birds to mount eradication and help lay the foundation for the a sufficient immune response. ILRI research in global eradication of PPR (Mariner et al., 2012). Ethiopia investigated the potential for improv- Social innovations in animal health institutions ing vaccination success in village chicken. ILRI were identified as key to capturing the benefits of conducted a study to compare the antibody technological developments such as the thermo- response to ND in intensively reared and back- stable rinderpest vaccine. This activity fitted well yard chicken. The latter mounted a much weaker with ILRI’s mandate in representing the inter- response. A follow-on experiment found that ests of developing countries and improving our veterinary treatment (especially worming) be- understanding of animal health institutions to fore vaccination could dramatically improve the mitigate the impact of disease. antibody response and hence the protection offered by the vaccine (Abera et al., 2017). Newcastle Disease Research impact: field epidemiology and control Newcastle disease (ND) is a highly contagious dis- ease of poultry and other birds caused by the New- In the early 2000s, desk research developed back- castle disease virus in the family Paramyxoviridae. It ground papers on ND in Ethiopia (Dessie and 288 D. Grace et al. Jobre, 2004) and southern Africa (McDermott While ND vaccines are readily available, uptake et al., 2001). ND was also addressed along with by poor farmers has been very limited. A project HPAI in South-east Asia as discussed in Chapter conducted a study in 2011 to investigate vac- 5 (this volume). cine adoption in Kenya and Tanzania in order to The first major research into poultry health understand the barriers and bridges to adoption in Ethiopia was launched in 2011. Its aim was to (Lindahl et al., 2019). In this study, two areas identify infectious diseases of Ethiopian village in Kenya and Tanzania were studied, where all poultry and to improve their control. ND was villages were eligible for government control identified by farmers as the highest priority dis- programmes, but some villages had received ease. A longitudinal study subsequently found additional support to get vaccination from a pro- that ND was responsible for the death of 843 ject (Tanzania) or non-governmental organiza- chickens of the cohort of 1358 birds (Jarso, 2015). tion (Kenya). Where vaccination support had Because of the lack of previous research, ILRI been given, 59% of households overall had used was also able to identify some diseases for the first vaccines against ND, which was significantly time. The poultry farm at the Ethiopian Institute of (p < 0.001) more than the 17% of households Agricultural Research (EIAR), Debre Zeit, repre- that had used the ND vaccine in areas with no sents an important centre for research, farmer additional support. However, many farmers training in poultry production and distribution of stopped using vaccines, and even those who did birds among smallholders. As such, it has vital use them often used them suboptimally and con- links with farmers, especially intensive and tinued to experience losses from ND. Despite this, semi-intensive producers. In 2012, respiratory dis- there were significantly fewer reported poultry ease was a major cause of morbidity and mortality. deaths in villages with support. This showed the ILRI used a combination of serological and mo- importance of additional support if vaccines are to lecular methods for the detection of pathogens, be taken up by poor farmers, and also the gap be- reporting for the first time variant infectious tween theoretical performance of disease control bronchitis virus (793B genotype), avian meta- and disease control in practice. Even with con- pneumovirus subtype B and Mycoplasma synoviae siderable external support, almost no farmers in poultry. This information was used for planning vaccinated their poultry in accordance with re- vaccination programmes (Hutton et al., 2017). commendations and losses from ND continued. The Ethiopian work later evaluated vaccine uptake. A contingent valuation method was Research impact: socio-economic studies conducted to elicit farmers’ willingness to pay for village poultry vaccine services. Two hypo- ILRI conducted an economic analysis in Nigeria thetical vaccine programmes were designed for and found that ND caused an estimated 25.5 mil- ND and Gumboro disease. The results showed lion poultry deaths each year nationwide, costing that farmers recognized the benefits of the vac- 8.9 billion naira. Control comprising vaccination, cine programme and that many would be will- sanitary measures and surveillance was estimated ing to pay for it (Terfa et al., 2015). to cost 1.8 billion naira per year (Fadiga et  al., 2011). Within this study, a related ex post analysis examined efforts to stop HPAI in Nigeria. A sto- chastic epidemiological model was used to para- Other Livestock Transboundary meterize counterfactuals of disease evolution with Diseases and without interventions. The results indicated that a programme of HPAI control versus a base- Rinderpest line of endemic, high-mortality HPAI gave a bene- fit:cost ratio of 1.75 over a 5-year period. Rinderpest is an infectious viral disease that has killed hundreds of millions of cattle over hun- Vaccine adoption dreds of years, often causing famine. Rinderpest was formally declared to have been eradicated in ND vaccines are widely used in intensive poultry 2011, becoming only the second disease (after production and are effective and inexpensive. smallpox) to have been eradicated. This is widely Transboundary Animal Diseases 289 regarded as the greatest veterinary achievement based on social and network risk factors (Or- of our time. It was the result of decades of effort tiz-Pelaez et  al., 2010), a method originally de- by a wide range of research institutes, donors, veloped for mapping FMD in the UK. This intergovernmental organizations, national gov- identified areas in the central and southern re- ernments, non-governmental organizations and gions of Somalia where veterinary authorities cattle-keepers. FAO had a key leadership role. could concentrate surveillance activities. ILRI made a small but strategic contribu- Participatory surveillance evolved during tion. First, economists developed and dissemin- the global eradication of rinderpest when the ated some of the first estimates of the very large tools used for participatory rural appraisal were benefits that could be obtained from control adapted to searching for rinderpest outbreaks. (Tambi et al., 1999). ILRI economists also con- The approach proved its utility by identifying oc- tributed to post hoc assessments of the cult foci of disease and providing appropriate in- socio-economic direct and indirect benefits of telligence to guide the eradication strategy rinderpest eradication (Roeder and Rich, 2009). (Mariner and Roeder, 2003). ILRI conducted an Rich et al. (2014) examined the ex post impact of evaluation of participatory surveillance to pro- rinderpest eradication in Chad and India. An im- vide guidance for appropriate use (Hannah et al., portant innovation of this study was methodo- 2012). This found that participatory surveil- logical, in terms of identifying impact from the lance was a useful epidemiological tool, most ap- producer level to national and international propriate for small-scale farmers and applied in levels. Farm impacts were examined through the complement to conventional surveillance. use of a herd demographic model and macro- economic impacts with a computable general equilibrium (CGE) model. Baseline benefit:cost Foot-and-mouth disease ratios associated with eradication in Chad were 4.0 over 1963–2002 (ranging from −5.8 to Foot-and-mouth disease (FMD) is often considered 47.2). In India, the final stage of eradication the most economically important global animal yielded a benefit:cost ratio of over 64. disease (see Chapter 5, this volume). Notable ILRI scientists also developed a mathemat- achievements in FMD research were: (i) an ical model for disease dynamics (Mariner et al., impact for specific regions; and (ii) participation 2005). Estimates from simulations suggested in GFRA. In 2010, GFRA performed a gap ana- populations of around 200,000 head of cattle lysis to identify areas where FMD research could were needed to sustain transmission. This meant have the greatest impact and to guide and coord- that communities smaller than 200,000 head inate future research efforts. The 2014 gap ana- did not need to be prioritized in the final stages of lysis was updated to include work published in eradication as the disease would naturally fade 2015 and was the subject of a special issue of out in these populations. Furthermore, this sup- Transboundary and Emerging Diseases. Most of the ported the view that, if the disease was con- research syntheses involved ILRI scientists (Vosloo trolled in cattle, it would fade out in wildlife and Knight-Jones, 2016). An earlier case study populations as the fragmented populations re- of the Philippines showed benefits of ILRI work. maining in Africa are not large enough to sus- In 1999, the National Foot and Mouth Disease tain infection even if biologically competent to Task Force of the Philippines requested ILRI’s do so. More importantly, modelling was shown support for a national FMD control and eradica- to be an effective communication tool to engage tion program. An ILRI team complied with the decision makers, illustrating concepts such as request and, in 2002, published the results of its fade out of disease from small populations and epidemiological and economic assessment of the how suboptimal vaccination could contribute to potential benefits of eradicating FMD from the virus persistence (Roeder et al., 2013). Philippines. The results clearly indicated signifi- Small contributions were made to field epi- cant potential economic benefits, particularly for demiology. By the 1990s, Somalia was one of Filipino swine producers, who were threatened the few remaining reservoirs for rinderpest, but by a virus subtype highly specific to pigs that was control there was hampered by political instabil- causing high piglet mortality, widespread pig ity. ILRI contributed to developing a risk map abortions, and infertility. The joint ILRI-Filipino 290 D. Grace et al. evidence clearly showed the proposed eradica- Lumpy skin disease tion program to be a worthwhile investment of public funds. This public-private partnership for Lumpy skin disease (LSD) is an economically im- control and eradication continued until 2011, portant TAD of cattle caused by the lumpy skin when the Philippines was declared free of foot disease virus in the genus Capripoxvirus. Although and mouth. This official disease-free status opened it is present in most African countries, the disease up markets for pork products from the Philip- is mild and inapparent, so difficult to detect. Fail- pines and motivated the country’s producers to ure to report subclinical cases of LSD due to lack upgrade their piggeries. of good diagnostic tools is a major limitation in LSD control. ILRI included LSD in several field epi- demiology surveys showing its importance in Classical swine fever Ethiopia and Tanzania. An economic analysis of the hides and skins value chain in Somaliland Classical swine fever (CSF) is a highly conta- found that LSD was a significant impediment to gious, potentially fatal viral disease of swine. It the export trade (Wanyoike et al., 2018). is endemic in much of Africa and Asia. In India, A major limitation to LSD control is the dif- pigs are most important in the north-eastern ficulty of detecting it. Despite the availability of states, and ILRI started to work with the sector molecular diagnostic methods that are appropri- in the early 2000s. ILRI conducted epidemio- ate for LSD diagnosis, there is no single rapid, logical and economic studies on CSF in the sensitive and inexpensive method. The BecA-ILRI I ndian states of Assam, Nagaland and Mizoram Hub tested a loop‐mediated isothermal amplifi- in 2011. These found that pig farmers incurred cation (LAMP) assay (Lamprey and Reid, 2004) huge losses, over 2 billion Indian rupees each for LSD diagnosis; this is a gene amplification year, from mortality, treatment and replace- procedure that can amplify a few copies of DNA ment costs. ILRI suggested interventions such to a large amount in less than 1 h using simple as vaccine-based control to the government and equipment. The tests suggested that LAMP private sector. As a result, the Government of would be an accurate and useful diagnostic tool India initiated a national swine fever control (Mwanandota et al., 2018). programme targeting north-east India. More- over, policy changed to better facilitate the licens- ing of vaccine production by public and private Infectious bursal disease institutes. A subsequent ILRI project trained community animal health workers and started Infectious bursal disease is an acute, highly conta- vaccination against pigs. No cases of disease gious, viral disease of young chickens. The disease were reported after vaccination. The State Gov- causes small-scale poultry farmers huge economic ernment of Nagaland then mainstreamed the losses, both from the many birds that die outright approach (Bett et  al., 2014). Although at an and from lost productivity among surviving birds. early stage, there is some evidence for ILRI re- With collaborators, ILRI conducted the first search leading to positive outcomes (Padmaku- molecular characterization of the infectious mar et al., 2017). bursal disease in Kenya. The Directorate of Veter- inary Services plans to use the findings to develop Porcine reproductive and respiratory improved vaccination, surveillance and control syndrome strategies for infectious bursal disease, such as procedures for virus diagnosis (ILRI, 2018). Porcine reproductive and respiratory syndrome (PRRS) is an important disease in pig production and is endemic in Vietnam. ILRI conducted the Transboundary Animal Diseases first nationwide studies of PRRS in Vietnam, in Systems identifying spatial and temporal patterns that are useful in planning control (Fig. 7.1) (Lee Transboundary animal diseases (TADs) do et al., 2019). not occur in isolation but as part of complex Transboundary Animal Diseases 291 N W E S First cluster Third cluster Fifth cluster Second cluster Fourth cluster PRRS outbreaks Cluster area 0 62.5 125 250 375 500 km Fig. 7.1. Space-time cluster analysis of PRRS outbreaks from 2008 to 2016 in Vietnam. (from Lee et al., 2019). ecosystems. Nearly all disease research focuses any naturally occurring animal population in on a single disease or few closely related dis- the world. This study generated a wide range eases, but in reality organisms are normally in- of scientific findings, published in high-impact fected with a number of more or less pathogenic journals. These included several ‘firsts’ in terms organisms at any one time. ILRI recruited a of pathogens, parameters and disease associ- large cohort of calves in western Kenya to inves- ations that had not been reported previously. tigate a wide range of diseases (over 100) and to Research tools were d eveloped and validated. follow up each calf with monthly clinical exam- Although many of the findings have implica- inations and laboratory testing for 51  weeks tions for disease control, their development over a period of 3  years (Fig. 7.2) (Bronsvoort impacts have not yet been evaluated. et al., 2013). This enormous undertaking was the Some of the more novel and important out- most ambitious veterinary epidemiological puts and findings were as follows: study to date carried out in a developing coun- try. It was scientifically significant as the first • The cohort experienced a high mortality attempt to describe the entire disease burden of rate of 16%, with at least 13% of this due to 292 D. Grace et al. IDEAL laboratory – Busia Study sublocations Sublocations within 45 km buffer Agro-ecological zones E.SIBOTI Low midland 1 KOKARE Low midland 2 KIDERA KARISA Low midland 3 KAMUNUOIT MABUSI Upper midland 3 OTIMONG N IGERO BULWANIBUKATI IKONZO LUANDA NAMBOBOTO BUMALA ‘A’ BUJWANGA SIMUR EAST YIRO WEST OJWANDO ‘B’ MAGOMBE EAST KODIERE 0 5 10 20 km Fig. 7.2. Study area in western Kenya. (From de Clare Bronsvoort et al., 2013.) infectious diseases. Over 50 pathogens were as a feed supplement had significantly lower detected in this population, with exposure deaths from helminths. This study gave to a further six viruses and bacteria. East empirical evidence on simple actions that Coast fever (ECF) was the main cause of could greatly reduce calf mortality (Thumbi death, accounting for 40% of all deaths, et al., 2014). haemonchosis 12% and heartwater disease • Heterologous reactivity is the influence of 7% (Thumbi et al., 2013). past or current infection on the outcome of • Following a cohort of animals allowed in- infection with another: this may be positive vestigation of coinfections. The study found or negative. Because calves were followed that these were common and that the risk of over time, researchers were able to obtain ECF death was itself significantly increased the first quantitative estimates of the effects by a high helminth burden and by coin- of heterologous reactivity for any parasitic fection with trypanosomiasis. Farmers disease. The study provided three strands of that provided crop residues to their animals evidence for heterologous protection against Transboundary Animal Diseases 293 ECF in a population of indigenous African • This study was the first to accurately describe cattle. A natural challenge analysis found the haematological parameters for any Af- that infection with less pathogenic Theileria rican breed of cattle. Unlike European cattle spp. reduced mortality from ECF; the case– breeds that experience a fall in red blood control study quantified the level of protec- cells following birth, the West African cattle tion (43% reduction in mortality); and the showed a rise. This is useful in understand- mathematical model of heterologous pro- ing what is normal, but also suggests a pos- tection successfully predicted key features of sible mechanism for disease resistance (van the epidemiology of ECF. This suggested that Wyk et al., 2013). Data from the study were heterologous protection may determine the also used to develop and validate a girth burden and distribution of many parasitic band to predict live weight of East African diseases in host populations, including hu- Shorthorn Zebu (Lesosky et al., 2012). This mans (Woolhouse et al., 2015). • is useful for research but also appropriate Many more farmers reported carrying out for dosing of animals. disease control measures such as tick control and worming than were actually observed. This suggests that farmers are answering what they think they should be doing or maybe Conclusion and Future Directions have done, but a significant proportion actu- ally then appeared to not carry out these meas- Because TADs can spread rapidly and cause ures over the course of our observations. This catastrophic losses, they are of extreme concern highlighted the need for caution in interpret- to the intensive livestock sector in high-income ing responses, especially from cross-sectional countries. In these industries, many TADs have data (Bronsvoort et al., 2013). been eradicated, and hence they are vulnerable • Investigation of antibodies to four tick-borne to reintroduction from developing countries. haemoparasites found that 90% of dams Because TADs were seen as having been well were seropositive for at least one of the studied, they were not initially prioritized by parasites, while 93% of calves had received ILRI. This has gradually changed as evidence colostrum. Surprisingly, there was no dis- has emerged that TADs can have deep impacts in cernible difference in mortality or growth poor countries. rate between calves that had taken colostrum A more pragmatic reason is that, because of and those that had not. These results are also their potential impact on high-income countries, important for interpretation of serosurveys there is often funding available for TAD control. of young calves following natural infection This introduces a certain tension between the or vaccination (Toye et al., 2013a). priorities of rich and poor countries, as non- • Bluetongue virus (BTV) and epizootic haem- epidemic animal diseases are almost certainly of orrhagic disease virus (EHDV) are members greater burden than the epidemic TADs. However, of the genus Orbivirus, transmitted by biting TAD research offers a bridge between low- and midges. BTV spread from Africa to Europe, middle-income countries and high-income coun- resulting in high economic losses. The tries research by providing a subject of common study found that BTV and EHDV are highly interest but differing relative advantage. Novel prevalent, with cattle being infected from an approaches, pioneered by ILRI and collaborators, early age. This was the first report of EHDV to investigating multiple diseases in cohorts of from East Africa (Toye et al., 2013b). animals can help us better understand the relative • Calf respiratory disease is a major problem importance of endemic and epidemic disease and globally but is little researched in Africa. how they interact. The study found that three viruses often im- External drivers, including demographic plicated (infectious bovine rhinotracheitis growth, climate change, biodiversity loss, urban- virus, bovine parainfluenza virus type 3 and ization, globalization and dietary change, have bovine viral diarrhoea virus) all have an potentially profound effects on TAD dynamics, estimated seroprevalence of around 20% which ILRI can influence, mitigate and take (Callaby et al., 2016). leverage of to fulfil its mission. 294 D. Grace et al. Laboratory-based work has produced some Students and graduate fellows have always notable achievements including generating been involved in ILRI health research, and this fundamental knowledge on genetics and phy- has been important for capacity development logenetics, new and improved diagnostics, and as well as establishing and expanding networks. promising vaccine candidates. Much of this To a lesser extent, training has been extended research is upstream and, by its nature, develop- to decision makers, implementers and value ment benefits will take several decades to be visible. chain actors. This is likely to be more important However, there is well-documented frustration in the future. with the technologies available for TAD man- As the research agenda has developed, ILRI agement and control, and much potential for and its investors have increasingly focused on improvement. impact today as well as tomorrow. Field studies Some of the identified priorities, which in- have shown that, while vaccines are theoretic- form future research, include: cheap, simple, ro- ally the best measure for disease control, poor bust field diagnostics; thermostable vaccines farmers have very little propensity to buy them that do not require a cold chain; DIVA vaccines and the public sector has very little capacity to that allow vaccinated and infected animals to be deliver them at scale, at least on a continuous as distinguished; vaccines that are inexpensive and opposed to a campaign basis. This has opened provide life-long immunity; vaccines that are new directions to explore. free of side effects; multivalent vaccines so that The eradication of rinderpest raises hopes one shot will immunize against many diseases; that progressive control may be the one best so- ways of telling whether a vaccine is authentic lution for those TADs that can be eradicated, and functional; and vaccines that are more ef- notably PPR. At the same time, there has been fective and do not ‘break down’ as the result of more attention on influencing farmers’ behav- challenges in administration and because ani- iour as a form of control. Research on several mals are immunocompromised. As this chapter TADs makes it clear that ‘training and informa- has summarized, ILRI is working actively in all tion’ are inadequate and that other incentives, these areas with some success and the promise whether social, financial or changes in choice of more to come. architecture (nudges), are needed. Developing innovative and impactful TAD Finally, there is a shift from scientific papers control requires a good understanding of how to field products. This means relying on markets both diseases and people behave. ILRI TAD research as well as public services, and public-private has had success in developing mathematical partnerships are an active area of research and models and understanding transmission dynam- engagement. Future research will continue to ics, i.e. how diseases move among hosts, vectors develop and improve diagnostics and vaccines and the environment. This information has been with increasing focus on applications. This central to important control successes in the past will be accompanied by efforts to create novel and will be in the future. However, good models re- approaches to TAD control that use market and quire good data on disease parameters, and these social forces. Understanding these needs a value are often not available or can only be gathered at chain and consumer focus, as well as an appreci- costs that investors are unwilling to pay. ation of equity issues. Note 1 The TickRisk project, led by Maxime Madder and Eva De Clerq and based largely in Benin, and the WecatiC project, led by Hassane Adakal, have provided the above data, mostly collected during 2012–2013. References Abera, B., Lynch, S., Duguma, R., Dessie, T., Bettridge, J., et al. (2017) Immunogenicity of the Newcastle disease virus vaccine La Sota, in introduced birds under intensive and extensive management condi- tions. Livestock Research for Rural Development 29, 110 Transboundary Animal Diseases 295 Abworo, E.O., Onzere, C., Amimo, J.O., Riitho, V., Mwangi, W., et al. (2017) Detection of African swine fever virus in the tissues of asymptomatic pigs in smallholder farming systems along the Kenya-Uganda border: implications for transmission in endemic areas and ASF surveillance in East Africa. Journal of General Virology 98, 1806–1814. Adeoye, S.A.O. (1985) Disease profiles of sheep and goats in two groups of villages in southwest Nigeria. In: Sumberg, J.E. and Cassaday, K. (eds) Proceedings of the Workshop on Small Ruminant Produc- tion Systems in the Humid Zone of West Africa, 23–26 January 1984. ILCA, Ibadan, Nigeria. Atherstone, C., Galiwango, R., Grace, D., Alonso, S., Dhand, N., et al. (2019) Analysis of pig trading networks and practices in Uganda. Tropical Animal Health and Production 51, 137–147. Aye, R., Mwirigi, M., Frey, J., Jores, J., Pilo, P. and Naessens, J. (2015) Cyto-adherence of Mycoplasma mycoides subsp. mycoides to primary bovine lung epithelial cells is a critical step in contagious bovine pleuropneumonia. BMC Veterinary Research 11, 27. Aye, R., Weldearegay, Y.B., Lutta, H.O., Chuma, F., Pich, A., et al. (2018) Identification of targets of monoclonal antibodies that inhibit adhesion and growth in Mycoplasma mycoides subspecies mycoides. Veterinary Immunology and Immunopathology 204, 11–18. Balamurugan, V., Hemadri, D., Gajendragad, M.R., Singh, R.K. and Rahman, H. (2014) Diagnosis and control of peste des petits ruminants: a comprehensive review. VirusDisease 25, 39–56. Barongo, M.B., Bishop, R.P., Fèvre, E.M., Knobel, D.L. and Ssematimba, A. (2016) A mathematical model that simulates control options for African swine fever virus (ASFV). PLoS One, 11, e0158658. Barongo, M.B., Ståhl, K., Bett, B., Bishop, R.P., Fèvre, E.M., et al. (2015) Estimating the basic reproductive number (R0) for African swine fever virus (ASFV) transmission between pig herds in Uganda. PLoS One, 10, e0125842. Bett, B., Deka, R.P., Padmakumar, V. and Sones, K.R. (2014) Prevention of classical swine fever – an impact narrative from Northeast India. ILRI Research Brief 8. ILRI, Nairobi. Bishop, R.P., Fleischauer, C., de Villiers, E.P., Okoth, E.A., Arias, M., et al. (2015) Comparative analysis of the complete genome sequences of Kenyan African swine fever virus isolates within p72 genotypes IX and X. Virus Genes 50, 303–309. Bonnet, P., Lesnoff, M., Thiaucourt, F., Workalemahu, A. and Kifle, D. (2005) Seroprevalence of contagious bovine pleuropneumonia in Ethiopian highlands (West Wellega zone, Bodji District). Ethiopian Veter- inary Journal 9, 85–93. de Clare Bronsvoort, B.M., Thumbi, S.M., Poole, E.J., Kiara, H., Auguet, O.T., et al. (2013) Design and descriptive epidemiology of the Infectious Diseases of East African Livestock (IDEAL) project, a lon- gitudinal calf cohort study in western Kenya. BMC Veterinary Research 9, 171. Callaby, R., Toye, P., Jennings, A., Thumbi, S., Coetzer, J., et al. (2016) Seroprevalence of respiratory viral pathogens of indigenous calves in Western Kenya. Research in Veterinary Science 108, 120–124. Chenais, E. (2016) Knowledge, attitudes and practices related to African swine fever within smallholder pig production in northern Uganda. Transboundary Emerging Diseases 64, 101–115. Costard, S., Wieland, B., de Glanville, W., Jori, F., Rowlands, R., et al. (2009) African swine fever: how can global spread be prevented? Philosophical Transactions of the Royal Society B: Biological Sciences 364, 2683–2696. Dessie, T. and Jobre, Y. (2004) A review of the importance and control of Newcastle disease in Ethiopia. Ethiopian Veterinary Journal 8, 71–81. de Villers, E.P., Carmina, G., Marisa, A., Melissa, S., Chris U., Raquel, M. and Bishop, R.P. (2010) Phylogenomic analysis of 11 complete African swine fever virus genome sequences. Virology, 400, 128–136. Dione, M.M., Ouma, E.A., Roesel, K., Kungu, J., Lule, P. and Pezo, D. (2014) Participatory assessment of animal health and husbandry practices in smallholder pig production systems in three high poverty districts in Uganda. Preventive Veterinary Medicine 117, 565–576. Dione, M.M., Akol, J., Roesel, K., Kungu, J., Ouma, E.A., et al. (2015) Risk factors for African swine fever in smallholder pig production systems in Uganda. Transboundary and Emerging Diseases 64, 872–882. Dione, M., Ouma, E., Opio, F., Kawuma, B. and Pezo, D. (2016a) Qualitative analysis of the risks and prac- tices associated with the spread of African swine fever within the smallholder pig value chains in Uganda. Preventive Veterinary Medicine 135, 102–112. Dione, M.M., Ochago, R., Ouma, E.A., Lule, P. and Birungi, R. (2016b) The gender dimensions of a pig disease: African swine fever in Uganda. In: Pyburn, R. and van Eerdewijk, A. (eds) A Different Kettle of Fish? Gen- der Integration in Livestock and Fish Research. LM Publishers, Volendam, The Netherlands. 296 D. Grace et al. Dione, M., Nantima, N., Mayega, L., Amia, W., Wieland, B. and Ouma, E. (2017) Enhancing biosecurity along Uganda’s pig value chains to control and prevent African swine fever. Livestock Brief 1. ILRI, Nairobi. Dione, M.M., Ochago, R., Lule, P. and Mayega, L. (2018a) Guideline for participatory training on African swine fever control for smallholder pig farmers in Uganda. ILRI Manual 28. ILRI, Nairobi. Dione, M.M., Masembe, C., Amia, C.W., Akol, J., Kungu, J. and Wieland, B. (2018b) The importance of on-farm biosecurity: Sero-prevalence AND RISK FACTORS of bacterial and viral pathogens in smallholder pig systems in Uganda. Acta Tropica 187, 214–221. Elsawalhy, A., Mariner, J.C., Chibeu, D., Wamwayi, H., Wakhusama, S., et al. (2010) Pan African strategy for the progressive control of peste des petits ruminants. Bulletin of Animal Health and Production in Africa 58, 185–193. Fadiga, M., Jost, C. and Ihedioha, J. (2011) Financial costs of disease burden, morbidity and mortality from priority livestock diseases in Nigeria: disease burden and cost-benefit analysis of targeted interven- tions. Nigeria Integrated Animal and Human Health Management Project Final Report. ILRI, Nairobi. Falquet, L., Liljander, A., Schieck, E., Gluecks, I., Frey, J. and Jores, J. (2014) Complete genome sequences of virulent Mycoplasma capricolum subsp. capripneumoniae strains F38 and ILRI181. Genome Announcements 2, e01041-1. FAO (2016) Peste des Petits ruminants Global Eradication Programme: contributing to food security, poverty alleviation and resilience. FAO and WHO, Rome. Available at: http://www.fao.org/3/a-i6316e. pdf (accessed 7 April 2020). FAO/AU-IBAR/ILRI (2017) Regional strategy for the control of African swine fever in Africa. FAO, Rome. Fischer, A., Shapiro, B., Muriuki, C., Heller, M., Schnee, C., et al. (2012) The origin of the ‘mycoplasma mycoides cluster’ coincides with domestication of ruminants. PLoS One 7, e36150. Fischer, A., Santana-Cruz, I., Giglio, M., Nadendla, S., Drabek, E., et  al. (2013) Genome sequence of Mycoplasma feriruminatoris sp. nov., a fast-growing mycoplasma species. Genome Announcements 1, e00216-12. Fischer, A., Santana-Cruz, I., Hegerman, J. and Gourlé, H, et al. (2015) High quality draft genomes of the Mycoplasma mycoides subsp. mycoides challenge strains Afadé and B237. Standards in Genomic Science 10, 89. Galindo, I. and Alonso, C. (2017) African swine fever virus: a review. Viruses 9, 103. Gallardo, C., Mwaengo, D.M., Macharia, J.M., Arias, M., Taracha, E.A., et al. (2009) Enhanced discrimin- ation of African swine fever virus isolates through nucleotide sequencing of the p54, p72, and pB602L (CVR) genes. Virus Genes 38, 85–95. Gallardo, C., Okoth, E., Pelayo, V., Anchuelo, R., Martin, E., et al. (2011) African swine fever viruses with two different genotypes, both of which occur in domestic pigs, are associated with ticks and adult warthogs, respectively, at a single geographical site. Journal of General Virology 92, 432–444. Gallardo, C., Soler, A., Nieto, R., Carrascosa, A., de Mia, G., et al. (2013) Comparative evaluation of novel African swine fever virus (ASF) antibody detection techniques derived from specific ASF viral geno- types with the OIE internationally prescribed serological tests. Veterinary Microbiology 162, 32–43. Gourgues, G. and Barr. (2016) Complete genome sequence of Mycoplasma mycoides subsp. mycoides T1/44, a vaccine strain against contagious bovine pleuropneumonia. Genome Announcements 14, e00263-16. Hannah, H., Pali, P., Rware, H., Bett, B., Randolph, T., et al. (2012) Participatory disease surveillance evaluation: findings from Africa (HPAI) and Pakistan (Rinderpest). Poster prepared for the 13th conference of the International Society for Veterinary Epidemiology and Economics, 2012, Maastricht, The Netherlands, 20–24 August, 2012. ISVEE, Durban, South Africa. Heller, M., Gicheru, N., Tjipura-Zaire, G., Muriuki, C., Yu, M., et al. (2016) Development of a novel cocktail enzyme-linked immunosorbent assay and a field-applicable lateral-flow rapid test for diagnosis of con- tagious bovine pleuropneumonia. Journal of Clinical Microbiology 54, 1557–1565. Holzer, B., Taylor, G., Rajko-Nenow, P., Hodgson, S., Okoth, E., et al. (2016) Determination of the minimum fully protective dose of adenovirus-based DIVA vaccine against peste des petits ruminants virus chal- lenge in East African goats. Veterinary Research 47, 20. Hutton, S., Bettridge, J., Christley, R., Habte, T. and Ganapathy, K. (2017) Detection of infectious bronchitis virus 793B, avian metapneumovirus, Mycoplasma gallisepticum and Mycoplasma synoviae in poultry in Ethiopia. Tropical Animal Health and Production 49, 317–322. ILRI (2018) ILRI corporate report 2016–2017. ILRI, Nairobi. Jarso, D. (2015) Epidemiology of village chicken diseases: a longitudinal study on the magnitude and determinants of morbidity and mortality – the case of Newcastle and infectious bursal disease. MSc thesis, Addis Ababa University, Debre Zeit, Ethiopia. Transboundary Animal Diseases 297 Jones, B.A., Rich, K.M., Mariner, J.C., Anderson, J., Jeggo, M., et al. (2016) The economic impact of eradi- cating peste des petits ruminants: a benefit-cost analysis. PLoS One 11, e0149982. Jores, J., Nkando, I., Sterner-Kock, A., Haider, W., Poole, J., et al. (2008) Assessment of in vitro interferon-γ responses from peripheral blood mononuclear cells of cattle infected with Mycoplasma mycoides ssp. mycoides small colony type. Veterinary Immunology and Immunopathology 124, 192–197. Jores, J., Meens, J., Buettner, F.F., Linz, B., Naessens, J. and Gerlach, G. (2009) Analysis of the immuno- proteome of Mycoplasma mycoides subsp. mycoides small colony type reveals immunogenic homo- logues to other known virulence traits in related Mycoplasma species. Veterinary Immunology and Immunopathology 131, 238–245. Jores, J., Fischer, A., Sirand-Pugnet, P., Thomann, A., Liebler-Tenorio, E.M., et al. (2013a) Mycoplasma feriruminatoris sp. nov., a fast growing Mycoplasma species isolated from wild Caprinae. Systematic and Applied Microbiology 36, 533–538. Jores, J., Mariner, J.C. and Naessens, J. (2013b) Development of an improved vaccine for contagious bovine pleuro- pneumonia: an African perspective on challenges and proposed actions. Veterinary R esearch 44, 122. Jores, J., Schieck, E., Liljander, A., Sacchini, F., Posthaus, H., et al. (2018) Capsular polysaccharide in Mycoplasma mycoides is a virulence factor. Journal of Infectious Diseases 219, 1559–1563. Kairu-Wanyoike, S.W., Kaitibie, S., Taylor, N.M., Gitau, G.K., Heffernan, C., et al. (2013) Exploring farmer preferences for contagious bovine pleuropneumonia vaccination: a case study of Narok District of Kenya. Preventive Veterinary Medicine 110, 356–369. Kairu-Wanyoike, S.W., Kaitibie, S., Heffernan, C., Taylor, N.M., Gitau, G.K., et al. (2014a) Willingness to pay for contagious bovine pleuropneumonia vaccination in Narok South District of Kenya. Preventive Vet- erinary Medicine 115, 130–142. Kairu-Wanyoike, S.W., Kiara, H., Heffernan, C., Kaitibie, S., Gitau, G.K., et al. (2014b) Control of contagious bovine pleuropneumonia: knowledge, attitudes, perceptions and practices in Narok district of Kenya. Preventive Veterinary Medicine 115, 143–156. Kairu-Wanyoike, S.W., Taylor, N.M., Heffernan, C. and Kiara, H. (2017) Micro-economic analysis of the potential impact of contagious bovine pleuropneumonia and its control by vaccination in Narok district of Kenya. Livestock Science 197, 61–72. Kama-Kama, F., Midiwo, J., Nganga, J., Maina, N., Schiek, E., et al. (2016) Selected ethno-medicinal plants from Kenya with in vitro activity against major African livestock pathogens belonging to the Myco- plasma mycoides cluster. Journal of Ethnopharmacology 192, 524–534. Kama-Kama, F., Omosa, L.K., Nganga, J., Maina, N., Osanjo, G., et al. (2017) Antimycoplasmal activities of compounds from Solanum aculeastrum and Piliostigma thonningii against strains from the Myco- plasma mycoides cluster. Frontiers in Pharmacology 8, 920. Krasteva, I., Liljander, A., Fischer, A., Smith, D.G., Inglis, N.F., et al. (2014) Characterization of the in vitro core surface proteome of Mycoplasma mycoides subsp. mycoides, the causative agent of contagious bovine pleuropneumonia. Veterinary Microbiology 168, 116–123. Lamprey, R.H. and Reid, R.S. (2004) Expansion of human settlement in Kenya’s Maasai Mara: what future for pastoralism and wildlife? Journal of Biogeography 31, 997–1032. Laval, G. (1999) Cost analysis of contagious bovine pleuropneumonia in Ethiopia. MSc thesis, Claude Bernard University, Lyon, France. Leblanc, N., Cortey, M., Fernandez Pinero, J., Gallardo, C., Masembe, C., et al. (2013) Development of a sus- pension microarray for the genotyping of African swine fever virus targeting the SNPs in the C-terminal end of the p72 gene region of the genome. Transboundary and Emerging Diseases 60, 378–383. Lee, H.S., Pham, T.L., Nguyen, T.N., Lee, M. and Wieland, B. (2019) Seasonal patterns and space-time clustering of porcine reproductive and respiratory syndrome (PRRS) cases from 2008 to 2016 in Viet- nam. Transboundary and Emerging Diseases 66, 986–994. Lesnoff, M., Laval, G., Bonnet, P., Abdicho, S., Workalemahu, A., et al. (2004) Within-herd spread of conta- gious bovine pleuropneumonia in Ethiopian highlands. Preventive Veterinary Medicine 64, 27–40. Lesosky, M., Dumas, S., Conradie, I., Handel, I.G., Jennings, A., et al. (2012) A live weight-heart girth relationship for accurate dosing of east African shorthorn zebu cattle. Tropical Animal Health and Production 45, 311–316. Lichoti, J.K., Davies, J., Kitala, P.M., Githigia, S.M., Okoth, E., et al. (2016) Social network analysis provides insights into African swine fever epidemiology. Preventive Veterinary Medicine 126, 1–10. Lichoti, J.K., Davies, J., Maru, Y., Kitala, P.M., Githigia, S.M., et al. (2017) Pig traders’ networks on the Kenya-Uganda border highlight potential for mitigation of African swine fever virus transmission and improved ASF disease risk management. Preventive Veterinary Medicine 140, 87–96. 298 D. Grace et al. Liljander, A., Yu, M., O’Brien, E., Heller, M., Nepper, J.F., et al. (2015) Field-applicable recombinase poly- merase amplification assay for rapid detection of Mycoplasma capricolum subsp. capripneumoniae. Journal of Clinical Microbiology 53, 2810–2815. Lindahl, J.F., Young, J., Wyatt, A., Young, M., Alders, R., et al. (2019) Do vaccination interventions have effects? A study on how poultry vaccination interventions change smallholder farmer knowledge, attitudes, and practice in villages in Kenya and Tanzania. Tropical Animal Health and Production 51, 213–220. Liu, B., Asseng, S., Müller, C., Ewert, F., Elliott, J., et al. (2016) Similar estimates of temperature impacts on global wheat yield by three independent methods. Nature Climate Change 6, 1130–1136. Mariner, J.C. and Roeder, P. (2003) The use of participatory epidemiology in studies of the persistence of rinderpest in East Africa. Veterinary Record 152, 641–647. Mariner, J.C., House, J.A., Sollod, A.E., Stem, C., van den Ende, M. and Mebus, C.A. (1990) Comparison of the effect of various chemical stabilizers and lyophilization cycles on the thermostability of a Vero cell-adapted rinderpest vaccine. Veterinary Microbiology 21, 195–209. Mariner, J.C., House, J.A., Mebus, C.A., Sollod, A. and Stem, C. (1991) Production of a thermostable Vero- cell-adapted rinderpest vaccine. Journal of Tissue Culture Methods 13, 253–256. Mariner, J.C., McDermott, J.J., Heesterbeek, J.A.P., Catley, A. and Roeder, P. (2005) A model of lineage-1 and lineage-2 rinderpest virus transmission in pastoral areas of East Africa. Preventive Veterinary Medicine 69, 245–263. Mariner, J.C., McDermott, J., Heesterbeek, J., Thomson, G., Roeder, P.L. and Martin, S. (2006a) A hetero- geneous population model for contagious bovine pleuropneumonia transmission and control in pas- toral communities of East Africa. Preventive Veterinary Medicine 73, 75–91. Mariner, J.C., McDermott, J., Heesterbeek, J., Thomson, G. and Martin, S. (2006b) A model of contagious bo- vine pleuropneumonia transmission dynamics in East Africa. Preventive Veterinary Medicine 73, 55–74. Mariner, J.C., House, J.A., Mebus, C.A., Sollod, A.E., Chibeu, D., et al. (2012) Rinderpest eradication: appropriate technology and social innovations. Science 337, 1309–1312. Mariner, J.C., Jones, B.A., Rich, K.M. Thevasagayam, S., Anderson, J., et al. (2016) The opportunity to eradicate peste des petits ruminants. Journal of Immunology 196, 2499–3506. Mariner, J.C., Gachanja, J., Tindih, S.H. and Toye, P. (2017) A thermostable presentation of the live, attenu- ated peste des petits ruminants vaccine in use in Africa and Asia. Vaccine 35, 3773–3779. McDermott, J., Coleman, P. and Randolph, T. (2001) Methods for assessing the impact of infectious diseases of livestock – their role in improving the control of Newcastle disease in Southern Africa. In: ACIAR Proceedings. ACIAR, Canberra, pp. 118–128. Muindi, P., Waithanji, E. and Bukachi, S. (2015) Gendered effects of contagious bovine pleuropneumonia (CBPP) occurrence and control in a pastoral community in Ijara sub county, northeastern Kenya. ILRI, Nairobi. Mujibi, F.D., Okoth, E., Cheruiyot, E.K., Onzere, C., Bishop, R.P., et al. (2018) Genetic diversity, breed com- position and admixture of Kenyan domestic pigs. PLoS One, 13, e0190080. Mulongo, M.M., Frey, J., Smith, K., Schnier, C., Wesonga, H., et al. (2013) Cattle immunized against the pathogenic l-α-glycerol-3-phosphate oxidase of Mycoplasma mycoides subsp. mycoides fail to gen- erate neutralizing antibodies and succumb to disease on challenge. Vaccine 31, 5020–5025. Mulongo, M., Frey, J., Smith, K., Schnier, C., Wesonga, H., et al. (2015) Vaccination of cattle with the N terminus of LppQ of Mycoplasma mycoides subsp. mycoides results in type III immune complex disease upon experimental infection. Infection and Immunity 83, 1992–2000. Mwanandota, J.J., Macharia, M., Ngeleja, C.M., Sallu, R.S., Yongolo, M.G., et al. (2018) Validation of a diagnostic tool for the diagnosis of lumpy skin disease. Veterinary Dermatology 29, 532-e178. Mwirigi, M., Nkando, I., Aye, R., Soi, R., Ochanda, H., et al. (2016a) Experimental evaluation of inactivated and live attenuated vaccines against Mycoplasma mycoides subsp. mycoides. Veterinary Immunology and Immunopathology 169, 63–67. Mwirigi, M., Nkando, I., Olum, M., Attah-Poku, S., Ochanda, H., et al. (2016b) Capsular polysaccharide from Mycoplasma mycoides subsp. mycoides shows potential for protection against contagious bovine pleuropneumonia. Veterinary Immunology and Immunopathology 178, 64–69. Nantima, N., Ocaido, M., Ouma, E., Davis, J., Dione, M., et al. (2015) Risk factors associated with occur- rence of African swine fever (ASF) outbreaks in smallholder pig farms in four districts along the Uganda- Kenya border. Tropical Animal Health and Production 47, 589–595. Nantima, N., Davies, J., Dione, M., Ocaido, M., Okoth, E., et al. (2016) Enhancing knowledge and aware- ness of biosecurity practices for control of African swine fever among smallholder pig farmers in four districts along the Kenya-Uganda border. Tropical Animal Health and Production 48, 727–734. Transboundary Animal Diseases 299 Naseem, S., Meens, J., Jores, J., Heller, M., Dbel, S., et al. (2010) Phage display-based identification and potential diagnostic application of novel antigens from Mycoplasma mycoides subsp. mycoides small colony type. Veterinary Microbiology 142, 285–292. Nkando, I., Ndinda, J., Kuria, J., Naessens, J., Mbithi, F., et al. (2012) Efficacy of two vaccine formulations against contagious bovine pleuropneumonia (CBPP) in Kenyan indigenous cattle. Research in Veter- inary Science 93, 568–573. Nkando, I., Perez-Casal, J., Mwirigi, M., Prysliak, T., Townsend, H., et  al. (2016) Recombinant Myco- plasma mycoides proteins elicit protective immune responses against contagious bovine pleuropneu- monia. Veterinary Immunology and Immunopathology 171, 103–114. Okoth, E., Gallardo, C., Macharia, J., Omore, A., Pelayo, V., et al. (2013) Comparison of African swine fever virus prevalence and risk in two contrasting pig-farming systems in South-west and Central Kenya. Preventive Veterinary Medicine 110, 198–205. Onono, J.O., Wieland, B., Suleiman, A. and Rushton, J. (2017) Policy analysis for delivery of contagious bovine pleuropneumonia control technologies in sub-Saharan Africa. Revue Scientifique et Technique 36, 195–205. Ortiz-Pelaez, A., Pfeiffer, D.U., Tempia, S., Otieno, F.T., Aden, H.H. and Costagli, R. (2010) Risk mapping of rinderpest sero-prevalence in Central and Southern Somalia based on spatial and network risk fac- tors. BMC Veterinary Research 6, 22. Ouma, E., Dione, M., Birungi, R., Lule, P., Mayega, L. and Dizyee, K. (2018) African swine fever control and market integration in Ugandan peri-urban smallholder pig value chains: an ex-ante impact assessment of interventions and their interaction. Preventive Veterinary Medicine 151, 29–39. Ouma, E., Dione, M., Lule, P.M., Pezo, D., Marshall, K., et al. (2015) Smallholder pig value chain assessment in Uganda: results from producer focus group discussions and key informant interview. ILRI Project Report. ILRI, Nairobi. Padmakumar, V., Deka, R.P. and Wright, I. (2017) The impact of evidence-based stakeholder policy facilitation on pig-sub sector growth in northeast India. ILRI Research Brief 80. ILRI, Nairobi. Perez-Casal, J., Prysliak, T., Maina, T., Wang, Y., Townsend, H., et  al. (2015) Analysis of immune re- sponses to recombinant proteins from strains of Mycoplasma mycoides subsp. mycoides, the causative agent of contagious bovine pleuropneumonia. Veterinary Immunology and Immunopathol- ogy 168, 103–110. Perry, B.D., Randolph, T.F., McDermott, J., Sones, K. and Thornton, P. (2002) Investing in animal health research to alleviate poverty. ILRI, Nairobi. Rich, K.M., Roland-Holst, D. and Otte, J. (2014) An assessment of the ex-post socio-economic impacts of global rinderpest eradication: methodological issues and applications to rinderpest control programs in Chad and India. Food Policy 44, 248–261. Rimi, N.A., Sultana, R., Ishtiak-Ahmed, K., Rahman, M.Z., Hasin, M., et al. (2016) Understanding the failure of a behavior change intervention to reduce risk behaviors for avian influenza transmission among backyard poultry raisers in rural Bangladesh: a focused ethnography. BMC Public Health 16, 858. Robinson, T., Thornton, P., Franceschini, G., Kruska, R., Chiozza, F., et al. (2011) Global livestock produc- tion systems. FAO, Rome, and ILRI, Nairobi. Roeder, P. and Rich, K.M. (2009) Conquering the cattle plague: the global effort to eradicate rinderpest. In: Spielman, D.J. and Pandya-Lorch, R. (eds) Millions Fed: Proven Successes in Agricultural Develop- ment. IFPRI, Washington, DC, pp. 109–116. Roeder, P., Mariner, J. and Kock, R.R. (2013) Rinderpest: the veterinary perspective on eradication. Philo- sophical Transactions of the Royal Society of London B: Biological Sciences 368, 20120139. Sacchini, F., Naessens, J., Awino, E., Heller, M., Hlinak, A., et al. (2011) A minor role of CD4+ T lympho- cytes in the control of a primary infection of cattle with Mycoplasma mycoides subsp. mycoides. Veterinary Research 42, 77. Sacchini, F., Luciani, M., Salini, R., Scacchia, M., Pini, A., et al. (2012) Plasma levels of TNF-α, IFN-γ, IL-4 and IL-10 during a course of experimental contagious bovine pleuropneumonia. BMC Veterinary Research 8, 44. Schieck, E., Lartigue, C., Frey, J., Vozza, N., Hegermann, J., et al. (2016) Galactofuranose in Mycoplasma mycoides is important for membrane integrity and conceals adhesins but does not contribute to serum resistance. Molecular Microbiology 99, 55–70. Schieck, E., Liljander, A., Hamsten, C., Gicheru, N., Scacchia, M., et al. (2014) High antibody titres against predicted Mycoplasma surface proteins do not prevent sequestration in infected lung tissue in the course of experimental contagious bovine pleuropneumonia. Veterinary Microbiology 172, 285–293. 300 D. Grace et al. Schnee, C., Heller, M., Jores, J., Tomaso, H. and Neubauer, H. (2011) Assessment of a novel multiplex real-time PCR assay for the detection of the CBPP agent Mycoplasma mycoides subsp. mycoides SC through experimental infection in cattle. BMC Veterinary Research 7, 47. Schubert, E., Sachse, K., Jores, J. and Heller, M. (2011) Serological testing of cattle experimentally infected with Mycoplasma mycoides subsp. mycoides Small Colony using four different tests reveals a variety of seroconversion patterns. BMC Veterinary Research 7, 72. Seixas, A., Alzugaray, M.F., Tirloni, L., Parizi, L.F. and Pinto, A.F.M. (2018) Expression profile of Rhipiceph- alus microplus vitellogenin receptor during oogenesis. Ticks and Tick-borne Diseases 9, 72–81. Sørensen, M.R., Ilsøe, M., Strube, M.L., Bishop, R., Erbs, G., et al. (2017) Sequence-based genotyping of expressed swine leukocyte antigen class I alleles by next-generation sequencing reveal novel swine leukocyte antigen class I haplotypes and alleles in Belgian, Danish, and Kenyan fattening pigs and Göttingen minipigs. Frontiers in Immunology 8, 701. Ssematimba, A., Jores, J. and Mariner, J. (2015) Mathematical modelling of the transmission dynamics of contagious bovine pleuropneumonia reveals minimal target profiles for improved vaccines and diag- nostic assays. PLoS One 10, e0116730. Ståhl, K., Ogweng, P., Okoth, E., Aliro, T., Muhangi, D., et  al. (2014) Understanding the dynamics and spread of African swine fever virus at the wildlife-livestock interface: insights into the potential role of the bushpig, Potamochoerus larvatus. Suiform Soundings 13, 24–28. Steinaa, L., Bishop, R.P., Okoth, E.A., Svitek, N. and Riitho, V. (2016) Pig vaccines and diagnostics for African swine fever in Uganda. Livestock and Fish Brief. ILRI, Nairobi. Sterner-Kock, A., Haider, W., Sacchini, F., Liljander, A., Meens, J., et al. (2016) Morphological characteriza- tion and immunohistochemical detection of the proinflammatory cytokines IL-1β, IL-17A, and TNF-α in lung lesions associated with contagious bovine pleuropneumonia. Tropical Animal Health and Pro- duction 48, 569–576. Tambi, E.N., Maina, O.W., Mukhebi, A.W. and Randolph, T.F. (1999) Economic impact assessment of rinderpest control in Africa. Revue Scientifique et Technique-Office International des Épizooties, 18, 458–477. Taylor, W.P. (1984) The distribution and epidemiology of peste des petits ruminants. Preventive Veterinary Medicine 2, 157–166. Terfa, Z., Garikipati, S., Dessie, T., Lynch, S., Wigley, P., et al. (2015) Farmers’ willingness to pay for a village poultry vaccine service in Ethiopia: prospect for enhancing rural livelihoods. Food Security 7, 905–917. Thomas, L.F., Bishop, R.P., Onzere, C., Mcintosh, M.T., Lemire, K.A., et al. (2016) Evidence for the presence of African swine fever virus in an endemic region of Western Kenya in the absence of any reported outbreak. BMC Veterinary Research 12, 192. Thumbi, S.M., de Clare Bronsvoort, B.M., Kiara, H., Toye, P.G., Poole, J., et al. (2013) Mortality in East African shorthorn zebu cattle under one year: predictors of infectious-disease mortality. BMC Veterin- ary Research 9, 175. Thumbi, S.M., de Clare Bronsvoort, B.M., Poole, E.J., Kiara, H., Toye, P.G., et al. (2014) Parasite co-infections and their impact on survival of indigenous cattle. PLoS One 9, e76324. Toye, P., Handel, I., Gray, J., Kiara, H., Thumbi, S., et al. (2013a) Maternal antibody uptake, duration and influence on survival and growth rate in a cohort of indigenous calves in a smallholder farming system in western Kenya. Veterinary Immunology and Immunopathology 155, 129–134. Toye, P.G., Batten, C., Kiara, H., Henstock, M., Edwards, L., et al. (2013b) Bluetongue and epizootic haemorrhagic disease virus in local breeds of cattle in Kenya. Research in Veterinary Science 94, 769–773. van Wyk, I.C., Goddard, A., Bronsvoort, B.M. de C., Coetzer, J.A., Booth, C., et al. (2013) Hematological profile of East African short-horn zebu calves from birth to 51 weeks of age. Comparative Clinical Pathology 22, 1029–1036. Vosloo, W. and Knight-Jones, T.J.D. (2016) GFRA Global foot-and-mouth disease research update and gap analysis. Transboundary and Emerging Diseases 63, 351–352. Waithanji, E., Mtimet, N. and Muindi, P. (2015) Delivery of the contagious bovine pleuropneumonia (CBPP) vaccine in northeastern Kenya. ILRI Discussion paper 31. ILRI, Nairobi. Wang, Q.H., Ren, W.J., Bao, J.Y., Ge, S.Q., Li, J.M., et al. (2018) The first outbreak of African swine fever was confirmed in China. Journal of China Animal Health Inspection 34, 1–5. Wanyoike, F.N., Mugunieri, L.G., Mtimet, N., Kiptoo, E. and Gulaid, I. (2018) An analysis of the hides and skins value chain in Somaliland. ILRI, Nairobi. Transboundary Animal Diseases 301 Weldearegay, Y.B., Pich, A., Liljander, A., Schieck, E., Gicheru, N., et al. (2015) Proteomic characterization of bovine pleural effusion, a specific host niche of Mycoplasma mycoides subsp. mycoides, the causative agent of CBPP. Journal of Proteomics 131, 93–103. Woolhouse, M.E., Thumbi, S.M., Jennings, A., Chase-Topping, M., Callaby, R., et al. (2015) Co-infections determine patterns of mortality in a population exposed to parasite infection. Science Advances 1, e1400026. Zsak, L., Borca, M., Risatti, G., Zsak, A., French, R., et al. (2005) Preclinical diagnosis of African swine fever in contact-exposed swine by a real-time PCR assay. Journal of Clinical Microbiology 43, 112–119. 8 Zoonoses Delia Grace1, Silvia Alonso2, Bernard Bett3, Elizabeth Cook3, Hu Suk Lee4, Anne Liljander5, Jeff Mariner6, Florence Mutua3, Hung Nguyen-Viet4, Ekta Patel3, Thomas Fitz Randolph3, Kristina Roesel3, Lian Thomas3, Phil Toye3, Fred Unger4 and Barbara Wieland7 1International Livestock Research Institute, Nairobi, Kenya and University of Greenwich, UK; 2International Livestock Research Institute, Jerusalem, Israel; 3Inter- national Livestock Research Institute, Nairobi, Kenya; 4International Livestock Re- search Institute, Hanoi, Vietnam; 5EUROIMMUN AG, Lübeck, Germany; 6Cummings School of Veterinary Medicine at Tufts University, Grafton, Massachusetts, USA; 7International Livestock Research Institute, Addis Ababa, Ethiopia Contents Executive Summary 303 The problem 303 ILRI’s role in the global context 303 Impacts of ILRI research 304 Capacity development 305 Introduction 305 Prioritization 306 Understanding drivers of disease emergence 308 Participatory disease surveillance 308 Zoonoses in Systems 309 Zoonoses in smallholder pig systems 311 Collaboration in international initiatives 312 Emerging Infectious Diseases 313 Rift Valley fever 313 Understanding transmission and burden in Kenya 315 Developing a more effective vaccine 315 Supporting response in East Africa 315 Highly pathogenic avian influenza 316 Supporting surveillance in Africa 316 Response models in Egypt and Indonesia 316 Identifying and supporting pro-poor interventions for disease control 317 Evaluation of control in Indonesia and Nigeria 317 Middle Eastern respiratory syndrome 318 Understanding disease reservoirs and transmission 318 Improving the national response 318 Ebola 318 First risk assessment in livestock systems 319 Antimicrobial resistance 319 © International Livestock Research Institute 2020. The Impact of the International 302 Livestock Research Institute (eds J. McIntire and D. Grace) Zoonoses 303 Discovery of anti-parasitic drugs 319 Understanding and improving the use of antibiotics 320 Understanding drug resistance emergence and transmission 320 The CGIAR Antimicrobial Resistance Hub 320 Neglected Zoonoses 321 Bovine tuberculosis 321 Cysticercosis 322 Human African trypanosomiasis 324 Brucellosis 325 Other neglected zoonoses 327 Conclusions and the Future 328 References 329 Executive Summary ILRI’s role in the global context The problem ILRI has worked in joint partnerships with sev- eral universities and research institutes to develop Zoonoses are diseases that are transmissible be- research activities on zoonoses. This area of tween humans and animals through either dir- research commenced in the 1990s with small ect contact or by way of food, water or the studies on neglected zoonoses. Initial work environment. Around 60% of all human dis- focused on bovine tuberculosis (TB), brucellosis, eases and around 75% of emerging human in- human African trypanosomiasis (commonly fectious diseases are zoonotic. Zoonoses have known as sleeping sickness) and rabies. From high impacts on human health, livelihoods, ani- the 2000s, an agenda on the pig tapeworm that mals and ecosystems. The first global syntheses causes the parasitic tissue infection cysticercosis on the impacts of zoonotic diseases, led by the emerged, while research on specific neglected International Livestock Research Institute zoonoses continued, including work on anthrax; (ILRI), estimated that in the least-developed the acute bacterial swine disease known as ery- countries, 20% of human sickness and death sipeloid (‘diamond skin disease’); the bacterial was due to zoonoses or diseases that had re- diseases leptospirosis, Q fever and streptococcosis; cently jumped species from animals to people and the parasitic (roundworm) disease trichinel- (Grace et  al., 2012a). Zoonoses sicken several losis, which humans can get when consuming billion people each year and kill millions, mostly undercooked or raw pork products. in low- and middle-income countries. While es- Work on emerging zoonoses started in the timates of the historical burden of zoonoses are 2000s as global attention heightened following the lacking, the World Bank has estimated that global pandemic of highly pathogenic avian influ- emerging zoonoses cost around US$7 billion a enza (HPAI, or avian influenza, caused by influenza year (World Bank, 2012). A virus subtype H5N1) and the high-p rofile out- Some zoonoses are considered neglected, breaks of bovine spongiform encephalopathy (BSE classical or endemic, and others as new or emer- or ‘mad cow disease’) and severe acute respiratory ging. Neglected zoonoses are mostly controlled syndrome (SARS). ILRI research on emerging or eradicated from high-income countries but diseases focused on livestock-related antimicro- impose a large burden on low- and middle- bial resistance, avian influenza, Ebola virus disease, income countries. Emerging zoonoses are a glo- Middle Eastern respiratory syndrome coronavirus bal threat, but most of the economic burden falls (MERS-CoV) and Rift Valley fever (RVF). on high-income countries. Disease categories As the programme matured, ILRI zoonoses are to some extent overlapping, and a disease research moved from a predominately veterinary may be endemic in one place and emerging in public health paradigm to embrace one-health, another. Many zoonoses, both neglected and which is predicated on the interdependence of emerging, are food-borne; these are discussed in human, animal and ecosystem health. As well Chapter 9 (this volume). This chapter focuses on as the research on high-priority zoonotic dis- zoonoses that are not transmitted primarily eases, ILRI research initiatives addressed specific through food. livestock systems, notably pastoral and urban. 304 D. Grace et al. In the last two decades, this work has extended alerted national medical services that some dis- from East Africa to India and South-east Asia, eases they did not know were present actually and in the last decade, ILRI’s zoonotic research were present and that other diseases they sus- was aligned to the livestock value chains priori- pected to be present were in fact not present. tized in the CGIAR Research Programmes (CRPs) Similarly, a study in Uganda found substan- on Livestock and Fish (2012–2016) and Live- tial underdiagnosis of sleeping sickness, another stock (2017–2021). Research on zoonoses has in western Kenya found extensive misdiagnosis included estimating the prevalence and burden of human brucellosis, and yet another in nor- of zoonoses, identifying factors or drivers of thern Kenya found substantial overdiagnosis of emergence of zoonoses, understanding the risk malaria. ILRI brought these findings to the at- factors of these infectious diseases, and strategies tention of local medical authorities. While the for reducing those risks and better managing benefits of this work have not yet been well these diseases. evaluated, this work is expected to lead to better treatment for tens of thousands of sick people. Initial work on zoonoses logically focused mostly on improving our understanding of the Impacts of ILRI research presence, prevalence, burden and drivers of zoo- notic disease. ILRI developed a prototype diag- Because large-scale zoonoses research at ILRI is nostic for cysticercosis, but this did not go to relatively new, ILRI’s scientific impacts in this scale because of lack of demand; tests for bovine area are more notable than its impacts on devel- tuberculosis diagnosis are ongoing at ILRI. A opment. The most-cited papers by any ILRI vaccine for RVF is under development by ILRI and authors or co-authors are in the realm of climate has potential for widespread use. Vaccination change and human disease, emerging zoonoses (using a commercial vaccine) for avian influenza and estimation of the human disease burden of was piloted in Java, Indonesia, with ILRI support; zoonotic diseases. Research analytics show that an ILRI evaluation showed this was effective but ILRI research was especially dominant in MERS- unsustainable purely by the market. Another CoV and RVF. There are also many important ILRI evaluation of vaccination for cysticercosis ILRI outputs on antimicrobial resistance, avian in Uganda came to the same conclusion. These influenza, brucellosis, cysticercosis and sleeping findings helped identify more practical solutions, sickness, but these diseases are the focus of con- some of them potentially involving public-private siderable global research efforts, and ILRI has been partnerships. a relatively minor player globally. Many synthe- Small pilots were conducted by ILRI to train ses chapters and papers on zoonoses have been farmers, extension agents, slaughterhouse work- authored or co-authored by ILRI scientists. An ers and street vendors to improve food safety: important scientific contribution has been devel- most showed benefits, at least in the short term. opment and deployment of tools as well as ILRI has also conducted ex ante and model-based methodological advances. Most notable were assessments to compare alternative response mathematical modelling, systematic prioritiza- strategies for zoonotic disease, notably for RVF tion of zoonoses, use of geospatial data and par- and sleeping sickness. ticipatory disease surveillance. ILRI has been involved in policy support for While many of ILRI’s research activities better controlling zoonoses at local, national, helped to advance our basic understanding of regional and international scales (UNEP and disease, which can be anticipated to contribute to ILRI, 2020). ILRI has partnered with the World long-term improvements in human health, some Health Organization (WHO) and other inter- had significant direct outcomes. In Uganda, ILRI national organizations on several initiatives to scientists identified the first cases of swine ery- better control neglected zoonoses generally and sipelas (‘diamond skin disease’), established that cysticercosis specifically. It has also partnered swine brucellosis did not occur and established with most of the major global one-health initia- that trichinellosis (pork measles) might be tran- tives. ILRI supported regional initiatives on rabies sitioning from a wildlife to a livestock cycle. This and cysticercosis. At national levels, ILRI has kind of new and surprising information usefully supported contingency planning and groups Zoonoses 305 working in different countries on specific zoonoses, prove sustainable). Zoonoses prioritizations con- including MERS, brucellosis and RVF, as well as ducted by ILRI have been directly connected to working groups and task forces focused more broadly funding decisions by donors. ILRI has also led on zoonoses and one-health. In cities and decentral- several evaluations of zoonoses control projects, ized countries, ILRI has worked with local authorities most notably on avian influenza, which either and in some cases has documented beneficial shifts endorsed national control programmes or led to in policy and approaches. When engaging in these improvements in how they were conducted. policy processes, ILRI provided evidence and advo- cated best practice, but ILRI’s contributions to health policy and the impacts of policies influenced by ILRI Capacity development have not been well documented. Estimating development impacts is more Dozens of graduate fellows have been trained by difficult. Moreover, many of the benefits of better ILRI and its partners as part of research into controlling zoonotic diseases come not from in- zoonoses. Many of these trained students have creasing incomes or improving nutrition of the poor taken up important roles in ILRI and other or- but rather from averting their losses, which is ganizations, including universities, ministries difficult to measure in the absence of a counterfac- and health services. ILRI scientists also super- tual. Certainly, individual projects demonstrated vised and taught a novel epidemiological educa- far-reaching improvements in capacity linked to tional course for the Field Epidemiology and development impacts. For example, projects in Laboratory Training Program within the Minis- South-east Asia piloted a village- based approach try of Health, Kenya, which has been evaluated to community control of rabies that was subse- positively. Veterinary staff members were trained quently extended to the entire island of Bali. by ILRI in several countries. ILRI also supported Millions of people have been reached by ini- the establishment of one-health centres in Thai- tiatives that aimed to protect human health and land and Indonesia and provided ongoing sup- reduce economic burdens as the result of timely port to the development of similar centres in detection of emerging diseases and an appropri- Côte d’Ivoire and Vietnam. These centres are ate response. For example, ILRI partnered several all operational (as of 2020), providing training stakeholders in Kenya to develop a decision- and conducting research in zoonoses. ILRI sup- support tool for better managing RVF. Together ported setting up and running a laboratory in with risk maps, this tool was incorporated into the town of Busia, in western Kenya, which has the country’s RVF contingency plan, which is helped improve understanding of livestock and the mainstay for mitigating the impacts of an human diseases and their links. In 2020, ILRI outbreak of the disease. We estimate that hun- launched a major initiative, the One Health dreds of millions of people lived in areas where Research, Education, Outreach and Awareness participatory disease surveillance was active. Centre. Its aim is to improve the health of humans, It is difficult to quantify the impact of these pre- animals and ecosystems through capacity build- paredness activities, but our best estimates suggest ing; strengthening local, regional and global that timely surveillance can reduce the impacts of networks; and evidence-based policy advice in a disease outbreak by 90% (Grace, 2014). Several the context of one-health by setting up a cen- of these programmes were operational for many tral facility for one-health in sub-Saharan Africa. years during the HPAI pandemic, and national programmes for RVF are still active in Kenya. In several cases, ILRI’s scientific outputs Introduction have been taken up by other development actors with probable benefits. For example, ILRI’s con- This chapter considers ILRI research on zoonoses tributions to research on the burden and distri- under two rubrics: (i) research on zoonoses bution of sleeping sickness in Uganda and on occurring in systems; and (ii) research on high- management options for better control of this priority emerging and neglected zoonotic disease influenced major donor investments that d iseases. Under the former, we prioritized diseases protected the health of hundreds of thousands and conducted studies on disease emergence and of cattle (although the intervention did not participatory disease surveillance, investigated 306 D. Grace et al. multiple zoonoses occurring in systems, and col- and SARS. Several ILRI scientists were involved laborated in regional and international initia- in emerging zoonoses research2. In 2011, fol- tives to control zoonoses. Under the latter, we lowing a research restructuring, zoonoses was conducted studies on antimicrobial resistance, once more the focus of a dedicated ILRI pro- avian influenza, Ebola virus disease, MERS-CoV gramme, namely, Animal Health, Food Safety and RVF, all of which are considered emerging and Zoonosis, led by Delia Grace. disease problems, and on key neglected zoonoses In 2018, another restructuring saw this including cysticercosis, brucellosis and sleeping programme joined with three other units work- sickness. ing on different aspects of health to constitute an Zoonoses were not within the research Animal and Human Health programme, jointly mandate of the International Laboratory for led by veterinary epidemiologist Delia Grace and Research on Animal Diseases (ILRAD), which, veterinary vaccine developer Vish Nene. Over the with the International Livestock Centre for years, ILRI has invested substantially in several Africa (ILCA), was a predecessor of ILRI. ILRAD high-priority emerging and neglected zoonoses, focused on African animal trypanosomiasis and including anthrax, avian influenza, brucellosis, East Coast fever. ILCA did not conduct dis- COVID-19, cysticercosis, Ebola, MERS, RVF and sleep- ease-specific research but frequently mentioned ing sickness. The outputs, outcomes and impacts of zoonotic diseases as key problems in Africa’s this work are summarized in the next sections. livestock systems and judged them important due to the harm they cause the continent’s live- stock sector and public health. Prioritization When ILRI was constituted in 1994, neglected zoonoses were considered diseases well researched Priority setting is essential for directing finite re- in high-income countries and therefore not high sources to activities that maximize benefits. As priorities for ILRI. However, opportunities for devel- discussed in Chapter 5 (this volume) on veterin- oping low-cost appropriate technologies for control- ary epidemiology, in 2002, ILRI was commis- ling neglected zoonoses in lower-income countries sioned to conduct the first study aiming to were not precluded from ILRI’s research agenda. determine the livestock diseases for which re- At that time (mid-1990s), emerging infectious search investments were most likely to alleviate diseases were not yet considered a global priority. poverty. Zoonoses were considered as a separate The first systematic approach to identify category, and ranking was done by experts at re- pro-poor animal health research priorities was gional workshops. The top three priorities were all conducted by ILRI in 2000 and identified several neglected zoonoses: brucellosis, cysticercosis and zoonoses as high priorities (Perry et al., 2002). leptospirosis. This landmark study influenced Following this, ILRI’s new strategy identified donor investment in research for development. impacts on the livestock livelihoods of the poor, In 2005, WHO and the UK Department for including those caused by neglected zoonoses, International Development (DFID) convened the as an important area of research. A programme first meeting to tackle neglected zoonoses, and on livestock impacts on human health was initi- several ILRI experts participated (WHO, 2006). ated in 2005, led by Tom Randolph. (In 2008, The meeting identified seven zoonoses – anthrax, this programme was incorporated into a broader bovine TB, brucellosis, cysticercosis, echinococ- agenda on livestock markets research.) cosis, rabies and zoonotic trypanosomiasis – as Because ILRI had limited expertise dedicated priorities. In 2012, ILRI developed the first glo- to zoonoses (as distinct from veterinary public bal assessment of zoonoses and poor livestock health), ILRI initiated a collaboration with the keepers (Fig. 8.1) (Grace et al., 2012b). This study Danish Bilharziasis Laboratory and the Swiss combined updated maps of poor livestock keep- Tropical Institute (STI) to develop a research ers with a literature review on the prevalence agenda on zoonoses1. Over the next decade, and impacts of zoonoses. The study assessed 56 work on neglected zoonoses continued. At the zoonoses, together responsible for around 2.5 same time, emerging zoonoses climbed rapidly billion cases of human illness and 2.7 million up the research agenda, driven by concern over human deaths a year. The most important diseases high-profile outbreaks of avian influenza, RVF in terms of burden and amenability to control Zoonoses 307 Number of poor livestock keepers per square kilometre 1–5 5–20 20–50 One or more people 50–100 or animals out of 100 Above 100 infected by one or more zoonotic diseases per year Fig. 8.1. Zoonoses and poor livestock keepers. (From Grace et al., 2012b.) were cysticercosis, leptospirosis and zoonotic animal disease, which will include zoonoses (Glo- gastrointestinal disease. This study was commis- bal Burden of Animal Diseases initiative). While sioned by DFID and was used to develop a major prioritization efforts have varied in their rigour funding initiative by DFID on zoonoses. and use of evidence, they have helped build a glo- From 2010, ILRI engaged with the Public bal consensus as to which of the many hundreds Health Foundation of India and was a member of of zoonoses are most important to poor people, the Roadmap to Combat Zoonoses in India initia- and this has also informed research globally on tive. This used a systematic and validated method neglected and emerging zoonoses. There are to prioritize a research agenda and identify pri- clear links between evidence produced by ILRI ority zoonoses as well as vulnerable populations and actions taken by donors in providing funding (Sekar et al., 2011). Subsequently, ILRI organized a for zoonoses research for development. series of one-health dialogues in India with the In 2015, ILRI scientists participated in a Indian Council of Agricultural Research, the systematic prioritization of zoonotic diseases for Public Health Foundation of India and other Kenya that was coordinated by the US Centers stakeholders (ILRI/ICAR, 2013; ILRI, 2014). for Disease Control and Prevention (CDC) and ILRI has continued to support evidence-based Kenya’s Zoonotic Disease Unit. The prioritization one-health approaches and coordination exercise identified the top five zoonotic diseases through workshops and training in India, and for the country: anthrax, brucellosis, rabies, RVF one-health continues to be supported by the Bill and trypanosomiasis (Munyua et  al., 2016a). & Melinda Gates Foundation as an important The prioritization tool has been used by govern- pathway to improving health care in India. ment researchers to allocate resources to specific ILRI scientists also contributed to the first surveillance, prevention and control campaigns. estimate of the global burden of food-borne dis- ILRI scientists have partnered with the Zoonotic ease (Havelaar et al., 2015), which estimated not Disease Unit to develop elimination plans for ra- only the high burden of food-borne disease but bies and to target research on brucellosis and also the importance of zoonoses and ani- RVF. ILRI scientists also sit on the Zoonosis mal-source foods. More recently, ILRI is a partner Technical Working Group, a panel of national in an initiative to develop a global assessment of experts in one-health. 308 D. Grace et al. Understanding drivers of disease Participatory disease surveillance emergence Participatory epidemiology is the systematic use ILRI conducted broad-based research to better of approaches and methods that facilitate the understand the drivers of zoonoses emergence. empowerment of people to identify and solve A foundational research paper by Jones et  al. their health needs. It should promote the direct (2008) had mapped all disease emergence events agency of individuals, leading to a shared learn- from the 1930s to the early 2000s. In collabor- ing environment that improves the understand- ation with Jones, ILRI developed an updated dis- ing of their risk perception, health risks, and ease emergence database focusing only on options for surveillance, control and health evalu- livestock (Grace et  al., 2012b). This reinforced ation in populations. Participatory disease sur- the importance of livestock in disease emer- veillance is a form of active clinical surveillance. gence. It also found that disease emergence ap- It involves the use of participatory approaches peared to be shifting from the western seaboard and is aimed at detecting clinical cases, which of Europe and the western seaboard of North can then be confirmed by specific biological America to low- and middle-income countries. tests. Participatory epidemiology evolved in the Another ILRI review summarized work link- 1990s as a new approach to working across cul- ing agricultural intensification to disease emer- tures in animal health surveillance and epidemio- gence (Jones et al., 2013). The review found strong logical research (Jost et al., 2007). evidence that modern farming practices and in- ILRI recruited scientists active in the devel- tensified systems were linked to disease emer- opment of participatory epidemiology from 2005, gence. However, the evidence was not sufficient who proceeded to introduce the approach into to judge whether the net effect of intensified the institute’s research programmes. ILRI has agriculture was more or less propitious to disease been involved in researching and supporting par- emergence than if land was left unused. Subse- ticipatory epidemiology and participatory disease quent fieldwork helped elucidate some of the surveillance for over a decade. Some examples relationships between land-use change and include the following: disease. Studies in Kenya found that degraded landscapes have more disease, but the relation- • Officials from 17 African countries received ship between biodiversity and disease is not training in participatory epidemiology and straightforward. Additional research addressed disease surveillance. the hypothesis that irrigation in dry areas would • In 2008, a participatory surveillance pro- increase the zoonoses burden by creating habi- gramme was introduced in Egypt to improve tats suitable for diseases; however, while irriga- animal health control activities through tion was associated with increased risk of some the use of participatory epidemiology. The zoonoses, adjacent pastoralist areas had in- programme eventually covered 53 districts creased risks of other zoonoses, suggesting a (30% of Egypt’s districts) in 15 governorates more complex interaction between ecosystems and at risk of avian influenza (Verdugo et  al., economic development (Bett et al., 2017a). 2016). ILRI scientists were also active in under- • A participatory disease surveillance programme standing the links between climate change in Indonesia began as a pilot in 12 districts with and disease emergence, often with a focus on 48 staff. It was rapidly scaled up, resulting in zoonoses. They were commissioned by the CRP more than 2000 practitioners in 31 provinces on Climate Change, Agriculture and Food Security by March 2009 (Azhar et al., 2010). to develop a paper on climate-sensitive livestock • Participatory disease surveillance was diseases that was incorporated into the plan- established in Pakistan as part of a rinder- ning processes of the United Nations Framework pest eradication campaign with support from Convention on Climate Change (Grace et al., 2015). the Food and Agriculture Organization of ILRI scientists are also members of the very the United Nations (FAO). influential Lancet Commission on Health and • ILRI has supported participatory epidemi- Climate Change, which has produced some of the ology and participatory disease surveillance most-cited papers from CGIAR (Watts et al., 2017, in Kenya for many years, including its use in 2018a,b). control of avian influenza and RVF. Zoonoses 309 An FAO review found that ILRI was the several CGIAR scientists presented papers. ILRI’s strongest contributor to participatory disease leadership thus facilitated development of a net- surveillance in terms of its projects and re- work subsequently sustained by its participants. search o utputs. Allepuz et  al. (2017) found Through PENAPH, ILRI advocated one- participatory epidemiology methods in 52 health applications of participatory epidemiology; countries. Countries where ILRI efforts were the techniques that emerged in animal health concentrated (Egypt, Ethiopia, Indonesia, were extended and adapted to public health Kenya, Nigeria and Pakistan) had more activ- (Mariner et  al., 2014). PENAPH core partners ities of that work (Fig. 8.2). include CDC and the African Field Epidemiology The international community of participatory Network. Participatory approaches to surveil- epidemiology practitioners recognized the utility lance for avian influenza were introduced in sev- of establishing a network to promote good practice eral countries in Africa, Egypt and Uganda in and to act as a training resource (Mariner et  al., particular. The approach has had lasting im- 2011). ILRI agreed to facilitate the process, and pacts on epidemiological institutions in ILRI’s the Participatory Epidemiology Network for Ani- partner countries. mal and Public Health (PENAPH) was  formed in 2008 with nine core partners, including inter- national agencies, non-governmental organiza- Zoonoses in Systems tions and universities with demonstrated commitment to participatory epidemiology. Ini- In 2006, ILRI started collaborating with STI tially, ILRI hosted the network and the secretariat (now the Swiss Tropical and Public Health Insti- with support from the Rockefeller Foundation tute) to address the strategic methodological for establishment and activities in two projects challenge of integrating veterinary and medical over 4 years. PENAPH continues to operate today as assessments of the impacts of zoonotic disease a self-sustaining network working across Africa burden on the livelihoods of the poor. STI and Asia. The secretariat is now hosted by Tufts had found that, because zoonoses imposed University, Massachusetts, and ILRI continues to costs in both the veterinary and health sectors be a core partner. PENAPH hosted its second inter- and because these were not integrated, the bene- national conference in Thailand in 2018, at which fits of zoonoses control were underestimated. United Kindom Belgium France TurkeyAArrmeennlaiar Japan Afghanistan Pakistan Nepal Egypt India Bangaladesh Taiwan Mali Laos Senegal Sudan Thailand Ghana Nigeria Ethiopia Cameroon Sri Lanka Congo Kenya InInddooneessiaia Tanzania, United Republic of Brazil Angola Zambia Zimbabwe Madagascar Number of PE activities Namibia 1 ––1 1 Australia South Africa 2 ––3 3 4 ––8 8 9 ––1 166 17 ––3 300 No PE activities Fig. 8.2. Participatory epidemiology activities by country. (From Allepuz et al., 2017.) 310 D. Grace et al. By integrating costs, they were able to show the TB was highest in better-off families (de Glanville ‘double benefits’ of brucellosis control in Mon- et  al., 2017). This is relevant to understanding golia ( McDermott et al., 2013). In collaboration the specific health benefits of reducing poverty. with ILRI, this approach was extended to RVF in Research on urban zoonoses started in the Kenya (Kimani et al., 2016). early 2000s with support from a CGIAR system- Africa is the least-irrigated continent, and it wide initiative. As little was known about urban is predicted that more irrigated areas will be de- zoonoses, this work focused first on generating veloped in the coming decades. However, there is evidence. Some findings led to further re- long-standing concern about the health impacts search-for-development activities. For example, of irrigation, especially in arid areas. A multi- Cryptosporidium, a protozoan zoonotic parasite, year project in Kenya investigated links between was found to be unexpectedly high in urban irrigation and zoonoses. A key finding was that dairies in Nairobi, which led to another project RVF virus (RVFV), West Nile virus and dengue focusing on Cryptosporidium. While generating virus were more prevalent in irrigated than in additional epidemiological evidence, the project pastoral areas while Leptospira spp. and Brucella staff also engaged with policy makers and devel- spp. bacteria were higher in pastoral than in irri- oped a behavioural approach to extension, one gated areas (Bett et al., 2017a). The same project based on encouraging farmers to implement investigated sociological aspects relevant to good practices out of concern for their social sta- control. For example, the standard health advice tus rather than out of fear about contracting an was to bury animals that had died of disease, but illness. This was evaluated as successful both in studies in communities found that this was improving farmer practices (Kang’ethe et  al., taboo, reporting, ‘We do not bury dead livestock 2012a,b) and in shifting policy in a pro-poor like human beings’; they preferred to eat them d irection (Nyangaga et  al., 2012). Because (Mutua et al., 2017). This showed that the com- cryptosporidiosis is a priority disease of im- mon approach to public health communication munosuppressed people, Kenyan medical au- of telling people to bury dead animals was inef- thorities had been wary about promoting live- fective and other methods were needed to change stock keeping among people with human behaviour. immunodeficiency virus/acquired immune defi- In 2015, ILRI, along with the University of ciency syndrome (HIV/AIDs). An important find- Edinburgh, UK, and other partners, established ing was that cryptosporidiosis in Nairobi was a field site in western Kenya for the study of zoo- transmitted more by person-to-person than by notic diseases. This allowed them to study neg- animal-to-human contact and that people with lected zoonoses in both livestock and the families HIV/AIDs receiving antivirals did not have an that keep them, and to gather the essential elevated risk of cryptosporidiosis. These findings baseline data about these infections and the helped safeguard their access to livestock. populations affected by them that are so severely In Uganda, a policy-relevant finding from lacking. This research identified risk factors for urban zoonoses research was that farmers who zoonotic diseases in people and animals (Fèvre had experienced more harassment from author- et al., 2017). The diseases investigated included ities had fewer good practices (Grace et al., 2008), cysticercosis, leptospirosis, Q fever, RVF and and the project contributed to a better policy taeniasis (tapeworm infection) (Wardrop et  al., environment for urban farming (Cole et  al., 2016; Cook et al., 2017). 2008). Many of the results from this early work The project identified slaughterhouse work- on urban zoonoses were summarized in a special ers as high-risk groups for zoonotic diseases; this edition of Tropical Animal Health and Production information is being used to explore interven- (Grace et al., 2012c). tions to reduce this occupational group’s risk of A more recent urbanization project focused exposure. The study demonstrated the existence on the emergence of zoonotic pathogens. This of a socio-economic gradient within households had several methodological innovations, includ- in rural Kenya, determining individual infec- ing the use of a livestock value chain approach tious disease risk. The risk of infection with that allowed identification of the stakeholders, amoeba, hookworm and malaria was highest in drivers and dynamics of urban livestock keep- poorer households and the risk of contracting ing, as well as the first-ever quantification of the Zoonoses 311 contribution that urban livestock keeping made to nutrition, economics and markets in the city Box 8.1. A successful intervention to control zoonoses in South-east Asia. (Alarcon et al., 2017; Carron et al., 2017). An- other innovation in this work was the use of bal- An ILRI-supported team in Indonesia built on an loons with cameras attached to them, snapping existing village cadres system to establish Village images every second as the balloon handler Rabies Working Groups for rabies control. These walked through urban slums to map, for the first groups consisted of paraprofessionals equipped time, food kiosks, mobile street vendors and haz- to raise awareness about rabies in schools, vil- ards such as rubbish dumps and open sewers. lage meetings and small groups in their own At the same time, another large project fo- homes. They also served as first responders to dog-bite cases, ensuring that victims received cused on urban zoonoses in India and Vietnam. rapid post-exposure prophylaxis, which is most These projects have had scientific impacts in effective when given soon after the bite. General terms of generating both new findings and new information on rabies and what it means to be ways of researching urban zoonoses and have a responsible dog owner encouraged commu- helped to build local capacity in urban zoonoses nities to register and vaccinate their dogs, two issues (Bett et al., 2019; Jakobsen et al., 2019); it is evidence-based ways of controlling rabies. The too early to observe any development outcomes. model was recognized by provincial-level leaders A large research programme in six countries as a promising community intervention. As a of South-east Asia explicitly adopted an Eco- result, a legal decree was made to adopt the vil- Health approach. Each country was supported lage rabies cadre system by officially appointing two persons to serve in this capacity in each of to systematically identify priorities: for example, the 723 villages in Bali. In addition, the ILRI team the Indonesian island of Bali identified rabies, partnered with the provincial-level leaders to the Laos team focused on pig zoonoses and the provide technical training for the rollout of Village South Vietnam team identified leptospirosis. This Rabies Working Groups in 30 villages that were was a strong departure from the usual mode of hotspots for rabies (Gilbert et al., 2014). zoonoses research, in which donors funded pro- jects according to their own priorities. Country the overall success of the use of the EcoHealth teams bringing together researchers and imple- approach in the region was demonstrated by the menters from animal health and social sciences scope and scale of activities collectively encom- were trained in research using EcoHealth prin- passed by the projects, programmes and initia- ciples, which require the involvement of the local tives reviewed. In a relatively short period of target communities and relevant policy makers. time, EcoHealth has been widely accepted and Outcome mapping was used to identify the changes has gained a remarkable amount of exposure in desired and to monitor success in achieving them. South-east Asia. This, in turn, has led to more In all countries, there was evidence of capacity effective and efficient health delivery. being built in these areas, and in most cases, teams were able to show community- and policy-level outcomes and impacts as well (Box 8.1). Assess- Zoonoses in smallholder pig systems ment showed that this novel approach was appreciated and found useful, but its longer- In 2012, the ILRI-led multicentre CRP on Live- term benefits and sustainability are less clear. stock and Fish started and identified two small- However, the project generated actionable scien- holder pig value chains among the nine livestock tific findings such as an identification of the value chains it considered most promising for most common pig diseases in Laos (Holt et  al., research that would benefit poor farmers. This 2019). led to comprehensive initiatives to assess and On a regional scale, ILRI played an import- manage pig-related zoonoses. Trichinellosis is a ant role in promoting the EcoHealth approach to parasitic disease caused by nematodes in the better control zoonoses and emerging diseases in genus Trichinella. This disease, commonly known South-east Asia. A review in 2015 (Nguyen-Viet as ‘pork measles’, historically caused outbreaks et al., 2016) traced the history of EcoHealth in of severe illness in people, but the disease nearly South-east Asia and showed the substantial role disappeared when control measures were devel- played by ILRI. It found that, in spite of barriers, oped and adopted by many pork-producing 312 D. Grace et al. countries. Countries new to pork-keeping, how- the day outside the home environment (Thomas ever, had not investigated its potential risk. ILRI et al., 2013). Pigs may interact with other live- conducted the first study on Trichinella spp. in stock and wild animal species, which may ex- East Africa, which confirmed that trichinellosis pose them to infectious agents. Further research was present and further suggested that trichinel- in the region demonstrated a high prevalence of losis might be shifting from a sylvatic transmis- non-typhoidal Salmonella spp. and Leptospira sion cycle, in which the pathogen cycles between spp. in pigs at slaughter, which poses a risk to wild animals and vectors, to a domestic cycle, in both slaughterhouse workers and consumers. which it cycles between domestic animals and A large study in pigs in Uganda found that, vectors, and thus might be in the process of spill- contrary to expectations, swine brucellosis was ing over from wildlife to establish itself in domes- not present. This meant that human health ser- tic pigs (Roesel et  al., 2016). Veterinary and vices could rule out swine brucellosis and remove a medical authorities were not aware of the pres- potential barrier to pig marketing and trade ence of trichinellosis or the risk that it was chan- (Erume et  al., 2016). Another example from ging its epidemiology, so this was valuable Uganda shows how taking a systems approach information. In Vietnam, where pig keeping is to zoonoses studies can lead to unexpected find- long established, Trichinella spp. antibodies were ings and eventually to development outcomes. detected at high levels (12%) in the serum of in- Systems approaches do not start by choosing a digenous pigs kept in the central highlands (Unger specific zoonotic disease to target but instead et  al., 2016). Studies in Laos on trichinellosis explore all potential disease problems. During revealed a seroprevalence in pigs of 14.4% (in participatory rural appraisals conducted in 2013, Luang Prabang) and 9.3% (in Savannakhet) farmers reported symptoms suggestive of diamond (Holt et al., 2016). skin disease. This is caused by the bacterium Er- Streptococcus suis is a leading cause of bac- ysipelothrix rhusiopathiae and is an economically terial meningitis in Vietnamese adults. The important disease of swine that had never been major risk factors have been identified as con- reported in Uganda. If transmitted to humans, it sumption of raw pig blood in a Vietnamese dish can cause erysipeloid, which manifests as skin of blood and cooked meat (tiêt canh) and occupa- lesions, and in more severe cases can have sys- tional exposure to pigs. ILRI conducted the first temic effects and even lead to death. The farmer study in northern and central Vietnam and reports triggered an epidemiological study in found S. suis type 2 in 1.4% of pig tonsils. pigs and pork. Overall, 67% of the pig sera car- Slaughterhouse workers were found habitually ried antibodies against E. rhusiopathiae and 45% to consume raw pig blood to an even greater of the fresh pork samples were contaminated with extent than consumers, and this was linked to E. rhusiopathiae (Musewa et  al., 2018). This, in excessive consumption of alcohol (Dang-Xuan turn, led to a study to determine the prevalence et  al., 2015). Identifying groups at high risk and factors associated with E. rhusiopathiae in- allows targeted interventions. Studies in Vietnam fection among raw pork handlers, which found investigated cysticercosis, dengue, leptospirosis, a prevalence of around 10%. This was the first rabies and other zoonoses, often for the first time time that the disease had been reported in humans in a specific region (Lee et al., 2020a,b). Models in East Africa. Because erysipeloid is an easily were also developed to inform control strategies diagnosed and treated disease, alerting patients (Bett et al., 2019; Lee et al., 2019). Information and health providers to its presence was necessary. was shared with authorities, and high impact ILRI paid for the treatment of all positive human factor papers produced but the outcomes in cases and all findings were communicated to terms of improvements in human health have stakeholders. However, these treatment impacts not been evaluated. were not formally assessed. A study on smallholder pig farmers in western Kenya was among the first to use tracer collars to understand the movement of pigs in free- Collaboration in international initiatives ranging pig production systems. This study found that pigs move considerable distances each day In September 2005, a memorandum of under- (on average 4340 m) and spend almost 50% of standing was signed by ILRI and WHO to better Zoonoses 313 understand links between livestock keeping and previously but are increasing in incidence or the health and general well-being of poor people geographical range (Morens et al., 2004). These in poor countries. This engagement with WHO diseases are caused by: (i) newly evolved patho- led to a major multi-stakeholder meeting in gens; (ii) pathogens that have spread to new Nairobi in 2007 hosted by ILRI. The meeting ap- areas or populations; and (iii) re-emerging patho- preciated that controlling, preventing and even- gens such as those associated with demographic, tually eliminating neglected zoonotic diseases environmental or other societal changes. Emer- would be highly cost-effective from a societal ging diseases are more likely than not to be zoo- point of view, taking into account both the notic in origin (Woolhouse and Gowtage-Sequeria, health and agricultural aspects. A plan of action 2005), i.e. acquired by humans from animal for implementing integrated control of neglected reservoirs, including livestock (Cleaveland et al., zoonotic diseases in Africa was recommended. 2001). Emerging infectious diseases are largely ILRI scientists also made major contributions, associated with ecological changes that destabilize including developing a methodology for priori- existing equilibria among pathogens, hosts and tizing diseases, to a WHO report on human in- vectors. These ecological changes are often the fectious diseases of poverty (WHO, 2013). result of human intervention, such as hunting, ILRI was a lead organization in developing land clearing for crops or for livestock, urbaniza- one-health approaches in low- and middle-income tion and irrigation. countries. As such, it engaged with international organizations promulgating one-health and Eco- Health and contributed to many of the global Rift Valley fever initiatives around one-health, including inter- national meetings on avian influenza, the Stone One new zoonotic disease is now emerging every Mountain Dialogue, the International One Health 4  months. While many are trivial, a minority Congresses and meetings at Bellagio, Italy, and have devastating health and economic impacts Chatham House, UK (Galaz et  al., 2015). One- (e.g. avian influenza, COVID-19, HIV/AIDS and health has clearly emerged as a powerful and Spanish flu). Agricultural expansion and intensifi- dominant approach for managing zoonotic dis- cation as well as climate change are driving an eases, with most evaluations of this approach accelerated emergence of these new diseases. citing positive impacts, although there are some Yet without proper assessment, undue emphasis concerns about gaps between theory and imple- on emerging infectious diseases may deflect mentation. Figure 8.3 shows how ILRI is embed- interest from diseases of more importance to ded within one-health research groups. the poor. RVF is a model emerging infectious dis- In Asia, ILRI is one of the active institutional ease because it causes severe disease in people members of the Vietnamese Government’s One but is not yet readily transmitted between people, Health Partnership for Zoonoses and a member it is sensitive to change in land use, and it has of Gestion des Risques Emergents en Asie du implications for trade and for the most poor and Sud-Est (GREASE, Management of Emerging vulnerable populations. ILRI’s geographical lo- Risks in Southeast Asia), initiated by the Centre cation also means that it has a comparative ad- de Coopération Internationale en Recherche vantage in studying this important disease. Agronomique pour le Développement (CIRAD). RVF is a mosquito-borne viral zoonosis that ILRI is also working with the Southeast Asia One mainly affects sheep, goats, cattle, buffaloes and Health University Network, based in Chang Mai, camels. Humans become infected following a Thailand, to develop the capacity of partners in bite from an infected mosquito or after close one-health work. contact with acutely infected animals or their infected tissues. In people, the disease manifests as a mild influenza-like syndrome in most cases (over 80%) or as a severe disease with haemor- Emerging Infectious Diseases rhagic fever, encephalitis or retinitis in a few cases (Njenga et al., 2009). In livestock, the dis- Emerging infectious diseases are those that have ease manifests as extensive abortions and peri- newly appeared in a population or have existed natal mortality. 314 D. Grace et al. Oklahoma State Univ Texas Anim Hlth Commiss Texas AnandM Univ Yale Univ CSIC UCLM JCCM Washiington State Univ Univ Texas Med Branch Univ Autonoma Tamaulipas Georgia Inst Technol USDA Grorgetown Univ Univ Minnesota Pacic Northwest National Laboratory Pacic NW Natl Lab US Centers for Disease Control and Prevention US Dept Def Sandia National Laboratories US Department of Health and Human Services National Center for Medical Intelligence Dierenartsen Zonder Grenzen Vet Sans Frontieres Univ Niamey Med Sans Frontieres Belgium US Department of Homeland Security Univ Calif Davis Univ Wesern Australia Cantholic Univ Louvain International Livestock Research Institute (ILRI) Vet Sans Frontieres Belgium Department of Health and Welfare of the Can Hassan II Agrovet Inst Damien Foundation Univ Salford Univ Bern National Institute of Sanitary Administration Inst Trop Med Univ Basel Cantonal Institute of Microbiology Swiss Trop and Publ Hlth Inst Inst Continuing Med Educ loannina Univ Hosp Ctr Canton Vaud Norwegiar Sch Vet Sci Univ London Univ Pretoria Univ Zurich Robert Koch Inst Univ Zimbabwe Univ Edinburgh Univ Zambia Cantonal Medical Office Fed Ubst Risk Assessment Vet Labs Agcy Royal Inrm Edinburgh NHS Trust Sokoine Univ Agr Rhein Westlal TH Achen Scotlands Rural College Bundeswehr Hlth Protect Scotland Centro de Investigaciony Tecnologa Aplicada (CITA) Queensland Inst Med Res Univ Clasgow Univ Navarra Univ Oklahoma Veterinary and Agrocheinical Research Centre Univ Liverpool Univ Liege Pirbright Inst Univ Peruana Cayetano Heredia Boston Univ INRA Philippine NGO Council Populat Hlth and Welf Univ Namur Riseal Niger Loeffler Inst Brock Univ Univ Khartoum Univ Alberta WHO Hlth Canada CLTS Foundation Agriculture and Agri-Food Canada Publ Hlth Agcy Canada nvironm Canada Alberta Environm Univ British Columbia Fig. 8.3. Network digram of co-publishing patterns among organizations having six or more co-authorship relationships (ILRI in blue). (Adapted from Galaz et al., 2015.) RVF outbreaks occur after periods of above- observed from 2008 to 2011 were associated normal precipitation associated with the warm with relentless and widespread seasonal rainfall phase of the El Niño/Southern Oscillation phe- and high soil saturation (Williams et al., 2016). nomenon (Bett et  al., 2017b). Outbreaks have In Mauritania, an outbreak was associated with occurred previously in Egypt, Kenya, Madagas- a fourfold increase in rainfall in a desert region car, Mauritania, Mayotte (French archipelago in in 2009–2010 and affected small ruminants, the Indian Ocean), Saudi Arabia, Senegal, South camels and people (Faye et  al., 2014). Similar Africa, Sudan, Zimbabwe and Yemen (Nanyingi outbreaks occurred in Senegal from 2013 to et  al., 2015). In South Africa, RVF outbreaks 2014, where the situation was exacerbated by Zoonoses 315 livestock movements that aided dissemination of Developing a more effective vaccine the virus (Sow et al., 2016). No licensed vaccines are currently available to protect humans against RVF and those widely Understanding transmission used in livestock have major safety concerns. and burden in Kenya A one-health vaccine co-developed for multiple Data from studies identifying rainfall patterns susceptible species is an attractive strategy for associated with increased risk of RVF provided RVFV. In partnership with the Jenner Institute, the basis for development of dynamic models by UK, and the Kenya Agricultural and Livestock ILRI and partners for evaluating the transmission Research Organization (KALRO), ILRI is devel- dynamics of this disease (Gachohi et al., 2017). oping an adenovirus-vectored vaccine called The model allowed researchers to determine al- ChAdOx1-GnGc to protect both livestock and ternative vaccination strategies for RVF. ILRI led humans from RVF. a related study, which revealed a positive associ- The vaccine has passed critical safety and ation between flood irrigation and endemic trans- challenge studies. A single-dose immunization mission of the virus in an arid and semi-arid area elicited high-titre neutralizing antibodies and of Kenya (Sang et al., 2016; Bett et al., 2017a; provided solid protection against the disease in Mbotha et al., 2017). The Kenya study demon- the most susceptible natural target species of the strated that standing water enhanced mosquito virus – sheep, goats and cattle (Warimwe et al., development, including the primary vectors of 2016). The vector being used is a replication- RVFV, which included Aedes mcintoshi, Aedes deficient chimpanzee adenovirus (ChAd), with a ochraceous, Culex univittatus and Culex pipiens. high capacity to insert genomic clones and plas- The effects of flood irrigation on RVFV mids that encode desired antigens. The clone endemicity had not been explored previously. used in this case had a genetic sequence encod- Earlier environmental impact studies of irrigation ing the RVFV Gn and Gc envelope glycoproteins. mainly considered malaria and schistosomiasis The methods used to identify the most appro- as case studies of irrigation and vector-borne dis- priate vector and to develop the clone are given eases (Ijumba and Lindsay, 2001; Keiser et  al., by Warimwe et al. (2013). Clinical trials are still 2005a,b). We estimated that RVF induced losses required to evaluate the vaccine before it is of more than US$32 million (2.1 billion Kenyan a pproved for use. Results to date suggest that shillings) on the Kenyan economy, based on its the vaccine has great potential given its safety negative impacts on agriculture and other sectors characteristics and ability to generate good (e.g. transport, services) alike (Rich and Wany- levels of neutralizing antibodies. The new vac- oike, 2010). cine is likely to achieve higher levels of uptake ILRI also made the first-ever assessment due to its safety margin being higher than the of the impact of RVF on human health in Kenya existing livestock vaccines, which may cause using a disability-adjusted life year (DALY) abortions. approach. (This is a standard human health Studies on RVF vaccines have focused on metric: one DALY corresponds to 1 year of lost the production processes, safety and efficacy healthy life.) This was used to model the cost- standards, but those on uptake and adoption effectiveness of livestock-based control of RVF levels are rare. ILRI conducted a study that iden- from a public health perspective in Kenya. This tified barriers faced by men and women farmers ILRI assessment found that improving livestock in the uptake of livestock vaccines (Mutua et al., vaccination coverage before a hypothetical out- 2019). break could avert close to 1200 DALYs. Improved vaccinations showed cost-effectiveness values Supporting response in East Africa of US$43–53 per DALY averted (Kimani et  al., 2016). These findings on the economic and human Studies conducted by ILRI in Kenya and Tanza- health impacts of RVF helped raise awareness of nia following a 2006/2007 RVF outbreak in its importance and were cited in subsequent na- these nations indicated that the impact of the tional policy documents and successful research outbreak was exacerbated by delays in recogniz- proposals. ing the risk and in implementing prevention and 316 D. Grace et al. control measures (Jost et al., 2010). Many stake- objective of increasing the capacity of veterinary holders later identified the need to develop a services in practical, community-focused, active framework to promote timely decision making and surveillance in 11 countries in West and East Af- to improve targeting of prevention and control rica. This project developed risk maps, which measures while encouraging closer collabor- were used to develop risk-targeted surveillance, ation among research and disease control insti- and conducted risk factor analyses to inform tutions. In partnership with the African vulnerability and control work. Union–Interafrican Bureau for Animal Re- There was a large capacity-building compo- sources (AU-IBAR), CDC, FAO and the Kenya nent. The training covered participatory epidemi- Department of Veterinary Services, ILRI has devel- ology, data management, risk mapping using oped a RVF risk map (Munyua et  al., 2016b), geographic information systems (GIS) software contingency plans and a decision model (Con- and use of risk maps. A subsequent independent sultative Group for RVF Decision Support, 2010) assessment of the impacts of the project’s cap- to be used jointly to determine optimal interven- acity building in participatory epidemiology and tions during an outbreak of RVF. The decision participatory disease surveillance on national model breaks down the epidemic cycle into phas- infectious disease surveillance was generally posi- es and identifies actions to manage the disease in tive. These capacity building materials were both animals and humans. ILRI has also been a integrated into regional field epidemiology and la- member of a RVF task force in Kenya, which co- boratory training programmes supported by the ordinates surveillance and response, especially CDC and used for additional surveillance and dis- during periods of heightened risk. ease investigations. A participatory surveillance A search for ‘Rift Valley fever’ in the Altmet- team diagnosed peste des petits ruminants in Ni- ric database (www.altmetric.com/; accessed 18 geria, which led to an effective emergency disease February 2020) shows that ILRI’s RVF projects control programme. In 2010, Nigeria chose to use have generated 91 outputs of the 632 outputs in participatory epidemiology practitioners trained this field overall, with the ILRI outputs being by the EDRSAIA project to investigate animal dis- mentioned 291 out of 1986 times and being eases in five regions (Mariner et al., 2014). quoted in nine policy documents. Response models in Egypt and Indonesia Highly pathogenic avian influenza Indonesia saw a major HPAI outbreak in 2004. By the end of 2005, it had spread to more than 23 Highly pathogenic avian influenza (HPAI) threatens provinces and more than 10 million birds had poultry industries and livelihoods worldwide as died. In January 2006, working with FAO, the well as human health. The HPAI Asian subtype government of Indonesia began a programme in H5N1 is especially deadly for poultry. The virus Participatory Disease Surveillance and Response was first detected in 1996 in geese in China, and (PDSR) for HPAI in poultry. This project was human cases were first detected in 1997. Be- partly driven by ILRI research on participatory cause of its ability to cause human cases and the epidemiology. possibility that it could evolve to cause a human By 2009, the FAO/Indonesia programme pandemic, the emergence of this disease in 2003 was operating in 27 out of 33 provinces of Indo- was of great concern. nesia. About 20,000 villages (30% of all villages in Indonesia) and 2.5 million backyard poultry Supporting surveillance in Africa producers were covered by surveillance, control and prevention activities, indicating the ability As HPAI threatened to become a pandemic, there of PDSR to deliver animal health cover at scale. was concern that countries in Africa were A subsequent external evaluation found that unprepared to cope. An early ILRI activity was PDSR did not appear to have had a significant im- providing training in relevant laboratory tech- pact on the prevalence of HPAI, partly because it niques. A project on Early Detection, Reporting focused on the backyard sector while much of the and Surveillance for Avian Influenza in Africa disease was driven by commercial firms. However, (EDRSAIA) ran from 2007 to 2012 with a main the PDSR programme did introduce two valuable Zoonoses 317 approaches: (i) information, education and com- ‘targeted’, substantially reducing the negative munication activities were well planned, sup- impacts of disease control on the poultry liveli- ported and executed, and most targeted people hoods of poor people. had good knowledge of HPAI; and (ii) participa- tory and pro-poor animal health services were Evaluation of control in Indonesia strengthened. and Nigeria Egypt has some of the highest poultry densities on the African continent, with most When HPAI H5N1 hit Nigeria, the first African birds concentrated along the River Nile. After country it occurred in, in early 2006, the fed- the first Egyptian outbreaks of HPAI H5N1 in eral government of Nigeria requested a World poultry in February 2006, there was wide- Bank/International Development Association spread culling of both commercial and house- credit of US$50 million, provided under the hold poultry. In 2008, a participatory disease Global Program for Avian Influenza and Human surveillance programme was introduced in Pandemic Preparedness and Response, to fight Egypt, with technical support from ILRI. By its ongoing outbreak. ILRI was asked to conduct 2014, it was covering 30% of Egyptian districts, an independent evaluation of Nigeria’s re- again demonstrating the potential to deliver im- sponse to avian influenza. A broad-reaching pact at scale. In 2011, when the overall HPAI multidisciplinary evaluation found overall posi- surveillance had slowed, the PDSR programme tive impacts but drew attention to areas that proved resilient, contributing over 50% of the needed to be addressed to ensure that the bene- HPAI confirmed cases in 2012. fits were sustainable. While cover and resilience of the initiative In Indonesia, an FAO-coordinated response were high, accuracy was moderate (Verdugo addressed the HPAI pandemic. However, as of et  al., 2016). Therefore, the results suggested 2007, efforts to control the disease in small-scale that the PDSR programme might best suit epi- commercial and backyard flocks had not demic situations or those where a high rate of proven effective. ILRI was asked to evaluate false positives is acceptable, given a high need to intervention strategies against HPAI in back- detect true positives. yard and small-scale commercial farms by assessing the feasibility of implementing the Identifying and supporting pro-poor interventions. ILRI’s evaluation consisted of a interventions for disease control longitudinal study on intervention options and specific studies targeted to epidemiological stud- An initial response to controlling HPAI world- ies. The longitudinal study found vaccination in wide was the culling of infected birds, an effective the backyard sector was effective in backyard approach in high-income countries. However, if poultry (Bett et  al., 2015). The incidence of smallholders in poor countries are not compen- HPAI declined by 12% in the HPAI-vaccinated sated for the loss of their culled birds, they may group and by 24% in the HPAI-plus-Newcastle be reluctant to report HPAI outbreaks and thus disease-vaccinated group. help spread the disease. In partnership with the However, vaccination appeared to be un- International Food Policy Research Institute feasible as an open-ended programme because (IFPRI), ILRI undertook research into pro-poor the cost of avoiding one poultry death was far interventions in Africa and South-east Asia. greater than the value of a bird (Lapar et  al., These interventions explicitly used a risk-based 2012). However, it was found that vaccination approach and built in capacity development work might have a role as a short-term targeted in the use of risk analysis. The research contrib- r esponse. This information was crucial for uted to a better understanding of how HPAI is p lanning public-sector control of HPAI. The tar- spread, what the critical control points for HPAI geted epidemiological studies also had useful re- risk mitigation are, and what would be cost- sults; for example, a study of village chicken found effective, pro-poor risk-reduction strategies. For that vaccination was needed every 4 months due example, in the Mekong Delta, this information led to high chicken population turnover; such infor- countries to coordinate HPAI control efforts and mation is, of course, essential for planning effect- to revise their culling strategies from ‘radical’ to ive vaccination campaigns (Unger et al., 2014). 318 D. Grace et al. Middle Eastern respiratory syndrome a reflection of the importance of ILRI’s role in studies in the region. ILRI scientists have been Middle Eastern respiratory syndrome-coronavirus involved in establishing MERS serodiagnostics at (MERS-CoV) is an emerging virus first identified the Biosciences eastern and central Africa (Be- in 2012 in Saudi Arabia. The case fatality rate for cA)-ILRI Hub laboratories and hosted a regional infected humans with overt respiratory symptoms is in FAO training programme for government veter- the region of 30%. As with many coronaviruses, inary staff to learn about MERS diagnostics. MERS-CoV is thought to have originated in bats, al- A search of the Altmetrics database for though camels appear to play a significant role in ‘MERS-CoV’ indicated that ILRI generated three maintaining the virus and transmitting it to humans. out of 325 outputs on this subject. The ILRI out- While the virus appears to be poorly transmissible to puts have been cited in four policy documents, humans, it can lead to human-to-human infections. including FAO’s EMPRES360 Animal Health Indeed, the largest outbreak of the disease was in bulletin (Issue No. 46, 2016) and the UK gov- South Korea in 2015, when human-to-human ernment’s Infectious Disease Surveillance and Moni- transmission resulted in hundreds of infections. toring System for Animal and Human Health (Summary, March 2016). In early 2020, ILRI started to plan a re- Understanding disease reservoirs search agenda around COVID-19, caused by and transmission SARS-CoV-2 (severe acute respiratory syndrome ILRI and partners have explored the role of camels coronavirus 2), a virus related to SARS-CoV, in the epidemiology and transmission of MERS- which causes MERS. ILRI has contributed to CoV, with a focus on Kenya. The first important evaluation of a diagnostic (Hoffman et al., 2020) work was to mine biobanks of historical camel and is undertaking studies to understand the samples as far back as the early 1990s, which con- impact of COVID-19 on farmers, value-chain firmed extensive exposure of camels to MERS or actors and consumers, their responses to the MERS-like viruses (Corman et  al., 2014; Muller pandemic, and their knowledge, attitude and et  al., 2014). High seroprevalences (up to 50%) practices related to COVID-19. were identified in herds in central Kenya (Deem et al., 2015) and in herds from locations around the whole country. Liljander et al. (2016) identi- fied one seropositive human in Kenya, which was Ebola the first reported human case in Africa. The Kenya studies were important in elucidat- Ebola was first described in 1976 with two simul- ing MERS epidemiology historically and in the pre- taneous outbreaks in the Democratic Republic of sent. A working hypothesis is that MERS is a Congo (Zaire at the time) and South Sudan. It has geographically widespread coronavirus that ac- since been responsible for several deadly out- quired an ability to infect humans in the Middle East, breaks, most recently from 2013 to 2016 in West but that this zoonotic trait has not spread through- and Central Africa and more recently (2019– out the range of the virus. ILRI researchers have 2020) in two outbreaks in eastern and north- been collaborating with the government of Kenya western regions of the Democratic Republic of the and the Emerging Pandemic Threats-2 programmes Congo. Many aspects of Ebola epidemiology re- of the US Agency for International Development main unknown, such as the virus reservoir and the (USAID) to isolate the MERS virus itself (a challenge role of wildlife and domestic animals in its transmis- given its transient nature in the camel host) and to sion. However, there has been long-standing con- carry out virus sequencing. This will allow place- cern that domestic livestock, in particular swine, ment of the East African virus in a global phylogeny might have a role in its disease epidemiology (Ath- and identification of its full host range. erstone et al., 2015). Pigs in the Philippines were shown to be infected with Reston virus, one of six Improving the national response known viruses within the genus Ebolavirus and re- lated to the Ebola virus found in Africa but that ILRI researchers have been members of the does not affect people (Marsh et al., 2011). Moreover, government of Kenya’s MERS Working Group, pigs have been experimentally infected with Zaire Zoonoses 319 Ebola virus (Kobinger et al., 2011) and antibodies to humans, and will provide the evidence needed to Ebola viruses have been found in pigs in Africa. to understand and manage Ebola outbreaks. First risk assessment in livestock systems Antimicrobial resistance Uganda has experienced four Ebola outbreaks Antimicrobial drugs are indispensable in man- since the disease was first identified, accounting aging infectious diseases in humans and animals. for 606 cases (probable and confirmed) and 283 If these drugs stop working, millions of human deaths (CDC, 2020). In Uganda, motivated by lives and livelihoods are at risk because of treatment interest to improve risk management, ILRI has failure. Drug resistence in the microbes infect- conducted ex ante assessments of the potential ing animals could lead to economic loss and re- health risks associated with the pig value chain duced yield of animal-source foods, fuelling in the country (Atherstone et al., 2014, 2015). global food insecurity. Drug-resistant pathogens This pioneering work is not only informing de- – viruses, bacteria, fungi and parasites – are velopment of healthy and risk-free pig produc- considered as emerging infections given that tion and pork value chains in the country but the genetic transformations that potentiate also providing an important basis for under- their emergence distinguish them from their par- standing missing links in the complex epidemi- ent generations. They also have the potential to ology of Ebola disease. spread between humans and animals and inter- ILRI’s research has produced a policy nationally through travel and trade, causing brief that summarizes Ebola dynamics, includ- higher numbers of infections globally. Resist- ing the role of bats (the main suspected reser- ance to antimicrobials – here defined as sub- voir of the Ebola virus) and a demonstration of stances of natural or synthetic origin that kill or how an integrated one-health approach is inhibit the growth of microorganisms – is not a being used to address the main research gaps new phenomenon; it has been a concern since (Smith et  al., 2015). Findings from ILRI’s re- the discovery of antimicrobials. However, the re- search on Ebola in Uganda will help protect the cent increases in microorganisms developing re- health of the more than 1.1 million Ugandan sistance to antimicrobials and the resulting households engaged in swine production. This increasing threat to public health (O’Neill, research will help policy makers develop re- 2015) warrant focused attention on the problem commendations to support safe systems for in both medical and agricultural sectors. ILRI pork production and supply with an emphasis research suggested that around twice as many on identifying suitable disease management antimicrobials are used in agriculture as in interventions for smallholder farmers, who treatment of humans (Grace, 2015). make up a large proportion of Uganda’s pig farmers. New collaborations with Australian, Canadian and German institutions have been Discovery of anti-parasitic drugs started to bring together expertise in develop- ing new diagnostics for surveillance of Ebola ILRI’s antimicrobial resistance research aims to viruses in livestock species. intensify livestock productivity while reducing Ongoing research will provide evidence on its negative externalities for human health. ILRI the role of pigs in Ebola transmission in Uganda assumes that the world’s poor farmers should and in contexts with similar agroecologies, will have access to antimicrobial drugs to safeguard identify places along the pork value chain that the health of their livestock, upon which so may increase Ebola transmission to humans and many depend. These principles have guided will evaluate areas in the country at high risk of work related to antimicrobials since the 1990s. Ebola. This work is the first to systematically charac- ILRI’s initial focus was on anti-parasitic drugs terize and describe Ebola viruses in pigs in Africa. and problems of emerging drug resistance in This information will inform policies around pork parasites given the importance of parasites to value chain development, may help to predict and smallholder livestock production. In addition to prevent the spread of emerging zoonotic disease research on drug use, ILRI scientists contributed 320 D. Grace et al. to the development and testing of new anti- led to increases in ILRI research on the use of parasitic compounds, such as plant compounds antibiotics in these systems. ILRI scientists have for the control of helminths. For example, Githi- closely collaborated with others to review global ori et al. (2002) evaluated the effectiveness of a antimicrobial use in livestock production sys- widely used natural compound derived from tems (van Boeckel et al., 2015) and have led dis- Cape myrtle (Myrsine africana) and Cape beech cussions highlighting the need for one-health (Rapanea melonophloeos) against the helminth approaches to controlling and preventing anti- Haemonchus contortus, a pathogen of sheep, microbial resistance (Robinson et al., 2016). These and showed that the effect on parasite egg publications have been heavily cited and have count was negligible. Another study of the an- helped shape the evolving research agenda on thelmintic effect of Albizia anthelmintica showed antimicrobial use in developing countries. Docu- that it, too, was not efficacious against H. con- mentation of antimicrobial resistance in the food tortus in sheep (Githiori et al., 2003). This led to value chains has become key to understand risks a widely consulted review of plants used in eth- to public health. In Ethiopia, a study in a slaugh- noveterinary sources and highlighted the need terhouse isolated multidrug-r esistant Escheri- for in vivo studies to take into account anthel- chia coli 0157 from goats, identifying the need for mintic properties and overall effects of these surveillance of antimicrobial resistance (Dulo plants on the performance of the parasitized et al., 2015). In general, ILRI has called for taking hosts through potential anti-nutritional effects more integrated and one-health approaches to (Githiori et al., 2006). work on antimicrobial resistance and use at the interface of people, animals and the environ- Understanding and improving the use of ment (Nguyen-Viet et al., 2016, 2019). ILRI was antibiotics the first research organization to draw attention to the worrying phenomenon of farmers in Initial antimicrobial work by ILRI focused on Uganda sharing their antiretroviral treatments understanding the knowledge, attitudes and with their livestock in the belief that it would practices of antimicrobial use among farmers make them gain weight and is currently re- and animal health service providers. Starting in searching this practice. the 1990s, significant ILRI research on trypano- cides had provided crucial evidence for improv- ing control of African animal trypanosomiasis Understanding drug resistance emergence and drew attention to trypanocide resistance and transmission (Mulugeta et al., 1997). Research to understand More recently, research on antimicrobial use is the use of trypanocides and the occurrence of being complemented by a growing bioscience resistance prompted more work on managing research portfolio. For example, scientists at the trypanosomiasis by evaluating different control BecA-ILRI Hub collaborated in genetic research strategies (McDermott et  al., 2001; Clausen on resistance mechanisms in Salmonella isolates et al., 2010) (see also Chapters 2 and 3, this vol- for quinolones and β-lactam antimicrobials ume, on trypanosomiasis). Methodologies to (Eguale et al., 2017). To understand what is driv- evaluate use of veterinary dugs, especially par- ing the emergence of antimicrobial resistance ticipatory approaches, remain of use and are at the human–livestock–environment interface, being developed further (Grace et  al., 2009). ILRI in collaboration with academic partners Further developing the capacity of farmers to has set up studies in urban settings to track anti- better use veterinary drugs was considered microbial resistance genes between species and promising (Grace et  al., 2008) but as a stand- to investigate the role of the environment in the alone intervention was found insufficient to fun- transmission of these genes. damentally change farmer behaviour. As these findings are relevant to antibiotic use, this line of The CGIAR Antimicrobial Resistance Hub research has shifted over the years to antibacter- ial drugs. ILRI’s research on antimicrobial resistance An ILRI review highlighting a dearth of focuses on developing, testing and evaluating inter- data on antimicrobial use in livestock production ventions that can mitigate livestock-associated systems in developing countries (Grace, 2015) antimicrobial resistance in livestock value Zoonoses 321 chains. ILRI is increasingly involved in policy most mammals, causing a general state of ill- work to stem the rise of antimicrobial resistance, ness and eventual death. Bovine TB is a zoo- such as supporting the development and imple- notic disease transmitted to humans mainly by mentation of national action plans and support- their consumption of contaminated milk. Zoo- ing capacity development in various stakeholder notic TB makes up only a small proportion of the groups. An increasing importance of livestock- overall human TB disease burden but ending the associated antimicrobial resistance is reflected at TB epidemic will not be possible without com- the CGIAR level with two CRPs having taken up bating zoonotic bovine TB. antimicrobial resistance research: Agriculture One of ILRI’s first research activities dedi- for Nutrition and Health focuses on reducing cated to bovine TB was evaluation of a new test risks to public health resulting from antimicro- for TB (interferon (IFN)-γ) in Ethiopia. The bial resistance arising in livestock systems, while cornerstone of bovine TB control is a rapid and Livestock Agri-Food Systems promotes rational accurate identification of infected animals and and effective use of antimicrobials in small- their removal from the herd. Conventional test- holder livestock systems to prevent future failure ing for bovine TB involved injecting cattle in the to treat disease in livestock and humans alike. neck with antigens to M. bovis and checking In both of these CRPs, ILRI scientists have after several days to see whether the animals had assumed leading roles in antimicrobial use- a swelling on their skin, indicating they were related research. In early 2019, a CGIAR Anti- positive for bovine TB. This test requires special- microbial Resistance Hub was launched at ILRI, ist personnel and is expensive and time consum- Nairobi. It is convening stakeholders with vari- ing. The newer IFN-γ test uses blood samples, ous interests in antimicrobial resistance. It will which are relatively easy and cheap to collect. work to influence a pro-poor antimicrobial use Developed in Australia in the 1980s, the IFN-γ agenda, provide an environment fostering col- test was approved by European Union legislation laboration and new research partnerships, and for use in cattle in 2002. Comparing the conven- streamline communications around agricultur- tional and new IFN-γ bovine TB tests, the ILRI ally associated antimicrobial resistance to support study found that both tests performed similarly evidence-based discussions. well and concluded that the choice of which to ILRI has produced 20 out of 7967 antimicro- use depended on their relative cost and simpli- bial resistance outputs in the Altmetrics data- city of use, as well as on livestock management base; one of them has been quoted in a policy and time factors (Ameni et al., 2000). This find- document. ing led to further evaluation of bovine TB diag- nostics in Ethiopia (Ameni et al., 2006). Neglected Zoonoses ILRI participated in a consortium investi- gating bovine TB in developing countries. ILRI’s Neglected zoonoses persist in communities with role was to determine cattle breed differences in complex and interrelated development problems immune responses to vaccination with BCG such as poverty, isolation, poor access to services, (bacille Calmette–Guérin vaccine, an attenuated insecurity, political marginalization, low literacy strain of M. bovis) followed by infection with rates, gender inequality, lack of sanitation, de- M. bovis. Previous studies and epidemiological graded natural resources and high dependence surveys had shown a difference in susceptibility on livestock. In these regions, ILRI research has to, and disease severity in, bovine TB in Zebu and focused on high-priority neglected zoonoses, in- Holstein cattle (Ameni et  al., 2006). The BCG cluding bovine TB, brucellosis, cysticercosis and v accination experiments confirmed that native Af- trypanosomiasis. rican Zebu cattle were more resistant to bovine TB than exotic Holstein animals (Vordermeier et al., 2012). Bovine tuberculosis ILRI studies of the kinetics of IFN-γ released in the peripheral blood of calves vaccinated with Bovine tuberculosis (bovine TB) is a chronic dis- the BCG vaccine showed a strong positive reaction ease of animals caused by Mycobacterium bovis, in the calves to tuberculin inoculation 15 weeks which is closely related to the bacteria that cause post-vaccination, demonstrating BCG’s ability to human and avian TB. This disease can affect induce the release of IFN-γ in the peripheral 322 D. Grace et al. blood and its role in protecting calves against in- and where latrine coverage and meat hygiene fection with M. bovis (Ameni and Tibbo, 2002). capacity may be low. Another area of ILRI research was surveys T. solium cysticercosis is often ranked among conducted to understand the presence, level and the top priorities of neglected zoonoses and was risk factors associated with TB. These included an early focus for ILRI research. Initial priorities, the following: formulated and led by Lee Willingham, were on • better defining the health and economic burden A study in Narok, which vindicated a long- of cysticercosis and supporting an important held official position that bovine TB was regional collaboration to better control it. When absent in Kenya (Koech, 2001). • Phil Toye joined ILRI, a multi-year effort focused A study of dairy farms in central Ethiopia, on developing lateral-flow diagnostics for this which found that 45% of cattle and 13% of disease, as well as testing simpler approaches milk samples were positive for bovine TB, to its diagnosis. Meanwhile, extensive epi- with a much higher prevalence in Holsteins demiological surveys were conducted in Kenya, than in local breeds (Ameni et  al., 2003). Mozambique, Tanzania and Uganda in Africa, This represented a very high risk to public and in India, Laos and Vietnam in Asia. The par- health. In contrast, a survey in north- ticipation of Lee Willingham and subsequently of eastern Ethiopia found that the prevalence Eric Fèvre in WHO-led initiatives has given ILRI of bovine TB seemed low, but there was lit- an influential position in the global one-health tle awareness of TB and a public health risk phalanx of efforts to control cysticercosis. was present (Hadush, 2015), while a study in A collaboration with the University of Guelph Gondar found a moderate but concerning in Canada included a component on training local prevalence of 8% (Shewatatek, 2015). • extension officials, farmers and pork butchers on A meta-analysis of studies in Tanzania found cysticercosis control, as well as ways to improve pig a prevalence of 1.8%, which was less than productivity. Engaging stakeholders in workshops that of brucellosis (8.2%) or trypanosomia- and one-on-one training sessions at the farm level sis (10.2%) (Alonso et al., 2016). • was associated with improved knowledge about A more recent project is investigating bovine cysticercosis (Wohlgemut et al., 2010). Three PhD TB in six cities in India, but the results are students were trained under the collaboration, and not yet available. several undergraduate students were exposed to ILRI collaborated in a study describing best relevant research activities in western Kenya. practices for diagnosing and assessing livestock Cysticercosis causes illness and productiv- productivity losses due to bovine TB and sum- ity losses in both people and livestock. Estimates marized the methodology of assessment for this of the burden of this disease provide essential, particular disease (Tschopp et al., 2009). evidence-based data for conducting cost–benefit and cost–utility analyses that should help secure political will and financial and technical resources. Cysticercosis ILRI led work on developing frameworks for assessing and combining health and economic Cysticercosis, a parasitic tissue infection caused information (Carabin et al., 2005). Several of the by larval cysts of a tapeworm (Taenia solium) that scientists involved went on to participate in land- infect brain, muscle and other tissue, is a major mark assessments of the global burden of dis- cause of adult-onset seizures in most low-income eases. The methods were also used to estimate countries. The adult tapeworm lives in the disease burdens in proof-of-concept studies. For human gut and the larval stages develop in pig example, an estimate of the impact in the Eastern muscle. However, if humans ingest the egg of Cape Province of South Africa found that there the adult tapeworm, then cysts may develop in were an estimated 34,662 cases of epilepsy due the human skin, eye, brain and other tissues to cysticercosis in 2004. The overall monetary resulting in serious illness. Endemic areas are cost was estimated to vary between US$18.6 located predominately across Latin America, million and US$34.2 million in a population of sub-Saharan Africa and South-east Asia where about 7 million people, depending on the method pigs are raised under ‘traditional’ extensive systems used to estimate productivity losses. The agricultural Zoonoses 323 sector contributed an average of US$5 million probability of pigs having access to human (Carabin et al., 2006). faeces and thus T. solium eggs (Waiswa Epidemiological studies on prevalence and et al., 2009). risk factors were conducted in several countries • A later study in Uganda found a prevalence generating evidence, often for the first time, on of 14% in the Lake Kyoga Basin (Nsadha et al., the presence and prevalence of cysticercosis. 2014). Subsequently, one of the largest and These findings led to further research and stimu- most rigorous and representative studies lated efforts to improve control. Among the found a high prevalence of T. solium in rural highlights were the following: production settings (10.8%) and an even • higher prevalence in urban settings (17.1%) In Tanzania, the prevalence of swine cysti- (Kungu et al., 2017). cercosis was found to be 7.6%, 8.4% and • A review suggested that cysticercosis 16.9% for Chunya District, Iringa Rural was problematic in Mozambique. Human District and Ruvuma Region, respectively. serological studies found that 15–21% of Risk factors were free-ranging of pigs, home apparently healthy adults were positive for slaughtering of pigs, pork not being inspected cysticercosis antibodies or antigen, while in before consumed, lack of latrines and bar- neuropsychiatric patients, seroprevalence becuing pork (Boa et al., 2006). • was as high as 51%. Slaughterhouse re-A study of 1051 epileptics in western Kenya cords indicated a countrywide occurrence found that one-third had observed nodules of porcine cysticercosis, while studies have in pork meat and one-half had observed shown that 10–35% of pigs tested were tapeworm segments in their own faeces, seropositive for cysticercosis antibodies or suggesting a possible role of cysticercosis antigen (Afonso et al., 2011). (Grace and Downie, 2011). Another study • A study in the state of Nagaland, in north- in Homa Bay confirmed that cysticercosis east India, found an alarmingly high preva- was endemic (Eshitera et al., 2012). • lence of cysticercosis in marketed pork A study in western Kenya found that (Fahrion et al., 2014). 34.4–37.6% of pigs at slaughter were posi- • A study in Laos found a low but concerning tive for Taenia spp. using an HP10 antigen prevalence of cysticercosis in people and enzyme-linked immunosorbent assay (Ag- pigs that was associated with poor hygiene ELISA). All pigs, however, were reported to (Holt et al., 2016). have passed routine meat inspection, rais- ing considerable concern over the effective- Field studies in Uganda identified additional ness of current public health measures risk factors for cysticercosis. Much pork is con- (Thomas et al., 2016). In contrast, an ear- sumed in ‘pork joints’, typically with alcohol. Be- lier study found that only 4% of pigs were cause pork fat is thought to absorb alcohol, some positive using a B158/B60 Ag-ELISA and no customers prefer undercooked pork, believing slaughtered pigs were positive (Kagira that when pork is fully cooked, the fat drips off et  al., 2010). A study in slaughterhouse and therefore will not fully ‘neutralize’ the alco- workers found a low but concerning preva- hol. This undercooked pork can expose custom- lence for both Taenia spp. and cysticercosis ers to several zoonotic pathogens, including T. (Cook, 2014). solium (Thomas et al., 2017). • While an abattoir study in Nairobi found The Cysticercosis Working Group of Eastern the prevalence of Taenia spp. was 8.5%, all and Southern Africa was founded in 2001 by the carcasses were passed for human con- scientists from Kenya, Mozambique, South Af- sumption. Furthermore, the abattoir had rica, Tanzania, Uganda, Zambia and Zimbabwe no facilities for handling T. solium-infected with the purpose of collaborating against cysti- carcasses (Akoko et al., 2016). cercosis. This cysticercosis working group had • A study in south-east Uganda found that considerable success in organizing workshops, 8.6% of pigs screened were seropositive for obtaining funding, conducting studies, agreeing cysticercosis. In addition, 26% of homes on a regional action plan and developing a did not have pit latrines, indicating a high widely used standard questionnaire about the 324 D. Grace et al. disease (Gabriël et  al., 2017). The group has needed to support public health campaigns. This been described as one of the notable successes of has led to interest in public–private or market- Danish aid (Johansen and Mukaratirwa, 2013) based solutions to controlling diseases. These are and is now recognized internationally, serving most likely to be effective for diseases such as as a model for other similar networks. cysticercosis that inflict costs on farmers and A promising avenue for control of cysticer- traders. ILRI collaborated with the Global Alli- cosis is the development of rapid and cheap diag- ance for Livestock Veterinary Medicines (GALV- nostics for the disease in live pigs, which could med) to evaluate an approach to cysticercosis be made readily available to farmers and meat control based on marketing pig vaccines and inspectors. This would allow infected animals to wormers to farmers in Uganda. A series of sur- be identified and rejected either before transport veys, auctions and economic studies showed to slaughter or at slaughter. Infected pigs could that farmers were unwilling to pay for cysticer- then be treated (e.g. with the anthelmintic ox- cosis prevention (vaccine and wormer), suggest- fendazole) and re-presented for slaughter when ing that these vaccines would have to be highly its cysts had become non-viable. This would safe- subsidized to make them attractive to pig keepers guard public health while allowing pig farmers (Dione et al., 2019). to retain an important source of their income. Given the failure to achieve the WHO goal ILRI undertook development and validation for a validated strategy for T. solium control and of a pen-side lateral-flow assay to detect porcine elimination by 2015, ILRI continued to work with cysticercosis, which showed that the assay per- international partners to develop alternative forms acceptably well compared with standard frameworks for the development of realistic con- laboratory-based assays. trol goals for endemic areas (Braae et al., 2019). A second round of funding was obtained for further development of the assay, in particular to convert to a format that would allow whole Human African trypanosomiasis blood to be tested instead of serum. The later- al-flow assay is now undergoing further develop- Human African trypanosomiasis, known ments; with an initial sensitivity of 73.84% and commonly as sleeping sickness, is caused by a specificity of 66.7%, and a κ index of 0.38, it has protozoan parasitic infection of public health clear potential to be developed and used for diag- importance. In East Africa, sleeping sickness nosis of cysticercosis in the field (Kivali et  al., presents as an acute syndrome caused by the 2013). ILRI scientists are now establishing a Trypanosoma brucei rhodesiense parasite, which is ‘gold standard’ sera bank collected from known maintained in an animal reservoir. Where wild- cysticercosis-positive and -negative pigs, achieved life is not abundant, domestic livestock species, through fine dissection of carcasses after slaugh- particularly cattle, are the main reservoir. West ter to identify cysts. This sera bank will be used African sleeping sickness is caused by T. b. gambi- for the validation of current and future diagnos- ense and is transmitted in a human–tsetse–human tics for cysticercosis. cycle. The role of an animal reservoir still needs Cysticercosis has traditionally been diagnosed to be clarified. ILRI work focused on East African by lingual palpation, a technique the utility of sleeping sickness and global control. which was confirmed in an early study by ILRI (Mu- Uganda and the Congo experienced a mas- tua et al., 2007). A subsequent meta-analysis found sive sleeping sickness outbreak from 1900 to that all areas with more than 10% of pigs having 1920 and a more recent outbreak from 1976 to cysts in their tongues had at least 30% seropreva- 1989. Building on its long history of trypano- lence of cysticercosis, and assessing the prevalence somiasis research, ILRI has collaborated with of tongue cyst-positive pigs is a potentially rapid local partners, the University of Edinburgh, UK, epidemiological tool for identifying areas at high and FAO to better understand the potential role risk of cysticercosis, although further refinement of veterinary interventions in controlling sleep- and validation is required using standardized data ing sickness in Uganda. sets (Guyatt and Fèvre, 2016). An early desk study produced the first- In resource-constrained developing coun- ever one-health assessment of the burden of tries, it is often difficult to obtain the funding tsetse- transmitted trypanosomiasis by combining Zoonoses 325 economic losses in livestock with impacts on called ‘Stamp out Sleeping Sickness’ that aimed human health measured in DALYs (Odiit et al., to slow the spread of sleeping sickness, which 2000). This led to studies of the risk factors for itself was driven by cattle movements. transmission of T. b. rhodesiense sleeping sick- The campaign helped private-sector animal ness in endemic regions and to identification of health providers to treat cattle with preventive new endemic areas (2004). Importantly, close- drugs and insecticides. Between 2006 and 2010, ness to a reporting health unit was a major de- nearly 200,000 cattle were treated, and the terminant for detecting early rather than intervention was shown to reduce sleeping sick- advanced cases (Odiit et  al., 2004). A related ness in people and to halt its expansion. However, study investigated late diagnosis of sleeping both the human and animal forms significantly sickness, which typically results in a worse out- increased after the intervention, and uptake by come for patients. This found long delays in farmers was low. These disappointing results diagnosis (and therefore in treatment), much were attributed to a lack of community concern of it due to the service provider failing to diag- about sleeping sickness, the cost of treating cattle, nose sleeping sickness among symptomatic in- insufficient incentives for private-sector disease dividuals (Odiit et  al., 2004). Following this control, a lack of control infrastructure, the sea- work, a model was developed to quantify un- sonality of the disease and poor targeting, among derdetection of sleeping sickness in Uganda other factors (Bardosh, 2018). during a 1988–1990 epidemic. This showed While the development impacts of this that, of 73 undetected deaths, 62 entered the large-scale, low-technology, private sector-based healthcare system but were not diagnosed and intervention were disappointing, ILRI’s evidence 11 died without seeking healthcare from a rec- on the burden, distribution, under-reporting ognized health unit (Odiit et  al., 2005). This and potential of managing sleeping sickness evidence of under detection stimulated an ini- through treatment of the livestock form of the tiative to upgrade 12 facilities to perform con- disease helped to draw attention to the import- firmatory testing for sleeping sickness, resulting ance of sleeping sickness and to increase efforts in a substantial reduction in the distance neces- to control it, which were globally spearheaded sary to travel to get a diagnosis for sleeping by WHO. sickness (Wamboga et al., 2017). The work was ILRI scientists contributed to the WHO Ex- also used by the WHO to support the argument pert Committee Report on Human African that zoonoses were neglected. Trypanosomiasis and the WHO stakeholder Further work helped to identify the distri- meetings on T. b. rhodesiense human African bution of villages at risk of sleeping sickness us- trypanosomiasis. Work continued with the first ing GIS and remote sensing (Odiit et al., 2006). geographically delimited estimation of the bur- Distribution maps based on archival and con- den of sleeping sickness disease at the subcoun- temporary data showed that the disease focus ty scale in Uganda (Hackett et  al., 2014). had moved from lakeshore Buganda (1905– Importantly, this work found that, whereas the 1920) to the Busoga and south-east districts relative burden of sleeping sickness was low at (Berrang-Ford et  al., 2006). This information the national level, in some districts it was a was used by medical services to better under- high-priority disease. Human sleeping sickness stand risk. was also considered in economic analyses of A model of T. b. rhodesiense infections trans- the control of African animal trypanosomiasis mitted by a single tsetse species between cattle and in East Africa (Shaw et  al., 2015). More re- humans was developed at ILRI using empirical cently, ILRI has been involved in evaluating data (Coleman et al., 1999) to assess the relative diagnostic tools for sleeping sickness (Lejon impact of mass treatment of cattle versus treat- et al., 2017). ment of human cases on the prevalence of T. b. rhodesiense sleeping sickness in south-eastern Uganda. Mass treatment of cattle with a cover- Brucellosis age of 80% was predicted to break the trans- mission to humans. This evidence contributed to Brucella spp. infect many animals, including cat- a major public–private partnership campaign tle, small ruminants, camels, water buffaloes, 326 D. Grace et al. yaks and pigs. Different Brucella spp. infect differ- A workshop held by ILRI and USAID in 2013 ent animal species, but most have the potential brought together participants from 16 countries to infect humans, with some species causing to identify gaps in brucellosis epidemiology, diag- more disease than others. Brucella spp. infection nosis, surveillance and control (USDA/USAID/ rates in some developing countries can reach ILRI, 2013). The workshop provided information more than 10% of the human population, mak- to help design brucellosis research programmes ing it a serious public health disease. In livestock, and intervention strategies at national and Brucella spp. cause diseases that reduce animal regional levels. production and cause abortions in females and A landmark study of brucellosis tested 825 reduced fertility in males. The most common patients for the disease at the Busia County Re- method by which humans are infected is ferral Hospital and the KEMRI Alupe Research through ingestion of unprocessed milk products Centre, located in the same area (de Glanville from infected animals, but direct contact with et al., 2017). The team used the regular govern- infected animals and meat can also be a source ment febrile antigen Brucella agglutination test of infection. and a second called the rose Bengal test, while Initial work at ILRAD and ILCA identified another two kits were used to confirm the results. brucellosis as one of many constraints to live- Of the 825 cases, 196 patients (19.6%) were stock production in Africa and Asia. Surveys on found positive for brucellosis from the regular brucellosis in livestock and marketed milk were tests. However, when the positive cases were tested conducted in the late 1990s in Ethiopia and with the second and other confirmatory tests, Kenya (Asfaw et  al., 1998; Kang’ethe et  al., only eight people (1%) of the total were found to 2000). These surveys confirmed the presence have been infected with Brucella spp. In this case, and importance of the Brucella pathogen and if not for the secondary and confirmatory tests, also revealed problems with the commonly used 188 people would have unnecessarily been put on diagnostics for this disease. In Kenya, Brucella the rigorous treatment to cure brucellosis. This abortus antibodies were not detected in raw milk high-profile study drew widespread attention sold in urban areas but were found at low levels and led to the formulation of improved guide- (2–5%) in milk sampled from consumers in rural lines for diagnosing brucellosis. In addition, ILRI areas, and at higher levels (25%) in pasteurized scientists have worked on evaluating improved milk (Omore et al., 2000). This finding added to tests (Falzon et al., 2019). the body of evidence being developed by ILRI Brucellosis is especially problematic to man- that formal milk is not always safe and informal age in countries where the culling of cattle is not milk is not always risky. These results were in- acceptable, such as in India. ILRI work com- strumental in a radical, pro-poor shift in Kenyan menced with a systematic review to clarify the dairy policy (Leksmono et al., 2006). contradictory evidence on prevalence. This has ILRI produced important syntheses of re- been reported to be as low as 1% and as high as search on brucellosis in developing countries. 60% by different researchers. The ILRI review The first assessment of brucellosis in Africa concluded that the disease’s overall prevalence found that the incidence of the disease was high- in the country was probably 12% or less (Deka est in pastoral production systems, that this inci- et  al., 2018), which is still high. This was fol- dence decreased with decreases in herd size and lowed by a number of prevalence studies to bet- size of landholding, that little was known of bru- ter understand the true picture of this disease cellosis in other species and that control of bru- (Lindahl et al., 2018), which confirmed that the cellosis was largely inadequate (McDermott and disease was highly heterogeneous. This informa- Arimi, 2002). A systematic review of the econom- tion is essential for targeting control efforts to ics of brucellosis control found that the benefits where they are needed most. of control always exceeded its costs (McDermott China’s Yunnan Province is at particular risk et  al., 2013). Interestingly, ex post assessments of brucellosis because ruminants are increasingly had relatively higher benefits than ex ante assess- introduced to the province from other parts of ments, and control in lower-income countries had the country in response to increasing demand relatively higher benefits than control in higher- for milk. ILRI developed an EcoHealth approach income countries. to control brucellosis. Zoonoses 327 ILRI conducted studies on brucellosis in a et al., 2018). ILRI is currently (2020) developing range of species, products, and countries: an anthrax risk map for the country for use in • developing prevention and control measures A study in camel-rearing areas of Ethiopia aimed at reducing the public health and economic found that around 6% of camels were sero- impact of anthrax. positive for Brucella spp. As camel milk here In many low- and middle-income countries, is rarely boiled before being consumed, this it is a common practice to consume animals that represents a significant public health risk have died of an illness because they are a valu- (Teshome et al., 2003). • able and scarce source of protein. An assess-Brucellosis was detected in just under half ment in Zambia found that the risk to humans of of the cows’ milk samples tested in Tanga, contracting illness due to consuming meat from Tanzania (Shija, 2013). • animals that have died of disease was rather low An anthropological study in Mali provided a and could be reduced by taking precautions, in- key insight – that brucellosis was common cluding careful butchery and ensuring that but that pastoralists believed milk to be in- meat is cooked properly (Simpson, 2015). This trinsically ‘pure’ and hence incapable of being example illustrates the complexity of work aim- a source of disease (Fokou et al., 2010). • ing to change behaviour, especially when people Brucellosis was not found in goats in west- do not have sufficient incentives to change their ern Kenya (Akoko et al., 2013). behaviour. Human cystic echinococcosis is a neglected zoonotic parasitic disease caused by the larval Other neglected zoonoses stage of the dog tapeworm, Echinococcus granulo- sus. Ungulates are intermediate hosts, but if a Neglected zoonotic diseases are commonly asso- human consumes tapeworm eggs, they may ciated with poverty and greatly impact in particu- develop into fluid-filled cysts in organs and tis- lar the lives and livelihoods of millions of poor sues. The infection is most common in pastoral livestock keepers, processors and consumers communities, where it causes considerable socio- of livestock products. WHO has identified a sub- economic impact. A systematic literature review group of eight ‘neglected zoonotic diseases’: an- showed its widespread occurrence in small thrax, bovine TB, brucellosis, cysticercosis, ruminants in Ethiopia (Asmare et al., 2016). hydatidosis, leishmaniasis, rabies and sleeping A study in abattoirs in Narok, Kenya, found that sickness. ILRI research has focused mostly on 16% of sheep had cysts, which was considered a bovine TB, brucellosis, cysticercosis, rabies and high level of infection constituting a public sleeping sickness. (ILRI’s work on rabies is noted health risk (Odongo et al., 2018). in Chapter 5, this volume.) Limited ILRI research Q fever, caused by Coxiella burnetii, is an old has been conducted on other zoonoses on the zoonotic disease believed to be widely present WHO list, as well as other zoonoses that fit a in ruminant populations worldwide. It is a com- broader definition of neglected. mon cause of abortion in ruminants, and in Anthrax is a bacterial zoonosis, caused by people can cause flu-like illness. There is little de- Bacillus anthracis. It primarily affects herbivores tailed knowledge of its presence in livestock sys- and is highly lethal to them. Humans contract tems in low- and middle-income countries. ILRI the disease from contact with infected or dead conducted or supported surveys of Q fever in animals, and infections can result in a high mor- Kenya and Tanzania. These showed a high tality rate for people as well, if not diagnosed and prevalence (20%) in camels in Laikipia, Kenya treated promptly. Anthrax is also of interest as a (Browne et  al., 2017), and in livestock (13%) ‘dreaded disease’, a climate-sensitive disease and and people (27%) in Tana River, Kenya (Mwolo- a possible ‘bioterrorist’ agent. When alarming lo, 2016). A moderate prevalence was found in outbreaks of anthrax occurred in Nakuru, Kenya, cattle (10%) and humans (2.5% in slaughter- ILRI was part of the team sent to investigate. house workers) in western Kenya (Cook, 2014; Participatory epidemiology was used to map Wardrop et al., 2016) and in cattle reported as current and previous outbreaks, with the results sick (15%) by farmers in northern Tanzania suggesting a long history of outbreaks (Muturi (Alonso et al., 2015). 328 D. Grace et al. A study of hospital patients in Kenya found recognizing the high dependence of the poor on acute Q fever in 16% of patients, a finding un- livestock. DALYs and economic costs of livestock suspected by the treating clinicians. A diagnostic and human disease are important metrics, but tool was developed based on symptoms and was good information is expensive and difficult to shown to be reasonably accurate (sensitivity collect, and even when available, we lack agreed 93.1%, specificity 76.1%) (Njeru et  al., 2016). ways of combining epidemiological and economic This tool has obvious potential to improve diag- metrics. Moreover, other aspects of the zoonoses nosis and hence healthcare, but its impact has burden, such as the cost of a high-impact, low- not been evaluated. Other studies have looked at probability and civilization-altering pandemic, farmer awareness of Q fever, finding it low are difficult to capture. ILRI is involved in new (Nyokabi et  al., 2017), while studies looking at and ongoing initiatives to measure the multiple the disease risk and economic impacts found burdens of zoonotic diseases. they were high (Oboge, 2016). It is generally accepted that zoonoses are best tackled in their livestock rather than human hosts. This avoids human suffering, has been shown to be more cost-effective in studied cases Conclusions and the Future (e.g. rabies) and historically was the main means of eliminating many neglected zoonoses from ILRI research has contributed to an emerging high-income countries. However, most developing consensus about neglected and emerging zoo- countries still lack comprehensive programmes notic diseases. It is clear that zoonotic diseases to tackle zoonoses in their livestock reservoirs. associated with livestock and livestock produc- To generate additional investment, more infor- tion are constraints to human health, well-being mation is needed on the costs, benefits, accept- and economic development. Many of these dis- ability and scalability of interventions, including eases pose risks to wildlife and to the ability of those targeting animals. Malaria and water re- ecosystems to provide services. However, for search show how credible, comparable informa- many of these diseases, basic epidemiological tion on control options can underpin action. data are unavailable. ILRI research has often The majority of emerging human diseases overturned conventional wisdom by finding zoo- are zoonotic, often following amplification in a noses absent where they were believed to be farmed animal host. The current (as of 2020) common and to be major problems when their COVID-19 pandemic is just one of a series of existence was unsuspected. emerging zoonoses over the last decades that has Establishing systematic data collection is resulted in enormous impacts, both health and the first step to manage zoonoses. Management economic. ILRI long-standing research into the is complicated by heterogeneity: zoonoses may drivers of disease emergence and into more have a significant and debilitating effect on some timely and effective surveillance and response communities but not on others. Understanding can contribute to a new urgency around the the spatial distribution of the burden of zoo- need to prevent and manage pandemics. noses is important to better focus control efforts. Zoonotic diseases are particularly complex A significant constraint is the lack of collabor- disorders involving the environmental sciences, ation between medical and veterinary author- agriculture and public health. Policy frameworks ities: institutionally speaking, zoonoses typically for managing zoonoses are sometimes weak, find themselves homeless and ignored. There is a and there are often gaps between policy and im- need for one-health thinking and research to plementation. Successful control of zoonoses overcome inter-sectoral barriers to effective requires a judicious legal and policy framework, control of zoonoses. well-functioning institutions, adequate finan- We can suggest some future directions. A first cing, rapid detection and an intervention imple- step is to develop cheap and efficient diagnostics in mentation plan. The failure of the public sector human and animal hosts to assist in under- to manage neglected zoonoses in developing standing zoonoses and in managing them. countries has led to interest in public–private Another important step is to develop metrics partnerships and market-based solutions. Al- that capture the societal burden of zoonoses, though harnessing market or social forces has Zoonoses 329 great potential in improving human health in further innovation is required. The role of resource-scarce environments, they have not yet behavioural sciences is therefore an active been shown to be sustainable or scalable, and area of ILRI. Notes 1 This work was led by Arve Lee Willingham and Esther Schelling, respectively, both of whom shared joint appointments at their home institutions and ILRI. 2 These included Silvia Alonso (Spain), Bernard Bett (Kenya), Delia Grace (Ireland), Joerg Jores (Germany), Lucy Lapar (Philippines), Anne Liljander (Germany), Johanna Lindahl (Sweden), Jeff Mariner (USA), Karl Rich (USA), Fred Unger (Germany) and Francis Wanyoike (Kenya). A later collaboration involved the University of Edinburgh, and later the University of Liverpool, led by Eric Fèvre (UK). References Afonso, S.M.S., Vaz, Y., Neves, L., Pondja, A., Dias, G., et al. (2011) Human and porcine Taenia solium infections in Mozambique: identifying research priorities. Animal Health Research Reviews 12, 123–129. Akoko, J.M., Kiyong’a, A.N., de Glanville, W.A., Thomas, L.F. and Fèvre, E.M. (2013) The seroprevalence of caprine brucellosis in western Kenya. Poster presented at the 47th annual scientific conference of the Kenya Veterinary Association, 24–27 April 2013, Mombasa, Kenya. ILRI, Nairobi. Akoko, J.M., MacLeod, E., Kang’ethe, E., Muinde, P., Alarcon, P., et al. (2016) Serological survey of porcine cysticercosis and associated risk factors in pigs slaughtered at Ndumbuini abattoir in Nairobi, Kenya. International Journal of Infectious Diseases 45 (Suppl. 1), 464. Alarcon, P., Fèvre, E.M., Muinde, P., Murungi, M.K., Kiambi, S., et al. (2017) Urban livestock keeping in the city of Nairobi: diversity of production systems, supply chains, and their disease management and risks. Frontiers in Veterinary Science 4, 171. Allepuz, A., de Balogh, K., Aguanno, R., Heilmann, M. and Beltran-Alcrudo, D. (2017) Review of participa- tory epidemiology practices in animal health (1980–2015) and future practice directions. PLoS One 12, e0169198. Alonso, S., Msalya, G., Toye, P., Grace, D. and Unger, F. (2015) Ancient farming styles and old zoonoses: brucellosis and Q fever among pastoralist and smallholder cattle herds in Tanzania. Poster prepared for the European College of Veterinary Public Health (ECVPH) annual scientific conference, 7–10 Oc- tober, Belgrade, Serbia. ILRI, Nairobi. Alonso, S., Dohoo, I., Lindahl, J., Verdugo, C., Akuku, I. and Grace, D. (2016) Prevalence of tuberculosis, brucellosis and trypanosomiasis in cattle in Tanzania: a systematic review and meta-analysis. Animal Health Research Reviews 17, 16–27. Ameni, G., Miorner, H., Roger, F. and Tibbo. M. (2000) Comparison between comparative tuberculin and γ-interferon tests for the diagnosis of bovine tuberculosis in Ethiopia. Tropical Animal Health and Production 32, 267–276. Ameni, G. and Tibbo, M. (2002) Kinetics of interferon-gamma (IFN-γ) release in the peripheral blood of calves vaccinated with BCG. Journal of Immunoassay and Immunochemistry 23, 245–251. Ameni, G., Bonnet, P. and Tibbo, M. (2003) A cross-sectional study of bovine tuberculosis in selected dairy farms in Ethiopia. Journal of Applied Research in Veterinary Medicine 1, 253–258. Ameni, G., Aseffa, A., Engers, H., Young, D., Hewinson, G. and Vordermeier, M. (2006) Cattle husbandry in Ethiopia is a predominant factor affecting the pathology of bovine tuberculosis and gamma inter- feron responses to mycobacterial antigens. Clinical and Vaccine Immunology 13, 1030–1036. Asfaw, Y., Molla, B., Zessin, K.H. and Azage, T. (1998) A cross-sectional study of bovine brucellosis and test performance in intra-and peri-urban production systems in and around Addis Ababa, Ethiopia. OAU/ STRC Bulletin of Animal Health and Production in Africa 46, 217–224. Asmare, K., Sibhat, B., Abera, M., Haile, A., Degefu, H., et al. (2016) Systematic review and meta- analysis of metacestode prevalence in small ruminants in Ethiopia. Preventive Veterinary Medicine 129, 99–107. 330 D. Grace et al. Atherstone, C., Picozzi, K. and Kalema-Zikusoka, G. (2014) Seroprevalence of Leptospira hardjo in cattle and African buffalos in southwestern Uganda. American Journal of Tropical Medicine and Hygiene 90, 288–290. Atherstone, C., Smith, E., Ochungo, P., Roesel, K. and Grace, D. (2015) Assessing the potential role of pigs in the epidemiology of Ebola virus in Uganda. Transboundary and Emerging Diseases 64, 333–343. Azhar, M., Lubis, A.S., Siregar, E S., Alders, R.G., Brum, E., et al. (2010) Participatory disease surveillance and response in Indonesia: strengthening veterinary services and empowering communities to prevent and control highly pathogenic avian influenza. Avian Diseases 54, 749–753. Bardosh, K.L. (2018) Towards a science of global health delivery: a socio-anthropological framework to improve the effectiveness of neglected tropical disease interventions. PLoS Neglected Tropical Dis- eases 12, e0006537. Berrang-Ford, L., Odiit, M., Maiso, F., Waltner-Toews, D. and McDermott, J.J. (2006) Sleeping sickness in Uganda: revisiting current and historical distributions. African Health Sciences 6, 223–231. Bett, B., McLaws, M., Jost, C., Schoonman, L., Unger, F., et al. (2015) The effectiveness of preventative mass vaccination regimes against the incidence of highly pathogenic avian influenza on Java Island, Indonesia. Transboundary and Emerging Diseases 62, 163–173. Bett, B., Said, M.Y., Sang, R., Bukachi, S., Wanyoike, S., et al. (2017a) Effects of flood irrigation on the risk of selected zoonotic pathogens in an arid and semi-arid area in the eastern Kenya. PLoS One 12, e0172626. Bett, B., Kiunga, P., Gachohi, J., Sindato, C., Mbotha, D., et al. (2017b) Effects of climate change on  the occurrence and distribution of livestock diseases. Preventive Veterinary Medicine 137, 119–129. Bett, B., Grace, D., Lee, H.S., Lindahl, J., Nguyen-Viet, H., et al. (2019) Spatiotemporal analysis of histor- ical records (2001–2012) on dengue fever in Vietnam and development of a statistical model for forecasting risk. PLoS One 14, e0224353. Boa, M.E., Mahundi, E.A., Kassuku, A.A., Willingham, A.L. and Kyvsgaard, N.C. (2006) Epidemiological survey of swine cysticercosis using ante-mortem and post-mortem examination tests in the southern highlands of Tanzania. Veterinary Parasitology 139, 249–255. Braae, U.C., Gabriël, S., Trevisan, C., Thomas, L.F., Magnussen, P., et al. (2019) Stepwise approach for the control and eventual elimination of Taenia solium as a public health problem. BMC Infectious Diseases 19, 182. Browne, A.S., Fèvre, E.M., Kinnaird, M., Muloi, D.M., Wang, C.A., et  al. (2017) Serosurvey of Coxiella burnetii (Q fever) in dromedary camels (Camelus dromedarius) in Laikipia County, Kenya. Zoonoses and Public Health 64, 543–549. Carabin, H., Budke, C.M., Cowan, L.D., Willingham, A.L. and Torgerson, P.R. (2005) Methods for assessing the burden of parasitic zoonoses: echinococcosis and cysticercosis. Trends in Parasitology 21, 327– 333. Carabin, H., Krecek, R.C., Cowan, L.D., Michael, L., Foyaca-Sibat, H., et al. (2006) Estimation of the cost of Taenia solium cysticercosis in Eastern Cape Province, South Africa. Tropical Medicine & Inter- national Health 11, 906–916. Carron, M., Alarcon, P., Karani, M., Muinde, P., Akoko, J., et al. (2017) The broiler meat system in Nairobi, Kenya: using a value chain framework to understand animal and product flows, governance and sani- tary risks. Preventive Veterinary Medicine 147, 90–99. CDC (Centers for Disease Control and Prevention). (2020) 40 years of Ebola disease around the world. https://www.cdc.gov.vhf/ebola/history/chronology.html (accessed 16 August 2020). Clausen, P.H., Bauer, B., Zessin, K.H., Diall, O., Bocoum, Z., et al. (2010) Preventing and containing trypano- cide resistance in the cotton zone of West Africa. Transboundary and Emerging Diseases 57, 28–32. Cleaveland, S., Laurenson, M. and Taylor, L. (2001) Diseases of humans and their domestic mammals: pathogen characteristics, host range and the risk of emergence. Philosophical Transactions of the Royal Society of London B: Biological Sciences 356, 991–999. Cole, D., Grace, D. and Diamond, M. (2008) Researcher approaches to evidence on urban agriculture and human health. Cole, D., Lee-Smith, D. and Nasinyama G. (eds) Healthy City Harvests: Generating Evidence to Guide Policy on Urban Agriculture. International Potato Center (CIP), Lima, and Maker- ere University Press, Kampala, pp. 31–47. Coleman, P., McDermott, J. and Mauldin, I. (1999) The public health impact of controlling Trypanosoma brucei rhodesiense in cattle. Paper presented at the 25th International Scientific Council on Trypano- somiasis Research and Control (ISCTRC) Meeting, Mombasa, Kenya, 1999. Zoonoses 331 Consultative Group for RVF Decision Support (2010) Decision-support tool for prevention and control of Rift Valley fever epizootics in the Greater Horn of Africa. American Journal of Tropical Medicine and Hygiene 83, 75–85. Cook, E.A. (2014) Epidemiology of zoonoses in slaughterhouse workers in western Kenya. PhD thesis, University of Edinburgh, Edinburgh, UK. Cook, E.A., Grossi-Soyster, E.N., de Glanville, W.A., Thomas, L.F., Kariuki, S., et al. (2017) The se- ro-epidemiology of Rift Valley fever in people in the Lake Victoria Basin of western Kenya. PLoS Neglected Tropical Diseases 11, e0005731. Corman, V.M., Jores, J., Meyer, B., Younan, M., Liljander, A., et al. (2014) Antibodies against MERS corona- virus in dromedary camels, Kenya, 1992–2013. Emerging Infectious Diseases 20, 1319–1322. Dang-Xuan, S., Bryant, J.E., Unger, F., Bao N.-T., Ngoc, B.V.T., et al. (2015) Risk pathways and preva- lence in slaughtered pig blood of Streptococcus suis in Vietnam. Poster presented at the 9th Euro- pean Congress on Tropical Medicine and International Health, 6–10 September, Basel, Switzerland. Hanoi School of Public Health, Hanoi. Deem, S.L., Fèvre, E.M., Kinnaird, M., Browne, A.S., Muloi, D., et  al. (2015) Serological evidence of MERS-CoV antibodies in dromedary camels (Camelus dromedarius) in Laikipia County, Kenya. PLoS One 10, 11–15. Deka, R.P., Magnusson, U., Grace, D. and Lindahl, J. (2018) Bovine brucellosis: prevalence, risk factors, economic cost and control options with particular reference to India – a review. Infection Ecology & Epidemiology 8, 1555448. de Glanville, W.A., Conde-Álvarez, R., Moriyón, I., Njeru, J., Díaz, R., et al. (2017) Poor performance of the rapid test for human brucellosis in health facilities in Kenya. PLoS Neglected Tropical Diseases 11, e0005508. Dione, M., Adediran, A.S., Colston, A., Ouma, E., Lule, P. and Grace, D. (2019) A market scoping study for porcine cysticercosis vaccine calls for a one-health approach to sustain the control of the disease in Uganda. GALVmed, Nairobi. Dulo, F., Feleke, A., Szonyi, B., Fries, R., Baumann, M.P. and Grace, D. (2015) Isolation of multidrug-resist- ant Escherichia coli O157 from goats in the Somali region of Ethiopia: a cross-sectional, abattoir-based study. PLoS One, 10, e0142905. Eguale, T., Birungi, J., Asrat, D., Njahira, M.N., Njuguna, J., et al. (2017) Genetic markers associated with resistance to β-lactam and quinolone antimicrobials in non-typhoidal Salmonella isolates from hu- mans and animals in central Ethiopia. Antimicrobial Resistance & Infection Control 6, 13. Erume, J., Roesel, K., Dione, M.M., Ejobi, F., Mboowa, G., et al. (2016) Serological and molecular investi- gation for brucellosis in swine in selected districts of Uganda. Tropical Animal Health and Production 48, 1147–1155. Eshitera, E.E., Githigia, S.M., Kitala, P., Thomas, L.F., Fèvre, E.M., et al. (2012) Prevalence of porcine cyst- icercosis and associated risk factors in Homa Bay District, Kenya. BMC Veterinary Research 8, 234. Fahrion, A.S., Jamir, L., Richa, K., Begum, S., Rutsa, V., et al. (2014) Food-safety hazards in the pork chain in Nagaland, North East India: implications for human health. International Journal of Environmental Research and Public Health 11, 403–417. Falzon, L.C., Traoré, S., Kallo, V., Assamoi3, J.-B., Bonfoh, B. and Schelling, E. (2019) Evaluation of the fluorescence polarization assay as a rapid on-spot test for ruminant brucellosis in Côte d'Ivoire. Fron- tiers in Veterinary Science 6, 287. Faye, O., Ba, H., Ba, Y., Freire, C.C.M., Faye, O., et al. (2014) Reemergence of rift valley fever, Mauritania, 2010. Emerging Infectious Diseases 20, 300–303. Fèvre, E.M., de Glanville, W.A., Thomas, L.F., Cook, E.A., Kariuki, S. and Wamae, C.N. (2017) An integrated study of human and animal infectious disease in the Lake Victoria crescent small-holder crop-live- stock production system, Kenya. BMC Infectious Diseases 17, 457. Fokou, G., Koné, B.V. and Bonfoh, B. (2010) “Mon lait est pur et ne peut pas rendre malade”: motivations des acteurs du secteur informel et qualité du lait local au Mali. Revue Africaine de Santé et de Pro- ductions Animales 8, 75–86. Gabriël, S., Dorny, P., Mwape, K.E., Trevisan, C., Braae, U.C., et al. (2017) Control of Taenia solium taenia- sis/cysticercosis: the best way forward for sub-Saharan Africa? Acta Tropica 165, 252–260. Gachohi, J.M., Njenga, M.K., Kitala, P. and Bett, B. (2017) Correction: modelling vaccination strategies against Rift Valley fever in livestock in Kenya. PLoS Neglected Tropical Diseases 11, e0005316. Galaz, V., Leach, M., Scoones, I. and Stein, C. (2015) The political economy of One Health research and policy. STEPS Working Paper 81. STEPS Centre, Brighton, UK. 332 D. Grace et al. Gilbert, J., Grace, D., Unger, F., Lapar, M.L., Asse, R., et al. (2014) Increasing awareness of zoonotic diseases among health workers and rural communities in Southeast Asia. ILRI Research Brief 15. ILRI, Nairobi. Githiori, J.B., Höglund, J., Waller, P.J. and Baker, R.L. (2002) Anthelmintic activity of preparations derived from Myrsine africana and Rapanea melanophloeos against the nematode parasite, Haemonchus contortus, of sheep. Journal of Ethnopharmacology 80, 187–191. Githiori, J.B., Höglund, J., Waller, P.J. and Baker, R.L. (2003) The anthelmintic efficacy of the plant, Albizia anthelmintica, against the nematode parasites Haemonchus contortus of sheep and Heligmo- somoides polygyrus of mice. Veterinary Parasitology 116, 23–34. Githiori, J.B., Athanasiadou, S. and Thamsborg, S.M. (2006) Use of plants in novel approaches for control of gastrointestinal helminths in livestock with emphasis on small ruminants. Veterinary Parasitology 139, 308–320. Grace, D. (2014) The business case for One Health. Onderstepoort Journal of Veterinary Research 81, E1–E6. Grace, D. (2015) Review of evidence on antimicrobial resistance and animal agriculture in developing countries. ILRI, Nairobi. Grace, D., Bett, B., Lindahl, J. and Robinson, T. (2015) Climate and livestock disease: assessing the vul- nerability of agricultural systems to livestock pests under climate change scenarios. CCAFS Working Paper No. 116. CCAFS, Copenhagen. Grace, D. and Downie, K. (2011) Cysticercosis in a population of epileptics in Western Kenya: relating human and pig risk factors. Presentation at the 1st International One Health Congress, 14–16 February, Melbourne, Australia. ILRI, Nairobi. Grace D, Gilbert J, Randolph T and Kang’ethe E. (2012a) The multiple burdens of zoonotic disease and an EcoHealth approach to their assessment. Tropical Animal Health and Production 44 (Suppl. 1), 67–73. Grace, D., Kang’ethe, E. and Waltner-Toews, D. (2012c) Participatory and integrative approaches to food safety in developing country cities. Tropical Animal Health and Production 44 (Suppl. 1), 1–2. Grace, D., Mutua, F., Ochungo, P., Kruska, R., Jones, K., et al. (2012b) Mapping of poverty and likely zoonoses hotspots. Zoonoses Project 4. Report to the UK Department for International Development. ILRI, Nairobi. Grace, D., Nasinyama, G., Randolph, T.F., Mwiine, F. and Kang’ethe, E. (2008) City dairying in Kampala: integrating benefits and harms. In: Cole, D., Lee-Smith, D. and Nasinyama, G. (eds) Healthy City Harvests: Generating Evidence to Guide Policy on Urban Agriculture. International Potato Center (CIP), Lima, pp. 193–210. Grace, D., Randolph, T., Affognon, H., Dramane, D., Diall, O. and Clausen, P.H. (2009) Characterisation and validation of farmers’ knowledge and practice of cattle trypanosomosis management in the cotton zone of West Africa. Acta Tropica 111, 137–143. Guyatt, H.L. and Fèvre, E.M. (2016) Lingual palpation for porcine cysticercosis: a rapid epidemiological tool for estimating prevalence and community risk in Africa. Tropical Medicine & International Health 21, 1319–1323. Hackett, F., Ford, L.B., Fèvre, E.M. and Simarro, P. (2014) Incorporating scale dependence in disease burden estimates: the case of human African trypanosomiasis in Uganda. PLoS Neglected Tropical Diseases 8, e2704. Hadush, A. (2015) Assessment of bovine tuberculosis in dairy farms and its public health importance in and around Adigrat District. MSc thesis, Mekelle University, Mekelle, Ethiopia. Havelaar, A.H., Kirk, M.D., Torgerson, P.R., Gibb, H.J., Hald, T., et al. (2015) World Health Organization global estimates and regional comparisons of the burden of foodborne disease in 2010. PLoS Medicine 12, e1001923. Hoffman, T., Nissen, K., Krambrich, J., Rönnberg, B., Akaberi, D., et al. (2020) Evaluation of a COVID-19 IgM and IgG rapid test; an efficient tool for assessment of past exposure to SARS-CoV-2. Infection Ecology & Epidemiology 10, 1754538. Holt, H.R., Inthavong, P., Boualam, K., Blaszak, K., Keokamphe, C., et al. (2016) Endemicity of zoonotic diseases in pigs and humans in lowland and upland Lao PDR: identification of socio-cultural risk factors. PLoS Neglected Tropical Diseases 10, e0003913. Holt, H.R., Inthavong, P., Blaszak, K., Keokamphe, C., Phongmany, A., et al. (2019) Production diseases in smallholder pig systems in rural Lao PDR. Preventive Veterinary Medicine 162, 110–116. Ijumba, J.N. and Lindsay, S.W. (2001) Impact of irrigation on malaria in Africa: paddies paradox. Medical and Veterinary Entomology 15, 1–11. Zoonoses 333 ILRI (2014) One Health, EcoHealth and agriculture associated diseases – report of a regional dialogue. New Delhi, 25 November 2013. ILRI, Nairobi. ILRI/ICAR (2013) Livestock research and development: summary report of the ICAR-ILRI partnership dialogue, New Delhi, 7 November 2012. ILRI, Nairobi. Jakobsen, F., Nguyen-Tien, T., Pham-Thanh, L., Bui, V.N., Nguyen-Viet, H., et al. (2019) Urban livestock- keeping and dengue in urban and peri-urban Hanoi, Vietnam. PLOS Neglected Tropical Diseases 13, e0007774. Johansen, M.V. and Mukaratirwa, S. (2013) The Cysticercosis Working Group of Eastern and Southern Africa (CWGESA). In Olsen, A., Ørnbjerg, N. and Winkel, K. (eds) A Success Story in Danish Devel- opment Aid – DBL (1964–2012). Narayana Press, Odder, Denmark. Jones, K.E., Patel, N.G., Levy, M.A., Storeygard, A., Balk, D., Gittleman, J.L. and Daszak, P., (2008). Global trends in emerging infectious diseases. Nature 451(7181), 990–993. Jones, B.A., Grace, D., Kock, R., Alonso, S., Rushton, J., et al. (2013) Zoonosis emergence linked to agri- cultural intensification and environmental change. Proceedings of the National Academy of Sciences 110, 8399–8404. Jost, C., Mariner, J.C., Roeder, P., Sawitri, E. and Macgregor-Skinner, G. (2007) Participatory epidemiology in disease surveillance and research. Revue Scientifique et Technique 26, 537–549. Jost, C.C., Nzietchueng, S., Kihu, S., Bett, B., Njogu, G., et al. (2010) Epidemiological assessment of the Rift Valley fever outbreak in Kenya and Tanzania in 2006 and 2007. American Journal of Tropical Medicine and Hygiene 83, 65–72. Kagira, J.M., Maingi, N., Kanyari, P.W.N., Githigia, S.M., Ng’ang’a, J.C. and Gachohi, J.M. (2010) Sero- prevalence of Cysticercus cellulosae and associated risk factors in free-range pigs in Kenya. Journal of Helminthology 84, 398–403. Kang’ethe, E.K., Arimi, S.M., Omore, A.O., McDermott, J.J., Nduhiu, J.G., et al. (2000) The prevalence of antibodies to Brucella abortus in marketed milk in Kenya and its public health implications. Paper presented at the 3rd All Africa Conference on Animal Agriculture (AACAA), 6–9 November, Alexan- dria, Egypt. ILRI, Nairobi. Kang’ethe, E.K., Kimani, V.N., McDermott, B., Grace, D., Lang’at, A.K., et al. (2012a) A trans-disciplinary study on the health risks of cryptosporidiosis from dairy systems in Dagoretti, Nairobi, Kenya: study background and farming system characteristics. Tropical Animal Health and Production 44, 3–10. Kang’ethe, E., Kimani, V., Grace, D., Mitoko, G., McDermott, B., et al. (2012b) Development and delivery of evidence-based messages to reduce the risk of zoonoses in Nairobi, Kenya. Tropical Animal Health and Production 44 (Suppl. 1), 41–46. Keiser, J., de Castro, M.C., Maltese, M.F., Bos, R., Tanner, M., et al. (2005a) Effect of irrigation and large dams on the burden of malaria on a global and regional scale. American Journal of Tropical Medicine and Hygiene 72, 392–406. Keiser, J., Utzinger, J. and Tanner, M. (2005b) The effect of irrigation and large dams on the burden of malaria on global and regional scale. Report prepared for the WHO commissioned study ‘Burden of water-related vector-borne diseases: An analysis of the fraction attributable to components of water resources development and management.’ WHO, Geneva, Switzerland. Kimani, T., Schelling, E., Bett, B., Ngigi, M., Randolph, T. and Fuhrimann, S. (2016) Public health bene- fits from livestock Rift Valley Fever control: a simulation of two epidemics in Kenya. EcoHealth 13, 729–742. Kivali, V., Thomas, L., Toye, P. and Fèvre, E. (2013) Diagnosis of porcine cysticercosis using lateral flow assay and HP10 Ag ELISA. Poster presented at the 47th Annual Scientific Conference of the Kenya Veterinary Association, 24–27 April, Mombasa, Kenya. ILRI, Nairobi. Kobinger, G.P., Leung, A., Neufeld, J., Richardson, J.S., Falzarano, D., et al. (2011) Replication, pathogen- icity, shedding, and transmission of Zaire ebolavirus in pigs. Journal of Infectious Diseases 204, 200–208. Koech, R.K. (2001) Epidemiology of tuberculosis in humans in Narok district with special interest on Myco- bacterium bovis. MSc thesis, University of Nairobi, Nairobi. Kungu, J.M., Dione, M.M., Ejobi, F., Harrison, L.J., Poole, E.J., et al. (2017) Sero-prevalence of Taenia spp. cysticercosis in rural and urban smallholder pig production settings in Uganda. Acta Tropica 165, 110–115. Lapar, M.L., Nuryartono, N., Toan, N.N., Rafani, I., Bett, B., et al. (2012) Are smallholders willing to pay for animal disease control? Empirical evidence from a study of mass vaccination for avian influenza in Indonesia. Asian Journal of Agriculture and Development 9, 74. 334 D. Grace et al. Lee, H.S., Thakur, K.K., Bui, V.N., Bui, A.N., Dang, M.V. and Wieland, B. (2019) Simulation of control scen- arios of porcine reproductive and respiratory syndrome in Nghe An Province in Vietnam. Transbound- ary and Emerging Diseases 66, 2279–2287. Lee, H.S., Bui, V.N., Nguyen, H.X., Bui, A.N., Hoang, T.D., et al. (2020a) Seroprevalences of multi-pathogen and description of farm movement in pigs in two provinces in Vietnam. BMC Veterinary Research 16, 15. Lee, H.S., Pham T.L. and Wieland, B. (2020b) Temporal patterns and space-time cluster analysis of foot-and-mouth disease (FMD) cases from 2007 to 2017 in Vietnam. Transboundary and Emerging Diseases 67, 584–591. Lejon, V., Ilboudo, H., Mumba, D., Camara, M., Kaba, D., et al. (2017) Diagnostic tools for human African trypano- somiasis elimination and clinical trials: the DiTECT-HAT project. BMJ Global Health 2 (Suppl. 2), A8. Leksmono, C., Young, J., Hooton, N., Muriuki, H. and Romney, D. (2006) Informal traders lock horns with the formal milk industry: the role of research in pro-poor dairy policy shift in Kenya. Working Paper 266. ILRI, Nairobi. Liljander, A., Meyer, B., Jores, J., Müller, M.A., Lattwein, E., et al. (2016) MERS-CoV antibodies in humans, Africa, 2013–2014. Emerging Infectious Diseases 22, 1086–1089. Lindahl, J.F., Kumar, N.G., Deka, R.P., Shome, R. and Grace, D. (2018) Serological evidence of Brucella infections in dairy cattle in Haryana, India. Infection Ecology & Epidemiology 8, 1555445. Mariner, J.C., Hendrickx, S., Pfeiffer, D.U., Costard, S., Knopf, L., et al. (2011) Integration of participatory approaches into surveillance systems. Revue Scientifique et Technique 30, 653–659. Mariner, J.C., Jones, B.A., Hendrickx, S., El Masry, I., Jobre, Y. and Jost, C.C. (2014) Experiences in participatory surveillance and community-based reporting systems for H5N1 highly pathogenic avian influenza: a case study approach. EcoHealth 11, 22–35. Marsh, G.A., Haining, J., Robinson, R., Foord, A., Yamada, M., et al. (2011) Ebola Reston virus infection of pigs: clinical significance and transmission potential. Journal of Infectious Diseases 204, S804–S809. Mbotha, D., Bett, B., Kairu-Wanyoike, S., Grace, D., Kihara, A., et al. (2017) Inter-epidemic Rift Valley fever virus seroconversions in an irrigation scheme in Bura, south-east Kenya. Transboundary and Emer- ging Diseases 65, e55–e62. McDermott, J.J. and Arimi, S.M. (2002) Brucellosis in sub-Saharan Africa: epidemiology, control and impact. Veterinary Microbiology 90, 111–134. McDermott, J.J., Kristjanson, P.M., Kruska, R.L., Reid, R.S., Robinson, T.P., et al. (2001) Effects of climate, human population and socio-economic changes on tsetse-transmitted trypanosomiasis to 2050. In: Seed, R. and Black, S. (eds) World Class Parasites, Vol 1: The African Trypanosomes. Kluwer Academic, Boston, pp. 25–38. McDermott, J., Grace, D. and Zinsstag, J. (2013) Economics of brucellosis impact and control in low-income countries. Scientific and Technical Review 32, 249–261. Morens, D.M., Folkers, G.K. and Fauci, A.S. (2004) The challenge of emerging and re-emerging infectious diseases. Nature 430, 242–249. Muller, M.A., Corman, V.M., Jores, J., Meyer, B., Younan, M., et al. (2014) MERS coronavirus neutral- izing antibodies in camels, eastern Africa, 1983–1997. Emerging Infectious Diseases 20, 2093–2095. Mulugeta, W., Wilkes, J., Mulatu, W., Majiwa, P.A.O., Masake, R. and Peregrine, A.S. (1997). Mechanisms and genetics of resistance to trypanocides in African trypanosomes. Acta Tropica 64, 205–217. Munyua, P., Bitek, A., Osoro, E., Pieracci, E.G., Muema, J., et al. (2016a) Prioritization of zoonotic diseases in Kenya 2015. PLoS One 11, e0161576. Munyua, P.M., Murithi, R.M., Ithondeka, P., Hightower, A., Thumbi, S.M., et al. (2016b) Predictive factors and risk mapping for Rift Valley fever epidemics in Kenya. PLoS One 11, e0144570. Musewa, A., Roesel, K., Grace, D., Dione, M. and Erume, J. (2018) Detection of Erysipelothrix rhusiopathi- ae in naturally infected pigs in Kamuli District, Uganda. Revue d’Élevage et de Médecine Vétérinaire des Pays Tropicaux 71, 31229. Mutua, E.N., Bukachi, S.A., Bett, B.K., Estambale, B.A. and Nyamongo, I.K. (2017) ‘We do not bury dead livestock like human beings’: community behaviors and risk of Rift Valley Fever virus infection in Barin- go County, Kenya. PLoS Neglected Tropical Diseases 11, e0005582. Mutua, F.K., Randolph, T.F., Arimi, S.M., Kitala, P.M., Githigia, S.M., et al. (2007) Palpable lingual cysts, a possible indicator of porcine cysticercosis, in Teso District, Western Kenya. Journal of Swine Health and Production 15, 206–212. Zoonoses 335 Mutua, E., de Haan, N., Tumusiime, D., Jost, C. and Bett, B. (2019) A qualitative study on gendered bar- riers to livestock vaccine uptake in Kenya and Uganda and their implications on Rift Valley fever control. Vaccines 7, 86. Muturi, M., Gachohi, J., Mwatondo, A., Lekolool, I., Gakuya, F., et al. (2018) Recurrent anthrax outbreaks in humans, livestock, and wildlife in the same locality, Kenya, 2014–2017. American Journal of Tropical Medicine and Hygiene 99, 833–839. Mwololo, D.K. (2016) Seroepidemiological survey of Q fever in livestock and humans in Bura, Tana River County, Kenya. MSc thesis, University of Nairobi, Nairobi. Nanyingi, M.O., Munyua, P., Kiama, S.G., Muchemi, G.M., Thumbi, S.M., et al. (2015) A systematic review of Rift Valley Fever epidemiology 1931–2014. Infection Ecology & Epidemiology 5, 1–12. Nguyen-Viet, H., Chotinun, S., Schelling, E., Widyastuti, W., Khong, N.V., et al. (2016) Reduction of anti- microbial use and resistance needs sectoral-collaborations with a One Health approach: perspectives from Asia. International Journal of Public Health 62 (Suppl. 1), 3–5. Nguyen-Viet, H., Grace, D. and McDermott, J. (2019) Integrated approaches to tackling health issues- related to agri-food systems. International Journal of Public Health 64, 5–6. Njenga, M.K., Paweska, J., Wanjala, R., Rao, C.Y., Weiner, M., et al. (2009) Using a field quantitative real-time PCR test to rapidly identify highly viremic rift valley fever cases. Journal of Clinical Microbiology 47, 1166–1171. Njeru, J., Henning, K., Pletz, M.W., Heller, R., Forstner, C., et al. (2016) Febrile patients admitted to remote hospitals in Northeastern Kenya: seroprevalence, risk factors and a clinical prediction tool for Q-Fever. BMC Infectious Diseases 16, 244. Nsadha, Z., Thomas, L.F., Fèvre, E.M., Nasinyama, G., Ojok, L. and Waiswa, C. (2014) Prevalence of porcine cysticercosis in the Lake Kyoga Basin, Uganda. BMC Veterinary Research 10, 239. Nyangaga, J.N., Grace, D., Kimani, V., Kiragu, M.W., Langat, A.K., et al. (2012) Outcome mapping for fostering and measuring change in risk management behaviour among urban dairy farmers in Nairobi, Kenya. Tropical Animal Health and Production 44 (Suppl. 1), S47–S51. Nyokabi, S., Birner, R., Bett, B., Isuyi, L., Grace, D. et al. (2017) Informal value chain actors’ knowledge and perceptions about zoonotic diseases and biosecurity in Kenya and the importance for food safety and public health. Tropical Animal Health and Production 50, 509–518. Oboge, H.M. (2016) Socio-economics and perceptions of Q-fever infection in a pastoralist system of Kajiado County. MSc thesis, University of Nairobi, Nairobi. Odiit, M., Coleman, P.G., Fèvre, E.M., Magona, J. and McDermott, J.J. (2000) Incorporating the burden of human sleeping sickness in an economic impact assessment of trypanosomiasis. In: Proceedings of the Ninth International Symposium on Veterinary Epidemiology and Economics, 7–11 August, Breck- enridge, Colorado. ISVEE, Breckenridge, Colorado, pp. 1300–1303. Odiit, M., Shaw, A., Welburn, S.C., Fèvre, E.M., Coleman, P.G. and McDermott, J.J. (2004) Assessing the patterns of health-seeking behaviour and awareness among sleeping-sickness patients in eastern Uganda. Annals of Tropical Medicine and Parasitology 98, 339–348. Odiit, M., Coleman, P.G., Liu, W.C., McDermott, J.J., Fèvre, E.M., et al. (2005) Quantifying the level of under- detection of Trypanosoma brucei rhodesiense sleeping sickness cases. Tropical Medicine & Inter- national Health 10, 840–849. Odiit, M., Bessell, P.R., Fèvre, E.M., Robinson, T.P., Kinoti, J., et al. (2006) Using remote sensing and geographic information systems to identify villages at high risk for rhodesiense sleeping sickness in Uganda. Transactions of the Royal Society of Tropical Medicine and Hygiene 100, 354–362. Odongo, D.O., Tiampati, C.M., Mulinge, E., Mbae, C.K., Bishop, R.P., et al. (2018) Prevalence and geno- typing of Echinococcus granulosus in sheep in Narok County, Kenya. Parasitology Research 117, 2065–2073. Omore, A.O., Arimi, S., Kang’ethe, E.K. and McDermott, J.J. (2000) Analysis of public health risks from consumption of informally marketed milk in Kenya. Paper presented at the Faculty of Veterinary Medi- cine Biennial Scientific Conference, 30–31 August, University of Nairobi, Kenya. ILRI, Nairobi. O’Neill, J. (2015) Antimicrobials in agriculture and the environment: reducing unnecessary use and waste. Review on Antimicrobial Resistance. Wellcome Trust. Perry, B.D., Randolph, T.F., McDermott, J.J., Sones, K.R. and Thornton, P.K. (2002) Investing in animal health research to alleviate poverty. ILRI, Nairobi. Rich, K.M. and Wanyoike, F. (2010) An assessment of the regional and national socio-economic impacts of the 2007 Rift Valley fever outbreak in Kenya. American Journal of Tropical Medicine and Hygiene 83 (Suppl.), 52–57. 336 D. Grace et al. Robinson, T.P., Bu, D.P., Carrique-Mas, J., Fèvre, E.M., Gilbert, M. et al. (2016) Antibiotic resistance is the quintessential One Health issue. Transactions of the Royal Society of Tropical Medicine and Hygiene 110, 377–380. Roesel, K., Nöckler, K., Baumann, M.P.O., Fries, R., Dione, M.M., et al. (2016) First report of the occurrence of Trichinella-specific antibodies in domestic pigs in Central and Eastern Uganda. PLoS One 11, e0166258. Sang, R., Lutomiah, J., Said, M., Makio, A., Koka, H., et al. (2016) Effects of irrigation and rainfall on the population dynamics of Rift Valley fever and other Arbovirus Mosquito vectors in the epidemic prone Tana River County, Kenya. Journal of Medical Entomology 54, 460–470. Sekar, N., Shah, N.K., Abbas, S.S. and Kakkar, M. (2011) On behalf of the Roadmap to Combat Zoonoses in India (RCZI) Initiative Research options for controlling zoonotic disease in India, 2010–2015. PLoS One 6, e17120. Shaw, A.P.M., Wint, G.R.W., Cecchi, G., Torr, S.J., Mattioli, R.C. and Robinson, T.P. (2015) Mapping the benefit–cost ratios of interventions against bovine trypanosomosis in Eastern Africa. Preventive Veter- inary Medicine 122, 406–416. Shewatatek, F. (2015) Bovine tuberculosis: prevalence, economic implications and diagnostic tests com- parison in smallholder dairy farms in selected districts of North Gondar Zone, Amhara Region. Thesis, University of Gondar, Gondar, Ethiopia. Shija, F. (2013) Assessment of milk handling practices and bacterial contaminations along the dairy value chain in Lushoto and Handeni districts, Tanzania. MSc thesis, Sokoine University of Agriculture, Moro- goro, Tanzania. Simpson, A. (2015) Consumption of fallen stock: a risk assessment. MSc thesis, Royal Veterinary College, University of London, London. Smith, E., Atherstone, C. and Grace, D. (2015) One Health approach recommended in investigating and com- municating the potential role of pigs in transmitting Ebola in Uganda. ILRI Policy Brief No 17. ILRI, Nairobi. Sow, A., Faye, O., Ba, Y., Diallo, D., Fall, G., et al. (2016) Widespread Rift Valley fever emergence in Sene- gal in 2013–2014. Open Forum Infectious Diseases 3, ofw149. Teshome, H., Molla, B. and Tibbo, M. (2003) A seroprevalence study of camel brucellosis in three camel-rearing regions of Ethiopia. Tropical Animal Health and Production 35, 381–390. Thomas, L.F., de Glanville, W.A., Cook, E.A. and Fèvre, E.M. (2013) The spatial ecology of free-ranging domestic pigs (Sus scrofa) in western Kenya. BMC Veterinary Research 9, 46. Thomas, L.F., Harrison, L.J.S., Toye, P., de Glanville, W.A., Cook, E.A.J., et al. (2016) Prevalence of Taenia solium cysticercosis in pigs entering the food chain in western Kenya. Tropical Animal Health and Production 48, 233–238. Thomas, L.F., de Glanville, W.A., Cook, E.A., Bronsvoort, B.M., Handel, I., et al. (2017) Modelling the risk of Taenia solium exposure from pork produced in western Kenya. PLoS Neglected Tropical Diseases 11, e0005371. Tschopp, R., Schelling, E., Hattendorf, J., Aseffa, A. and Zinsstag, J. (2009) Risk factors of bovine tubercu- losis in cattle in rural livestock production systems of Ethiopia. Preventive Veterinary Medicine 89, 205–211. Unger, F., Bett, B., Siregar, E.S., Idris, S. and Randolph, T.F. (2014) Integrated HPAI control in kampong chicken in Indonesia – an overview on ILRI’s research and lessons learned. Poster presented at the Tropentag 2014 Conference on Bridging the Gap between Increasing Knowledge and Decreasing Resources, 17–19 September, Prague, Czech Republic. ILRI, Nairobi. Unger, F., Long, C.T.M. and Khong, N.V. (2016) Prevalence of trichinellosis and cysticercosis in indigen- ous pigs from ethnic minorities for selected communes in the Central Highlands (Dak Lak). Summary report. ILRI, Nairobi. United Nations Environment Programme and International Livestock Research Institute. (2020). Prevent- ing the next pandemic: Zoonotic diseases and how to break the chain of transmission. Nairobi, Kenya: UNEP. https://cgspace.cgiar.org/handle/10568/108707 USDA/USAID/ILRI (2013) An integrated approach to controlling brucellosis in Africa: workshop report. ILRI, Nairobi. van Boeckel, T.P., Brower, C., Gilbert, M., Grenfell, B.T., Levin, S.A., et al. (2015) Global trends in antimicrobial use in food animals. Proceedings of the National Academy of Sciences USA 112, 5649–5654. Verdugo, C., El Masry, I., Makonnen, Y., Hannah, H., Unger, F., et al. (2016) Sensitivity and specificity es- timation for the clinical diagnosis of highly pathogenic avian influenza in the Egyptian participatory disease surveillance program. Avian Diseases 60, 805–809. Zoonoses 337 Vordermeier, M., Ameni, G., Berg, S., Bishop, R., Robertson, B.D., et al. (2012) The influence of cattle breed on susceptibility to bovine tuberculosis in Ethiopia. Comparative Immunology, Microbiology and Infec- tious Diseases 35, 227–232. Waiswa C., Fèvre, E.M., Nsadha, Z., Sikasunge, C.S. and Willingham, A.L. (2009) Porcine Cysticercosis in Southeast Uganda: seroprevalence in Kamuli and Kaliro Districts. Journal of Parasitology Research 2009, 375493. Wamboga, C., Matovu, E., Bessell, P.R., Picado, A., Biéler, S. and Ndung’u, J.M. (2017) Enhanced passive screening and diagnosis for gambiense human African trypanosomiasis in north-western Uganda – moving towards elimination. PLoS One 12, e0186429. Wardrop, N.A., Thomas, L.F., Cook, E.A.J., de Glanville, W.A., Atkinson, P.M., et  al. (2016) The sero- epidemiology of Coxiella burnetii in humans and cattle, western Kenya: evidence from a cross-sectional study. PLOS Neglected Tropical Diseases 10, e0005032. Warimwe, G.M., Gesharisha, J., Carr, B.V., Otieno, S., Otingah, K., et al. (2016) Chimpanzee adenovirus vaccine provides multispecies protection against Rift Valley Fever. Scientific Reports 6, 20617. Warimwe, G.M., Lorenzo, G., Lopez-Gil, E., Reyes-Sandoval, A., Cottingham, M.G., et al. (2013) Immuno- genicity and efficacy of a chimpanzee adenovirus-vectored Rift Valley fever vaccine in mice. Virology Journal 10, 349. Watts, N., Adger, W.N., Ayeb-Karlsson, S., Bai, Y., Byass, P., et al. (2017) The Lancet Countdown: tracking progress on health and climate change. Lancet 389, 1151–1164. Watts, N., Amann, M., Ayeb-Karlsson, S., Belesova, K., Bouley, T., et al. (2018a) The Lancet Countdown on health and climate change: from 25 years of inaction to a global transformation for public health. Lan- cet 391, 581–630. Watts, N., Amann, M., Arnell, N., Ayeb-Karlsson, S., Belesova, K., et al. (2018b) The 2018 report of the Lan- cet Countdown on health and climate change: shaping the health of nations for centuries to come. Lancet 392, 2479–2514. Williams, R., Malherbe, J., Weepener, H., Majiwa, P. and Swanepoel, R. (2016) Anomalous high rainfall and soil saturation as combined risk indicator of Rift Valley fever outbreaks, South Africa, 2008–2011. Emerging Infectious Diseases 22, 2054–2062. Wohlgemut, J., Dewey, C., Levy, M. and Mutua, F. (2010) Evaluating the efficacy of teaching methods regarding prevention of human epilepsy caused by Taenia solium neurocysticercosis in Western Kenya. American Journal of Tropical Medicine and Hygiene 82, 634–642. Woolhouse, M.E. and Gowtage-Sequeria, S. (2005) Host range and emerging and reemerging pathogens. Emerging Infectious Diseases 11, 1842–1847. World Bank (2012) People, pathogens and our planet: the economics of One Health. World Bank, Washington, D.C. WHO (2006) The control of neglected zoonotic diseases: a route to poverty alleviation: report of a joint WHO/ DFID-AHP meeting, 20 and 21 September 2005, WHO Headquarters, Geneva, with the participation of FAO and OIE. WHO, Geneva. Available at: www.who.int/neglected_diseases/zoonoses/9789241594301/ en/ (accessed 12 February 2020). WHO (2013) Research priorities for the environment, agriculture and infectious diseases of poverty: technical report of the TDR Thematic Reference Group on Environment, Agriculture and Infectious Diseases of Poverty. Technical report series no. 976. WHO, Geneva. 9 Food Safety and Nutrition Delia Grace1, Silvia Alonso2, Bernard Bett3, Johanna Lindahl4, Ekta Patel3, Hung Nguyen-Viet5, Kristina Roesel6, Fred Unger5, and Paula Dominguez-Salas1 1International Livestock Research Institute, Nairobi, Kenya and University of Green- wich, UK; 2International Livestock Research Institute, Jerusalem, Israel; 3International Livestock Research Institute, Nairobi, Kenya; 4International Livestock Research Institute, Nairobi, Kenya, and Uppsala University, Sweden; 5International Livestock Research Institute, Hanoi, Vietnam; 6International Livestock Research Institute, Frankenberg, Saxony, Germany Contents Executive Summary 338 The problem 338 ILRI’s role in the global context 339 Impacts of ILRI’s research 339 Scientific impacts 339 Development impacts 339 Policy impacts 340 Capacity building 340 Introduction 340 Why food safety matters 340 The History of Food Safety Research at ILRI and in CGIAR 341 Food safety pathways to impact 343 Empirical evidence 344 Developing research methods and tools 345 Developing and testing innovations for application 346 Capacity building activities by ILRI 347 Influence on international, regional and national policies 348 Impact that scales 351 Training and enabling informal sector agents 351 Human Nutrition Research at ILRI 354 The Future 360 References 360 Executive Summary and food systems develop, concerns about food safety grow. The emergence of food safety science The problem responds to those concerns. Food safety science – drawing on health, agri- ‘Is our food safe?’ is a fundamental concern of culture, technology, marketing and psychology – consumers. Moreover, as populations urbanize emerged as a separate discipline in the latter half © International Livestock Research Institute 2020. The Impact of the International 338 Livestock Research Institute (eds J. McIntire and D. Grace) Food Safety and Nutrition 339 of the last century. Food safety is relevant to safety, and towards consumer willingness to pay domestic and international markets and involves for food attributes, including safety. private and public sectors as well as civil society. In the early 2000s, ILRI began a programme Recent evidence suggests that the health burden on improving human health through livestock re- of food-borne disease (FBD) is comparable to that search in three areas: (i) animal-source foods for of three major diseases – malaria, human immuno- nutrition; (ii) zoonoses (diseases transmitted be- deficiency virus/acquired immune deficiency tween animals and people); and (iii) FBD. This was syndrome (HIV/AIDS) and tuberculosis. Most of the first CGIAR group with an explicit food safety the unsafe food health burden is due to contam- mandate (rather than focusing on specific hazards) inated fresh foods purchased from informal mar- and with expertise in using research methods for kets. Livestock products – milk, meat, offal and food safety rather than diseases in general. ILRI eggs – are especially risky. As our understanding was also one of the first groups to focus on food of the importance of FBD, and its complicated safety in the ‘informal markets’ of developing links with livestock development, has increased, countries, and by the 2010s, had become the lead so too has research conducted by the Inter- research institute globally in this emerging area. national Livestock Research Institute (ILRI) and other research organizations in this area. Impacts of ILRI’s research ILRI’s role in the global context Scientific impacts ILRI developed, contributed to, adapted and tested Food safety was historically a minor part of tools, methods and metrics, including participa- CGIAR research. This was partly due to a lack of tory risk assessment, systematic literature reviews of awareness that FBD was a major development food safety in informal markets, systems dy- issue but also because FBD was conceptualized namics and food safety system performance assess- as an aggregation of specific diseases rather than ment. Technology development and testing was a as a systems problem. Donor investments in food growing area with a focus on appropriate tech- safety were small compared with the scale of the nologies such as disinfectants and pest control. problem, with investments in comparable diseases Many publications were produced, often regarding and with the potential return on investments tools and technologies. In addition, publications (GFSP, 2019). Most investments have focused on covered other aspects of evidence generation in- trade rather than on ensuring the health of con- cluding reviews, reports of surveys, risk factor ana- sumers in low- and middle-income countries. In lyses, and interaction between food safety and the early 2000s, ILRI conducted some work on other development issues, such as gender equity. the health aspects of trade in livestock but did not have a major programme in food safety. Development impacts Rather, food safety was seen mainly as a po- tential barrier to market access by poor livestock ILRI’s initial work on food safety focused on keepers. It was addressed to some extent in dairy adapting methodologies for developing coun- projects and initiatives such as the Debre Zeit tries, assessing the extent, nature and drivers of Dairy Technology Centre. As such, ILRI and its FBD, and piloting potential solutions. Only with predecessors, the International Livestock Centre the advent of the CGIAR Research Programmes for Africa (ILCA) and the International Laboratory (CRPs) (2012–2016) did the focus shift to achiev- for Research on Animal Diseases (ILRAD), were ing wide-scale development impacts. However, one of many research and development institutes some development potential and realized impacts seeking to increase the quality of agricultural can be discerned. In summary, pilot projects identi- products through research, training and capacity fied various promising technologies. Moving to building. An evolving focus on food value chains the intermediate scale, food safety research has in these institutions, however, helped shift the been embedded in high-potential livestock value research agenda towards the food attributes de- chains identified by the CRP on Livestock. These sired by consumers, which often included food CRP initiatives reached hundreds of value chain 340 D. Grace-Randolph et al. agents, thousands of farmers and tens of thousands work on food safety to draw conclusions about of consumers, although the impact on food safety its actual and potential impacts. Unlike other outcomes is more difficult to estimate. In Kenya aspects of agricultural research, food safety is a and the Indian state of Assam, there is some relatively new area for CGIAR, and we can easily evidence that food safety interventions went to scale trace the emergence and growth of its research and were sustained after the end of projects. We agenda. The research agenda represents a de- estimate that 6.5 million people benefited. parture from traditional CGIAR research in two main ways: (i) food consumers rather than food Policy impacts producers are the focus of food safety research; and (ii) the prime motivator of food safety research The first assessment of the global health burden is improving human health rather than improv- of FBD found that the burden was unexpectedly ing farm productivity, food security or natural high and borne mostly by low- and middle- resource management. income countries (Havelaar et al., 2015). ILRI was one of the few research institutes with a substan- Why food safety matters tial track record and publications in this area. As such, it was requested or commissioned to produce evidence syntheses for several intergovernmental Food-borne disease (FBD) includes any illness organizations and donor agencies, substantially caused by ingesting contaminated or naturally influencing their policies and activities. At a hazardous food or drink. Food produced in devel- national level, ILRI has had a major impact on food oping countries often contains high levels of bio- safety policies in Kenya and Vietnam, as well as logical and chemical hazards and is prone to in India’s Assam state, and has had a moderate adulteration (Grace, 2015a,b), therefore creat- impact in several other countries, including ing conditions in which FDB thrives. Cambodia, Uganda, Tanzania and Ethiopia. Only recently has good evidence on the bur- dens of FBD in developing countries started to emerge. The best assessment was published by Capacity building the World Health Organization (WHO) in 2015, the culmination of nearly 10 years of work by Given the previous neglect of food safety in the dozens of experts (Havelaar et al., 2015; Gibb domestic markets of low- and middle-income et al., 2015). A conservative estimate found that countries, ILRI research had a strong emphasis the health burden of unsafe foods (a combination on capacity building. Most food safety research of morbidity and mortality) was comparable to projects included students at undergraduate, that of malaria, HIV/AIDS or tuberculosis, mak- MSc and PhD levels. In several countries, such as ing FBD a major public health priority. The first Tanzania and Vietnam, all qualified risk analysis part of the study, focusing on 31 hazards for professionals have been trained by ILRI. In add- which there was enough information to gener- ition, ILRI developed and delivered a range of ate global estimates, found that around 98% of 1–2-week trainings aimed at policy makers and the FBD burden fell on developing countries, and implementers. Many thousands of value chain 97% was due to microbes, parasites or viruses, agents were trained in individual projects. This with the remainder due to chemical hazards. was done to develop appropriate training and to FBD from these hazards caused 600 million ill- test approaches because the role of CGIAR was nesses and 420,000 deaths in 2010. The second seen to be that of developing material, approaches, part of the study, using a less conservative meth- delivery systems and incentives that could sup- odology, found four heavy metals resulted in an port the training of value chain agents rather additional 1 million illnesses and 56,000 deaths than to conduct the training itself. in 2015. FBDs are estimated to cost the USA US$15–80 billion a year (Scharff, 2012; Hoff- mann et al., 2015), which would be as high as Introduction 0.4% of estimated 2020 US gross domestic prod- uct (GDP). A recent World Bank/ILRI study esti- What is the role of international agriculture re- mated that FBD costs developing countries at search in food safety? This chapter looks at ILRI’s least US$100 billion a year (Jaffee et al., 2019). Food Safety and Nutrition 341 The WHO study on FBD identified the haz- 1986 at Debre Zeit, about 65 km south-east of ards responsible for most illness and death. In Addis Ababa (ILCA, 1987), which aimed to developed countries, most of the FBD burden is d evelop milk-processing methods adapted for attributable to microbes, especially those of zoo- smallholders. This unit produced manuals that notic origin; in developing countries, macropar- covered hygienic milk handling but did not focus asites are relatively important in addition to the on milk safety. The episodic production of dairy microbes controlled in developed countries (such manuals and training material continued over as those responsible for cholera and brucellosis) the next decades: a major achievement was seen (Havelaar et al., 2015). It is more difficult to as- in 2006 when dairy boards in Kenya, Rwanda, certain which food is responsible. In developed Tanzania and Uganda endorsed generic training countries, most of the burden is due to animal- material for informal milk traders in the eastern source food and fresh produce, and this seems to and central Africa region. During this time, dairy be the case in developing countries (Hoffmann research for development work continued in et al., 2017; Grace, 2015a). Botswana, Burundi, Ethiopia, Ghana, Kenya, Aside from its health burden and associated Malawi, Mali, Nigeria, Senegal and Zimbabwe, economic costs, FBD is important as a barrier to among others; in the early 2000s, this was extended market access. Food export markets, formal mar- to Latin America and India. kets and provisioning programmes already require Food safety, as opposed to food technology, food to meet certain sanitary and phytosanitary research started at ILRI after the institute widened standards and, as a result, tend to exclude small- the focus of its predecessors to cover a broader holder, women, less educated and more remote range of livestock issues following the merger of farmers, who have less ability than others to ILRAD and ILCA in 1995. The first food safety meet these standards (Unnevehr and Ronchi, research started in the late 1990s. It was con- 2014). As concern over FBD increases, meeting ducted within a veterinary public health frame- food safety standards is likely to become an ever work and focused on milk safety in Kenya (Aboge more important constraint to smallholder pro- et al., 2000; Kang’ethe et al., 2000, Mwangi et al., duction. These health, economic and equity 2000; Omore et al., 2000). This work was ex- concerns show how relevant food safety issues tended to Ghana, Tanzania and Uganda in the are to pro-poor agricultural research for devel- early 2000s. Food-borne zoonoses were specific- opment. ally considered among other zoonotic diseases in This chapter first summarizes the history of a landmark ILRI volume prioritizing the livestock food safety research at ILRI and CGIAR, describ- diseases whose control would most significantly ing how the discipline grew, became a research reduce poverty (Perry et al., 2002). Around the agenda and evolved from an ad hoc and haz- same time, another strand of research started on ard-based agenda to one that was more system- economic aspects of food safety, especially the atic and risk based. The next section sets out the trade-offs between safe food and other develop- theory of change linking food safety research to ment objectives (Omore et al., 2001). This led to economic and health benefits. It identifies two the development of an ILRI programme on Ani- main pathways: evidence that counts and im- mal Health and Food Safety for Trade, which had pact that scales. The following sections summar- the objective of addressing food safety as a bar- ize ILRI progress along both pathways, and we rier to smallholder market access rather than as end with conclusions and recommendations for a constraint to human health. Congruent with new food safety research. the economic perspective, there was research on consumer demand for safety and quality. Ten studies from seven countries in Asia and Africa The History of Food Safety Research were brought together in an influential report at ILRI and in CGIAR (Jabbar et al., 2010) In 2003, for the first time, ILRI initiated a ILCA was established in 1974, and by 1977 had programme – Livestock Keeping and Human developed a research programme on smallholder Health Impacts – with an explicit focus on improv- production in the eastern African highlands. An ing human health through livestock. This marked ILCA Dairy Technology Unit was founded in the start of ILRI research employing a risk 342 D. Grace-Randolph et al. analysis framework and focusing on improving sector alternatives and often offer services (such food safety outcomes rather than subsuming food as immediate payment to farmers and provision safety under market issues or veterinary public of credit to consumers) that the formal sector health. Following an external review (Science does not provide. Food is perceived by consumers Council/CGIAR, 2008), food safety was again to be fresh, healthy, natural, convenient and less placed in an economic programme: in hindsight, expensive (Roesel and Grace, 2014; Zhong et al., this was a retrograde move given the broad trends 2020). With these advantages, it is not surpris- of agricultural research towards greater emphasis ing that the formal sector share of animal-source on human health. Subsequently, the ILRI pro- food markets is less than 10% in most of sub- gramme on Animal Health and Food Safety for Saharan Africa and South Asia (Gomez and Trade became Animal Health, Food Safety and Ricketts, 2013). In southern and East Africa, in- Zoonoses, and finally Food Safety and Zoonoses, formal markets currently supply 85–95% of mar- as it became clear that world export markets ket demand and are predicted to still supply were less important to poor people and that FBD 50–70% of market demand in 2040 (Tschirley was more important than had been realized. In et al., 2015). In South Asia, traditional food 2017, the wheel came full circle when research retail occupies 95% of the market, in South-east groups working on different aspects of human Asia 71% and in South America 54% of the food and animal health in four separate ILRI pro- retailed. In this context, informal markets are grammes across ILRI’s two directorates (bio- likely to remain important for at least several sciences and integrated sciences) were brought more decades. together in a new Animal and Human Health The relative neglect of informal markets Programme. Food safety was one of four major compared with other CGIAR research areas im- areas in this programme (the others were zoonoses plies greater marginal utility of research invest- and emerging infectious disease, herd health, ments. ILRI is almost unique in having a large and vaccines and diagnostics). research programme focused on food safety in The ILRI food safety research agenda focuses informal markets, with a strong focus on gener- its attention on traditional ‘informal markets’, ating actionable, high-quality evidence. As such, where most smallholder and poor farmers sell the group is responsible for much of the research their livestock products. Traditional processing, information in this area. Importantly, the group products and prices predominate in these infor- produced the first book on food safety in informal mal or ‘wet’ markets, which tend to escape effect- markets (Roesel and Grace, 2014), in addition to ive health and safety regulation, go untaxed and dozens of journal papers, theses, posters, confer- unlicensed, and sell food at lower prices than ence papers, research briefs, policy briefs, videos, formal markets. Informal markets are also closer infographics and blogs. ILRI also conducted to and more accessible for poor consumers than numerous training courses for policy makers, formal markets. researchers and value chain agents. The results An ILRI review of food safety and informal of the various research and training activities markets largely categorized the attitude of officials are all available in open-access formats from the and donors towards informal markets as one of ILRI document repository (CGSpace: https://cg- either neglect or unhelpful attention (Roesel and space.cgiar.org/; accessed 19 February 2020) Grace, 2014). Much attention has been paid to and other sites. the role of informal markets in maintaining and In parallel with the evolution of food safety transmitting diseases but little to their role in at ILRI, there have been developments in the role supporting livelihoods (especially for women) and of food safety in CGIAR, in which ILRI has been nutrition. Informal markets are often seen as a major player. Food safety was not an initial outdated and unsafe, destined to be replaced by focus of CGIAR research, with the first official industrial production and modern retail. The mention of food safety in 2000 (Technical Ad- ongoing COVID-19 pandemic has accentuated visory Committee, 2000). However, eight CGIAR this belief among many stakeholders, especially centres had started small-scale research related those not familiar with wet markets. to food safety in the following areas: breeding Nevertheless, informal outlets are much more staple crops resistant to pests (so farmers can re- common and widely distributed than formal duce pesticide use), breeding staple crops resistant Food Safety and Nutrition 343 to aflatoxins, controlling aflatoxins using other Box 9.1. Justification for incorporation of food organisms (biocontrol), breeding ergot (fungus) safety research at ILRI. resistance in sorghum, reducing cyanide levels in cassava, and improving milk quality and safety Food safety research is a relatively new area for (Kassam and Barat, 2003). Only research in the CGIAR and ILRI. It can be seen as a response last area assessed health outcomes. In 2011, an- to changing agri-food systems and as evidence other survey of CGIAR food safety research was of the ability of CGIAR to take on new challenges. conducted, with more centres reporting food Like most organizations, ILRI periodically revisits safety research. Aflatoxin research dominated, its priorities and strategies, but this is typically but there was an expansion of risk assessment done based on donor interest, popular wisdom, consultation and expertise rather than by using and prioritization activities and substantial pro- a systematic framework. Important exceptions grammes on the safety of perishables (vegetables at ILRI were: (i) an institutional prioritization in and animal-source foods), on zoonotic diseases, 1998 based on ex ante returns to research, which on occupational hazards and on water-associated tightened ILRI’s focus but led to a reduction of diseases. As this list suggests, food safety research work in Latin America and with pastoralists; (ii) a was almost entirely supply led, with centres look- geographic information system (GIS)-based ing at problems in the commodities they special- mapping of poor livestock keepers, suggesting ized in and with no overall alignment to health that Asian poor livestock keepers were neglected outcomes. The research effort and budget were clients (Thornton et al., 2002); (iii) an accompany- very small compared with the overall CGIAR ing identification of research priorities based on expert opinion of pro-poor impacts, which research portfolio. showed the importance of zoonoses (Perry Food safety research became more prom- et al., 2002); (iv) a non-systematic identification inent with the development of the CRP on of priority countries for CRPs, which for the first Agriculture for Nutrition and Health (A4NH), time focused on consumers as well as produ- one of 15 CGIAR multicentre research pro- cers; and (v) a mapping of zoonoses and poverty, grammes (Box 9.1). The Nutrition and Health which suggested that FBDs were among the programme was originally conceived as a joint most important zoonoses (Grace et al., 2012c). venture between ILRI and the International These exercises in general provided justification Food Policy Research Institute (IFPRI). How- for increased focus on FBD. In particular, com- ever, the CGIAR Consortium (now the CGIAR pared with other animal health issues, FBD is relatively important, neglected and tractable, System Office) refused a jointly led CGIAR characteristics suggesting that it is a relatively research programme, and, because most of the promising area for research investment. research in this programme focused on nutri- tion, it was agreed that IFPRI should lead the programme. A4NH had four main themes, or Food safety pathways to impact flagships, three focused on nutrition and one on the diseases associated with agriculture, The 2017–2021 strategy for food safety research i ncluding FBD. A4NH brought together port- in A4NH identified two impact paths: evidence folios on aflatoxin research led by the International that counts and impact that scales (A4NH, Institute of Tropical Agriculture (IITA), the 2016). The first pathway, evidence that counts, International Crops Research Institute for the posits that ILRI evidence, published in peer- Semi-Arid Tropics (ICRISAT), IFPRI and ILRI, reviewed journals and actively communicated to and a portfolio of research on animal-source users in ways that are clear, compelling and foods led by ILRI (A4NH, 2011). After two suc- actionable, will lead to better decisions and these cessful external evaluations (Sridharan et al., will lead to positive impacts. The second path- 2015; Compton et al., 2015), in which the way, impact that scales, is based on ILRI dis- research was deemed to be highly relevant, to covering, developing or contributing to novel have generated important evidence and to technologies or institutions that improve food have generally met expectations, it was de- safety for millions of people. Our big-idea impact cided to make food safety a new, stand-alone pathway is the triple path to improving food flagship in the second phase of A4NH, starting safety in mass domestic markets by working in 2017. with informal traders through a combination of 344 D. Grace-Randolph et al. increasing their capacity to sell safe food and practice is desirable, and that research can through training and technologies, providing also tackle the process problem of insufficient motivation for behaviour change (e.g. by im- reliance on evidence in decision making. As a proving business and marketing skills) and pro- result, important research-for-development don- viding a more enabling operating environment, ors rely increasingly on evidence. The food so that authorities, instead of ignoring or pun- safety work at ILRI, which strongly drew on epi- ishing informal traders, encourage them to pro- demiology, was and is well placed to meet this fessionalize their work. demand. CGIAR is an important generator of agri- cultural research evidence in developing coun- tries. Surveys have found that CGIAR science Empirical evidence outputs compare well with advanced research institutes in production of evidence (Elsevier, The first pathway – evidence that counts – is 2014). However, there is less information on how well within the traditional research sphere. Our this evidence is used or linked to development theory of change is that, for ‘evidence to count’, impact. In general, the implementation of research the right information must be conveyed to the evidence is not straightforward. A review of the relevant people through appropriate channels. use of public health evidence in developed coun- Research efforts can also build capacity of the tries found that there was no reliable evidence relevant people so that they can make good use on the extent of its use and that its impact was of the evidence generated and better align their often indirect, competing with other influences incentives with action to improve food safety. (Orton et al., 2011). The same review suggested Recent decades have seen an increasingly that barriers to the use of research evidence in- systematic and systemic approach to using evi- cluded: decision makers’ perceptions of research dence across a broad range of fields; ILRI seeks evidence, the gulf between researchers and deci- to apply this to the issue of food safety in infor- sion makers, the culture of decision making, mal markets. Much of the interest can be traced competing influences on decision making and back to the evidence-based medicine movement, practical constraints. which started in the 1990s in Canada. Evidence- Food safety research is more likely to have based medicine was defined as ‘a systemic ap- an impact if the following are true: proach to analyse published research as the basis of clinical decision making’ (Claridge and • The research is of objectively high quality. Fabian, 2005). The approach quickly spread to Our food safety research seeks to drive up allied health fields, such as dentistry, and then to quality by publication in high-impact- areas such as education and housing. factor journals, shifting from less to more Evidence-based approaches explicitly weight rigorous protocols and following best prac- different types of evidence. In the evidence hier- tice guidelines for conducting and report- archy, scientific evidence trumps anecdote or opin- ing studies. ion, and scientific evidence itself is considered • Stakeholders are involved. For example, they weaker or stronger depending on defined char- may take part in the design of the research, acteristics. For example, evidence from a multi- serve as advisory members or visit research centre randomized controlled trial is stronger than sites. ILRI’s food safety research has often evidence from a cohort study, which in turn is involved national ‘champions’ who were stronger than evidence from a cross-sectional identified as key promoters and dissemin- study. While the best research evidence is in- ators of the research findings. tended to be the major factor in medical deci- • The research is produced by scientists in whom sions, it is acknowledged that research evidence decisions makers have confidence. For example, is only one factor, often a minor one, in develop- Kenyan policy makers want to see studies on ment decision making. However, there is a aflatoxins in feed from Kenya, even if studies consensus in the literature that, especially in from Tanzania are likely to be almost as rele- developing countries, a more evidence-based, or vant. ILRI’s food safety research has taken at least evidence-informed, approach to policy place in 27 countries as of 2020. Food Safety and Nutrition 345 • The research is important but non-obvious. For countries. Applying this food safety approach example, our finding in Vietnam that pork in was an important innovation of the programme supermarkets was less safe than pork sold (Grace et al., 2008, 2010, 2011, 2012a,b; Grace in wet markets contradicted policy makers’ and Randolph, 2009). The approach was subse- preconceptions. They initially resisted the quently used in Tanzania, Uganda, Vietnam and information, but when they saw the reasons elsewhere, and its strengths and weaknesses, as for this finding, it made more of an impres- well as the recommendations generated, were sion on them than research findings that captured in peer-reviewed publications (Häsler matched their preconceptions. et al., 2018; Nguyen-Viet et al., 2019; Roesel • The evidence is timely, coming when decision et al., 2019). makers need to do something. For example, Within this risk analysis framework, other research on training dairy traders in methods and innovations were developed, in- north-east India provided a solution for cluding a global mapping of zoonotic diseases decision makers dealing with public concern and poverty. This involved an updating of the over milk safety. global maps of poor livestock keepers, a system- We have found that food safety evidence atic prioritization of zoonotic diseases likely to leading to impacts generally occurs as one of be relevant to the poor, a systematic literature three kinds: (i) developing the methods and tools review of the prevalence of these zoonoses in needed to generate evidence of food safety in in- people, livestock and food products, and combin- formal markets; (ii) developing and testing in- ing these in global maps (Grace et al., 2012c). novations with potential for widespread use; and This was subsequently used to inform a major (iii) influencing policy. call for research on zoonotic diseases funded by the UK Department for International Development (DFID) and British research councils, which sub- sequently generated important research find- Developing research ings across a range of projects. methods and tools Economic assessment is another key tool to improving food safety. Collaborative research Faced with the challenge of informal food haz- by ILRI over a number of years on the demand ards but little understanding of their risks to for livestock products in Ethiopia, Kenya and human health, ILRI identified the need for new Tunisia in Africa, in Bangladesh and India in tools and methods for conducting food safety South Asia, and in Cambodia and Vietnam in research in a development context. The over- South-east Asia provided strong empirical evi- arching framework for food safety work was an dence on food safety (Jabbar et al., 2010). The approach that ILRI called ‘Participatory Risk study identified ‘wet markets’ as the typical point Analysis’. Over the past several decades, risk of purchase of animal products. The quality and analysis has been accepted as the ‘gold standard’ safety of livestock food products were mostly for assuring food safety. It has been adopted by defined according to how these attributes were the international community and underpins perceived by consumers: by their taste, colour, trade in foods and livestock. However, risk ana- flavour and smell. Developing-country consumers lysis has not had much success in the informal also judge quality and safety by what they per- markets of developing countries, where most of ceive to be the nutritional attributes of the foods, the poor buy and sell their food. Conventional such as freshness, absence of adulteration, fat risk analysis is often expensive and time con- content (milk) and fat cover (meat), and various suming, requires considerable amounts of data aspects of appearance, packaging, geographic and quantitative analysis, and is typically led by origins, indicators of expired shelf life, a govern- technocrats. By taking the core concepts of risk ment inspection stamp and the cleanliness of analysis and combining them with proven devel- the premises selling the products. The same opment analytic methods such as participatory consumers are aware of microbial, chemical rural appraisal and gender analysis, an approach and physical hazards in animal-source foods. emerged that could be applied successfully In  general, quality and safety issues were not to the food safety challenges in developing always clearly demarcated: consumers tended to 346 D. Grace-Randolph et al. associate some attributes with both while in and applied to the countries belonging to the other cases the differences were clearer. Organization for Economic Cooperation and De- One ILRI innovation was an adaptation of velopment (OECD). Currently, ILRI is providing system dynamics – a model that maps resource technical support to develop the world’s first flows and management processes within a ‘Food Safety Index’, which the African Union complex system – to informal food systems (see intends to include in the Malabo Declaration Chapter 6, this volume, for an adaptation to East process. This means that all African Union coun- Coast fever). This was used to investigate inter- tries will have an obligation to report on food ventions in the pork chain in Vietnam. Desk safety and be mutually accountable, driving up studies have combined information on the health food safety in Africa. ILRI is also a partner in the burden of FBD, the foods responsible and macro- international Global Burden of Animal Diseases economic models to predict future trends in FBD initiative. in terms of health burden and economic cost (Kristkova et al., 2017). In India, the number of FBD cases is expected to rise from 100 million to Developing and testing innovations 150–177 million in 2030 compared with 2011, for application and an economy-wide model predicted that this would incur costs equivalent to 0.5% of the GDP. Another suite of ILRI research focuses on gener- CGIAR identified gender as a cross-cutting ating outputs or products intended for use by issue that should be mainstreamed in research. value chain agents and implementers, including However, most food safety research does not technologies, approaches and surveillance. have a gender perspective. We adapted and ap- plied gender analysis tools to understanding • Technologies. Food safety technologies are food safety and documented this in several pa- technical approaches to improving food pers (Kimani et al., 2012; Grace et al., 2015d; safety. Nearly all of the technologies re- Kiama et al., 2016). searched by ILRI food safety scientists are Similarly, although food safety and nutri- adaptations of products developed by tion are biologically coupled, they are not often others. For example, we adapted the insecti- well integrated in agricultural development. cide-treated bed nets widely used in the This can be problematic because interventions control of malaria to reduce flies in infor- intended to improve food safety can work against mal markets. In other cases, ILRI had no nutrition and vice versa. We developed a frame- role in the development of the technology work for a rapid assessment of food safety and but tested it in order to assess its suitability nutrition and applied it to several of the livestock and/or to suggest improvements to make it value chains where the CRP on Livestock and more useful (e.g. use of ozone in disinfec- Fish was working (Eltholth et al., 2014; Hoa et tion). None of the technologies developed, al., 2014; Häsler et al., 2019) and, along with tested or adapted is being delivered at scale the lead UK think tank at Chatham House, devel- but several are considered to have potential oped a widely disseminated evidence synthesis for widespread use. on animal-s ource foods in the first 1000 days of • Approaches. These comprise processes or dif- life, covering nutrition and food safety (Grace ferent ways of doing things. Many are oriented et al., 2018a). around capacity building in new practices What cannot be measured cannot be man- or providing information. We can consider aged. When ILRI started work on food safety, that one approach is having impact at scale: there was little understanding of suitable this is the triple-path approach to informal m etrics and indicators for food safety in low- and traders comprising capacity building, enab- middle-income countries. ILRI led a working ling environment and motivation. group with broad expert inclusion to develop the • Surveillance. The third category of innov- first synthesis and analysis of food safety metrics ations is concerned with disease detection, for these countries (Grace et al., 2018b). It also reporting and response, such as the use of developed a tool to measure ‘food safety system information technology for reporting from performance’, inspired by a similar tool developed slaughterhouses. Food Safety and Nutrition 347 Table 9.1. Food safety product lines. Green shading indicates that the development is on track, yellow indicates that there are some delays or problems, orange indicates that there are significant delays or problems, and red indicates that it has been cancelled. (Constructed by authors). Stages of development Proof of Evaluate to Category Product Research concept development Delivery Technologies Aflatoxin binders for animal feed Aflatoxin probiotics for food and feed Insecticide-treated netting for food safety in wet markets Ozone disinfection for wet markets Chlorine disinfection for wet markets Filter paper tests for bacterial load in food Mazzican for less mastitis and safer milk Boiling milk to improve safety Fermenting milk to improve safety Off-ground slaughter using a metal grid Approaches Training slaughterhouse workers Training and certification of traders GAP light – simplified Good Agricultural Practice SMS messages for changing farmer behaviour Biosecurity Surveillance Bidirectional e-surveillance system for slaughterhouses Participatory epidemiology for outbreak investigation e-Surveillance for disease reporting Participatory disease surveillance for managing large-scale outbreaks The most important ILRI food safety prod- suggests that these projects are indeed produ- ucts in these categories are summarized in Table cing outputs that go beyond research papers and 9.1. This summary of product lines gives an that plausibly will help to ensure that ‘evidence overview of ILRI evidence generation. More in- counts’. sight into potential impact can be gained by Outcome and impact assessments carried looking at specific research projects and topics. out by specific projects can also illustrate the po- To give a concrete example, we analysed re- tential use and benefits of evidence generated on search outputs on aflatoxins posted on the CG- food safety projects (Box 9.2). Space document repository. Over a period of 6  years and with the input of one or two full-time-equivalent ILRI scientists per year Capacity building activities by ILRI working with students and partners, we pro- duced 29 journal articles accompanied by 50 Capacity building was integral to all of ILRI’s science outreach items (conference presenta- food safety research and is an important tions, reports), 14 policy outreach items (briefs, dimension of CGIAR research. Between 2012 technical packages) and 50 public outreach and 2015, the following training activities items (videos, infographics, press conference, were carried out by ILRI’s food safety research blog articles). In addition, we communicated the programme (Table 9.2): (i) training of value research results to all the farmers and value chain agents for the purpose of developing and chain agents participating in this research. This testing models and approaches; (ii) training 348 D. Grace-Randolph et al. Box 9.2. Evaluation of a multi-country food safety project. Safe Food, Fair Food was the first major ILRI research project to focus on food safety. It conducted a peer-to-peer project assessment whereby teams from participating countries visited another country to conduct a structured evaluation. This was generally positive. For example, the project had five major components, the first being a situational analysis of food safety. The main impacts of this situational analysis work were: (i) raised awareness on food safety in informal markets among food safety stake- holders; and (ii) a coming together of different sectors (especially medical and veterinary) to discuss the common issue of food safety. A semi-quantitative analysis of the situational analysis identified four success criteria, and a peer evaluation was conducted when each of the seven country teams evaluated another country against these criteria (maximum score of 5). The average across seven countries was 16.4 out of 20, equivalent to 82 out of 100, suggesting good overall impact. Assessment points (maximum score of 5 for each category) Average score 1. Did the participants of situational analysis represent 4.7 stakeholders of food safety in the country well? 2. What information does the situational analysis provide? 4.1 3. Was there any delay in conducting situational analysis? 3.6 4. Were there measurable impacts from the situational analysis? 4.0 Subtotal 16.4 Percentage 82.1 Table 9.2. ILRI capacity development in food safety, 2012–2015. (Compiled by authors from ILRI archives). (unpublished data, ILRI). Value chain Officials and Researchers Graduate Year agents policy makers and students fellows 2012 70 110 52 26 2013 524 77 42 69 2014 304 146 161 101 2015 1460 37 192 42 government officials and policy makers to in- Influence on international, regional crease their capacity to understand and make and national policies good decisions, creating an enabling policy en- vironment; and (iii) training researchers to build International and regional agriculture and health their capacity but also to influence future imple- organizations are considered crucial to develop- menters and decision makers. Although we do ment and this implies that ILRI engagement with not have denominator data, we believe this rep- them can have far-reaching impacts. Some ILRI resents a majority of the food safety graduate fel- inputs were specific to food safety high- level pro- lows and researchers in the countries in which cesses (e.g. its participation in WHO, 2013), ILRI worked, a substantial proportion (around while others incorporated food safety dimensions half) of relevant government officials and policy into broader livestock or development initiatives makers, and a much smaller proportion (much (e.g. food safety as an aspect of sustainable live- less than 1%) of value chain agents. stock development). Another distinction is be- The benefits of training have been documented tween initiatives led by ILRI and initiatives where to some extent by projects that conducted out- ILRI scientists were part of a broad range of sci- come studies, including Safe Food, Fair Food entists. Some of the most notable contributions (Box 9.2), which worked in multiple countries in are shown in Tables 9.3 and 9.4. sub-Saharan Africa, and PigRisk, a project work- The following summary gives examples of ing in Vietnam. where ILRI food safety research has contributed Food Safety and Nutrition 349 Table 9.3. Food safety research led by ILRI. Commissioner Food safety aspect Outcomes from the report Source DFID FBD in low- and middle-income Contributed to a funding Grace (2015b) countries call on food safety OIE FBD as neglected livestock OIE communiqué issued Grace et al. (2015a) diseases UNEP Aflatoxins Featured in UNEP annual Harvey et al. (2016) report EAC Aflatoxins in feed and livestock Developed technical briefs Grace et al. (2015b,c) products used for setting policy at East African Community level World Bank Food safety in Vietnam Contributed to a major funding World Bank (2017) initiative and to national policy USAID Food safety in developing Contributed to initiation of first Grace (2017) countries food safety Innovation Laboratory LCIRAH Food safety metrics Contributed to food safety Grace et al. (2018a) tracking by African Union DFID/BMGF Food safety Investment report influenced Grace et al. (2018b) major funding call OIE, Office International des Epizooties (World Organisation for Animal Health); UNEP, United Nations Environment Programme; EAC, East African Community; USAID, US Agency for International Development; LCIRAH, Leverhulme Centre for Integrative Research on Agriculture and Health; BMGF, Bill & Melinda Gates Foundation. Table 9.4. Food safety initiatives to which ILRI contributed. Commissioner Food safety aspect Outcomes from the report Source IFPRI Aflatoxins 2020 briefs – an influential Unnevehr and Grace (2013) series of communications WHO Food safety burden Co-author on the FERG Havelaar et al. (2015) report WHO Trade and human Chapter in WHO book Hawkes et al. (2015) health IFPRI Emerging economies Paper in Global Food Policy Grace and McDermott (2015) Report HLPE Food safety as an Co-author in ‘Sustainable HLPE (2016) element of Livestock’ report sustainable livestock systems FAO Food safety as an Included in International Grace (2017) element of a healthy Conference of Nutrition food environment agenda World Bank Food safety Contributions to two major Jaffee et al. (2019); GFSP reports (2019) FAO/WHO Food safety First FAO/WHO/AU conference http://www.fao.org/3/ on food safety issued a CA3225EN/ca3225en.pdf communiqué WTO Food safety Speaker at pre-panel event https://www.wto.org/english/ economics tratop_e/sps_e/ faowhowtoapril19prog_e.htm HLPE, High-Level Panel of Experts; FAO, Food and Agriculture Organization of the United Nations; FERG, Foodborne Disease Burden Epidemiology Reference Group. 350 D. Grace-Randolph et al. to policy. A more detailed explication can help il- and fish grown/caught in wastewater; (ii) assess lustrate the specific contributions. the health risks related to antibiotic residues in WHO undertook the first global assess- pork; and (iii) disseminate research results and ment of FBDs through its Foodborne Disease advocate for risk assessment as a tool for food Burden Epidemiology Reference Group (FERG). safety management. The health risks from these This showed the high burden of FBD and is case studies were assessed quantitatively, and likely to lead to increased funding in this neg- risk communication and management strategies lected area. The WHO’s burden of disease stud- were developed. Achievements of the task force ies were highly influential in determining the included the training of policy makers, managers global health agenda and especially in directing and researchers; the publication of case studies billions of dollars in funding to the ‘big three’ of risk assessment in a special edition of a Viet- diseases (Maudlin et al., 2009). It is therefore namese journal; and the publication of policy plausible that the FERG study will also have briefs. The task force was also requested to run widespread impacts. training courses for veterinary professionals of A High-Level Panel of Experts (HLPE) is the ministries. The process, outcomes, challenges and science–policy interface of the Committee on potential impacts of the task force have been World Food Security (CFS), the foremost inter- documented by Nguyen-Viet et al. (2018). national platform for food security. In October IITA coordinated the development of tech- 2014, the CFS requested the HLPE to prepare a nical packages for the East African Community report on sustainable agricultural development comprising technical papers on aflatoxin situ- for food security and nutrition, including the role ational analysis, the scientific basis for aflatoxin of livestock (HLPE, 2016). An important plan- control and policy recommendations for afla- ning meeting was held at ILRI, where ILRI’s toxin control. These technical packages aimed to Delia Grace served as one of ten members of the assemble the best scientific thinking on the topic HLPE livestock project team. HLPE reports are as the basis for policy recommendations. Through widely used as reference documents within and A4NH, ILRI scientists drafted two of these pack- beyond CFS and the United Nations system, by ages, which were submitted to the East African the scientific community as well as by political Community (Grace et al., 2015b,c) and officially decision makers and stakeholders, and at inter- launched in 2018. national, regional and national levels. ILRI was commissioned by the US Agency A World Bank-supported task force on risk for International Development (USAID) to develop assessment for food safety comprising researchers a white paper on the potential need and role of a and policy makers was formed in 2013 to build new Feed the Future Innovation Lab on Food capacity for food safety management in Vietnam. Safety (Grace, 2017). The report recommended ILRI scientists were involved in the task force and this, which contributed to the initiation of the A4NH provided funding (Nguyen-Viet, 2012). laboratory in 2019. The task force consisted of researchers in Viet- ILRI was asked by the Global Food Safety nam working on risk assessment and food safety Partnership (GFSP; a World Bank hosted public– with representatives of the Vietnamese Ministry private initiative for supporting food safety cap- of Health and Ministry of Agriculture and Rural acity building) to participate in a study on previous Development. The task force first analysed the food safety investment in Africa and to make re- situation of food safety policy in Vietnam. Key commendations for future directions (GFSP, 2019). constraints and areas where research and devel- This led to engagement with the East African opment interventions could assist policy were Community (EAC) and three-way collaboration identified. Stakeholder workshops were con- between the EAC, GFSP and ILRI to support EAC ducted to determine the scope of activities and to in developing food safety strategy. prioritize food safety issues. Training sessions ILRI was asked by the World Bank to be a with a focus on case studies of risk assessment partner and co-author of the Eat Safe Initiative, for food safety were organized to strengthen the which sets out global strategy for improving food risk assessment capacity of task force members safety and developed the first estimate of the cost and of policy makers. Case studies were con- of foodborne disease in low- and middle-income ducted to: (i) assess the health risks of vegetables countries (Jaffee et al., 2019). Food Safety and Nutrition 351 In 2015, the African Union (AU) launched Specifically, ILRI’s food safety research part- the Comprehensive Africa Agriculture Devel- ners include the following: opment Programme (CAADP) Biannual Review (BR) to monitor progress on agricultural devel- • Researchers. Important research partners in opment in the continent. The CAADP BR encom- ILRI food safety are the veterinary, agricul- passed 43 indicators, seven of which tracked ture and, to a lesser extent, medical univer- nutrition, but none captured food safety. In dis- sities, national agriculture and medical cussion with the AU, ILRI partnered to help de- research systems and centres of excellence velop the first African Food Safety Index (AFSI). in the countries in which we work. Advanced The AFSI was launched as part of the 2019 research institutes are important partners, CAADP BR, and 50 out of 55 AU Member States especially Free University Berlin, Liverpool reported in at least one of its three elements. University, Uppsala Agricultural University, the University of Florida and the University of Sydney. The CGIAR centres IFPRI, IITA and World Fish have been major partners. Impact that scales • Value chain agents. Most of ILRI’s food safety research engagements have been with small- International agricultural research has always scale value chain agents, often via intermedi- aimed for widespread impact, first by improving aries such as trader associations, but there food production in developing countries and has been increased interest in medium-sized later by widening its focus on livelihoods and on formal businesses. We have also worked the health and environmental externalities of with public–private partnerships such as agriculture. Impact assessments show large and the Global Alliance for Livestock Veterinary well-documented benefits to CGIAR research on Medicines (GALVmed). crop genetic improvement, most notably rice, • Development programme implementers. maize and wheat, and especially in Asia. There is Development-implementing partners of ILRI much less evidence, however, for large-scale bene- include non-governmental organizations fits from global agricultural research in the fields such as Veterinarians without Borders and of policy, natural resource management and large-scale development projects funded by livestock (Renkow and Byerlee, 2010; Jutzi and the World Bank, USAID and others. Rich, 2016). • Enablers. The international and regional en- There are different models for under- ablers include: the Africa Union–Interafrican standing how innovations in agri-food systems, Bureau for Animal Resources(AU-IBAR), whether technologies or institutions, could Association of Southeast Asian Nations have widespread, sustained impact. In developing (ASEAN), EAC, Economic Community of countries, agricultural extension services and West African States, Food and Agriculture development initiatives are important but often Organization of the United Nations (FAO), have limited reach. In recent years, interest has Intergovernmental Authority on Develop- grown in other dissemination actors, especially ment, United Nations Environment Pro- the private sector and collective action and in gramme (UNEP), World Bank, WHO and novel dissemination pathways such as social the Office International des Epizooties (OIE, media. The food safety research agenda explores World Organisation for Animal Health). the potential of different partnerships to achieve We also work with policy makers and im- impact at scale. plementers at the country level, including ILRI food safety research partnered with national ministries, state veterinary services four broad categories of individuals or organ- and municipal authorities. izations: researchers, agents in value chains, development programme implementers and en- Training and enabling informal sector ablers. The relative level of involvement of these agents groups varies – it will grow, reduce or stay the same – based on the particular stage of given Demand for fresh foods is growing rapidly in research. developing countries and most of this demand 352 D. Grace-Randolph et al. must be met by markets. A study in southern The traders were trained in hygienic milk handling and East Africa found that most food is already and business practices and at the end of their obtained from markets (54% in 2010, predicted training could apply for a certificate from the to reach 70% in 2040) and that the informal Kenya Dairy Board that entitled them to legally sector currently supplies 85–95% of market de- sell milk (Box 9.3). mand and 51–57% of total demand (Tschirley Participant tests before and after the train- et al., 2015). ing showed that trader knowledge and practices ILRI pioneered a ‘triple-pathway’ approach improved, and microbiological tests showed that to improving food safety in informal markets by there was a substantial and significant decrease professionalizing rather than penalizing the in unsafe milk. A later economic evaluation found informal sector, with the aims of supporting an important reduction in transaction costs at- smallholder market access, safeguarding the tributable to less harassment by authorities, less supply of cheap nutritious food to the poor and confiscated equipment and fewer bribes paid but reducing the burden of FBD. In the early 2000s, also fewer losses of milk to spoilage. There was a training and certification scheme was designed anecdotal evidence of improved business per- and launched in Kenya to improve the quality and formance. A more recent evaluation found that, safety of informal dairy markets by improving although the scheme had encountered some the practices of traders, while also supporting challenges, it was still operational. Eight years the livelihoods of the dairy value chain agents. after the project officially ended, many traders The scheme was taken up by a large propor- have continued in the scheme; we estimated that tion of eligible traders (with project support). up to 5 million consumers are benefiting from Box 9.3. Smallholder dairy training and certification initiative. In Kenya, dairy products are a significant expenditure in poor households. The informal, small-scale milk sector dominates the milk marketing chain, with some 60 - 70 % of the raw milk market. Milk sold informally from door to door or in milk bars reaches poor consumers who pay a lower price for it than for factory-packaged milk; it also generally provides farmers with higher prices than they can get in the formal sector. However, prior to policy change in 2004, informal vendors, including mobile milk traders and bar vendors, milk transporters and small-scale milk producers (many of them women), were not officially recognized. They were unable to obtain a licence and were frequently harassed by powerful dairy mar- ket players, who sought to protect their own interests while professing concern over the safety and quality of milk sold in the informal sector. Efforts to revise the dairy policy were spearheaded by ILRI’s Smallholder Dairy Project. Imple- mented along with the Kenya Agricultural Research Institute (KARI) and the Kenya Ministry of Livestock and Fisheries Development, the project generated research-based evidence to reveal the economic significance of the informal milk sector and highlight the potential for improved handling and hygiene practices to ensure milk quality. As part of the ongoing development of pro-poor strategies for small-scale milk market develop- ment, the Dairy Traders Association of Kenya was officially launched in September 2009. Its aims and activities include self-regulation based on the training and certification concept originally developed by the Smallholder Dairy Project and further scaled up by other projects. Around 4000 milk traders, offering employment to over 10,000 people, have been trained and certified by the Kenya Dairy Board through the association. Field regulators also ensure that licensed outlets and premises operated by milk traders meet conditions for milk hygiene, testing requirements and sanitation, and that operators know how to comply with these conditions. A key supporting aspect of the Smallholder Dairy Project was the development of modules for training (milk handling, processing and marketing) and certification of vendors to improve milk quality. This training, along with simple innovations such as wide-necked milk cans, were shown to improve the safety of milk significantly. The proportion of milk with high levels of contamination fell from 71% to 55% among traders using plastic containers and from 48% to 42% among those using metal containers. Without the intervention, policy change would have been unlikely (WRENmedia, 2010). Food Safety and Nutrition 353 milk provided by trained traders and tens of There was, however, a lack of systematic sup- thousands of dairy farmers from market access port to this initially successful project. The ori- through trained traders. ginal assumption that vendors would pay An evaluation of the Kenya-ILRI collabora- private business development services to provide tive Smallholder Dairy Project was conducted in training was not valid. However, other develop- 2008 (Kaitibie et al., 2010a,b; see Chapter 17, ment actors did use the modules to provide this volume). This showed significant economic one-off trainings. More critically, changes in the benefits derived from changes in dairy policy institutional and political context were not fa- resulting in lower transaction costs. Some 73% vourable to the informal sector and a subsequent of national benefits accrued to producers and follow up found that, while traders expressed a consumers with the balance going to traders very favourable opinion to training, there was and input suppliers. Related evidence showed no systematic training programme in place and improvements in milk quality (Omore and Baker, moreover milk sold by trained traders was no 2011), although it was not possible to link such safer (Alonso et al., 2018). improvements to changes in market prices. Moreover, the approach used in the Small- Key lessons from the Smallholder Dairy holder Dairy Project was never evaluated to see Project were as follows: whether health benefits were obtained from safer milk. Although marketing skills were taught, • The scheme was successful in improving the there was no emphasis on teaching vendors how quality and safety of milk, at least in the short to promote the nutritional benefits of milk. The term, and the focus on quality seems to have capacity-building initiative did not benefit from improved business performance. a gender perspective in design or implementa- • The scheme reduced milk marketing costs tion, notwithstanding the importance of women and was appreciated by both traders and as milk producers, traders and consumers. consumers. Sustainability and scalability challenges had not • The traders provided information to con- been fully overcome. These deficits are being ad- sumers and can be a practical node for dissem- dressed in a project under way in 2020 (www. ination of nutritional change and promotion ilri.org/research/projects/moremilk-making- of milk consumption to consumers as part of most-milk); accessed 1 August 2020. a marketing intervention. The trader intervention is a model for im- • Training in business skills, including a proving food safety when approaches based on greater consumer orientation, can improve regulation do not work (Johnson et al., 2015). business performance. The model has been adapted and tested in other Key policy lessons were the following: contexts, including dairy (India and Tanzania) and meat (Ethiopia, Nigeria and Senegal). In two • Policies seeking to exclude the informal of the three cases, evaluations documented that sector are unlikely to improve food safety or participating value chain agents increased their nutritional quality and may paradoxically knowledge and skills and improved their food- decrease food safety and reduce the accessi- handling practices. In some cases, better milk bility of food. quality and higher incomes were found (Lapar • Food safety and nutrition programmes et  al., 2014) and significant economic benefits should also help to reform anti-informal were generated (Kaitibie et al., 2010a,b). In the sector policies. Merely reducing inappropri- case of Nigeria, the intervention could plausibly ate regulatory pressure on small businesses be linked with a reduction in diarrhoea and sav- has the potential to increase small business ings in reduced healthcare expenditure worth capacities and to create incentives for them many times the cost of training butchers (Grace to improve the quality of their product. et al., 2012a). However, follow-up research • ‘Light-touch’ interventions centred around 9 years later revealed a marked deterioration in training can deliver substantial improve- meat quality as the result of lack of follow-on ments in product quality, even in the absence training and, more importantly, a shift from of major technological or infrastructure enabling to disabling environment (Grace et al., upgrades. 2019). 354 D. Grace-Randolph et al. Based on results from early studies, a formal around 2 years of age, are considered an espe- theory of change was developed by Johnson et al. cially crucial nutritional period: setbacks during (2015). This identified three components that this period are hard to recover from by later they considered essential for success. The so-called attempts to ‘catch up’. Undernutrition, while ‘triple-path’ model included the following: declining, remains at high levels in vulnerable • communities, while diseases associated with too Training and technologies. Informal sector much food consumption trend upwards. agents needed tools to deliver safe food. This An initiative in 1984 brought together 12 usually meant training, awareness raising CGIAR centres at ILCA, in Ethiopia, to discuss and simple technologies such as disinfect- how the centres were addressing human nutri- ants. Training in business skills was often tion. At that time, ILCA was including nutritional included. • status in its field research, while ILRAD viewed Enabling environment. Regulatory authorities its contribution to better nutrition as an indirect had to be on board with the intervention one made by tackling serious livestock diseases and there had to be some mechanism for (Doyle, 1984). institutionalization (e.g. a locally or nation- During the 1960s and 1970s, insufficient ally recognized certificate) and a means of energy was thought to be the most serious diet- quality assurance. • ary constraint to improved human nutrition. As Motivation and incentives. Incentives were a result of research during the 1980s and 1990s essential for behaviour change but were very and improving levels of energy consumption, context specific. In one case, certificates pro- attention shifted to micronutrient deficiencies in tected traders against harassment from the diets of the poor. Because milk, eggs and meat authorities; in another, the training enabled are among the richest dietary sources of vita- traders to improve their bargaining power mins and minerals, in addition to protein, this with the public sector. It was originally hy- created a new appreciation for the contribution pothesized that trained traders would be that livestock products can make to ensuring able to charge a premium for safer food, but nutritious and diverse diets. in no project were they able to charge more In the late 1990s, ILRI conducted its first for food, although some may have increased empirical studies investigating links between their market share. livestock keeping and human nutrition. A study This triple-path approach is sometimes called from Ethiopia (using data from 1989 to 1998) ‘Training, Certification and Marketing’, or TCM, found that introducing cross-bred cows could where ‘training’ refers to the capacity building as- improve human health and nutritional status pect, ‘certification’ to the enabling environment (Thornton and Odero, 1998); similar findings and ‘marketing’ to the provision of incentives for were reported from coastal Kenya (Nicholson behaviour change. et al., 1999). Another Ethiopian study, in 1997 and Table 9.5 presents evidence for the outcomes 1999, indicated that market-oriented livestock and impacts of food safety interventions, based activities moderately reduced poverty and improved on five relatively well-evaluated projects. food security and nutrition of smallholder house- holds (Ahmed et al., 2003). Econometric models applied to data from coastal and highland Kenya Human Nutrition Research at ILRI in the late 1990s found positive impacts of dairy cattle ownership on chronic malnutrition in Many rural poor people worldwide subsist on coastal Kenya (Nicholson et al., 2003). substandard diets consisting largely of the same A major event to bring together nutrition cheap cereal and tuber staples day in and day researchers and stimulate nutrition research out. When they move to cities, their intake of in CGIAR was held in 2000 in the Philippines cheap, highly processed foods high in sugar, salt (Pinstrup-Andersen, 2000). Discussions at this and fats increases. Nutritional deficiencies in meeting explicitly addressed the role of highly such diets are common and are associated with nutritious foods, including livestock products. a range of poor health and development outcomes. The meeting concluded that ILRI efforts to increase The first 1000  days of life, from conception to the supply of livestock products to the poor could Food Safety and Nutrition 355 Table 9.5. ILRI food safety interventions in informal markets. Particulars Kenya Senegal Ibadan, Nigeria Assam state, India Kampala, Uganda Value chain Informal milk sector Goat restaurants Butchers Informal milk sector Butchers Year range 1997–2006 2010–2011 2009–2011 2009–2013 Number of 25,000–30,000 Several hundred in three Around 900 in the market Around 300 traders and 600 traders slaughterhouses producers in the main milkshed Number of In 2010, 4200 traders Around 100 trained 80 directly by the project and 265 traders and 480 producers 50% of butchers market registered nationally; in around 420 by peer-to-peer have been trained agents pilot areas, 85% of traders training trained had been trained Consumers Around 0.5–5 million Nearly 1 million Around 360,000 Around 1.5 million Around 0.5 million reached Gender Not explicit; women made Not included: all workers Targets for women participation Not explicit; nearly all traders aspects up about one-third of the were men and gender dimensions and farmers were men traders researched Intervention Training in hygiene and Training in hygiene, raising Peer-to-peer training on basic In-depth training needs Training in business practices, awareness on food hygiene; provision of analysis; training of trainers; hygiene, provision of hygienic dairy safety equipment, banners and training covering hygiene and equipment, cans, with a certificate promotional material; use of business skills; traders posters, given to successful butchers’ associations to motivated by better relations certificates trainees, reducing their monitor performance and with officials and positive harassment by officials ensure compliance publicity and farmers by visible reduction in mastitis Documented Improved KAP after training; No change in KAP after Reduction of unacceptable meat Improved KAP after training; Improved KAP impact improved milk safety after training; management from 97.5% to 78.5% (p<0.001); significantly higher milk after training; training with reduction in provided no soap or significant improvements in KAP production after training and satisfaction with unacceptable coliforms other necessities and after training; cost of training was tendency for reduced mastitis; training from 71% to 42%; This were rather indifferent to US$9 per butcher and estimated significant sector benefits in project gave training and practices, and there was gains through diarrhoea averted several sites certification programs for no obvious incentive for was US$780 per butcher informal milk traders, behaviour change enabling thousands to be licensed and resulting in national economic benefits having a net present value of US$230 million. Continued 356 D. Grace-Randolph et al. Table 9.5. Continued. Particulars Kenya Senegal Ibadan, Nigeria Assam state, India Kampala, Uganda Policy High: legislation changed None Low: only engagement with High: new institutions but no Some: linked to influence and new institutions market authorities change to legislation broader ILRI policy processes Current Training and certification are None: one-off training The pilot was intended to Training and monitoring are Training is being status episodic and project-led, investigate efficacy and ongoing and supported by supported by of the but trained vendors have an acceptability and did not have the government donors initiative important share of the a strategy for sustainability market Reference(s) Kaitibie et al. (2010a,b); Submitted Grace et al. (2012a) Lapar et al. (2014); Lindahl Ongoing Omore and Baker (2011); et al. (2018) Alonso et al. (2018) KAP, knowledge, attitudes and practices. Food Safety and Nutrition 357 be presumed to have nutritional benefits while ILRI leveraged external expertise to re- acknowledging that there had been insignificant establish nutrition work. This included collab- efforts to measure these benefits (Bouis, 2000). orations with senior nutritionists at Emory Delgado  et al.  (2001) examined the effects University in Georgia, IFPRI, the London School of income growth on diets using Chinese panel of Hygiene and Tropical Medicine, UK, and data. As incomes improved, Chinese consumers Washington State University. Exploratory work shifted from high-carbohydrate foods towards and pilots were conducted in several field sites. high-fat, energy-dense foods, with these changes Highlights include the following: varying by income levels. These income effects suggested that increased incomes could affect • An ILRI study conducted with households diet and body composition in ways detrimental representing low, medium and high levels to health; moreover, the biggest harm would fall of dairy intensification in rural Kenya indi- on low-income groups due to their increasing cated that women’s increased labour de- incomes. The study argued that higher incomes mands as households intensified their dairy might reverse health gains achieved in the preceding production were associated with poorer nu- two decades if diet-related non-communicable tritional outcomes for their young children; diseases could not be controlled (Delgado et al., in contrast, children in households of high 2001). dairy intensity received more milk than chil- In 2003, for the first time, an ILRI programme dren in lower-intensity households (Njuki was initiated with an explicit focus on improving et al., 2015). human health through livestock by considering • ILRI produced the first reported study show- both the associated benefits and risks of livestock ing a link between aflatoxin in milk and to people’s health. The new ILRI Livestock Keep- child stunting (children who are too small ing and Human Health Impacts programme for their age) in two low-income areas in focused on nutrition, zoonoses and food safety. Nairobi (Kiarie et al., 2016). This programme sought to leverage expertise • ILRI conducted a project to develop tools for through partnerships, and commissioned some rapid, integrated assessment of food safety in important evidence syntheses. These concluded value chains. Studies in five countries docu- that the available evidence suggested that inter- mented the potential importance of livestock ventions to promote livestock were generally products to nutrition and how these were positive for nutrition, although few high-quality being eroded by poor food safety (El-Tholth studies took into account the complex links et al., 2018; Häsler et al., 2018, 2019; Roesel between livestock and nutrition, and most had et al., 2019; Nguyen-Viet et al., 2019). substantial methodological weaknesses (Leroy • ILRI conducted an analysis of the demand et al., 2006). The project also developed an influ- for livestock products, the drivers of this ential conceptual framework (Fig. 9.1) articulating demand and the barriers to consuming the links among livestock, nutrition and human livestock products among poor households health (Randolph et al., 2007). These links are in Nairobi. Price was found to be the most context specific. To begin teasing out the roles of important barrier to consumption, while taste different species, a study conducted in Ethiopia was reported as the main driver for con- demonstrated that ownership of small stock did not sumption. Estimated demand elasticities in- contribute to improved child nutrition within the dicated that increases in total food household, whereas poultry might provide direct expenditure would lead to the greatest in- benefits through egg consumption (Good, 2009). crease in demand for beef meat. Price re- An external review (Science Council/CGIAR, ductions would increase the demand 2008) recommended that human nutrition not relatively more for fish, other meats and be a focus for ILRI. This led to fragmentation of dairy products (Cornelsen et al., 2016). ILRI’s first human health programme, and for • A systematic review suggested that food several years little research was done at ILRI scares linked to livestock disease outbreaks relevant to human nutrition. However, the launch and FBD could harm nutrition due to con- of A4NH in 2012 provided an opportunity to re- sumers avoiding the implicated foods (Green vive this important area of research. et al., 2017). 358 D. Grace-Randolph et al. Land allocation to feed Food crop Traction, nutrient production cycling Animal and Food crop product sales sales Animal production Food crop Animal purchases owned Labour allocated to HH livestock income Health inputs Chronic disease HH ASF risk consumption (Female) Water Environmental toxin ASF caregiver Probability of transport concentration purchases income zoonotic disease Food-borne diseases Nutrient interactions HH crop Health status (Child) dietary consumption Nutritional (growth) intake status Cognitive performance Level of Wage labour by care/feeding (female) caregiver Total labour behaviour demands Hired labour Labour demands on (female) caregiver Fig. 9.1. Impact pathways among livestock keeping and human nutrition and health outcomes among the poor. (Adapted from Randolph et al., 2007.) ASF, animal-source foods; HH, household. Food Safety and Nutrition 359 • A study in Tanzania suggested that participa- companies providing mobile phone-based health, tion in a pro-poor agricultural intervention to nutrition and agriculturally based information improve milk production may improve women’s services to the poor. ILRI helped to build the milk consumption (Mishkin et al., 2018). capacity of local partners to develop appropriate • A Women Empowerment in Livestock Index, nutrition messages and to ensure the quality of based on a widely used index to measure the messages (CABI, 2017). More than 5 million empowerment of women in agriculture and people were reached with these nutrition mes- adapted to livestock keepers, incorporated sages. There was evidence of some behaviour nutrition and was used to identify dimen- change due to implementing this service, but it sions of empowerment associated with diet- proved difficult to develop business models to ary diversity and food security (Galiè et al., keep the service going because people were gen- 2018). erally unwilling to pay for mobile phone-based • Work with FAO on the challenges of ensur- health information. A rigorous external evalu- ing livestock interventions in the Sahel had ation of this project is under way. Preliminary positive nutritional benefits and led to a results indicate that aspects of the approach are reformulation of relevant FAO guidelines attractive to mothers, but considerable techno- (Dominguez-Salas et al., 2019). logical and sociological barriers challenge access and uptake (https://perma.cc/7QSA-Z9DF; ac- As ILRI also endeavoured to engage with the cessed 19 August 2020). Millennium Development Goals and the subse- Another large ILRI-led project focused on quent Sustainable Development Goals, there behavioural communication change messages were increasing efforts to understand the appro- to promote dietary diversity, including livestock priate contributions of livestock products to products, in Kenya. This project gave more than human diets, especially given the wide and some- 5000 women training in nutritional issues and times conflicting concerns about undernutrition, reached over 50,000 infants via nutritional overnutrition, the environmental externalities messages to their mothers (Kiome et al., 2019). of livestock systems, livestock-associated human This was not a research project and the impact diseases and animal welfare. A series of papers is not clear. Another project in Rwanda aimed looked at some of the synergies and trade-offs to evaluate the nutritional impacts of a social among these societal goals (Enahoro et al., 2018; and behavioural change communication inter- Salmon et al., 2018; Sirma et al., 2018). ILRI vention combined with a government initiative increasingly engaged in broad platforms that dubbed ‘One Cow per Poor Family’ (Flax et al., addressed all these issues. These included live- 2017). The final results of this project are not stock initiatives taking on greater nutritional yet available, but initial results confirm that focus, such as the multi-stakeholder Global Agenda families who are given a free cow had lower for Sustainable Livestock partnership, the Live- stunting prevalence than families who were eli- stock Data for Decisions project, the Global Live- gible but had not yet received a free cow (Flax stock Agenda to 2020 initiative and the Global et al., 2019). Livestock Advocacy for Development project. The ILRI projects have also been the entry point links among livestock, livestock-associated disease for other nutrition projects. The ILRI-led African and human nutrition were also set out in several Chicken Genetic Gains (ACGG) project in Ethi- influential publications that ILRI authored or opia has partnered with a Food, Agriculture and co-authored (Grace, 2015a, 2016, 2017; ILRI, Natural Resources Policy Analysis Network 2019). ILRI’s collaboration with Chatham House (FANRPAN) project which will promote chicken produced a widely cited and evidenced-based and egg consumption in the ACGG households synthesis of livestock-enhanced diets in the first benefiting from ACGG provision of 25 imported 1000 days of life (Grace et al., 2018a). tropically adapted chicken strains and locally A few ILRI projects have aimed to improve developed indigenous strains. Again, work is nutrition through consumption of livestock ongoing and findings are yet to emerge. p roducts as opposed to better understanding this In conclusion, the contribution that live- issue or advocating for it. ILRI participated in an stock make to human nutrition has evolved at mNutrition initiative that involved mobile phone ILRI from an assumed but unexamined premise, 360 D. Grace-Randolph et al. to an active area of research, to relegation outside h ypothesis that food safety is an important and of ILRI and finally back to renewed recognition probably growing constraint to smallholder that this should be an important focus of ILRI’s value chains because of its multiple burdens on agenda. The very small investments in this area to human health, livestock production and prod- date have necessarily constrained its impacts. Re- uct marketing. Over the same period, our search studies did produce useful information on understanding of the global burden of FBD in links among livestock keeping, livestock product developing countries has greatly increased, val- consumption and nutrition. There were also idating ILRI’s emphasis in this area, especially methodological advances in tools for assessing the importance of zoonotic disease and ani- nutrition in value chains, for formulating diets mal-source foods, areas where ILRI is man- and for measuring women’s empowerment. ILRI dated to research. advice has also been incorporated in many guide- ILRI research on FBD has resulted in many lines. While recent decades have seen livestock science outputs, including some genuinely in- production coming under increasing criticism in novative tools and approaches, and has already high-income countries because of environmen- demonstrated outcomes at community, national tal, health and animal welfare concerns, the in- and regional levels. These include substantial creasing numbers of high-level reports and global inputs into global, regional and national strat- engagements on nutrition and livestock issues are egies and national training programmes. The likely to draw attention to the importance of live- major development-oriented approach – the stock and livestock-derived products for nutrition- triple-path for training, motivating and enabling ally vulnerable populations of informal market agents – has been shown to be both scalable and sustainable. While ques- tions remain about its lasting effects on food The Future safety and its application outside those few countries where its success has been demon- ILRI and partners have been studying food strated, the next few years should bring further safety in informal markets for more than a evidence about this, with benefits lasting for decade. This work has helped confirm the many decades to come. References A4NH (2011) Agriculture for improved nutrition and health: full proposal. IFPRI, Washington, DC. A4NH (2016) Agriculture for nutrition and health: proposal for phase II, 2017–2022. IFPRI,Washington, DC. Aboge, G.O., Kang’ethe, E.K., Arimi, S.M., Omore, A.O., McDermott, J.J., et al. (2000) Antimicrobial agents detected in marketed milk in Kenya. Paper presented at the 3rd All Africa Conference on Animal Agri- culture, 6–9 November, Alexandria, Egypt. Smallholder Dairy Project (SDP), Nairobi. Ahmed, M.M., Bezabih, E., Jabbar, M.A., Tangka, F. and Ehui, S. (2003) Economic and nutritional impacts of market-oriented dairy production in the Ethiopian highlands. Socio-economics and Policy Research Working Paper 51. ILRI, Nairobi. Alonso, S., Muunda, E., Ahlberg, S., Blackmore, E. and Grace, D. (2018) Beyond food safety: socio-economic effects of training informal dairy vendors in Kenya. Global Food Security 18, 86–92. Bouis, H.E. (2000) Improving human nutrition through agriculture: the role of international agricultural research. Conference summary and recommendations. Food and Nutrition Bulletin 21, 550–567. CABI (2017) Lessons learned from the content development stream of the mNutrition initiative: Quality assurance and quality control processes as a part of mNutrition’s aim to produce high-quality content. mNutrition Brief 4. CAB International, Wallingford, UK. Claridge, J.A. and Fabian, T.C. (2005) History and development of evidence-based medicine. World Journal of Surgery 29, 547–553. Compton, J., McClean, D., Emmens, B. and Balagamwala, M. (2015) Independent CRP-commissioned external evaluation of the CGIAR Research Program on Agriculture for Nutrition and Health (A4NH) background papers. IFPRI, Washington, DC. Cornelsen, L., Alarcon, P., Häsler, B., Amendah, D.D., Ferguson, E., et al. (2016) Cross-sectional study of drivers of animal-source food consumption in low-income urban areas of Nairobi, Kenya. BMC Nutrition 2, 70. Food Safety and Nutrition 361 Delgado, C.L., Rosegrant, M.W. and Meijer, S. (2001) Livestock to 2020: the livestock revolution continues. Paper presented at the annual meeting of the International Agricultural Trade Research Consortium (IATRC), 18–19 January, Auckland, New Zealand. Dominguez-Salas, P., Kauffmann, D., Breyne, C. and Alarcon, P. (2019) Leveraging human nutrition through livestock interventions: perceptions, knowledge, barriers and opportunities in the Sahel. Food Security 11, 777–796. Doyle, J.J. (1984) ILRAD’s research program – its relevance to improved human nutrition. In: Pinstrup- Andersen, P., Berg, A. and Forman, M. (eds) International Agricultural Research and Human Nutrition. Proceedings from a workshop held at the International Livestock Centre for Africa, 29 February–2 March. IFPRI, Washington, DC, and United Nations Administrative Committee on Coordination/ Sub-Committee on Nutrition, Rome, pp. 153–156. Elsevier (2014) Research performance of CGIAR Research Programs. A report prepared by Elsevier for CGIAR. Eltholth, M., Fornace, K., Häsler, B. and Rushton, J. (2014) Rapid integrated assessment of nutrition and health risks associated with tilapia value chains in Egypt. ILRI project report. ILRI, Nairobi. Enahoro, D., Lannerstad, M., Pfeifer, C. and Dominguez-Salas, P. (2018) Contributions of livestock-derived foods to nutrient supply under changing demand in low- and middle-income countries. Global Food Security 19, 1–10. Flax, V., Ouma, E., Kamana, O. and Twine, E. (2017) Enhancing milk quality and consumption for improved income and nutrition in Rwanda. Presented at the Project Inception Workshop, Kigali, Rwanda, 7 March. ILRI, Nairobi. Flax, V., Ouma, E.A., Schreiner, M., Izerimana, L. and Niyonzima, E. (2019) Does the Girinka program improve the nutrition of young children and can the impact be increased through social behavior change communication? Feed The Future Innovation Lab For Livestock Systems project on enhancing milk quality and consumption for improved income and nutrition in Rwanda. ILRI, Nairobi. Galiè, A., Teufel, N., Korir, L., Baltenweck, I., Girard, A.W., et al. (2018) The Women’s Empowerment in Livestock Index: indicators for the start of a global, badly needed, conversation. ILRI, Nairobi. Gibb, H., Barchowsky, A., Bellinger, D., Bolger, M., Carrington, C., et al. (2015) Estimates of the 2015 global and regional disease burden from four foodborne metals – arsenic, cadmium, lead and methylmercury. Environmental Research 174, 188–194. Global Food Safety Partnership (GFSP) (2019) Food safety in Africa: past endeavors and future directions. World Bank, Washington, D.C. Gomez, M. and Ricketts, K. (2013) Food value chain transformations in developing countries. Food Policy 42, 139–150. Good, S. (2009) Animal source foods and nutrition during early life: an evaluation of the possible link be- tween livestock keeping, food intake and nutritional status of young children in Ethiopia. PhD thesis, Eidgenössische Technische Hochschule, Zürich, Switzerland. Grace, D. (2011) Assessment of risks to human health associated with meat from different value chains in Nigeria: using the example of the beef value chain. Nigeria Integrated Animal and Human Health Management Project Draft Report. ILRI, Nairobi. Grace, D. (2015a) Food safety in developing countries: an overview. Evidence on Demand, Hemel Hempsted, UK. Grace, D. (2015b) Food safety in low- and middle-income countries. International Journal of Environmental Research and Public Health 12, 10490–10507. Grace, D. (2016) Influencing food environments for healthy diets through food safety. In: Influencing Food Environments for Healthy Diets. FAO, Rome, pp. 14–19. Grace, D. (2017) Food safety in developing countries: research gaps and opportunities. White paper. ILRI, Nairobi. Grace, D. and McDermott, J. (2015) Food safety: Reducing and managing food scares. In: International Food Policy Research Institute. 2014–2015 Global Food Policy Report. IFPRI, Washington, DC, pp. 41–50. Grace, D. and Randolph, T. (2009) Development of a participatory methodology to prioritise milk borne dis- ease in data-scarce environments. In: Newton, J.R. and Pfeiffer, D. (eds) Proceedings of the annual conference of the Society for Veterinary Epidemiology and Preventive Medicine, 1–3 April. Society for Veterinary Epidemiology, London, pp. 188–196. Grace, D., Randolph, T., Olawoye, J., Dipelou, M. and Kang’ethe, E. (2008) Participatory risk assessment: a new approach for safer food in vulnerable African communities. Development in Practice 18, 611–618. Grace, D., Makita, K., Kang’ethe, E.K. and Bonfoh, B. (2010) Safe food, fair food: participatory risk analysis for improving the safety of informally produced and marketed food in sub-Saharan Africa. Revue A fricaine de Santé et de Productions Animales 8 (Suppl.), 3–11. 362 D. Grace-Randolph et al. Grace, D., Dipeolu, M., Olawoye, J., Ojo, E., Odebode, S., et al. (2012a) Evaluating a group-based inter- vention to improve the safety of meat in Bodija Market, Ibadan, Nigeria. Tropical Animal Health and Production 44 (Suppl. 1), 61–66. Grace, D., Kang’ethe, E. and Waltner-Toews, D. (2012b) Participatory and integrative approaches to food safety in developing country cities. Tropical Animal Health and Production 44 (Suppl. 1), S1–S2. Grace, D., Mutua, F., Ochungo, P., Kruska, R., Jones, K., et al. (2012c) Mapping of poverty and likely zoonoses hotspots. Zoonoses Project 4. Report to the UK Department for International Development. ILRI, Nairobi. Grace, D., Songe, M. and Knight-Jones, T. (2015a) Impact of neglected diseases on animal productivity and public health in Africa. Paper presented at the 21st Conference of the World Organisation for Animal Health (OIE) Regional Commission for Africa, 16–20 February, Rabat, Morocco. Grace, D., Kang’ethe, E., Lindahl, J., Atherstone, C., Nelson, F. and Wesonga, T. (2015b) Aflatoxin: im- pact on animal health and productivity. Building an aflatoxin safe East African community. Technical Policy Paper 4. IITA, Dar es Salaam, Tanzania. Grace, D., Lindahl, J., Atherstone, C., Kang’ethe, E., Nelson, F., Wesonga, T. and Manyong, V. (2015c) Aflatoxin standards for feed. Building an aflatoxin safe East African community. Technical Policy Paper 7. Dar es Salaam, Tanzania: IITA. Grace, D., Roesel, K., Kang’ethe, E., Bassirou, B. and Theis, S. (2015d) Gender roles and food safety in 20 informal livestock and fish value chains. IFPRI Discussion Paper No. 1489. IFPRI, Washington, DC. Grace, D., Dominguez-Salas, P., Alonso, S., Lannerstad, M., Muunda, E., et al. (2018a) The influence of livestock-derived foods on the nutrition of mothers and infants during the first 1,000 days of a child’s life. ILRI Research Report 44. ILRI, Nairobi. Grace, D., Dominguez-Salas, P., Alonso, S., Fahrion, A., Haesler, B., et al. (2018b) Food safety metrics relevant to low and middle income countries: Technical brief. Agriculture, Nutrition and Health Academy Food Safety Working Group. Innovative Methods and Metrics for Agriculture and Nutrition Actions Programme, London. Grace, D., Dipeolu, M. and Alonso, S. (2019) Improving food safety in the informal sector: nine years later. Infection Ecology & Epidemiology 9, 1579613. Green, T., Dominguez-Salas, P. and Grace, D. (2017) Can food scares shift health and nutrition outcomes in low- and middle-income countries? Poster presented at the 2nd annual Agriculture, Nutrition and Health (ANH) Academy Week, 9–13 July, Kathmandu, Nepal. London School of Hygiene and Trop- ical Medicine, London. Harvey, J., Macdevette, M., Mutiga, S., Mutuku, J., Eldridge, T., et al. (2016) Poisoned chalice: toxin accu- mulation in crops in the era of climate change. In: UNEP Frontiers 2016 Report: Emerging Issues of Environmental Concern. UNEP, Nairobi, pp. 54–63. Häsler, B., Msalya, G., Garza, M., Fornace, K., Eltholth, M., et al. (2018) Integrated food safety and nutri- tion assessments in the dairy cattle value chain in Tanzania. Global Food Security 18, 102–113. Häsler, B., Msalya, G., Roesel, K., Fornace, K., Eltholth, M., et al. (2019) Using participatory rural appraisal to investigate food production, nutrition and safety in the Tanzanian dairy value chain. Global Food Security 20, 122–131. Havelaar, A.H., Kirk, M.D., Torgerson, P.R., Gibb, H.J., Hald, T., et al. (2015) World Health Organization glo- bal estimates and regional comparisons of the burden of foodborne disease in 2010. PLoS Medicine 12, e1001923. Hawkes, C., Grace, D. and Thow, A.M. (2015) Trade liberalization, food, nutrition and health. In: Smith, R., Blouin, C., Mirza, Z., Beyer, P. and Drager, N. (eds) Trade and Health: Towards Building a National Strategy. WHO, Geneva, Switzerland, pp. 92–116. HLPE (2016) Sustainable agricultural development for food security and nutrition: what roles for livestock? A report by the High-Level Panel of Experts on Food Security and Nutrition of the Committee on World Food Security. HLPE report 10. FAO, Rome. Hoa, N.T.T., Nguyen-Viet, H., Dang, X.S., Pham, D.P., Nguyen, T.T. and Luu, Q.T. (2014) Rapid inte- grated assessment on food safety related to pork consumers in Hung Yen and Nghe An provinces. Journal of Practical Medicine 5, 230–233. Hoffmann, S., Maculloch, B. and Batz, M. (2015) Economic burden of major foodborne illnesses acquired in the United States. Economic Information Bulletin No. EIB-140. US Department of Agriculture, Eco- nomic Research Service, Washington, DC. Hoffmann, S., Devleesschauwer, B., Aspinall, W., Cooke, R., Corrigan, T., et al. (2017) Attribution of global foodborne disease to specific foods: Findings from a World Health Organization structured expert elicitation. PLoS One 12, e0183641. Food Safety and Nutrition 363 ILCA (1987) Annual Report 1986/87. ILCA. Addis Ababa. ILRI (2019) Options for the livestock sector in developing and emerging economies to 2030 and beyond. Meat: the Future Series. World Economic Forum. Geneva, Switzerland. Jabbar, M. A., Baker, D. and Fadiga, M. L. (eds) (2010) Demand for livestock products in developing coun- tries with a focus on quality and safety attributes: evidence from case studies. Research Report 24. ILRI, Nairobi. Jaffee, S., Henson, S., Unnevehr, L., Grace, D. and Cassou, E. (2019) the safe food Imperative: acceler- ating progress in low- and middle-income countries. World Bank, Washington, DC. Johnson, N.L., Mayne, J., Grace, D. and Wyatt, A. (2015) How will training traders contribute to improved food safety in informal markets for meat and milk? A theory of change analysis. IFPRI Discussion Paper 1451. IFPRI, Washington, D.C. Jutzi, S.C., and Rich, K.M. (2016) An evaluation of CGIAR Centers’ impact assessment work on live- stock-related research (1990–2014). Standing Panel on Impact Assessment (SPIA), CGIAR Inde- pendent Science and Partnership Council (ISPC), Rome. Kaitibie, S., Omore, A., Rich, K. and Kristjanson. P. (2010a) Kenya dairy policy change and economic im- portance. World Development 38, 1494–1505. Kaitibie, S., Omore, A., Rich, K., Salaysa, B., Hooton, N., et al. (2010b) Influence pathways and economic impacts of policy change in the Kenyan dairy sector. ILRI Research Report. Nairobi, Kenya: ILRI. Kang’ethe, E.K., Arimi, S.M., Omore, A.O., McDermott, J.J., Nduhiu, J.G., et al. (2000) The prevalence of antibodies to Brucella abortus in marketed milk in Kenya and its public health implications. Paper presented at the 3rd All Africa Conference on Animal Agriculture (AACAA), 6–9 November, Alexan- dria, Egypt. ILRI, Nairobi. Kassam, A., and Barat, S. (2003) Food safety considerations for CGIAR research. Journal of Agricultural & Food Information, 5, 27–71. Kiama, T.N., Lindahl, J.F., Sirma, A.J., Senerwa, D.M., Waithanji, E.M., et al. (2016) Farmer perception of moulds and mycotoxins within the Kenya dairy value chain: a gendered analysis. African Journal of Food, Agriculture, Nutrition and Development 16, 11106–11125. Kiarie, G.M., Dominguez-Salas, P., Kang’ethe, S.K., Grace, D. and Lindahl, J. (2016) Aflatoxin exposure among young children in urban low-income areas of Nairobi and association with child growth. African Journal of Food, Agriculture, Nutrition and Development 16, 10967–10990. Kimani, V.N., Mitoko, G, McDermott, B., Grace, D., Ambia, J., et al. (2012) Social and gender determinants of risk of cryptosporidiosis, an emerging zoonosis, in Dagoretti, Nairobi, Kenya. Tropical Animal Health and Production 44, 17–23. Kiome, R.M., Bamanya, B., Wamwere-Njoroge, G., Rao, J., Audi, P., et al. (2019) The Accelerated Value Chain Development program national conference report. ILRI, Nairobi. Kristkova, Z.S., Grace, D. and Kuiper, M. (2017) The economics of food safety in India – a rapid assess- ment. Wageningen University & Research, Amsterdam, Netherlands. Lapar, M.L.A., Deka, R., Lindahl, J. and Grace, D. (2014) Quality and safety improvements in informal milk markets and implications for food safety policy. Presentation to the Eighth International Conference of the Asian Society of Agricultural Economists, 14 October, Dhaka, Bangladesh. Leroy, J.L., Nicholson, C.F., Demment, M.W., Randolph, T.F., Pomeroy, A.M. and Frongillo, E.A. (2006) Can livestock production ameliorate under nutrition? ILRI, Nairobi. Lindahl, J.F., Deka, R.P., Melin, D., Berg, A., Lundén, H., et al. (2018) An inclusive and participatory ap- proach to changing policies and practices for improved milk safety in Assam, northeast India. Global Food Security 17, 9–13. Maudlin, I., Eisler, M.C. and Welburn, S.C. (2009) Neglected and endemic zoonoses. Philosophical Trans- actions of the Royal Society of London B: Biological Sciences 364, 2777–2787. Mishkin, K., Raskind, I., Dominguez-Salas, P., Baltenweck, I., Omore, A. and Girard, W.A. (2018) The effect of participation in a pro-poor dairy development project on milk consumption among reproductive age women in rural Tanzania. African Journal of Food, Agriculture, Nutrition and Development 18, 12992–13008. Mwangi, A., Arimi, S.M., Mbugua, S., Kang’ethe, E.K., Ouma, E.A., et al. (2000) Application of HACCP to improve the safety of informally marketed raw milk in Kenya. In: ISVEE 9: Proceedings of the 9th Symposium of the International Society for Veterinary Epidemiology and Economics, Breckenridge, Colorado, p. 504. Nguyen-Viet, H., Dang-Xuan, S., Pham-Duc, P., Roesel, K., Huong, N.M., et al. (2019) Rapid integrated assessment of food safety and nutrition related to pork consumption of regular consumers and mothers with young children in Vietnam. Global Food Security 20, 37–44. 364 D. Grace-Randolph et al. Nicholson, C.F., Mwangi, L., Staal, S.J. and Thornton, P.K. (2003) Dairy cow ownership and child nutri- tional status in Kenya. Selected paper presented at the American Agricultural Economics Asso- ciation Annual Meetings, 27–30 July, Montréal, Québec, Canada. AAEA, Québec, Canada. Nicholson, C.F., Thornton, P.K., Mohammed, L., Muinga, R.W., Mwamachi, D.M., et al. (1999) Smallholder dairy technology in coastal Kenya. An adoption and impact study. ILRI Impact Assessment Series No. 5. ILRI, Nairobi. Njuki, J.M., Wyatt, A., Baltenweck, I., Yount, K., Null, C., et al. (2015) An exploratory study of dairying in- tensification, women’s decision making, and time use and implications for child nutrition in Kenya. Euro- pean Journal of Development Research 28, 722–740. Omore, A., and Baker. D. (2011) Integrating informal agents into the formal dairy industry in Kenya through training and certification. In: Towards Priority Actions for Market Development for African Farmers. Proceedings of an International Conference, 13–15 May 2009, Nairobi, Kenya. AGRA and ILRI, Nairobi, pp. 281–291. Omore, A.O., McDermott, J.J., Staal, S.J., Arimi, S.M., Kang’ethe, E.K. and Ouma, E.A. (2000) Analysis of public health risks from consumption of informally marketed milk in sub-Saharan African countries. In: ISVEE 9: Proceedings of the 9th Symposium of the International Society for Veterinary Epidemiology and Economics, Breckenridge, Colorado, p. 502. Omore, A., Staal, S., Kurwijila, L., Osafo, E., Aning, G., et al. (2001) Indigenous markets for dairy products in Africa: trade-offs between food safety and economics. Paper presented at the 12th Symposium on Trop- ical Animal Health and Production, Dairy Development in the Tropics, organized by the Faculty of Veter- inary Medicine, Utrecht University, 2 November, Utrecht, The Netherlands. ILRI, Nairobi, pp. 19–24. Orton, L., Lloyd-Williams, F., Taylor-Robinson, D., O’Flaherty, M. and Capewell, S. (2011) The use of re- search evidence in public health decision making processes: systematic review. PLoS One 6, e21704. Perry, B.D., Randolph, T.F., McDermott, J.J., Sones, K.R. and Thornton, P.K. (2002) Investing in animal health research to alleviate poverty. ILRI, Nairobi. Pinstrup-Andersen, P. (2000) Improving human nutrition through agricultural research: overview and object- ives. Food and Nutrition Bulletin 21, 352–355. Randolph, T.F., Schelling, E., Grace, D., Nicholson, C.F., Leroy, J.L., et al. (2007) Invited review: Role of livestock in human nutrition and health for poverty reduction in developing countries. Journal of Animal Science 85, 2788–2800. Renkow, M. and Byerlee, D. (2010) The impacts of CGIAR research: a review of recent evidence. Food Policy 35, 391–402. Roesel, K. and Grace, D. (2014) Food Safety and Informal Markets: Animal Products in sub-Saharan Africa. Routledge, London. Roesel, K., Ejobi, F., Dione, M., Pezo, D., Ouma, E., et al. (2019) Knowledge, attitudes and practices of pork consumers in Uganda. Global Food Security 20, 26–36. Salmon, G., Teufel, N., Baltenweck, I., Wijk, M. van, Claessens, L. and Marshall, K. (2018) Trade-offs in livestock development at farm level: different actors with different objectives. Global Food Security 17, 103–112. Scharff, R.L. (2012) Economic burden from health losses due to foodborne illness in the United States. Journal of Food Protection 75, 123–131. Science Council/CGIAR (2008) Report of the Second External Programme And Management Review (EPMR) of the International Livestock Research Institute. Science Council of the Secretariat, Rome. Sirma, A.J., Lindahl, J.F., Makita, K., Senerwa, D., Mtimet, N., et al. (2018) The impacts of aflatoxin stand- ards on health and nutrition in sub-Saharan Africa: the case of Kenya. Global Food Security 18, 57–61. Sridharan, S., Tschirley, D. and Stark, K. (2015) CRP-Commissioned External Evaluation of the Food Safety Research at the CGIAR Research Program on Agriculture for Nutrition and Health. CGIAR, Montpellier, France. Technical Advisory Committee (2000) A food secure world for all: towards a new vision and strategy for the CGIAR. TAC Secretariat, FAO, Rome. Thornton, P.K. and Odero, A.N. (1998) Compendium of ILRI research impact and adoption, 1975–98. ILRI Impact Assessment Series 1. ILRI, Nairobi. Thornton, P.K., Kruska, R.L., Henninger, N., Kristjanson, P.M. and Ndegwa, T. (2002) Mapping poverty and livestock in the developing world. ILRI, Nairobi. Tschirley, D., Reardon, T., Dolislager, M. and Snyder, J. (2015) The rise of a middle class in East and Southern Africa: implications for food system transformation. Journal of International Development 27, 628–646. Food Safety and Nutrition 365 Unnevehr, L. and Grace, D. (eds) (2013) Aflatoxins: finding solutions for improved food safety. IFPRI 2020 Focus Brief 20. IFPRI, Washington, DC. Unnevehr, L.J., and Ronchi, L. (2014) Food safety and developing markets: research findings and research gaps. IFPRI, Washington, DC. WHO (2013) Research priorities for the environment, agriculture and infectious diseases of poverty: technical report of the TDR Thematic Reference Group on Environment, Agriculture and Infectious Diseases of Poverty. Technical Report Series No. 976. WHO, Geneva, Switzerland. World Bank (2017) Vietnam Food Safety Risks Management: Challenges and Opportunities, Vol. 2: Annexes. World Bank, Washington, D.C. WRENmedia (2010) Policy change: milking the benefits for small-scale vendors. DFID, Kenya. Zhong, S., Crang, M. and Zeng, G. (2020) Constructing freshness: the vitality of wet markets in urban China. Agriculture and Human Values 37, 175–185. 10 Ticks and Their Control Peter Willadsen Brisbane, Australia Contents Executive Summary 366 The problem 366 ILRI’s role in the global context 367 Scientific impacts 367 Development impacts 368 Policy influence and advice 368 Capacity development and partnerships 368 Introduction 368 Practical Tick Control 369 Development of Research Capacity 373 The ILRI Tick Unit 373 Artificial feeding systems 373 Genomics and molecular biology 374 Application of Research Capacity 375 Ticks as vectors, with a focus on R. appendiculatus as the vector of T. parva 375 The genetic complexity of R. appendiculatus 377 Tick vaccines 378 Recent Developments 380 The Future 381 References 382 Executive Summary the irritation resulting from their feeding activity. Other negative effects include the injection of The problem toxins, transmission of endemic and emerging diseases (such as heartwater, African swine fever Ticks are bloodsucking external parasites. They and Congo-Crimean haemorrhagic fever) and tick- are responsible for decreased productivity due to associated disease (such as dermatophilosis). blood loss from the host animal and ‘tick worry’, Tick-borne pathogens affect 80% of the world’s © International Livestock Research Institute 2020. The Impact of the International 366 Livestock Research Institute (eds J. McIntire and D. Grace) Ticks and Their Control 367 cattle population and are ubiquitous in the trop- or control – was soon involved in diverse areas of ics and subtropics. Countering these negative tick research. effects requires expensive control measures. Glo- Following flagship projects at ILRAD, ILRI bal costs associated with ticks and tick-transmitted conducted important research on tick biology, pathogens in cattle alone were estimated years tick population dynamics, the impact of ticks ago at above US$13.9 billion and US$18.7 and tick control using chemicals. Apart from b illion, respectively (de Castro, 1997) and costs ECF, a major ILRI research theme, other haemo- have doubtless increased in this century. Climate parasites and, more recently, viral pathogens change and transboundary trade in livestock were studied. have recently begun to drive ticks into new areas where animals have less resistance to both ticks and tick-borne diseases. In Africa, ticks and tick- Scientific impacts borne disease appear at the top of several rank- ings of important livestock diseases; notably, Sustainable strategies for the control of ticks and diseases transmitted by parasites, such as tryp- tick-borne disease involve a complex interplay anosomiasis and East Coast fever (ECF), ranked between parasite and host species, available con- high in International Livestock Research Insti- trol technologies and a range of environmental tute (ILRI) prioritizations of livestock disease factors. In this challenging situation, ILRI’s con- across regions and production systems of tributions were a continuum from laboratory- sub-Saharan Africa (Perry et al., 2002). based science, through field experimentation, to The most common control methods are the practical advice and policy recommendations. use of genetically resistant animals and the ILRI made early and ongoing contributions to application of acaricides. Acaricides may be our understanding of ‘endemic stability’. This oc- applied through dips, sprays or pour-on formu- curs when rates of infection are sufficient to main- lations as well as intra-ruminal boluses, ear tags tain a level of acquired immunity that minimizes and footbaths. Resistance to acaricides is the clinical disease in a population. The concept was ability in a strain of ticks to tolerate doses of developed to describe patterns of tick-borne disease acaricides that would prove lethal to most indi- in cattle but has since been applied more broadly viduals in a normal population of the same spe- in veterinary and human health. cies, and this is a major and growing problem. ILRI’s field experiments in Kenya, Ethiopia An anti-tick vaccine is commercially available and Uganda led to recommendations for on- for only a single tick species. Pasture manage- farm tick control. For example, in Kenya, acari- ment also has a role in integrated control. cide treatment improved weight gains where keeping unvaccinated cattle without acaricides was uneconomical. Rotation of acaricides could ILRI’s role in the global context potentially mitigate resistance, so ILRI scientists conducted the first field examination of acari- Because of its geographical location, abundant cide rotation, finding it was advantageous. infrastructure and technical expertise, ILRI was More laboratory-based research into tick and remains in a powerful position to contribute biology was greatly facilitated by the existence of to the global understanding of African ticks and the Tick Unit, described above. ILRI developed an tick-borne diseases. In 1979, the International artificial feeding system for ticks and methods Laboratory for Research on Animal Diseases for maintaining tick colonies (four stocks for (ILRAD) Tick Unit was established as a resource – over 30 years, and six for more than 20 years). a provider of skills and materials that could be used In addition, tick-breeding research established in ILRI’s research and, potentially, by others as high- and low-infectivity lines for ECF: seven dif- well. The unit was built primarily to support re- ferent stocks and lines of ticks from eastern and search on ECF. The control of ECF and other southern Africa, which include low (refractory) such diseases is inextricably linked to ticks and and high (susceptible) genetic tick lines. Vector their control, so ILRI, although it lacked a sys- biology studies included understanding popula- tematic tick research programme – a programme tion genetics of ticks in East Africa and molecu- aimed at some specific component of tick biology lar taxonomy of Afro-tropical ticks, tick ecology, 368 P. Willadsen and disease dynamics at the wildlife and live- for farmers but since uptake of this information stock interface. ILRI developed maps of tick is indirect, via extension services, its final impact d istribution and models that explained the is not known. spread and the subsequent disappearance of an e specially problematic tick species after its intro- duction to Zimbabwe. Policy influence and advice The Tick Unit has adapted protocols for detecting and quantifying chemical resistance ILRI generated evidence on the adverse conse- in ticks to backstop similar efforts by regional quences of stopping state-supported tick control veterinary services. Pen and field trials of new acari- in Zimbabwe and advised on better control cide formulations are being tested as alternatives approaches. A consultancy to the Director of to existing compounds to mitigate acaricide re- Veterinary Services in Zimbabwe recommended sistance in tick vectors of veterinary importance. tick control strategies at a critical time for live- ILRI has played a significant role in the stock production in that country. international effort to apply a range of molecu- lar technologies to improve the scientific under- Capacity development and partnerships standing of ticks and future means of their ILRI has built capacity in tick research and con- control. As genomic technologies emerged glo- trol for students, extension workers and farmers. bally, ILRI’s scientists played a valuable part in Supported by the Wellcome Trust, ILRI and sequencing tick genomes as they did for ECF. Sig- government officials set up meetings between nificant research was conducted on anti-tick farmers and researchers, to identify problems vaccine evaluation and antigen identification, associated with ticks and tick-borne disease, in- vector genomics and vector–pathogen inter- cluding tick sampling with farmers. In addition, actions. The involvement in the identification of tick scientists from Sudan, Kenya, M alawi and novel tick antigens continues. Ghana have come to ILRI to be trained on tick According to Altmetric (www.altmetric.com/; dissections and tick management. To strengthen accessed 24 February 2020), ILRI contributed future research capacity, the institute has to 2% of the research outputs on ticks. trained more than 20 PhD and MSc students and Fellows. The close link on the ILRI campus Development impacts between advanced molecular sciences and prac- tical tick culture and control have often been a key competitive advantage. The most visible development achievement is an National and international partners have ECF vaccine, which is covered in Chapter 6 (this included the Wellcome Trust, the UK Biotech- volume). The Tick Unit and the tick research nology and Biological Sciences Research Coun- have been essential components of this major cil (BBSRC), the Brazilian Federal Agency for commitment by ILRI. The organization has Support and Evaluation of Graduate Education played a key role in the development of a recom- (CAPES), Genesis Labs, the University of Bern binant ECF vaccine and an important part in the and the Directorate of Veterinary Services in development of the infection-and-treatment Kenya. regime, which has been shown to have a signifi- cant impact in Tanzania. A significant and prac- tical outcome of the Tick Unit and ILRI’s research includes sharing knowledge and advice with the Introduction Centre for Ticks and Tick-Borne Disease (CTTBD), an animal vaccine production facility in Malawi. From the early days of ILRI, the impact of Afri- Studies of heartwater, an important tick- can tick species on animal productivity, the com- borne disease affecting small ruminants, gener- plexity of their control in the field and their ated evidence on economic impacts and a potential activity as vectors of diseases such as ECF, ba- market for a vaccine. besiosis, anaplasmosis and heartwater have all ILRI developed recommendations on tick engaged ILRI’s scientists. Supporting these prac- and tick-borne disease control and acaricide usage tical issues has been research into fundamental Ticks and Their Control 369 aspects of tick biology. The aggregate contribu- laboratory-based research has focused on devel- tion of ILRI’s scientists to tick research over the oping research capacity through the ILRI Tick decades has been considerable. However, from Unit and the acquisition of techniques such as in the beginning, the major focus of animal health vitro tick feeding and genomic technologies. In research in ILRI has been on a small number of turn, this has led to application in the develop- tightly defined areas such as vaccination against ment of vaccines for ECF, an understanding of ECF and trypanosomiasis. Perhaps for that reason, the genetic diversity of the major ECF vector spe- ILRI has typically lacked a single, focused tick cies, Rhipicephalus appendiculatus, and progress programme, and its engagement in tick research in the development of anti-tick vaccines. Each of has often seemed to be ambivalent. Increasing these areas will be addressed in turn. this ambivalence may have been a reluctance to A note on nomenclature: Boophilus ticks were intrude where other African institutions could long classified as a genus and in all of the litera- claim involvement. An additional consideration ture pre-2001 are cited as such. Approximately may have been the fact that tick control depended 15  years ago, the suggestion was made, on the largely on either the use of indigenous livestock basis of molecular evidence, that the genus should by smallholders or the application of chemical be considered a subgenus of Rhipicephalus (Barker acaricides manufactured by global animal health and Murrell, 2004). The coherence of the genus companies. Hence, the role of an institution like Rhipicephalus as it stands has also been questioned. ILRI was not so clear. Hence, in the literature a species may be described The lack of a coherent strategy had two broad as Rhipicephalus decoloratus, Rhipicephalus (Boophi- consequences. First, as the research activities lus) decoloratus or Boophilus decoloratus. In what that were undertaken reflected either the specific follows, to cover all possibilities, the Boophilus ticks needs of other projects or a scientist’s personal will be referred to as Rhipicephalus (Boophilus) spp. interest, ILRI’s contributions to tick research or R. (Boophilus) spp. cover an extremely diverse set of topics. In con- trast with other, more focused areas of activity within ILRI, it is therefore harder to point to major, quantifiable impacts. This is the downside Practical Tick Control of the somewhat piecemeal approach ILRI took to tick research. On the upside, as the contribu- The impact of ticks as vectors of disease has tions were mostly a result of individual initia- been quantified by groups working on these dis- tives, they were typically dependent on external eases, principally, in the case of ILRI, ECF (Min- collaborations. They therefore constitute an jauw and McLeod, 2003). The direct effect of ticks excellent example of ILRI’s capacity to engage themselves has been less examined, although it productively in multi-organizational projects. The is a major concern in countries such as Australia number of collaborating institutions involved in and much of Central and South America. There ILRI’s published tick research is large. In what were, however, several early attempts to quantify follows, the focus is on areas where, subjectively production losses due to ticks. De Castro et al. assessed, ILRI’s contributions have been major (1985a) used Boran cattle immunized against or at least essential to the final result. Theileria spp. to examine the effects of acaricide In reviewing this diverse effort, it is conveni- treatment. Five tick species were found to infect ent to break it generally into two components. The the cattle, although in smaller numbers on first is work that has been largely field based, fo- acaricide-treated animals, which also showed cusing on the practicalities of tick control. The higher live weight gains. Overall, however, the second is work that has been more laboratory conclusion was that it was possible to keep Zebu based. Significant contributions have been made cattle without chemical tick control, once im- to practical tick control by quantifying the im- munized against Theileria spp. Estimates of the pact of ticks, in understanding the dynamics of direct impact of R. appendiculatus on productiv- tick populations on a local and landscape scale, ity were low. In another study published in and in providing input to the effective use of 1985, only relatively small and transient effects pesticides (acaricides), including the economic were found from the infestation of cattle with up cost and benefit of their application. The more to 400 R. appendiculatus a week for 24  weeks. 370 P. Willadsen There was some suggestion of acquisition of im- controlled efficiently through a programme of munity to the ticks (de Castro et al., 1985b). intensive dipping. The low incidence of tick-borne Morzaria et al. (1988) returned to the question, disease led indirectly to overgrazing in commu- finding that, for cattle vaccinated against ECF, nal areas. Then, between 1973 and 1978, a acaricide treatment improved weight gains, weakening of the dipping programme was fol- while keeping unvaccinated cattle without lowed, typically after 1–3 years, by increases in acaricide treatment was simply uneconomical. the population of R. (Boophilus) decoloratus and In Africa, as throughout the rest of the hence outbreaks of babesiosis and anaplasmosis. world, if tick control is consciously applied, it is Cattle numbers plummeted; then, with reduced most likely to be through the application of syn- animal numbers, grazing pressure reduced, grass thetic acaricides. The scientific challenge is to cover increased and R. appendiculatus and Am- identify the optimal acaricide treatment regime, blyomma hebraeum re-established in these areas. one that balances the negatives of cost and po- This, in turn, was followed by outbreaks of tential environmental and health impacts with theileriosis and heartwater (Norval, 1979; Law- the positives of the control of tick-borne disease rence et al., 1980). These disease outbreaks were and the minimization of production losses attributed to a loss of immunity to the tick-borne through the direct effects of ticks themselves. diseases, i.e. a loss of endemic stability because Within this question lies an extremely complex set of a long-term lack of exposure to ticks and the of issues. Are the animals to be treated – usually diseases they transmit. Subsequently, serological cattle – indigenous or exotic breeds? Which are surveys in 1981–1982 suggested that endemic the most important tick species and what is their stability was being regained in cattle on commu- distribution, both spatially and temporally? nal land, and the suggestion was made not to re- Which acaricides are available and what is their introduce intensive dipping (Norval et al., 1992). efficacy? The choice of acaricide may be affected Over the years, ILRI has made a number of by issues such as the cost and the occurrence of contributions to such complex situations. Al- acaricide resistance. ‘Acaricides’ will include not though there has been little direct involvement only commercial products of certified quality but in the evaluation of acaricides in straightfor- also market-derived treatments of uncertain ori- ward field situations, there are exceptions. For gin, quality and efficacy. This, in turn, is affected example, in 2003, there was a 1-year longitu- by national and regional registration proced- dinal study of 92 smallholder dairies in Kenya ures, which may be rigorous or disturbingly lax. to evaluate the efficacy of deltamethrin in the All these issues may be regarded as essentially control of the major tick-borne diseases and scientific or technical. The final set of complex- trypanosomiasis. Four application regimes were ities relate to infrastructure and human behav- compared – biweekly, monthly and bimonthly iour: the availability of dips or other treatment treatment and untreated controls – and the con- facilities, the distance between the farmer and clusion was reached that monthly treatment could the treatment facility, and the fact that acaricide reduce the incidence of tick-borne diseases to a application, regardless of the efficacy of the statistically significant degree (Muraguri et al., chemical itself, is frequently done so poorly that 2003). In 2003 and 2004, the efficacy of cyper- treatment is ineffective. methrin on four tick species was investigated in An event in recent African history provides Ethiopia and it was found that application every a good example of the benefits and risks of tick 3  weeks provided good protection (Mekonnen control, especially in the absence of a scientific- et al., 2004). ally sound and enforceable policy. Andy Norval, The complex issue of optimal frequency of who was head of the ILRI Tick Unit from 1987 to acaricide treatment in the absence of pesticide 1989, had in earlier work noted that the most resistance has been examined by ILRI and their common tick in Zimbabwe in overgrazed tribal collaborators in a variety of production settings. areas was R. (Boophilus) decoloratus, while in The population dynamics of four major tick spe- well-managed commercial farms, the diversity cies on indigenous and cross-bred cattle in a dry of species was greater but the economically most or semi-arid area of Uganda were examined for important one was R. appendiculatus. Until the cattle given biweekly, monthly or no acaricide early 1970s, ticks and tick-borne diseases were treatment. There were effects of lactation and Ticks and Their Control 371 year-on-year variation but only on the cattle of two different acaricides in a controlled way. given biweekly treatment were the overall num- The results were encouraging (Thullner et al., bers of ticks reduced significantly (Okello-Onen 2007). However, the work reported by Kamidi et al., 1999). In a companion paper published in and Kamidi (2005) is perhaps the only published 1998, the variable costs of acaricides, drugs and example where this has been examined in a field labour, and the benefits in live weight gains, were situation. The work described the consequences together used to calculate the most economically of three different acaricide treatments on a sin- beneficial treatment frequency for indigenous gle smallholder farm with exotic dairy cattle cattle. The conclusion was that a biweekly dipping over 7.5 years. Initially, ticks were controlled by strategy did not offer benefits commensurate with weekly spraying with amitraz. Then, when re- costs, while the monthly treatment gave clear sistance developed after 3 years, this was replaced economic benefit (Okello-Onen et al., 1998), a by fortnightly treatments with an organophos- conclusion that seemed somewhat at variance phate, although it was known that resistance to with the finding on tick numbers. The authors this chemical group occurred in the district. A returned to the question in 2003, again examin- dramatic increase in tick-borne disease followed. ing the effect of biweekly and monthly dipping The final strategy was a regime of fortnightly on indigenous breeds over a period of almost spraying, using amitraz and the organophosphate 3 years. The biweekly dipping improved milk off- in rotation. This resulted in the lowest incidence take and pre-weaning growth rates (Okello-Onen of tick-borne disease of the three strategies. et al., 2003). As they currently stand, such dis- These observations raise interesting questions. crepant results would be difficult to translate First, it is striking that, although both acaricides into a clear message for farmers. were expected to achieve at best partial tick con- Scientifically more interesting was a paper trol used separately, the rotation seemed to be derived from an ILRI–Kenya Agricultural Research effective. Second, the cost of acaricide was, inev- Institute (KARI) collaboration that addressed an itably, very high, although perhaps less than unresolved issue in acaricide usage (Kamidi and might have been incurred with an increased in- Kamidi, 2005). Boophilus spp. in particular have cidence of tick-borne disease. Finally, of course, a pronounced capability to develop resistance to the long-term financial and biological sustain- a range of acaricides. The example of resistance ability of intensive pesticide treatment in the in Rhipicephalus (Boophilus) microplus is aston- face of known resistance must be doubtful. ishing and very well documented. It is known Several publications have attempted to take that, with the possible exception of amitraz, re- the issue of ticks from the field to the landscape sistance once acquired can persist for a very long scale. During the 1990s, the evolving situation time, even in the absence of chemical selection in Zimbabwe was a focus, and the approaches in- pressure. Commonly, the market offers an odd cluded: (i) examination of the field distribution mix of different acaricides that are used by farmers of particular species, often supplemented by in a fairly random way. On occasion, attempts ecological and climatic modelling; (ii) studies on have been made to regulate a chaotic situation current acaricide usage and practices; and (iii) by enforcing the serial use of acaricides with the economic analysis of preferred strategies. For idea of exhausting one before applying another. example, factors contributing to the spread of An alternative approach would be to use R. appendiculatus in Zimbabwe after its introduction acaricides in a structured rotation, although the and its subsequent disappearance were exam- benefits of this for tick control have not been ined using a climate model. It was concluded that established. the chief causal factor was a wet–dry climate Rotation of pesticides from different chem- cycle (Norval and Perry, 1990). ical groups has been used in the management of On a number of occasions, these concerns some crop pests to reduce the probability of the evolved into policy advice. The focus on Zim- emergence of strong resistance. This has never babwe may have been a result of the dramatic been systematically studied with acaricides for effects of a failure in tick control during the civil tick control. There has been one published labo- unrest. In 1990, Perry and Mukabeni of ILRI, ratory-based test of the idea that the emergence of and Norval, who had recently shifted to the Uni- resistance could be delayed by the rotational use versity of Florida, together with Barrett, in the 372 P. Willadsen Department of Veterinary and Tsetse Control Z imbabwe displayed anomalous features: the Services, submitted a report containing recom- ticks occurred in areas of lowest predicted cli- mended tick control strategies to Zimbabwe’s matic suitability for survival and development Director of Veterinary Services (Perry et al., and in areas where the densities of cattle, the 1990a). Twenty-five years later, the report still most important domestic host, were lowest. The encapsulates many of the typical issues: political only factor favouring the survival of the species imperatives, changing disease epidemiology, the in the Lowveld habitats in which they occurred cost of acaricides, the paucity of methods to was the presence of alternative wildlife hosts for control tick-borne disease, and so on. To briefly the adult stage. Norval et al. (1994) concluded: recap this history: prior to the independence war of the 1970s, intensive acaricide treatment had Their absence from more climatically favourable led to the eradication of ECF and the control of Highveld habitats appears to have been the result of intensive acaricide treatment of cattle other tick-borne diseases. The highly regulated over a long period and a historic absence of system of treatment was, however, increasingly significant numbers of wildlife hosts. Eradica- unpopular. It collapsed during the war, as noted tion of A. hebraeum and A. variegatum by previously, with consequent severe outbreaks intensive acaricide treatment of cattle can be of disease and the death of approximately 1 mil- achieved in the absence of significant numbers lion cattle. Subsequently, there was evidence of alternative hosts, because of the long that endemic stability to these diseases was attachment and feeding periods of the adults of r e-e stablished or in the process of being re- these tick species. However, eradication becomes established, and there seemed to be the oppor- impossible when alternative hosts for the adult tunity to pair satisfactory disease control with stage are present, because a pheromone emitted by attached males attracts the unfed nymphal reduced use of expensive (state-purchased) and adult stages to infested hosts. The unfed acaricides. The catastrophic loss of cattle, how- ticks are not attracted to uninfested hosts, such ever, had reawakened a demand for intensive as acaricide-treated cattle. dipping, which was reintroduced for political reasons. By the time of the ILRI consultancy, the In the face of a probable reduction in intensive costs were becoming unsustainable. acaricide treatments, due to the cost to the gov- The policy advice that flowed from this con- ernment, the authors suggested two potential sultancy was complex. It divided communal alternative strategies: to establish a buffer zone to lands into four categories, based largely on the restrict disease spread, or to allow the ticks to probable but different impacts of babesiosis, an- spread and to control heartwater by immunization. aplasmosis, heartwater and theileriosis. A reduc- By 1998, there was evidence that the heart- tion in acaricide treatment with a move, where water vectors were, in fact, spreading (Peter et al., possible, to minimal acaricide application was 1998). A study reporting the results of a survey proposed. The target was to achieve endemic sta- in 1995–1996 (Perry et al., 1998) returned to bility to the various diseases, supported by vac- the question of A. hebraeum, A. variegatum and cination. Significant cost-saving was envisaged, heartwater. The dynamics of tick control were although the costs of vaccinations were uncer- changing. The government, under financial tain (Norval et al., 1992). pressure, had abandoned intensive dipping, and These ideas were revisited again in 1994. wildlife species were moving back into the High- The concern was specifically with heartwater veld. Suggestions were made on how to deal with the and its vectors, the ticks A. hebraeum and Ambly- increased disease risk. The overuse of acaricides omma variegatum (Norval et al., 1994). A. hebrae- remained a concern. It was noted that intensive um was then widely distributed in the dry acaricide treatment (i.e. more than 30 times a southern Lowveld with some foci in the wetter year) was associated with a higher risk of heartwa- areas of the Highveld, while A. variegatum oc- ter than on farms using more strategic dipping. curred in the Zambezi Valley and surrounding Given the very strong focus in ILRI on ECF, dry Lowveld areas. The distribution of A. hebrae- the distribution of R. appendiculatus was a long- um had changed over the preceding 70  years, standing interest. The example of the spread of while that of A. variegatum had remained static. R. appendiculatus in Zimbabwe is given above. The distribution patterns of both species in A broader consideration of the distribution of Ticks and Their Control 373 R. appendiculatus in Africa based on climate logical resources. Currently, the ILRI Tick Unit and vegetation was published by Perry et al. has in culture representatives of three genera of (1990b). The starting point in this research was ticks, including the subgenus of Boophilus. These a predictive model of tick distribution based on genera are Rhipicephalus, Rhipicephalus (Boophilus), climatic factors. The model was originally devel- Hyalomma and Amblyomma. The species in cul- oped by the Commonwealth Scientific, Industrial ture are R. appendiculatus, Rhipicephalus zambez- and Research Organisation (CSIRO) to better iensis, Rhipicephalus evertsi, A. variegatum, R. understand the distribution of R. (Boophilus) (Boophilus) decoloratus, R. (Boophilus) microplus microplus in Australia (Sutherst, 2003). Here, it and Hyalomma anatolicum. The diversity of R. ap- was to be applied to R. appendiculatus. The pendiculatus cultures is a focus of the collection, authors found that the ecoclimatic indices of with seven different lines from eastern and suitability correlated well with known tick distri- southern Africa, including selected lines derived butions in eastern Africa but were insufficient from the Kiambu stock that differ in susceptibil- as a single-factor explanation in central and ity to T. parva infection and lines from South and southern Africa. Cold stress and vegetation eastern Africa selected for experiments on the microenvironments were also considered to be impact of diapause on disease transmission, as important. Factors such as cattle and wildlife well as geographically diverse isolates, including distributions and acaricide control (if any) were two from Zimbabwe and three from Zambia. The hypothesized to be relevant, although the data to expertise of the ILRI Tick Unit currently is in tick test that hypothesis were insufficient. Predictably, culture, performance of animal trials under con- too, this focus on tick distribution and climatic trolled conditions and the process of producing factors was combined into an attempt to under- sporozoites for ECF projects. These skills continue stand the epidemiology of Theileria parva using to underpin basic research into T. parva (e.g. additional information from geographic infor- Henson et al., 2012). mation systems (Lessard et al., 1990). Ticks share with many other parasites the irritating characteristic that, while they are ex- ceedingly difficult to kill in the wild, they are Development of Research Capacity often very difficult to maintain under controlled conditions. Hence, it is an achievement that A mix of fundamental methodology, facilities, of the R. appendiculatus stocks, four have been equipment and hands-on experimental expertise maintained for over 30  years and six for more underpins all scientific research. ILRI played a than 20  years. Two of the other species have significant role in developing such scientific been in culture for over 30  years and one for capacity in at least two areas of tick research. more than 20 years. Admirable though this is, The first was through the establishment of a it is also a cause for concern. Experience with Tick Unit for the development and maintenance R. (Boophilus) microplus in Australia showed that of tick colonies together with the acquisition a much-used reference tick colony, maintained of techniques needed for tick and tick-borne dis- for even longer, showed declining viability and ease research. The second was via the involve- then sudden collapse. ment of ILRI staff in tick genomics and other molecular technologies. Artificial feeding systems The ILRI Tick Unit To understand and, if necessary, dissect the pro- cess of tick feeding on a mammalian host (cattle) The ILRI Tick Unit was built around experimen- and thus acquiring a disease organism such as tally valuable infrastructure: fly- and tick-proof T. parva, it would be experimentally useful to be animal isolation rooms (pens) with a capacity able to feed ticks in a less physiologically and bio- for 16 cattle, tick culture and incubation rooms, logically complex situation than on the natural small-animal facilities and associated laboratory host. For R. appendiculatus, this problem was space. More important were and are the bio- tackled in the 1990s. Slightly earlier work at 374 P. Willadsen ILRI with A. variegatum showed that it was pos- smaller genomes, that of the North American sible to feed the ticks on rabbit or cattle skin tick, Ixodes scapularis (Hill and Wikel, 2005), membranes, with high carbon dioxide concen- which was of only indirect relevance to African trations and a temperature of 37°C being issues. By 2016, this first tick genome had been i mportant for success. Adult female and male thoroughly explored and now undoubtedly ticks were fed to engorgement over a period of up r epresents an important resource for ongoing to 16 days. All stages of the tick fed successfully research into all tick species (Gulia-Nuss et al., and although engorgement weights were less than 2016). By 2006, high in the priority list for tick on natural hosts, egg laying was successful. Ticks genomes was R. (Boophilus) microplus because of fed in vitro successfully transmitted Theileria its economic importance. In this case, the size of mutans and Ehrlichia ruminantium to cattle (Voigt the genome, considerably larger than the human et al., 1993). genome, was daunting (Ullmann et al., 2005; For a number of reasons, use of an artifi- Guerrero et al., 2006). Nevertheless, by 2010, cial membrane rather than animal skin is desir- considerable advances had been made in assem- able, but success with such membranes is easier bling sections of the genome. In this international to achieve for ticks with long mouth parts than effort, ILRI’s scientists played a valuable part for ticks with short ones, such as R. appendic- (Guerrero et al., 2010). ulatus (Young et al., 1996a). Nevertheless, in In the specific area of genome sequencing, 1995, successful feeding of nymphal R. appen- African tick species lagged behind. Nevertheless, diculatus on an artificial membrane was described. large quantities of interesting sequence data for When the system was used to feed nymphal African species were being accumulated. The ticks on blood infected with T. parva piroplasms, focus usually was on the salivary gland, a logical the prevalence of infection was high and com- choice. The pharmacologically and physiologic- parable with results achieved with ticks feed- ally active factors necessary for maintaining the ing on blood donor cattle (Waladde et al., tick’s attachment site pass through the salivary 1995). The status of such feeding systems was gland. The salivary gland is also the route of reviewed (Waladde et al., 1996; Young et al., transmission of tick-transmitted diseases as 1996a). well as often being a site for their development. In 2002, a complementary DNA (cDNA) library was generated from the salivary gland of feeding A. variegatum females. Sequencing of random Genomics and molecular biology clones from the library gave more than 2000 non-redundant sequences, 39% of which could By the middle of the first decade of the 21st cen- be tentatively identified with known proteins tury, the biological revolution precipitated by based on sequence similarities. Abundant fam- full-genome sequencing of many organisms was ilies of cement proteins, anti-haemostatics and well under way. Just as important as the genome also possible anti-inflammatories were identi- sequences themselves were the new experimen- fied, all proteins expected to be important to the tal approaches and techniques that evolved, success of tick feeding (Nene et al., 2002). drawing on the enormous quantities of new in- Somewhat more targeted approaches were formation. For ticks, the start was slower than adopted in two subsequent studies. In 2004, for many other organisms, including a variety of cDNA libraries were prepared from the salivary other parasites and disease organisms. The lower glands of R. appendiculatus infected with T. parva priority attached to veterinary diseases, particu- and from uninfected salivary glands. Over 9000 larly of developing economies, and the smaller sequences were collected from each sample, size of the research community were probably with the intention of identifying genes specific- two reasons for this. More significant, however, ally up- or downregulated as a result of the T. parva was the sheer size of the tick genomes. Depend- infection. No major differences between the ing on species, size estimates varied from one- abundantly expressed genes in the two samples third that of the human genome to more than were found, although the sequences themselves twice the size. The first full tick genome sequen- represent a repository of useful information cing project to be launched was for one of the (Nene et al., 2004). Ticks and Their Control 375 Secreted proteins are often identified by a piece of scientific equipment needed for such signal sequence, a relatively short peptide se- analyses is complex and expensive, while the quence necessary for their export from cells. A skill needed to operate it is considerable. The in- novel method using a ‘signal sequence trap’ was tuitive response to the idea of using MALDI-TOF used to identify possible secreted proteins from to identify ticks could well be that the idea is both R. appendiculatus and A. variegatum. Given scientifically interesting but impractical. How- the already extensive databases of known pro- ever, preparation of samples for MALDI-TOF teins existing in 2005, it was interesting that of analysis is simple and cheap, the instruments 61 R. appendiculatus sequences, only 15 could be are high throughput and analyses are com- tentatively identified. For A. variegatum, the pro- monly carried out in a specialized, central facil- portion was just one out of seven (Lambson et al., ity (in the case of this study in Japan). These 2005). This underlines, if such emphasis was factors together have the potential to change needed, the paucity of fundamental information the feasibility of using this technology. As re- about tick genes and proteins. ported, a collection of 398 African ticks of the The reason for the sheer size of the tick gen- genera Rhipicephalus, Rhipicephalus (Boophilus), omes continues to be of interest. A recent ILRI Hyalomma and Amblyomma, in total ten spe- shed some light on this question. Relatively small cies, were identified morphologically and by segments of high-molecular-weight DNA from cytochrome C oxidase DNA sequencing. Of R. appendiculatus were shown to have large num- these, 48 individual ticks were then used to con- bers of degenerate transposable elements. By struct a reference database of mass spectra. e xtrapolation, the authors suggested that there Subsequently, the remaining ticks were identi- could be about 65,000 copies of a single family fied by mass spectrometry using this database. of these repetitive elements; in basic terms, this Overall, a sensitivity of 96.1% and a specificity means that the size of the tick genomes could be, of 99.7% were achieved. The potential useful- in part, due to the accumulation of such elements ness of this methodology for field surveys is of unknown (if any) function over evolutionary considerable. time spans (Sunter et al., 2008). A recent and rather different ‘molecular’ study offers a new solution to an old problem. This chapter has already described historical Application of Research Capacity changes in tick distribution that affect the occurrence of tick-borne disease. These are not Ticks as vectors, with a focus on isolated instances. Rather, the speed, scale and R. appendiculatus as the vector of T. parva potential impact of such changes in tick distri- bution appear to have increased, while climate Ticks and their biology are inextricably linked change, animal movements, trade and the with the epidemiology, impact and control of tick- multiplicity of associated factors will probably borne diseases. The special complex of R. appen- ensure that such changes continue. diculatus and T. parva (i.e. ECF) has been central Coping with changes in tick distributions to ILRI’s animal health research from the insti- demands reliable knowledge of the identity and tute’s beginning. distribution of ticks. This requires the collection ECF is the subject of a major chapter in of large numbers of ticks across broad areas this book (Chapter 6). Given that, this section and then the accurate identification of species. will deal only with a few specific questions This has previously been done either morpho- where the dominant research issue concerned logically or using DNA-based analysis or a com- the biology of the tick vector rather than that bination of both. The former requires a trained of T. parva. Some aspects have been described scientist to distinguish closely related species. already. Disease epidemiology depends on DNA-based methods tend to demand expensive knowledge of the incidence and distribution of equipment and reagents, time and expertise. the tick vectors; ECF control in most areas relies Rothen et al. (2016) described an alternative, the heavily on tick control via the use of acaricides. use of matrix-a ssisted laser desorption/ionization ILRI’s contributions to both these areas have (MALDI-T OF) mass spectrometry. The essential been discussed. 376 P. Willadsen A major practical achievement by ILRI has stocks of the ILRI Tick Unit. Subsequently, the been its contribution to the development of an competence of seven different stocks of R. appen- infection-and-treatment method (ITM) vaccine diculatus and R. zambeziensis as vectors of two for ECF. The ability to produce such a vaccine de- different stocks of T. parva was examined (Ochan- pends on quality-controlled tick colonies and the da et al., 1998). Reproducible differences were technical ability to perform all stages of vaccine found. production (Patel et al., 2016). ILRI remains a In a slightly earlier paper with significance resource of knowledge and hands-on expertise for ECF epidemiology, the transmission of T. par- for this vaccine production. va by nymphal and adult R. appendiculatus was As with many other vector-borne diseases, examined using two isolates of T. parva. With in the case of ECF, the tick is not merely a bio- both, infection levels were much higher in adult logical syringe. Rather, it is the site of complex than nymphal ticks, a contributory factor being developmental changes in the T. parva organism the large difference in the number of requisite involving a range of specific and poorly understood structures, type III acini, in the salivary glands interactions between the two species. A number of the different tick instars (Ochanda et al., 1996). of papers in the early- to mid-1980s addressed The authors suggested that transmission by various aspects of the tick–T. parva interaction. nymphal ticks, representing a lower challenge Some were practical. Irvin et al. (1981) estab- dose of T. parva, might cause milder infections lished a rapid method for staining salivary and lead to immunity in the cattle host rather glands of R. appendiculatus infected with T. parva, than death, a factor in the establishment of en- and then used it to follow the dynamics of infec- demic stability. Field evidence from Zimbabwe tion and parasite maturation. Later, a method of was consistent with this hypothesis. transplanting T. parva kinetes to establish infec- Then, in 2009, there was a re-examination tion in a single tick salivary gland acinus was de- of the differences between ticks using the same scribed, a method that the authors suggested Muguga and Kiambu isolates, although this time could be applied to the development of cloned the differences were examined by the modern parasites (Fujisaki et al., 1988). Other research techniques of quantitative and nested polymer- was more fundamental. The site of T. parva ase chain reaction (PCR) (Odongo et al., 2009). developmental changes, the type III acinus in These techniques were shown to be more sensi- the tick salivary gland, was first examined (Faw- tive than traditional microscopy, and it was con- cett et al., 1981) before the ultrastructure of the firmed that, for the Kiambu isolate, a higher sporogony of the parasite in the salivary gland proportion of ticks became infected and parasite was investigated (Fawcett et al., 1982). Then, in numbers within adult salivary glands were also 1993, it was shown that both a crude extract of higher. It is interesting that these differences tick salivary gland and interleukin-2 as a single, between tick isolates persisted through an add- pure cytokine could enhance the susceptibility itional 20 years of continuous tick culture. These of bovine lymphocytes to infection by T. parva observations would make possible the examin- sporozoites (Shaw et al., 1993). In a general sense, ation at the level of gene expression of factors this ability of salivary gland material from the tick that might explain the differences in T. parva vector to facilitate infection was subsequently susceptibility. rediscovered for a number of tick-transmitted Finally, an older paper exemplifies the use viral diseases (Nuttall et al., 2008). that can be made of large databases in under- The fact that vector competence of R. appen- standing disease. Much information had been diculatus varies with tick isolate and that these recorded on the transmission of the Muguga sta- differences are heritable has been examined bilate of T. parva to Boran cattle using two iso- several times. In 1995, estimates of heritability lates of R. appendiculatus. Between 1986 and were obtained for T. parva infecting two tick stocks, 1991, 1241 records of tick batches harvested Kiambu and Muguga (Young et al., 1995), and from 286 infected cattle had been compiled, the point was made that the heritabilities were from which a total of 812 records were selected high enough to allow selection of tick strains for as suitable for analysis. This database was inter- high and low susceptibility to infection. This was, rogated using statistical modelling procedures in fact, done, as noted in the discussion of the tick searching for relationships between a long list of Ticks and Their Control 377 factors and the prevalence, abundance and in- 12S ribosomal RNA and cytochrome C oxidase tensity of salivary gland infection in both female subunit 1, as well as a third gene that, ultim- and male ticks (Young et al., 1996b). Twenty- ately, was not informative. Analysis of sequence four factors or interactions were found to be stat- variation in these genes in ticks sampled from istically significant. Some factors might have different geographical locations, from different been intuitively expected; others were more hosts and from laboratory and field isolates (the unexpected. Consistently among the most im- three chief variables examined) identified 28 portant factors, however, were the piroplasm haplotypes clustering into two haplogroups. levels on the day of harvesting ticks and the These groups could not be defined by geograph- month in which the ticks were harvested. The lat- ical origin, host species or the laboratory/field ter finding is surprising, although a tentative divide. Based on these observations, the au- correlation between time of nymphal engorge- thors suggested that two major genetic groups ment and the climatic conditions (temperature of R. appendiculatus existed in Kenya, with pos- and humidity) at the time was noted. The final sible broader distribution in eastern and south- models were complex and with significant error ern Africa. It was considered possible that the margins, with the consequence that they had existence of two genetic groups could have limited predictive value. There does not appear had implications for the spread of the tick and to have been subsequent follow-up using this for the transmission dynamics of ECF (Kanduma statistical approach. et al., 2016a). The authors returned to the issue in a second publication in the same year (Kanduma et al., 2016b). Ticks from ten locations in Kenya The genetic complexity were collected, associated with three grazing of R. appendiculatus systems: cattle, cattle and wildlife co-grazing, and wildlife without livestock. Numerous individual As described above, it had long been known that ticks from ten closed laboratory colonies plus strains of R. appendiculatus had various capaci- ticks of five other Rhipicephalus species were in- ties to act as vectors of ECF and that this cap- cluded in the analysis. On this occasion, the con- acity was heritable. This raises the question of clusions appeared to be somewhat different. There the genetic complexity of this tick species and was a low degree of genetic differentiation in the the extent to which it relates, for example, to field samples, and no relationship, as before, be- geographical distribution. tween the genetic diversity and either geograph- Kanduma et al. (2012) utilized the database ical location or host species. The ten laboratory of expressed genes coding for salivary gland pro- strains, as well as the other species, were strongly teins developed by Nene et al. (2004) to identify differentiated. In addition, some of the labora- 29 polymorphic markers. These they then used tory strains were differentiated from current to discriminate among populations of R. appen- field samples taken from the same geographical diculatus and among R. appendiculatus and four point of origin as the original laboratory strain. other Rhipicephalus spp. using ten field popula- The key conclusion, in terms of disease trans- tions and ten laboratory stocks (Kanduma mission and control, was that there was in fact et al., 2012). Distinguishing R. zambeziensis from little genetic diversity in the R. appendiculatus R. appendiculatus was difficult, although the populations, perhaps as a result of livestock and other species were satisfactorily separated. Within wildlife movements through the country. This R. appendiculatus, clustering into two popula- conclusion can also be compared with the results tions occurred, which did not, however, correl- of analysis of the diversity of protein sequences ate with a field/laboratory culture division. The in a single protein, Ra86, which will be discussed authors proposed two preferred sets of markers, in the following section on vaccines. Even more one optimal for distinguishing between species recently, an array of gene markers was used to and the second for intraspecies comparisons. look at the broader phylogeny of Rhipicephalus Subsequently, the question of genetic diver- and R. (Boophilus) species, with the results dem- sity in R. appendiculatus populations was addressed onstrating once again that much remains to be again, this time using two mitochondrial genes, resolved (Kanduma et al., 2019). 378 P. Willadsen Tick vaccines sanguineus (Sabadin et al., 2017). This is not only positive news for control of the tick of greatest There has been occasional interest in the devel- interest to ILRI but also another example of the opment of tick vaccines since the early days of unpredictability of cross-immunity among tick ILRAD, although initially this was as a minor species, something explored in greater detail by contributor to antigen discovery research initi- ILRI and other laboratories with respect to the ated elsewhere. Certainly, this was the case with Bm86 antigen. an anti-tick vaccine effort driven by the Inter- Rather more effort went into exploring the national Centre of Insect Physiology and Ecol- potential of Bm86 homologues in the control of ogy (ICIPE) in Nairobi (Mongi et al., 1986). endemic African tick species. This was (and From the late 1990s onwards, there was a remains) a tantalizing scientific and practical more consistent interest in the field. Three factors issue. It was shown early on that the sequence of account for this. First, as had long been the case the Bm86 molecule was quite strongly conserved for all parasite vaccines, the ability to express across tick genera and species, although with recombinant proteins offered at least a potential variation (Willadsen, 2004). Even within isolates pathway to practical application of antigen dis- of R. (Boophilus) microplus itself, sequence differ- covery programmes. This was as true of ticks as ences of up to about 5% have been reported. it was of ECF. Second, the steady accumulation Sequence differences did not, however, translate of tick genomic information, referred to previ- into differences in protection against tick infest- ously, was a useful source of novel information. ation in any predictable way. The most striking Third, in 1988 and 1989, an Australian group example was Rhipicephalus (Boophilus) annulatus, patented an antigen from the tick R. (Boophilus) a species closely related to R. (Boophilus) mi- microplus that, when expressed as a recombinant croplus. Vaccination with recombinant Bm86 protein in a commercially viable way, gave useful typically gave 80–90% protection against R. protection against field infestations of ticks. Soon (Boophilus) microplus on cattle, but effectively two commercial vaccine developments were under total protection against R. (Boophilus) annulatus, way, one in Australia and the second in Cuba, protection that was also sustained for a longer leading to the release in the early 1990s of two period. The reason is still unclear. Practically, vaccines, TickGARD and GAVAC (Willadsen, 2004). however, the obvious question was: what is the There were several aspects to ILRI’s re- protection of Bm86 against R. (Boophilus) de- sponse. The Australian work gave a stimulus to coloratus or R. appendiculatus? antigen discovery in a number of laboratories First, it was shown by Odongo et al. (2007) and countries. ILRI collaborated, although as that antisera to Bm86 reacted with the R. a minor partner, in the evaluation of several (Boophilus) decoloratus tick gut, the location of antigens, for example, work on the recombinant the antigen in R. (Boophilus) microplus, and that p29 antigen from Haemaphysalis longicornis R. (Boophilus) decoloratus feeding on vaccinated (Mulenga et al., 1999) and other work that re- cattle showed vaccination effects similar to those mains unpublished. ILRI has also engaged in recorded for R. (Boophilus) microplus on Tick- de novo antigen discovery. An early effort led to GARD-vaccinated cattle: reduced tick numbers the identification of a cement protein, RIM36, and reduced overall fertility, measured as total from R. appendiculatus (Bishop et al., 2002). This egg production, reductions of 46 and 61%, re- was shown to be the target of a strong antibody spectively. Although less than the effects seen response by cattle naturally exposed to the tick but, with R. (Boophilus) microplus, these results were apparently, was of little use as a protective anti- certainly encouraging. The presence of two vari- gen. ILRI has contributed to ongoing, basic re- ants of Bm86 was shown, Bd86-1 and Bd86-2, search on tick proteins (Costa et al., 2017; Seixas with amino acid sequences that were 86% and et al., 2018), some of which had no obvious con- 85% identical to the original Bm86. Recombin- nection to vaccine research. Recently and more ant Bd86-1 reacted strongly with antisera from encouragingly, it was shown that recombinant TickGARD-vaccinated cattle. Two linear peptide glutathione S-transferase from H. longicornis in- epitopes shared between the Bd86 molecules duced 67% protection against R. appendiculatus, and Bm86 were identified that, it was suggested, although it was ineffective against Rhipicephalus could be the basis of a peptide-based vaccine. Ticks and Their Control 379 Unfortunately, the Bm86 vaccine had no ef- levels of ticks fed on T. parva-infected cattle fect on feeding R. appendiculatus. The obvious v accinated with Ra86 compared with non- question was whether a sequence difference be- Ra86-vaccinated controls. Clearly, a more tween the Bm86 antigen used for the vaccination efficacious vaccine would be necessary, as such trials and the homologues in R. appendiculatus, a vaccine, even if theoretically useful, would be named Ra86, was sufficient to explain the lack of unattractive to farmers. protection against that tick species. One problem A recent study examined the potential of an was the presence of at least two variants of Ra86, antigen cocktail to give protection against tick which themselves showed only 80% amino acid infestation and/or T. parva transmission (Olds sequence identity, exceptionally divergent for the et al., 2016). The cocktail used antigens selected ‘same’ protein (Kamau et al., 2011). Evidence was from the literature: the tick antigens chosen obtained that both variants could be present in a were subolesin, TRP64 (the cement protein from single genome but that when both were present, one R. appendiculatus) and three tick histamine-binding of the two would be transcriptionally dominant. proteins, while from T. parva, the sporozoite anti- Further research showed the situation to be gen p67C was selected. All of these antigens had even more complex. Nineteen Ra86 sequences separately shown promise in vaccination trials, were obtained from the laboratory Muguga although the tick species, host and challenge strain of R. appendiculatus, defining two alleles model all varied. In this trial, cattle were vaccin- differentiated by insertions or deletions (indels) ated, produced good antibody responses and and different in length by 39 amino acids. Then were challenged with the normal Muguga iso- a further 20 sequences of Ra86 were obtained late of R. appendiculatus and the Muguga ‘low- from each of four field sites in central and west- line’ ticks that had been infected with T. parva. ern Kenya, revealing a further three different No significant effects on either tick engorgement size types, differentiated by 39–49 amino acid and fertility or disease transmission were found. indels and hence a total of five indel-defined There were, however, some differences in the genotypes. The longest sequence was found susceptibility of the two isolates of tick used, only in the laboratory strain. Analysis clustered emphasizing the potential importance of isolate all Ra86 sequences and Bm86 into four major variation in the field. The authors stressed the clades based on amino acid substitutions. Al- desirability, in future work, of early assessment though there was evidence that selection con- of any tick antigen in the natural tick–host tributed to the sequence variation, there was no relationship. evidence that the groupings correlated with geo- Another aspect of this deserves to be noted. graphical separation of tick populations (Kamau The idea of antigen cocktails as a means to im- et al., 2016). prove the efficacy of vaccines is frequently With such diversity – which has not been touted, not only for tick vaccines but for many reported in R. (Boophilus) microplus in Austra- other anti-parasite vaccines as well, to the extent lia or Central and South America – is the idea that it has become an important but poorly of vaccination using a Bm86 homologue im- acknowledged rationale for much antigen re- practical? Olds et al. (2012) vaccinated cattle search. The concept is experimentally testable, with a mixture of two recombinant forms of but reports of tests are scarce and those of suc- Ra86 and challenged them with Muguga cess even scarcer (Willadsen, 2008). The fact strain ticks. There was no significant effect on that ILRI carried out such a test, and reported adult mortality, engorgement weight or weight failure, is a worthwhile contribution. of eggs laid, although there was an interesting The tantalizing dream of tick vaccine re- decrease in egg hatching, which grew more search is a vaccine that protects against multiple pronounced in late-laid eggs. There was a species. The partial cross-protection induced by slight but statistically significant decrease in Bm86, despite the variability of protection across moulting of nymphs to adults. These figures, species, seemed to offer some hope. This was ex- incorporated into a tick population model, plored further in a study that used a peptide from showed the potential for a useful, although ex- Bd86 to raise monoclonal antibodies that were tremely gradual, decline in tick numbers. found to cross-react with R. (Boophilus) microplus, There was also a slight decline in the infection R. (Boophilus) decoloratus, R. appendiculatus and 380 P. Willadsen Hyalomma anatolicum anatolicum (Kopp et al., records of R. (Boophilus) microplus in Africa, 2009). The degree to which such cross-reactivity all in the south-east. West Africa was still con- translates into protection is unknown. sidered free of the tick. Then, in 2007, its discov- ery was reported in Ivory Coast as a result of an accidental introduction (Madder et al., 2007). Recent Developments A second independent introduction was found to have occurred in Benin in about 2005. Since As the 21st century progressed, ILRI’s interest in then, it has continued to spread (Madder et al., ticks, with the exception of molecular approaches 2011, 2012; de Clercq et al., 2012), apparently to tick biology and vaccine development, appeared displacing other Boophilus spp. as it invades. The to decline. The result was that pointers to the fu- tick has now spread throughout Ivory Coast and ture, when they came, were from other institutions far into the north of Benin. In November 2012, and scientists. Three issues emerged: the impact R. (Boophilus) microplus was discovered as in- of climate change, the spread of R. (Boophilus) tense field infestations in south-western Burkina microplus through East and southern Africa and Faso and Mali (Adakal et al., 2013). It continues its accidental introduction into West Africa, to spread within Burkina Faso. Its spread has and finally the evidence of significant acaricide been accompanied by decreased milk production, resistance in African tick species. uncontrolled tick populations, inappropriate Today, any discussion of the future of disease acaricide use and cattle deaths (Madder et al., control is likely to begin with the anticipated im- 2011; Adakal et al., 2013). pacts of climate change. The idea, supported by Surveys in Cameroon in 2013 at a number a growing body of evidence, that shifts in vector of sites, involving the identification of 20,000 populations will greatly change the epidemi- ticks, showed that R. (Boophilus) microplus ology of many diseases has been well explored. seemed not to have reached that country. How- Changes to tick distributions and hence, for ex- ever, there appeared then to be no ecological ample, to ECF are expected (Grace et al., 2015). reason why it would not eventually (and prob- ILRI had earlier played a role in understand- ably rapidly) spread through much of West ing the dynamic nature of ticks and tick-borne Africa, south of the Sahel (de Clercq et al., disease. The strongest indication that this would 2013, 2015)1. This expectation has regret- be an ongoing if not escalating problem came tably been confirmed. Recent evidence, which from R. (Boophilus) microplus. In contrast to other has involved several ILRI scientists, has shown tropical and subtropical parts of the world includ- that the species is now present in large num- ing Australia, Central and South America and bers throughout much of Cameroon, apparently large parts of south and eastern Asia, Africa was displacing R. (Boophilus) decoloratus as it spreads until recently fortunate in being spared any major (Silatsa et al., 2019a,b). impact from this tick species. The situation has In more detail, R. (Boophilus) microplus has changed over the last two decades. From small become the dominant tick species on cattle in populations in eastern South Africa (Natal), the south-western Burkina Faso and southern tick has spread through much of South Africa, Benin. A full year survey at three sites in Benin displacing endemic R. (Boophilus) decoloratus and three in Burkina Faso showed that the abun- (Tønnesen et al., 2004). In Tanzania, tick surveys dance of R. (Boophilus) microplus was over 50% conducted between 1998 and 2001 were com- of all ticks collected of all species at five of the pared with historical (40-year-old) data, with the six sites. When tick counts were averaged over results showing that, while R. (Boophilus) decol- the full year, the daily numbers of R. (Boophilus) oratus had largely retreated to high-altitude areas microplus collected were Gogounou (38), Ouan- in northern and central Tanzania, R. (Boophilus) golodougou (113), Farnifaso (164), Okpara microplus had invaded all but the driest and cold- (249), Kpinnou (465) and Kimini (710). At est parts of the country (Lynen et al., 2008). This Gogounou, in the northern part of Benin, A. var- seemed not to have happened by 2007 in iegatum was the most abundant species at 26% of Rwanda (Bazarusanga et al., 2007). the total count, although the relative abundance In a review published in 2006, Estrada- of R. (Boophilus) microplus was almost identical Pena et al. (2006) identified only limited confirmed at 25%, suggesting that R. (Boophilus) microplus Ticks and Their Control 381 is continuing to expand its range towards drier Clearly, under the conditions of cattle farming in areas. Ongoing spread along the tropical, wetter Uganda, this has not been the case. near-coastal regions is, of course, almost certain to occur. No scientific studies have as yet been carried The Future out to measure the impact of such tick numbers on local cattle. However, based on Australian Given the diversity of tick-related problems, experience, the blood loss from a daily tick count where might ILRI best make its future research of 100 or above would have significant effects on contributions? Its greatest competitive advan- productivity, and the higher numbers could easily tage over most other research providers is the result in mortality, particularly at times of nutri- close juxtaposition of facilities for large-animal tional stress. Anecdotal evidence from farmers experimentation and parasite culture with la- confirms this expectation. boratories capable of advanced molecular re- The next evolving problem is that of the search. This is more than a huge experimental spread of acaricide resistance. In work led from advantage. In a more indefinable but still im- Obihiro University of Agriculture and Veterin- portant way, it brings together the molecular sci- ary Medicine in Hokkaido, Japan, and involving entist and the livestock target of the scientist’s a number of collaborating institutions in research. This juxtaposition could be utilized in Uganda, ticks were collected from 30 farms many ways, but three examples suffice. First, as across Uganda, all having a history of acaricide is clear throughout this chapter, current tick failure (Vudriko et al., 2016). Such reported and tick-borne disease control depends heavily failures can be due to acaricide resistance, as is on synthetic acaricides. Resistance to these is an usually assumed, but other factors are com- increasing problem. No effective resistance man- monly part of the explanation. The major tick agement strategy can be developed until resist- species were R. appendiculatus and R. (Boophi- ance can be rapidly and accurately diagnosed, lus) decoloratus. Acaricide resistance assays which itself presupposes an understanding of showed that 90% of the samples were resistant resistance mechanisms. Current methods of to synthetic pyrethroids and 60% of the total diagnosis have not changed significantly in dec- were highly resistant, i.e. the acaricides had ades, are often slow and are too frequently in- effectively no lethal activity. Resistance was high accurate. To do better is a challenging research against a combined organophosphate/syn- problem. thetic pyrethroid product and was significant The second area where the close associ- against organophosphates and amitraz. Resist- ation of scientist/laboratory/experimental ani- ance to multiple acaricides was detected in mal is beneficial is in vaccine research. This about half of the samples. Over a 2-year period, is as true of anti-tick vaccines as it is of ECF three-q uarters of the farms had used two or vaccines. Interest in this approach to tick more acaricides, and of these, over half had control for African ticks is developing. Despite used chemical rotations that made no sense ILRI’s worthwhile contributions to this area, scientifically. the institute has always lacked a tick vaccine The work is significant for a number of programme. As described above, the wealth reasons. Most obvious is the worryingly high of genomic information and associated experi- frequency of resistance and the occurrence of mental techniques together offer abundant new multiple resistance to diverse chemical groups. Se- opportunities and approaches (de la Fuente cond, the study underlines the importance of in- et al., 2016). If the promise of these is yet to be correct acaricide usage and the frequency with convincingly realized, it is undoubtedly true which it happens. Third, there is the fact that re- that the combination of new science, trad- sistance to acaricides was found in over 40% of itional biology and large-animal evaluation is a the R. appendiculatus samples. It was once a com- powerful combination that ILRI is well placed mon speculation that resistance would evolve to exploit. slowly, if at all, on multi-host ticks compared with A third possibility is the investigation of the single-host Boophilus spp., with the alternative the genetics of within-breed and between-breed hosts perhaps providing a chemical-free refuge. variation in the ability of livestock, especially 382 P. Willadsen cattle, to become resistant to ticks. This has borne disease requires an understanding of tick been a major component of the management distributions and economic impacts; the current of R. (Boophilus) microplus both in Australia and future effects of climate change; and the and South America. In the African context, regulatory system and the production environ- questions of fundamental scientific interest ment in which tick control is to be applied. ILRI and practical i mportance are unanswered. For has skills in all of these areas and a strong focus example, to what degree are indigenous cattle on at least one target group, the smallholder resistant to ticks? Are there between-animal livestock farmer. Thus, ILRI has, in principle, not differences in heritability high enough to be only the potential to develop new technologies useful? To what degree is the acquisition of but also the infrastructure and experience to resistance effective against multiple tick spe- facilitate their effective adoption. cies? ILRI possesses the scientific skills, the biological resources and, importantly, the in- frastructure and field stations to address such questions. Acknowledgement Each of these options would be about devel- oping technologies, at best partial solutions to The author thanks Naftali Githaka and Stephen practical problems. As has been described above, Mwaura for their assistance with information a robust solution to the control of ticks and tick- concerning tick research in ILRI. Note 1 The TickRisk project, led by Maxime Madder and Eva De Clerq and based largely in Benin, and the We- catiC project, led by Hassane Adakal, have provided the above data, mostly collected during 2012–2013. References Adakal, H., Biguezoton, A., Zoungrana, S., Courtin, F., de Clercq, E.M. et al. (2013) Alarming spread of the Asian cattle tick Rhipicephalus microplus in West Africa – another three countries are affected: Burkina Faso, Mali and Togo. Experimental and Applied Acarology 61, 383–386. Barker, S.C. and Murrell, A. (2004) Systematics and evolution of ticks with a list of valid genus and species names. Parasitology 129 (Suppl. 1), S15–S36. Bazarusanga, T., Geysen, D., Vercruysse, J. and Madder, M. (2007) An update on the ecological distribution of Ixodid ticks infesting cattle in Rwanda: countrywide cross-sectional survey in the wet and the dry season. Experimental and Applied Acarology 43, 279–291. Bishop, R., Lambson, B., Wells, C., Pandit, P., Osaso, J., et al. (2002) A cement protein of the tick Rhipi- cephalus appendiculatus, located in the secretory e cell granules of the type III salivary gland acini, induces strong antibody responses in cattle. International Journal for Parasitology 32, 833–842. Costa, E.P., Façanha, A.R., Cruz, C.S., Silva, J.N., Machado, J.A., et al. (2017) A novel mechanism of func- tional cooperativity regulation by thiol redox status in a dimeric inorganic pyrophosphatase. Bio- chimica et Biophysica Acta 1861, 2922–2933. de Castro, J.J. (1997) Sustainable tick and tickborne disease control in livestock improvement in developing countries. Veterinary Parasitology 71, 77–97. de Castro, J.J., Young, A.A.S., Dransfield, R.R.D., Cunningham, M.P.M. and Dolan, T.T.T. (1985a) Effects of tick infestation on Boran (Bos indicus) cattle immunised against theileriosis in an endemic area of Kenya. Research in Veterinary Science 39, 279–288. de Castro, J.J., Cunningham, M.P., Dolan, T.T., Dransfield, R.D., Newson, R.M., et al. (1985b) Effects on cattle of artificial infestations with the tick Rhipicephalus appendiculatus. Parasitology 90, 21–33. de Clercq, E.M., Vanwambeke, S.O., Sungirai, M., Adehan, S., Lokossou, R., et al. (2012) Geographic dis- tribution of the invasive cattle tick Rhipicephalus microplus, a country-wide survey in Benin. Experimental and Applied Acarology 58, 441–452. de Clercq, E.M., Estrada-Peña, A., Adehan, S., Madder, M. and Vanwambeke, S.O. (2013) An update on distribution models for Rhipicephalus microplus in West Africa. Geospatial Health 8, 301–308. Ticks and Their Control 383 de Clercq, E.M., Leta, S., Estrada-Peña, A., Madder, M., Adehan, S., et al. (2015) Species distribution mod- elling for Rhipicephalus microplus (Acari: Ixodidae) in Benin, West Africa: comparing datasets and modelling algorithms. Preventive Veterinary Medicine 118, 8–21. de la Fuente, J., Kopácek, P., Lew-Tabor, A. and Maritz-Oliver, C. (2016) Strategies for new and improved vaccines against ticks and tick-borne diseases. Parasite Immunology 38, 754–769. Estrada-Peña, A., Bouattour, A., Camicas, J.-L., Guglielmone, A., Horak, I., et al. (2006) The known distri- bution and ecological preferences of the tick subgenus Boophilus (Acari: Ixodidae) in Africa and Latin America. Experimental and Applied Acarology 38, 219–235. Fawcett, D.W., Doxsey, S. and Büscher, G. (1981) Salivary gland of the tick vector (R. appendiculatus) of East Coast Fever. II. Cellular basis for fluid secretion in the type III acinus. Tissue Cell 13, 231–253. Fawcett, D.W., Büscher, G. and Doxsey, S. (1982) Salivary gland of the tick vector of East Coast Fever. III. The ultrastructure of sporogony in Theileria parva. Tissue Cell 14, 183–206. Fujisaki, K., Irvin, A.D., Voigt, W.P., Leitch, B.L. and Mozaria, S.P. (1988) The establishment of infection in the salivary glands of Rhipicephalus appendiculatus ticks by transplantation of kinetes of Theileria parva and the potential use of the method for parasite cloning. International Journal for Parasitology 18, 75–78. Grace, D., Bett, B., Lindahl, J.F. and Robinson, T.P. (2015) Climate and livestock disease: assessing the vulnerability of agricultural systems to livestock pests under climate change scenarios. CCAFS Work- ing Paper No. 116. CCAFS, Copenhagen. Guerrero, F.D., Nene, V.M., George, J.E., Barker, S.C. and Willadsen, P. (2006) Sequencing a new target gen- ome: the Boophilus microplus (Acari: Ixodidae) genome project. Journal of Medical Entomology 43. 9–16. Guerrero, F.D., Moolhuijzen, P., Peterson, D.G., Bidwell, S., Caler, E., et al. (2010) Reassociation kinetics- based approach for partial genome sequencing of the cattle tick, Rhipicephalus (Boophilus) microplus. BMC Genomics 11, 374. Gulia-Nuss, M., Nuss, A.B., Meyer, J.M., Sonenshine, D.E., Roe, R.M. (2016) Genomic insights into the Ixodes scapularis tick vector of Lyme disease. Nature Communications 7, 10507. Henson, S., Bishop, R.P., Morzaria, S., Spooner, P.R., Pelle, R., et al. (2012) High-resolution genotyping and mapping of recombination and gene conversion in the protozoan Theileria parva using whole genome sequencing. BMC Genomics 13, 503. Hill, C.A. and Wikel, S.K. (2005) The Ixodes scapularis genome project: an opportunity for advancing tick research. Trends in Parasitology 21, 151–153. Irvin, A.S., Boarer, C.D., Dobbelaere, D.A., Mahan, S.M., Masake, R., et al. (1981) Monitoring Theileria parva infection in adult Rhipicephalus appendiculatus ticks. Parasitology 82, 137–147. Kamau, L., Skilton, R.A., Odongo, D.O., Mwaura, S., Githaka, N., et al. (2011) Differential transcription of two highly divergent gut-expressed Bm86 antigen gene homologues in the tick Rhipicephalus appen- diculatus (Acari: Ixodida). Insect Molecular Biology 20, 105–114. Kamau, L.M., Skilton, R.A., Githaka, N., Kiara, H., Kabiru, E., et al. (2016) Extensive polymorphism of Ra86 genes in field populations of Rhipicephalus appendiculatus from Kenya. Ticks and Tick-borne Dis- eases 7, 772–781. Kamidi, R.E. and Kamidi, M.K. (2005) Effects of a novel pesticide resistance management strategy on tick control in a smallholding exotic-breed dairy herd in Kenya. Tropical Animal Health and Production 37, 469–478. Kanduma, E.G. Mwacharo, J.M., Sunter, J.D., Nzuki, I., Mwaura, S., et al. (2012) Micro- and minisatellite- expressed sequence tag (EST) markers discriminate between populations of Rhipicephalus appen- diculatus. Ticks and Tick-borne Diseases 3, 128–136. Kanduma, E.G. Mwacharo, J.M., Githaka, N.W., Kinyanjui, P.W., Njuguna, J.N., et al. (2016a) Analyses of mitochondrial genes reveal two sympatric but genetically divergent lineages of Rhipicephalus appen- diculatus in Kenya. Parasites & Vectors 9, 353. Kanduma, E.G. Mwacharo, J.M., Mwaura, S., Njuguna, J.N., Nzuki, I., et al. (2016b) Multi-locus genotyping reveals absence of genetic structure in field populations of the brown ear tick (Rhipicephalus appen- diculatus) in Kenya. Ticks and Tick-borne Diseases 7, 26–35. Kanduma, E.G., Bishop, R.P., Githaka, N.W., Skilton, R.A., Heyne, H., et al. (2019) Mitochondrial and nuclear multilocus phylogeny of Rhipicephalus ticks from Kenya. Molecular Phylogenetics and Evolution 140, 106579. Kopp, N., Diaz, D., Amacker, M., Odongo, D.O., Beier, K., et al. (2009). Identification of a synthetic peptide inducing cross-reactive antibodies binding to Rhipicephalus (Boophilus) decoloratus, Rhipicephalus 384 P. Willadsen (Boophilus) microplus, Hyalomma anatolicum and Rhipicephalus appendiculatus BM86 homologues. Vaccine 28, 261–269. Lambson, B., Nene, V., Obura, M., Shah, T., Pandit, P., et al. (2005) Identification of candidate sialome components expressed in ixodid tick salivary glands using secretion signal complementation in mammalian cells. Insect Molecular Biology 14, 403–414. Lawrence, J.A., Foggin, C.M. and Norval, R.A. (1980). The effects of war on the control of diseases of livestock in Rhodesia (Zimbabwe). Veterinary Record 107, 82–85. Lessard, P., L’Eplattenier, R., Norval, R.A., Kundert, K., Dolan, T.T., et al. (1990) Geographical information systems for studying the epidemiology of cattle diseases caused by Theileria parva. Veterinary Record 126, 255–262. Lynen, G., Zeman, P., Bakuname, C., Di Giulio, G., Mtui, P., et al. (2008). Shifts in the distributional ranges of Boophilus ticks in Tanzania: evidence that a parapatric boundary between Boophilus microplus and B. decoloratus follows climate gradients. Experimental and Applied Acarology 44, 147–164. Madder, M., Thys, E., Geysen, D., Baudoux, C. and Horak, I. (2007) Boophilus microplus ticks found in West Africa. Experimental and Applied Acarology 43, 233–234. Madder, M., Thys, E., Achi, L., Touré, A. and de Deken, R. (2011) Rhipicephalus (Boophilus) microplus: a most successful invasive tick species in West-Africa. Experimental and Applied Acarology 53, 139–145. Madder, M., Adehan, S., de Deken, R., Adehan, R. and Lokossou, R. (2012) New foci of Rhipicephalus microplus in West Africa. Experimental and Applied Acarology 56, 385–390. Mekonnen, S., Kgasi, A., Mureithi, W., Zena, G., Tekle, T., et al. (2004) In vivo and in vitro evaluation of the efficacy of cypermethrin high-cis (ECOTOMIN) against economically important cattle ticks in Ethiopia. Ethiopian Veterinary Journal 8, 29–38. Minjauw, B. and McLeod, A. (2003) Tick-borne diseases and poverty. The impact of ticks and tick-borne diseases on the livelihood of small-scale and marginal livestock owners in India and eastern and southern Africa. Research report. DFID Animal Health Programme, Centre for Tropical Veterinary Medicine, University of Edinburgh, UK. Mongi, A.O., Shapiro, S.Z., Doyle, J.J. and Cunningham, M.P. (1986) Characterization of antigens from extracts of fed ticks using sera from rabbits immunized with extracted tick antigen and by successive tick infestation. Insect Science and Its Application 7, 479–487. Morzaria, S.P., Irvin, A.D., Wathanga, J., d’Souza, D., Katende, J., et al. (1988) The effect of East Coast fever immunisation and different acaricidal treatments on the productivity of beef cattle. Veterinary Record 123, 313–320. Mulenga, A., Sugimoto, C., Sako, Y., Ohashi, K., Mozaria, S., et al. (1999). Molecular characterization of a Haemaphysalis longicornis tick salivary gland associated 29 kilodalton protein and its vaccine effect against tick infestation in rabbits. Infection and Immunity 67, 1652–1658. Muraguri, G.R., McLeod, A. and McDermott, J.J. (2003) Efficacy of a deltamethrin-based pour-on in the control of tick-borne diseases and trypanosomosis in Kwale District, Kenya. Insect Science and Its Application 23, 69–74. Nene, V., Lee, D., Kang’A, S., Skilton, R., Shah, T., et al. (2004) Genes transcribed in the salivary glands of female Rhipicephalus appendiculatus ticks infected with Theileria parva. Insect Biochemistry and Molecular Biology 34, 1117–1128. Nene, V., Lee, D., Quackenbush, J., Skilton, R., Mwaura, S., et al. (2002) AvGI, an index of genes transcribed in the salivary glands of the ixodid tick Amblyomma variegatum. International Journal for Parasitology 32, 1447–1456. Norval, R.A. (1979) Tick infestations and tick-borne diseases in Zimbabwe Rhodesia. Journal of the South African Veterinary Association 50, 289–292. Norval, R.A.I. and Perry, B.D. (1990) Introduction, spread and subsequent disappearance of the brown ear- tick, Rhipicephalus appendiculatus, from the southern lowveld of Zimbabwe. Experimental and Ap- plied Acarology 9, 103–111. Norval, R.A.I., Perry, B.D. and Hargreaves, S.K. (1992) Tick and tick-borne disease control in Zimbabwe: what might the future hold? Zimbabwe Veterinary Journal 23, 1–15. Norval, R.A., Perry, B.D., Meltzer, M.I., Kruska, R.L. and Booth, T.H. (1994) Factors affecting the distribu- tions of the ticks Amblyomma hebraeum and A. variegatum in Zimbabwe: implications of reduced acaricide usage. Experimental and Applied Acarology 9, 383–407. Nuttall, P.A., Labuda, M. and Bowman, A.S. (2008) Saliva-assisted transmission of tick-borne pathogens. In: Bowman, A.S. and Nuttall, P.A. (eds) Ticks: Biology, Disease and Control. Cambridge University Press. Cambridge, pp. 205–219. Ticks and Their Control 385 Ochanda, H., Young, A.S., Wells, C., Medley, G.F. and Perry, B.D. (1996) Comparison of the transmission of Theileria parva between different instars of Rhipicephalus appendiculatus. Parasitology 113, 243–253. Ochanda, H., Young, A.S., Medley, G.F. and Perry, B.D. (1998) Vector competence of 7 rhipicephalid tick stocks in transmitting 2 Theileria parva parasite stocks from Kenya and Zimbabwe. Parasitology 116, 539–545. Odongo, D., Kamau, L., Skilton, R., Mwaura, S., Nitsch, C., et al. (2007) Vaccination of cattle with TickGARD induces cross-reactive antibodies binding to conserved linear peptides of Bm86 homologues in Boophilus decoloratus. Vaccine 25, 1287–1296. Odongo, D.O., Ueti, M.W., Mwaura, S.N., Knowles, D.P., Bishop, R.P., et al. (2009) Quantification of Theileria parva in Rhipicephalus appendiculatus (Acari: Ixodidae) confirms differences in infection between selected tick strains. Journal of Medical Entomology 46, 888–894. Okello-Onen, J., Mukhebi, A., Tukahirwa, E., Musisi, G., Bode, E., et al. (1998) Financial analysis of dipping strategies for indigenous cattle under ranch conditions in Uganda. Preventive Veterinary Medi- cine 33, 241–250. Okello-Onen, J., Tukahirwa, E.M., Perry, B.D., Rowlands, G.J., Nagda, S.M., et al. (1999) Population dynamics of ticks on indigenous cattle in a pastoral dry to semi-arid rangeland zone of Uganda. Ex- perimental and Applied Acarology 23, 79–88. Okello-Onen, J., Tukahirwa, E.M., Perry, B.D., Rowlands, G.J., Nagda, S.N., et al. (2003) The impact of tick control on the productivity of indigenous cattle under ranch conditions in Uganda. Tropical Animal Health and Production 35, 237–247. Olds, C., Mwaura, S., Crowder, D., Odongo, D., van Oers, M., et al. (2012) Immunization of cattle with Ra86 impedes Rhipicephalus appendiculatus nymphal-to-adult molting. Ticks and Tick-borne Diseases 3, 170–178. Olds, C.L., Mwaura, S., Odongo, D.O., Scoles, G.A., Bishop, R., et al. (2016) Induction of humoral immune response to multiple recombinant Rhipicephalus appendiculatus antigens and their effect on tick feeding success and pathogen transmission. Parasites & Vectors 9, 484. Patel, E., Mwaura, S., Kiara, H., Morzaria, S., Peters, A., et al. (2016) Production and dose determination of the Infection and Treatment Method (ITM) Muguga cocktail vaccine used to control East Coast fever in cattle. Ticks and Tick-borne Diseases 7, 306–314. Perry, B.D, Mukhebi, A.W., Norval, R.A.I. and Barrett, J.C. (1990a) A preliminary assessment of current and alternative tick and tick-borne disease control strategies in Zimbabwe. Report to the Director of Vet- erinary Services, Zimbabwe. ILRAD, Nairobi. Perry, B.D., Lessard, P., Norval, R.A.I., Kundert, K. and Kruska, R. (1990b) Climate, vegetation and the distribution of Rhipicephalus appendiculatus in Africa. Parasitology Today 6, 100–104. Perry, B.D., Chamboko, T., Mahan, S.M., Medley, G.F., Minjauw, B., et al. (1998) The economics of in- tegrated tick and tick-borne disease control on commercial farms in Zimbabwe. Zimbabwe Veterinary Journal 29, 21–29. Perry, B.D., Randolph, T.F., McDermott, J., Sones, K.R. and Thornton, P.K. (2002) Investing in Animal Health Research to Alleviate Poverty. ILRI, Nairobi. Peter, T.F., Perry, B.D., O’Callaghan, C.J., Medley, G.F., Shumba, W., et al. (1998) Distributions of the vectors of heartwater, Amblyomma hebraeum and Amblyomma variegatum (Acari: Ixodidae), in Zimbabwe. Experimental and Applied Acarology 22, 725–740. Rothen, J., Githaka, N., Kanduma, E.G., Olds, C., Pflueger, V., et al. (2016) Matrix-assisted laser desorp- tion/ionization time of flight mass spectrometry for comprehensive indexing of East African Ixodid tick species. Parasites & Vectors 9, 151. Sabadin, G.A., Parizi, L.F., Kiio, I., Xavier, M.A., da Silva Matos, R., et al. (2017) Effect of recombinant gluta- thione S-transferase as vaccine antigen against Rhipicephalus appendiculatus and Rhipicephalus sanguineus infestation. Vaccine 35, 6649–6656. Seixas, A., Alzugaray, M.F., Tirloni, L., Parizi, L.F. and Pinto, A.F.M. (2018) Expression profile of Rhipiceph- alus microplus vitellogenin receptor during oogenesis. Ticks and Tick-borne Diseases 9, 72–81. Shaw, M.K., Tilney, L.G. and McKeever, D.J. (1993). Tick salivary gland extract and interleukin-2 stimulation en- hance susceptibility of lymphocytes to infection by Theileria parva sporozoites. Infection and Immunity 61, 1486–1495. Silatsa, B.A., Kuiate, J.-R., Njiokou, F., Simo, G., Feussom, J.-M.K., et al. (2019a) A countrywide molecular survey leads to a seminal identification of the invasive cattle tick Rhipicephalus (Boophilus) microplus in Cameroon, a decade after it was reported in Cote d’Ivoire. Ticks and Tick-borne Diseases 10, 585–593. 386 P. Willadsen Silatsa, B.A., Simo, G., Githaka, N., Mwaura, S., Kamga, R.M., et al. (2019b) A comprehensive survey of the prevalence and spatial distribution of ticks infesting cattle in different agro-ecological zones of Cameroon. Parasites & Vectors 12, 489. Sunter, J.D., Patel, S.P., Skilton, R.A., Githaka, N., Knowles, D.P. et al. (2008). A novel SINE family occurs frequently in both genomic DNA and transcribed sequences in ixodid ticks of the arthropod sub-phylum Chelicerata. Gene 415, 13–22. Sutherst, R.W. (2003) Prediction of species geographical ranges. Journal of Biogeography 30, 805–816. Thullner, F., Willadsen, P. and Kemp, D. (2007) Acaricide rotation strategy for managing resistance in the tick Rhipicephalus (Boophilus) microplus (Acarina: Ixodidae): laboratory experiment with a field strain from Costa Rica. Journal of Medical Entomology 44, 817–821. Tønnesen, M.H., Penzhorn, B.L., Bryson, N.R., Stoltsz, W.H. and Masibigiri, T. (2004) Displacement of Boophilus decoloratus by Boophilus microplus in the Soutpansberg region, Limpopo Province, South Africa. Experimental and Applied Acarology 32, 199–208. Ullmann, A.J., Lima, C.M.R., Guerrero, F.D., Piesman, J. and Black, W.C. (2005) Genome size and organ- ization in the black-legged tick, Ixodes scapularis, and the southern cattle tick, Boophilus microplus. Insect Molecular Biology 14, 217–222. Voigt, W.P., Young, A S., Mwaura, S.N., Nyaga, S.G., Njihia, G.M., et al. (1993) In vitro feeding of instars of the ixodid tick Amblyomma variegatum on skin membranes and its application to the transmission of Theileria mutans and Cowdria ruminatium. Parasitology 107, 257–263. Vudriko, P., Okwee-Acai, J., Tayebwa, D.S., Byaruhanga, J., Kakooza, S., et al. (2016) Emergence of multi-acaricide resistant Rhipicephalus ticks and its implication on chemical tick control in Uganda. Parasites & Vectors 9, 4. Waladde, S.M., Young, A.S., Mwaura, S.N., Njihia, G.N. and Mwakima, F.N. (1995) Optimization of the in vitro feeding of Rhipicephalus appendiculatus nymphae for the transmission of Theileria parva. Parasitology 111, 463–468. Waladde, S.M., Young, A.S. and Morzaria, S.P. (1996) Artificial feeding of ixodid ticks. Parasitology Today 12, 272–278. Willadsen, P. (2004) Anti-tick vaccines in ‘Ticks, Disease and Control’. Parasitology Supplement 129, S367–S388. Willadsen P. (2008) Antigen cocktails: valid hypothesis or unsubstantiated hope? Trends in Parasitology 24, 164–167. Young, A.S., Dolan, T.T., Mwakima, F.N., Ochanda, H., Mwaura, S.N., et al. (1995) Estimation of heritability of susceptibility to infection with Theileria parva in the tick Rhipicephalus appendiculatus. Parasit- ology 111, 31–38. Young, A.S., Waladde, S.M. and Morzaria, S.P. (1996a) Artificial feeding systems for ixodid ticks as a tool for study of pathogen transmission. Annals of the New York Academy of Sciences 791, 211–218. Young, A.S., Dolan, T.T., Morzaria, S.P., Mwakima, F.N., Norval, R.A.I., et al. (1996b) Factors influencing infections in Rhipicephalus appendiculatus ticks fed on cattle infected with Theileria parva. Parasitology 113, 255–266. Preface to Part II: Research Spending and Publications on Primary Production This preface first shows the estimated spending primary production investigations (Fig. PII.1)1. in the principal fields corresponding to the four ILCA scientists (1974–1994) worked on primary chapters in the domain of primary production– production in four broad branches – rangeland rangeland ecology, forage diversity, planted forages production systems, forage diversity, planted and multidimensional crops. forages and multidimensional crops. Subclasses The preface then presents ‘scientific impact’ such as the nutritional quality of feeds and as a function of International Livestock Centre farming practices related to plant productivity for Africa (ILCA)/International Laboratory for would have been included within ‘rangeland Research on Animal Diseases (ILRAD)/Inter- production systems’ or in ‘planted forages’. This national Livestock Research Institute (ILRI) pub- work incurred some US$26 million (in 2015 lications and citations extracted from the Scopus US$) in the 20 years of ILCA’s existence, or 4.1% and Google Scholar databases using search key- of the ILCA/ILRAD total of US$636 million. words relevant to the four fields, and software Spending at ILRAD on plant production from from Aria and Cuccurullo (2017). 1974 to 1994 was negligible. ILRI lifetime (1975–2018) spending on primary production was some US$111 million, or roughly 6.4% of Research Spending on Primary the lifetime total. Production Data from the financial and annual reports of Bibliometrics ILCA, ILRAD and ILRI were used to compile a spending database for 1975–2018 (www.ilri. About 280 papers were produced by ILCA in the org/dataportal/impact/finance). Current spend- four branches of the primary production do- ing for each year and institution was assigned to main. The mean number of citations per paper scientific domains using spending detail by pro- in that domain was 29 and the median was 12 ject, by scientists’ fields of expertise and, occa- during the ILCA era. The scientific return on pri- sionally, from cost accounting by the institutions. mary production research was 11 papers and Current annual spending in US$ was converted 309 citations per US$ million in the ILCA era. to constant annual spending in 2015 US$ using The mean number of citations per paper was the global Manufacturers’ Unit Value Index. 22 and the median was ten during the lifetime of ILRI and one of its predecessors, ILCA, ILRI (1975–2018) in a sample of 1558 papers. made comparatively modest investments in The scientific return on lifetime (1975–2018) 387 388 Research Spending and Publications on Primary Production 400 200 0 1975–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Period Animal science Economics and policy Management and technical Domain Capacity development Livestock systems Primary production Total spending Fig. PII.1. Primary production research has been a small share of total ILRI spending, 1975–2018. (Data from ILCA/ILRAD/ILRI Annual Reports and Financial Reports.) primary production work was 14 papers and b ovine immunology, veterinary epidemiology 308 citations per US$ million of spending. and the interactions of wildlife with domestic livestock, is indirectly related to rangelands. The bibliometrics must therefore be interpreted Rangeland Ecology carefully. Using the Altmetric database, which covers Published work by ILCA, ILRAD and ILRI on social and academic media, ILRI has contributed rangelands comprises the disparate domains some 1.3% of global research papers identified of field investigations proper, such as those by the search terms ‘grazing’, ‘rangelands’, ‘pas- done in the 1970s and 1980s in Mali, Ethiopia ture’ and ‘grassland’ (www.altmetric.com/; ac- and Kenya and later in Niger, and the land- cessed 2 April 2020). The corresponding ILRI mark work in this century on mixed wildlife– share of global Altmetric citations on those livestock systems in Kenya and Tanzania. At terms is less than 1.5%. The ILRI share of such the same time, much of the work on animals papers in Africa, including North Africa, is 37%, held by pastoralists, including studies of with a similar fraction for citations. Spending in millions of 2015 US$ Research Spending and Publications on Primary Production 389 The major papers on rangelands tend to be mean number of citations per paper of 25 and a older, including the work of Sandford (1983) on median of 11 (Fig. PII.2). An approximation of pastoral development, King (1983) on water, the importance of ‘climate change’ papers Coppock (1994) on Borana in Ethiopia, Solo- within the domain of rangelands is the finding mon Bekure et al. (1991) on Kajiado county in of some 182 ILRI lifetime papers treating some Kenya, and Reid and Ellis (1995) on South Tur- aspect of climate change within rangelands kana, Kenya. More recent work, such as Thorn- research, with a mean number of citations per ton et al. (2009), focuses on climate change, or paper of 45 and a median of 14. wildlife–livestock management (Lamprey and The recent burst of new research (as shown Reid, 2004; Hobbs et al., 2008). After the studies in Fig. PII.2) on the broadly defined field of of the colonial era and the decades of the 1960s ‘rangelands’ has occurred outside ILRI and is and 1970s, ILCA and ILRI research held a dom- largely a function of studies of climate change as inant position in rangelands work in both East it affects, and is affected by, the use of global and West Africa. rangelands (see Chapter 16, this volume). Des- ILRI produced some 975 rangelands/grass- pite landmark work in this field by ILCA/ILRI sci- lands papers over its lifetime (1975–2018), a entists, the complexity and scale of African Papers 1,000 500 0 Citations 10,000 5,000 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Global Fig. PII.2. ILRI work has been a large share of global publications on rangelands and related problems, 1977–2018. ILRI sample = 974 papers; global sample = 2,351. (Data from www.scopus.com/.) Counts 390 Research Spending and Publications on Primary Production rangeland problems, and their roles in global cli- B ecause that value is better expressed by the dis- mate and livelihoods, are far greater than the tribution of materials, the scientific impact of the limited resources spent on those problems by ILRI forage gene bank, developed, maintained ILRI and its predecessors; the scarcity of re- and used in close collaboration with the gene search resources in this field will of course be banks of the International Center for Tropical made worse by the growing impact of climate Agriculture (CIAT) and the International Center change in arid areas. for Agricultural Research in the Dry Areas (ICARDA) for more than 35 years, has been ex- tensive (see Chapter 12, this volume), even though it cannot be monetized by commercial sales. Forage diversity and planted forages ILRI’s publications record in planted for- ages has two parts. The first ‘global’ part is glo- Publications in the field of forage diversity are bal and regional analyses of the roles of various chiefly the work of the forage gene banks. As resources, including planted forages, in livestock such, published work (Fig. PII.3) does not cap- feed systems. The second ‘experimental’ part ture well the value of the forage gene banks, comprises plot and laboratory work on forage which is mainly held in the plant collections. yields and forage characteristics, including feed Papers 6,000 4,000 2,000 0 Citations 80,000 60,000 40,000 20,000 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Global Fig. PII.3. ILRI work has been a small share of global publications on planted forages, 1977–2018. ILRI sample = 927 papers; global sample = 12,032. (Data from www.scopus.com/.) Counts Research Spending and Publications on Primary Production 391 digestibility. The ‘experimental’ part has become ILCA/ILRAD (1975–1994) produced about 163 less important in ILRI’s portfolio in this century papers with a mean of 33 citations per paper and hence is less capable of supporting technology and a median of 11. Over its history, ILRI development in forages and in contributing to ex- (1975–2018) produced about 928 papers in the planations of why introduced forages have not broad domain of forage diversity and planted generally succeeded on smallholdings in the tropics. forages, with a mean of 21 citations per paper ILRI’s work in this area has become a small- and a median of 9. ILRI produced about 8.3 er part of global work in this century. ILRI pa- papers per US$ million, generating about 174 pers in forage diversity and planted forages were citations per US$ million. In addition to the about 6.3% of global papers and about 5.6% of falling ILRI shares of total papers and total global citations in that domain over the period citations, mean citations per paper fell from 33 1975–2018. One reason is that the boom in pa- citations per paper in the period 1975–1994 pers on plant biomass as an energy source has (median of 11) to 21 (median of 9) over the life- depressed ILRI’s share of the global spread of work time period 1975–2018 at a time when mean in all subfields related to primary production. citations per paper were growing in most fields. Papers 4,000 3,000 2,000 1,000 0 Citations 50,000 40,000 30,000 20,000 10,000 0 1977–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Global Fig. PII.4. ILRI has made niche contributions to multidimensional crops research, 1977–2018. ILRI sample = 506 papers; global sample = 7,105. (Data from www.scopus.com/.) Counts 392 Research Spending and Publications on Primary Production Multidimensional Crops During the ILCA/ILRAD period, there were nearly 100 papers, with a mean number of cit- ILRI’s published record in multidimensional ations per paper of 32 and a median of 10. There crop research is substantial. Most of the relevant were approximately 506 papers with at least one ILRI output in this field is due to the dedication ILRI author on some aspect of multidimensional of Ercole Zerbini and the late Michael Blümmel, crops published between 1975 and 2018. The who collaborated for many years with scientists long-term mean citations per paper in multidi- of the International Crops Research Institute for mensional crop research was 19 (median of 8). the Semi-Arid Tropics (ICRISAT) in Hyderabad, Despite the accomplishments of ILRI re- India, on fodder improvement in pearl millet, search on multidimensional crops, the global sorghum, maize and grain legumes (Fig. PII.4). importance of ILRI in that field has fallen over That work has created the empirical basis of a time. The ILCA shares of global papers and cit- new model of multiple traits – food and fodder – ations were 16% and 17%, respectively, from in single crops for both grasses and legumes. 1975 to 1994. The long-term ILRI shares Rangeland Planted forages All multidimensional crops 500 400 600 100 300 400 200 50 200 100 0 0 0 First second third fourth top 5% First second third fourth top 5% First second third fourth top 5% Quantiles of merged ILRI and global institution samples Institution ILRI Global Fig. PII.5. Frequency of citations of ILRI and global institutions in primary production research by quantile, 1977–2018. Papers published from 1977 to 2018 with more than nine citations. ILRI (global) rangeland papers = 478 (1,045); ILRI (global) planted forage papers = 402 (6,021); ILRI (global) multidimensional papers = 230 (3,489). (Data from www.scopus.com/.) Counts of papers >9 citations Research Spending and Publications on Primary Production 393 (1975–2018) were 5.4% and 5.8%. The de- citations in a given field, compared with the clining weight of ILRI papers in this field is due share of all papers – ILRI plus global – in the top to the recent global boom of papers on the use 5% (Fig. PII.5). This subset was limited to papers of biomass, including cereal crop residues, for having at least ten citations. ILRI’s global leader- energy. ship in studies of extensive rangelands systems is evident on this metric. The ILRI share of all rangelands papers was about 32%; the ILRI High citation papers share of papers in the top 5% of citations was about 40%. The corresponding shares of ILRI One measure of major ILRI scientific contribu- papers in planted forages and multidimensional tions is the share of ILRI papers in the top 5% of crops were in the range of 5–7%. Note 1 Spending data from the International Center for Tropical Agriculture (CIAT) and the International Center for Agricultural Research in the Dry Areas (ICARDA) were not available. References Aria, M. and Cuccurullo, C. (2017) bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics 11, 959–975. Coppock, D.L. (1994) The Borana Plateau of Southern Ethiopia: Synthesis of Pastoral Research, Develop- ment and Change, 1980–91. Systems Study No. 5. ILCA, Addis Ababa, Hobbs, N.T., Galvin, K.A., Stokes, C.J., Lackett, J.M., Ash, A.J., et al. (2008) Fragmentation of rangelands: implications for humans, animals, and landscapes. Global Environmental Change 18, 776–785. King, John M. (1983). Livestock water needs in pastoral Africa in relation to climate and forage. ILCA Research Report No. 7. ILCA, Addis Ababa. Lamprey, R.H. and Reid, R.S. (2004) Expansion of human settlement in Kenya’s Maasai Mara: what future for pastoralism and wildlife? Journal of Biogeography 31, 997–1032. Reid, R.S. and Ellis, J.E. (1995) Impacts of pastoralists on woodlands in South Turkana, Kenya: livestock- mediated tree recruitment. Ecological Applications 5, 978–992. Sandford, S. (1983) Management of Pastoral Development in the Third World. Wiley, Chichester. Solomon Bekure, de Leeuw, P.N., Grandin, B.E. and Neate, P.J. H. (1991) Maasai herding: an analysis of the livestock production system of Maasai pastoralists in eastern Kajiado District, Kenya. ILCA, Addis Ababa. Thornton, P.K., van de Steeg, J., Notenbaert, A. and Herrero, M. (2009). The impacts of climate change on livestock and livestock systems in developing countries: a review of what we know and what we need to know. Agricultural Systems 101, 113–127. 11 Rangeland Ecology Polly Ericksen1, Pierre Hiernaux2, Augustine Ayantunde3, Philip K. Thornton4, Jason Sircely5 and Lance Robinson6 1International Livestock Research Institute, Nairobi, Kenya; 2Caylus, France; 3International Livestock Research Institute, Ouagadougou, Burkina Faso; 4CGIAR Research Programme on Climate Change, Agriculture and Food Security (CCAFS) and International Livestock Research Institute, Nairobi, Kenya; University of Edinburgh, Edinburgh, UK; 5International Livestock Research Institute, Addis Ababa, Ethiopia; 6Ruwaza Sustainable Development, Stratford, Ontario, Canada Contents Executive Summary 396 The problem 396 ILRI’s role in the global context 396 Impacts of ILRI’s research 396 Scientific impacts 396 Development impacts 397 Introduction 397 Main Trends in African Range Systems 397 Human populations 397 Livestock populations 398 Climate 398 Vegetation 398 Mobility and grazing access 399 The Central Problems of Tropical Range Ecology 399 The technical and economic feasibility of measures to raise primary productivity under arid conditions 399 The short- and long-term effects of climate 400 The effects of rangeland activities 400 Equilibrium and non-equilibrium models 400 The African Rangelands Research Environments 401 East Africa 401 West Africa (central Sahel in Mali, Inner Delta, Gourma, western Niger, northern Nigeria, Senegal and Gambia) 402 Key predecessors and partners 403 The main research questions in ILRI 404 Quantitative assessment and monitoring 404 Grazing regimes 404 Access to grazing resources 405 Interactions among people, livestock and rangelands 405 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 395 396 P. Ericksen et al. Principal Findings 405 East Africa 405 The conflict between wildlife and livestock 406 Modelling 407 West Africa: assessment and monitoring of rangelands 409 Effects of grazing regimes on animal and rangeland performance 411 Mobility and access to grazing 412 Conclusions and the Future 413 Range governance 413 Monitoring rangeland conditions 413 Mitigating climate change and the rangelands 413 Adapting to climate change in the rangelands 414 Rangeland productivity 414 References 414 Executive Summary were the bimodal rainfall areas of Kenya and Ethiopia. A novel feature of ILRI research since The problem the 1990s has been integrative studies on the rangelands of southern Kenya, in areas of inten- Sub-Saharan Africa rangelands are vital to the sive livestock–wildlife interactions in Amboseli livestock economy of the subcontinent. They National Park, Nairobi National Park, the Mara cover roughly 9 million km2 or about 40% of the Reserve and the Serengeti in Tanzania (with col- areas with potential for livestock production. The laborative tasks on rangeland management in majority of African livestock – chiefly cattle, sheep, southern Ethiopia/Borana region). goats, camels and donkeys – live on rangelands at some point during the annual production cycles. Impacts of ILRI’s research Most of the range is managed by mobile groups who constitute perhaps one-fifth of poor livestock Scientific impacts owners in sub-Saharan Africa. Grazing areas have limited vegetative production because of low and There were a number of scientific impacts of variable rainfall – they receive less than 600 mm ILRI’s range research: annual rainfall – shallow and eroded soils, and in- • Integrated studies of domestic livestock, adequate forage due to reduction and fragmenta- vegetation, water, crop and management tion of the rangeland areas resulting from crop- interactions with original data from field land and urban expansions. For all these reasons, it conditions. has been difficult to make sustained improvements • Identification of determinants of plant and in range productivity, even under controlled man- animal productivity under arid and semi- agement, despite several decades of research and arid conditions. development interventions. • Definition of constraints to growth of pastoral systems. • Development and refinement of research ILRI’s role in the global context methods – for animal counts, techniques for vegetation surveys, participatory house- The International Livestock Research Institute hold surveys, application of remote sensing (ILRI) and its predecessor, the International to African environments, and joining of Livestock Centre for Africa (ILCA, 1974–1994) biophysical process models to agent-based have worked for more than 40 years on range- household models. land ecologies of sub-Saharan Africa. ILCA’s • Extension of range science methods and rangelands research focused on the monomodal models to problems of climate change. rainfall areas of the West African Sahel in Mali, • Extension of range science methods to the inner delta of the Niger River in Mali, Niger problems of interactions between wildlife and Senegal. ILRI’s historical sites in East Africa and livestock. Rangeland Ecology 397 • Elucidation of the economic and biological inner delta of the Niger River (Macina) in the rationale of transhumant pastoralism. mid-1970s, which led to an understanding of • Refinement of estimates of greenhouse gas seasonal vegetation growth under the influence emissions from range components (animals, of rainfall, flooding, grazing and cropping. The plants, soils, water and infrastructure). impact of the most severe drought in history that hit the Sahel in 1982–84 was assessed on the pastoral systems of the Gourma region in Development impacts northern Mali in 1983 and was followed by a The chief development impact of range science monitoring of the rangeland recovery over a followed from the defence of the economic and decade. Later West African studies documented biological rationale of extensive pastoralism. This the impacts of grazing on pastures, including defence, made on theoretical grounds following through recycling of nutrients via excretions, the emergence of density-independent models in based on long-term research in the Fakara re- the 1980s, and on empirical grounds following gion of Niger. Related studies of herder decisions the ILCA/ILRI studies of pastoralism, has pre- about animal movements, in response to changes vented bad policies (such as certain forms of in rainfall, cropping intensity and wage opportun- group ranges and grazing reserves) from being ities, built on the characterizations of rangeland imposed on pastoral groups. As it has done so, it productivity and its determinants. has preserved pastoral livelihoods and limited rising inequality; an example of this scientific and development impact can be seen in the Ken- Main Trends in African Range Systems yan Mara Community Conservation Planning Framework. Map 2 (see p. xviii) shows the main research areas of ILRI and its predecessors. General trends around those areas can be described for human Introduction populations, livestock populations, climate, vege- tation, mobility and grazing access. Arid and semi-arid pastoral systems are This chapter summarizes the impact of research 1 found across much of sub-Saharan Africa. In by ILRI and collaborators on rangeland ecology Kenya, Mali, Niger, Nigeria and Sudan, they are in East and West Africa. the largest livestock systems by land area. Although In East Africa, the research topics included sometimes considered ‘marginal’ in terms of pub- identifying the drivers of rangeland vegetation, lic expenditure priorities, these systems contrib- such as rainfall, animal density, bush and crop en- ute significantly to agricultural gross domestic croachment, and increased landscape fragmenta- product, household food and nutritional security, tion. A second group of topics was establishing the and to ecosystem maintenance in dry areas. In importance of wildlife–livestock interactions in Kenya, over 80% of domestic meat consumption Kenya and Tanzania, given the significant and di- is produced in pastoral systems; Sudan has been verse wildlife populations in those countries. This exporting around 1.8 million animals annually led to a process to foster innovations in conserva- for decades; and in the Horn of Africa, livestock tion area management, better incorporating pas- and meat export values may have exceeded toral livestock production with wildlife. Simultan- US$500 million by 2010 (Catley et al., 2013, p. 7). eously, modelling of coupled systems was ongoing In West Africa, these pastoral systems contrib- to better capture multiple drivers of change in ute many of the young males and interact with rangelands and examine both the impacts of land- more intensive and sedentary livestock produc- scape fragmentation and the growth of agricul- tion systems via seasonal mobility. ture on livelihoods and ecosystem services. These results highlighted the need to view humans, cat- tle and wildlife as integrated components of eco- Human populations systems when making management decisions In West Africa, ILCA research focused on Significant increases in human populations are comparative evaluations of pastoral and agro- noted in East and West Africa alike. In West pastoral systems in central Mali (Niono) and the Africa, Touré et al. (2012) estimated an increase 398 P. Ericksen et al. of 2.6% per year in the overall rural population In East Africa, with bimodal rainfall in of the drylands between 2005 and 2010, with some areas, dry-season movements are shorter 3.6 times more people in 2010 than in 1960. de and more opportunistic. Increased fragmenta- Haan et al. (2016) assumed a 2.5–3% annual tion of rangelands due to expansion of crop- increase for the future. ping and other activities and exclusion from dry-season water and grazing further limit seasonal mobility. The observation of range Livestock populations fragmentation was a motivation for much of ILRI’s research as scientists sought to under- After the severe losses during the droughts of the stand its impact on rangeland, livestock and early 1970s and early 1980s (Toulmin, 1987) wildlife productivity. livestock populations have increased rapidly as well, with the Food and Agriculture Organiza- tion Corporate Statistical Database (FAOSTAT) showing rates of growth of between 3.1% and Vegetation 4.4% from 1980 to 2010. In both East and West Africa, the small-ruminant population has grown Long-term trends are extremely difficult to detect faster than cattle. Detailed data on national level in the East African rangelands. First, the cost of species distributions are lacking in most places; long-term field studies of vegetation and land however, unique data from Kenya based on use make them infeasible in some instances. 30  years of aerial surveys shows that cattle Second, the high temporal and spatial variability numbers have declined as camel and small rumin- of precipitation and vegetation make it problem- ants have increased (Ogutu et al., 2011). The atic to project from shorter periods and smaller same study found significant declines in wildlife study areas (Homewood, 2008; Oba, 2013). Third, numbers. changing drivers such as drought frequency, re- cent fragmentation and the spread of invasive plant species disturb historical patterns of land Climate use. As Homewood (2008, pp. 65–72) explains, dryland rangelands go through periods of vege- West Africa is characterized by a south–north tation fluctuation in response to different drivers, climate gradient ranging from humid/subhumid in particular climate variability. In East Africa, on coast lines to semi-arid/arid in the north. there is no consistent long-term evidence or Transhumant pastoralism has dominated in trend in rainfall, although heavy grazing from the arid Sahelian zones, with herders migrating sedentarization or restriction of mobility can among rangelands and water sources. At the degrade rangelands (Galvin et al., 2002), and end of the rainy season, herders move to areas with hence increased fragmentation is a threat to the available grass or crop residues, often southwards sustainability of rangelands. or in wetlands but also more opportunistically. Long-term dry matter (DM) production per Over the past three decades, ‘these movements hectare above the 600 mm isohyet in West have become longer and more dispersed’ (Touré Africa is in the order of 600–2400 kg DM/ha. et al., 2013, p. 14) and conflicts with sedentary Rainfall and vegetation improved after the farmers and in protected areas have increased droughts of the 1970s and 1980s, and the sub- (Turner et al., 2011). This is for various reasons, sequent ‘regreening’ of parts of the Sahel has including changes in traditional land uses around been notable (Hiernaux et al., 2009c; Dardel water points and increased livestock numbers. et al., 2014b) including for the woody popula- Although many transhumance corridors have tion in rangelands (Hiernaux et al., 2009b; been legally recognized since colonial times, trans- Brandt et al., 2016a) and agroforestry parklands humance corridors are not always respected (Reij et al., 2009; Sendzimir et al., 2011). (Niamir-Fuller, 1999; Turner et al., 2016) and Galvin et al. (2002) summarized the re- failure to respect them sometimes leads to violent search, concluding that woody plants have conflicts among herders and between herders expanded across most rangelands globally, in and crop farmers. response to both grazing management and Rangeland Ecology 399 global increases in carbon dioxide and nitrogen the seasonal, annual and spatial mobility of emissions. livestock? • What is the appropriate use of equilibrium and non-equilibrium models as explanatory Mobility and grazing access frameworks and policy guides? This chapter first presents these issues schemat- Associated with the growth in population dens- ically before discussing the evidence produced ity in East and West Africa has been an expan- about them by the studies of ILRI, its predeces- sion of cropping. The area expansion of cropped sors and its collaborators in sub-Saharan Africa. fields reduces grazing areas on average and typically limits seasonal grazing, such as on wetlands, even more sharply (Haywood, 1981; Marie, 2000; Schlecht et al., 2001). The technical and economic feasibility Increased private ownership and enclosure of measures to raise primary productivity of rangelands, particularly in East Africa, includ- under arid conditions ing ‘land grabbing’, and the erosion of some traditional practices governing common lands Efforts to raise pasture yields in dry areas with- has restricted grazing even more. The loss of out added water have failed systematically. Pratt mobility and grazing access through public use and Gwynne (1977, pp. 100–128) showed that planning, enclosures and urbanization (Galaty, the economics of range improvement – by water 2013a) ultimately threatens rangeland integ- development, fencing, overseeding of the range, rity (Galvin et al., 2008a,b). Formal tenure is managed grazing and seeding with new species now more important as a tool for pastoralists to of grasses – were unfavourable in drier condi- secure rights to access lands; however, this poses tions and risky in wetter ones. These findings were challenges, because Hobbs et al. (2008) has generally confirmed in Le Houérou’s (1980) argued that fragmentation results from modern book on browse and in Sandford’s (1983) global land tenure. review of pastoral development. Le Houérou (1989, pp. 149–155) synthe- sized the Sahelian experience with native pasture improvement over many years and concluded: The Central Problems of Tropical Range Ecology • Irregular rainfall, a long dry season and competition from native grasses and forbs A summary of the central problems of tropical made the introduction of higher-yielding range ecology would be the following questions: plant species unsuccessful. • There had been no commercial success in • What are the technical and economic feasi- reseeding pastures at rainfall around the bility of measures to raise plant productivity 550 mm isohyet, with local or introduced under arid conditions? grasses or with legumes; the same was true • What is the technical and economic feasi- of ‘rotational or deferred grazing’. bility of raising livestock at given levels of • ‘…of the 80 herbaceous tropical arid zone feed resources? species of forages which have been tried at • What are the short- and long-term effects Niono (550 mm) in Mali in 1977–1980, not of climate and climate change in primary one single species became established and and secondary range productivity? amenable to produce a grazing impact’. • What are the effects of rangeland use for live- • There was one success – with Acacia tortilis stock, crops, wildlife and tourism on soils, and Acacia senegal near M’Bidi, Senegal – in nutrient flows and transfers, vegetation, establishing browse species at rainfall less water and greenhouse gas emissions? than the 400 mm isohyet in West Africa. • How have land rights evolved under pres- • The high cost of fencing, firebreaks and sure from competing uses in crops and water made it nearly impossible to achieve other sectors, and what are their effects on higher plant yields, even if browse and 400 P. Ericksen et al. pasture production could be improved under Major livestock losses do occur, with severe experimental conditions (see also Montgolf- droughts, which is one reason why herd sizes need ier-Kouèvi and Le Houérou, 1980). to be large enough so that production can return to trend within 3–5 years (Ellis and Galvin, 1994). These droughts are often cumulative, with several successive ‘failed’ seasons ultimately The short- and long-term effects leading to major hardship and loss of animals, as of climate lack of feed and water plus diseases lead to mortal- ity (Toulmin, 1987). Drought recovery is mostly Climate is the key determinant of rangeland related to levels of herd die-off, as population is productivity, especially in the arid and semi-arid mostly related to lagged rainfall variation; how- areas where extensive animal production pre- ever, drought intervals are also important (Lesnoff dominates. In East Africa, these systems receive et al., 2012). Long-term shifts in climate, com- no more than 600 mm of rainfall annually and pounded by persistent growth of human popula- often only 200–300 mm (Ellis and Swift, 1988; tions and of cultivated areas into rangelands, Oba, 2013). Higher aridity means greater rain- explain some of the shift in West Africa into small fall variability in time and space, and in much of stock in western Niger (Turner, 1999). East Africa, the rainfall is bimodal. Inter-annual variability in precipitation is a major driver of herd and grazing management. In East Africa, The effects of rangeland activities rainfall is the primary driver of vegetation, with vegetation growth closely following rain- Pastoralists have for centuries managed the fall amount, frequency and duration (Coppock, spatial and temporal variability of water and 1994; Coppock et al., 2017). pasture by moving animals annually and sea- In West Africa, the monsoon drives a climate sonally, as shown in the early compilations of gradient from subequatorial humid in the south Monod (1975) and ILCA (1975). Seasonal to desert margins in the north. In the arid and mobility in West Africa follows the monsoon semi-arid Sahel, rainfall varies from 100 to 600 along the south to north gradient, while in East mm annually in a period of 1–4 months, with Africa, the movements are more opportunistic the more southerly regions receiving from 600 and local owing to variability in altitude and in mm to more than 1200 mm over periods of up to the bimodal rainfall regime. Pastoralists also 8 months. Although the monsoon occurrence is move to avoid seasonal disease outbreaks, often predictable, it occurs only in one season and the caused by vectors influenced by rainfall patterns. distribution of rain across this season is unpre- Traditionally, pastoral and agro-pastoral produ- dictable (Hiernaux and Le Houérou, 2006). There cers managed access to and availability of grazing is a marked contrast in the Sahel between fodder and water through complex rules and agree- quality in the wet and dry seasons. ments, sometimes including specialized institu- Forage availability and quality are the tions, so that in drought years reserves would primary drivers of variability in livestock pro- still be available (Gallais, 1984; ODEM/CIPEA, duction. Primary production, in turn, is highly 1983). In addition to the ecological drivers, variable over time and space because the major security and access to markets also influence determinant of plant growth is available soil movements. Homewood (2008) noted that the moisture and fertility. Shortages in forage avail- need for flexibility made pastoral systems vul- ability arise from drought, constrained access nerable to land loss and exclusion from custom- (e.g. due to disease, distance and access to water, ary ranges; losses of land access to fragmentation or to conflict) or changes in palatability. Most threaten the viability of pastoralism. rangelands include a mix of vegetation types in- cluding an herbaceous layer and scattered woody plants. These rangelands return quite rapidly to Equilibrium and non-equilibrium models peak production after drought once the rains re- turn, as documented by long-term research Equilibrium models from temperate systems (on (Hiernaux et al., 2009a,c; Miehe et al., 2010). which many of the papers in ILCA’s first major Rangeland Ecology 401 book relied; ILCA, 1975) influenced the early The African Rangelands Research studies of tropical arid and semi-arid rangelands. Environments These equilibrium models focused on livestock densities and carrying capacities, and blamed East Africa (southern Ethiopia, livestock for overgrazing and environmental Kenya, Tanzania) degradation. The early ILCA studies in Mali (Hiernaux, The rangelands of Kenya, Tanzania and south- 1983; Wilson et al., 1983; Wilson, 1986; Wilson ern Ethiopia are noted for bimodal rainfall, which et al., 1988) and later work in the Gourma in nor- heightens the seasonal, inter-annual and spatial thern Mali (de Leeuw et al., 1992), and further in variability characteristic of arid and semi- western Niger (Hiernaux et al., 2009a) formed arid areas. The short rains occur in October– some of the empirical basis of non-equilibrium December, with the long rains in March–May. models, which became dominant in the 1980s and The lowlands of Ethiopia and northern Kenya 1990s (e.g. Ellis and Swift, 1988; Behnke and are arid and semi-arid, while southern Kenya and Scoones, 1993; Scoones, 1995). northern Tanzania are semi-arid to dry–subhumid The ‘Synthesis Paper’ of Ellis and Swift with patches of greater humidity. (1988) (Table 11.1) contrasted equilibrium Historical rangelands research at ILCA and non-equilibrium systems in four domains. focused on the Borana Plateau of southern Ethi- Equilibrium systems operate through biotic feed- opia (Coppock, 1994). The Borana areas have a backs, such that rangelands have a density- semi-arid climate (400–600 mm of rain) with a dependent carrying capacity. Non-equilibrium complex mix of vegetation consisting mainly of systems are density independent and respond to perennial grasses and shrubs. stochastic events such as rainfall and fire, rather ILCA’s initial range research in Kenya than to livestock populations. Semi-arid range- was in Maasailand in Kajiado County of lands in bimodal rainfall systems such as in south-eastern Kenya (Solomon Bekure et al., southern Ethiopia behave more as equilibrium 1991). ILRI’s later work concentrated on the systems (Coppock et al., 2017), while the behav- rangelands of southern Kenya, especially areas iour of Sahelian rangelands under monomodal of intensive livestock–wildlife interactions in rainfall regime clearly affiliates to non-equilibri- Amboseli National Park, Nairobi National Park, um (Hiernaux, 2004). the Mara Reserve and the Serengeti in Tanzania. Vetter (2005) summarized the equilibrium/ The rangelands of northern Tanzania and south- non-equilibrium debate over arid and semi-arid ern Kenya are more varied in terms of rainfall rangelands. She concluded that: ‘most arid and (400–1200 mm), with national parks occupy- semi-arid rangelands encompass elements of ing some of the wetter areas (Amboseli National both equilibrium and non-equilibrium be- Park and Maasai Mara National Reserve). These haviour, and management must account for parks are among the most popular for wild- the high temporal variability and spatial life tourism, and the challenges of integrating heterogeneity’. Table 11.1. Contrasts between equilibrium and non-equilibrium systems. (From Ellis and Swift, 1988.) Domain Equilibrium systems Non-equilibrium systems Abiotic patterns Vegetative conditions relatively Highly variable conditions; typically, constant; wetter drier; monomodal rainfall Plant–herbivore interactions Deterministic from livestock to Weak determinism from livestock to vegetation with limited feedback vegetation; plants under ‘abiotic control’ Population patterns Density-dependent carrying Density independent; ‘abiotic cycles’; capacity; animal ‘populations carrying capacity random track carrying capacity’ Ecosystem characteristics Typically more humid and densely Typically more arid and less densely populated populated than equilibrium system 402 P. Ericksen et al. wildlife tourism with livestock production re- West Africa (central Sahel in Mali, Inner main significant. Vegetation in the Mara is more Delta, Gourma, western Niger, northern productive than in more arid Kajiado, but both Nigeria, Senegal and Gambia) areas are predominantly tall- and short-grass plains interspersed with woodlands and shrubs. Rangeland production in the Sahel occurs under Some of ILRI’s research focused on the Athi- monsoon conditions and is characterized by high Kaputiei Plains, in northern Kajiado district and spatial, seasonal and annual variability (Ayan- bordering Nairobi National Park (Reid et al., tunde et al., 1999; Hiernaux and Le Houérou, 2008). Rainfall ranges from 500 to 800 mm. 2006). The Mali studies of the late 1970s to the The soils are derived from phonolitic lava and early 1990s would have occurred in a period of contribute to a nutrient-rich savannah with exceptionally low Sahel rainfall, which would both grasses and trees that can support consid- not have been true of the East Africa studies erable wildlife biomass. For at least 400 years, (Nicholson, 2000; Nicholson et al. 2018). Mean Maasai have occupied this area. Over the 20th primary production ranges from 600 kg DM/ha century, this pastoral–wildlife system has become in the northern Sahel with 200 mm of rainfall to more fragmented and compressed, which was 2400 kg DM/ha in the southern Sahel with 600 the subject of ILRI’s research. mm rainfall, but there is no linear relationship Further south in Kajiado, on the border between rainfall and herbage production (Hier- with Tanzania, the Amboseli ecosystem has long naux et al., 2009c; Dardel et al., 2014a). been considered a conservation jewel. However, Herbage production is characterized by a its ongoing transition from extensive pastor- wide local variation due to differences in soil alism to intensive pastoralism carried out on type, runoff water patterns based on topography individual land parcels is threatening both and geomorphology, and dominant plant species wildlife in and around the Amboseli National (Hiernaux and Le Houérou, 2006). The feed Park and livestock production by the communi- quality of herbage is often inversely proportional ties (Burnsilver et al., 2008). The constraints to the amount of water infiltrated in the soil dur- on mobility and fragmentation of the resource ing the growing season, at least for a given soil base has altered the landscape, modifying wild- texture and fertility (Breman and de Wit, 1983). life habitat and increasing competition among Free-ranging ruminants are selective in choos- wildlife and livestock for pasture and water. ing their diets (Ayantunde et al., 2007), and Over to the west in Narok County, the therefore spatial heterogeneity in herbage mass Mara ecosystem is one of the wettest pastoral and quality affects the spatial distribution of savannahs in East Africa and is the northern grazing animals (Turner et al., 2005; Schlecht site of wildebeest, zebra and Thomson’s gazelle et al., 2006). migration from the Serengeti in Tanzania. The ILRI’s pastoral research in West Africa has Maasai Mara National Reserve is limited to focused on the Sahel, in view of the importance wildlife tourism but is surrounded by group of these arid rangelands to livestock production. ranches to the north and east. It receives rela- ILRI research on rangeland ecology started in tively high rainfall (up to 1200 mm annually) Mali in the 1970s in Niono (central Mali) and and is a productive range that supports a high then expanded to Macina (the inner delta of the density of wildlife, especially from July to Octo- Niger River). In 1983, 25 rangeland sites were ber when animals migrate north from Tanzania. established in Gourma and were monitored by Some research also occurred in the Ngorongoro ILCA from 1984 to 1993, and then irregu- Conservation Area adjacent to the Serengeti, larly until 1998. The site monitoring continued which is a designated multiple-use area with from 2000 to 2009 under the African Monsoon important objectives for both wildlife conserva- Multidisciplinary Analysis (AMMA) project tion and human welfare. Over 50,000 Maasai (Mougin et al., 2009) and is still going on under pastoralists and their livestock live in the area the AMMA-CATCH (Coupling the Tropical (Boone et al., 2002). The Tarangire–Simanjiro– Atmosphere and the Hydrological Cycle) research Manyara pastoral ecosystem, south-east of the network (www.amma-catch.org; accessed 25 Serengeti in Tanzania, was also a site for some March 2020) although the observations had of ILRI’s work. to be scaled down from 2012 because of civil Rangeland Ecology 403 insecurity. In Niger, 71 vegetation monitoring Key predecessors and partners sites were established in Fakara and were moni- tored from 1994 to 2006 under ILRI research The early ILCA work in Niono, Mali, was devel- activities. The monitoring of these sites is con- oped alongside a Malian–Dutch research pro- tinuing within the AMMA-CATCH network gramme called ‘Primary Production in the Sahel’ (Cappelaere et al., 2009). In addition, there were (Penning de Vries and Djiteye, 1991). Pierre grazing studies conducted on-station at the Hiernaux was among the first to set up system- International Crops Research Institute for the atic monitoring of rangeland sites for long-term Semi-Arid Tropics (ICRISAT) Sahelian Centre in analysis, complementing others in Senegal Sadoré, Niger, and at the Niger government cat- (Valenza, 1984; Gaston and Boerwinkel, 1982; tle ranch in Toukounous. Miehe et al., 2010). ILCA researchers contrib- The Gourma monitoring sites are spread uted early studies on livestock productivity, along the south–north bioclimatic gradient mobility and transhumance within the Macina from Boulakessy at the border with Burkina flood plain (ODEM/CIPEA, 1983) and in central Faso to Gourma Rharous on the Niger River via Mali. Toulmin et al. (2002) and Cotula et al. Hombori and Gossi. Gourma is in the Sahelian (2004) conducted research on pastoralist mobil- agroecological zone with average annual rainfall ity and conflicts over land, especially between of between 200 and 500 mm (Hiernaux et al., herders and farmers, in the 1990s, which high- 2009b,c). It is dominated by annual herbaceous lighted changing land-use dynamics in response plants and a few tussock perennials towards the to different legal and administrative changes. driest end, with widely scattered shrubs and Turner (1992, 1999, 2017) contributed to this small trees. line of research by examining how community Fakara is a small natural region of west- resource management to ‘clarify rules’ of ac- ern Niger covering about 6000 km2 between cess actually runs the risk of increasing local the confluent valleys of the Niger River to the ecological and economic vulnerabilities, given West and the fossil valley of the Dallol Bosso to both the agro- ecological realities and the range the East (Hiernaux and Ayantunde, 2004). of resources needed to sustain both crop and The study site covers 500 km2 (included livestock production. Turner highlighted the within latitude 13°20′–13°35′N and longi- trend of a greater year-round presence of live- tude 2°35′–2°52′E) all falling within the central stock in the former ‘agricultural’ zone as a sig- Sahel bioclimatic zone with annual precipita- nificant factor changing resource access and tion of between 400 and 500 mm. needs and leading to conflict. The study site in Toukounous is a 4474 ha In East Africa, key researchers who were ranch situated in the central Sahel, 14°33′N and working at the same time as ILRI (and in some 33°17′E, at an altitude of 290 m above sea level. cases were ILRI collaborators) on issues of land The ranch is fenced into 30 paddocks, varying in use and livestock included Peter Coppolillo, Lane size from 49 to 283 ha. The local climate is typical Coppock, Solomon Desta and Katherine Home- Sahelian semi-arid tropical with monomodal rain- wood. Galaty (1994a, 2013b) noted in particular fall from July to September. The long-term annual the negative trends towards excluding livestock rainfall (mean±standard deviation) is 336±105 or selling lands for other, crop-based, purposes. mm. Like any other Sahelian site, the vegeta- McCabe found similar issues in Tanzania in long- tion is dominated by annual herbaceous species. term research on Maasai livelihood diversifica- The study on the effect of grazing regimes on tion (McCabe, 2003; McCabe et al., 2010), as did cattle performance was conducted between 1995 Homewood (2008) and collaborators from ILRI. and 1997 (Ayantunde et al., 1999). Coppolillo (2000) described factors affecting the The study site in Sadoré (13°14′N, 2°16′E) distribution of animals across landscapes and at the ICRISAT Sahelian Centre in Niger is situ- found that the distribution of dry-season water ated about 45 km south-east of Niamey. The cli- was a major influence throughout the year, not mate is typical south Sahelian with an annual only in the dry season. He also noted variability precipitation of between 450 and 600 mm. Most in herder practice and cattle productivity. grazing studies were carried out at an enclosure Desta and Coppock (2002, 2004) summar- that was established in 1982. ized the long-term data on cattle populations, 404 P. Ericksen et al. rainfall and changing access to rangeland (ii)  effects of grazing regime and intensity; resources in the Borana region of southern Ethi- and (iii) grazing access and mobility; and opia. Their 2002 study convincingly demon- (iv)  interactions among people, livestock and strated the ‘boom-and-bust’ cycle of livestock rangelands. losses associated with periodic rainfall deficits. Their 2004 study documented a trend of declines Quantitative assessment and monitoring in per-capita livestock holdings due to resource pressure/population increase and reduced access Quantitative assessment and monitoring of range- to key grazing areas (similar to trends documented land resources included detailed mapping in the in southern Kenya by Solomon Bekure et al., large study zones of central Mali, the Macina 1991, and Galaty, 1994b, among others). This flood plains, the Niono ranch, Gourma and later was leading some families to try cultivation as in the Fakara of Niger. The assessment and well as privatizing some grazing areas. They sug- subsequent monitoring in the 1970s and 1980s gested that these patterns are broadly predictable were important to document how vegetation and hence development interventions should behaved in response to rainfall and soil proper- reflect these system dynamics and respond to ties. These studies also contributed to a broader growing resource pressure. understanding of the variation between range- Homewood et al. (2001) summarized long- land types in different parts of Africa, again in term changes in land cover in Serengeti–Mara relation to soil properties (fertility) and rainfall. focused on wildlife rather than livestock per se. To date, the monitoring and assessment of range- Comparing the Kenyan side to the Tanzanian land resources in Gourma, Mali (1984–2006), side, they found a marked decline in species and and in Fakara, Niger (1994–2006) (which have rapid land-cover change only on the Kenyan side. then continued under AMMA and AMMA- Their analysis concluded that the Kenya–Tanza- CATCH ) are the only such long-term studies in nia differences were primarily due to landowners the West African Sahel. Other investigations responding to market opportunities for mechan- were mainly short-term studies of grazing or of ized agriculture and less to cattle numbers or conflict between farmers and herders. to population growth. Homewood went on to collaborate with ILRI and other colleagues on Grazing regimes the book Staying Maasai: Livelihoods, Conserva- tion and Development in East African Rangelands A second area of research sought to explain the (Homewood et al., 2009), which focused more effects of grazing regimes and grazing intensity on livelihood diversification in the context of on animal performance, rangeland vegetation changing land-use dynamics. and forage production, and on the physical and Gufu Oba conducted long-term research in chemical properties of soil. ILRI researchers southern Ethiopia and northern Kenya on sought to understand the relationships between rangeland dynamics, exploring herder man- animal productivity and rangeland management, agement of grazing areas and herder percep- including the impacts of livestock grazing on tions of land-use change and responses to these rangeland properties and cropped areas (e.g. changes (Angassa and Oba, 2007, 2008; Oba through nutrient redistribution). The answers to et al., 2000). His research contributed evidence the first two questions also contributed empirical on the interactions among rainfall, plants and evidence to the debates about whether arid and grazers, and documented herder manage- semi-arid African rangelands behave as equilib- ment of rangelands for animal and rangeland rium or non-equilibrium systems (Behnke et al., performance. 1993). These debates and the evidence under- lying them has significant implications for rangeland management (see Homewood, 2008, The main research questions Chapter 4, and Vetter, 2005). They also contrib- in ILRI uted to the related debates about whether the Sahel was under a process of desertification and The research was organized into four main areas: if livestock systems had particular responsibil- (i) quantitative assessment and monitoring; ity in the process (Hiernaux et al., 2016). Rangeland Ecology 405 Access to grazing resources The general goal was to promote biodiversity con- servation and prevent land degradation through The third area was research on access to grazing research to provide tools for understanding key resources and livestock mobility in relation to changes. The network specifically sought to vegetation resources and the livelihoods of pas- ‘examine the causes and consequences of land toralists and agropastoralists. These studies led degradation and the biophysical systems that to an understanding of how livestock keepers underlie changing patterns of land use within used mobility to actively manage their herd East Africa land-cover/land-use changes in re- productivity, given the high temporal and spatial sponse to multiple drivers’ (www.lucideastafrica. distribution of vegetation in the rangelands. In org; accessed 25 February 2020). Three foci West Africa, reciprocal relationships with croppers were: (i) the drivers and impact of agricultural ex- have historically been important for the mutual pansion into grazing areas; (ii) the transition of benefits that pastoralists received by grazing agro-pastoral systems in semi-arid areas; and (iii) their animals in wetter areas and croppers the impacts of land-use change on biodiversity. received from the manure deposited by transhu- Over 48 working papers were published mant animals. covering 30 years of land-use change across East Africa, primarily in subhumid and semi-arid Interactions among people, livestock zones where cropping, livestock production, and rangelands protected areas and wildlife commonly over- A common area of research in East and West lapped. The main findings on the agro-pastoral Africa was to document interactions among systems transition and the impacts of land-use people, livestock and the range. This research change on biodiversity (Maitima et al., 2004) built upon the characterization, assessment and included: monitoring work in West Africa to take into ac- count the new dynamics, especially changes in • Causes of land-use change in agro-pastoral cropping patterns, increased fragmentation, live- systems: the research found that the fluid- stock population dynamics and the impact of ity of land-tenure arrangements and quite declines in previously reciprocal arrangements major changes around who has rights to and the failures of formal land tenure. In East use land and decide how it should be used Africa, the research combined modelling with was most marked in agro-pastoral areas. ground observations, resulting in key findings The trend was a movement away from com- about the impacts of fragmentation (in particular) mon management to private and individual on rangeland, wildlife and livestock performance arrangements such as subdivision of for- (Galvin et al., 2008c). This also resulted in the mer group ranches (Olsen et al., 2004), in- development of several modelling exercises to cluding fencing and smaller grazing areas, better understand the impacts of changing range- as well as more crop-based agriculture. land dynamics, and a new model. In addition, • Beneficial impacts of grazing on rangeland Reid et al. (2004, 2008) sought to understand vegetation: moderate grazing was found to complementarities between livestock production support native vegetation and local d iversity. and wildlife management in southern Kenya, • Impacts of the increased competition for leading to significant innovations in land man- key land and water resources: this was agement around the Nairobi National Park and resulting in the restriction of herding and the Maasai Mara National Reserve. an increase in cropping and transition of some households to agro-pastoral produc- tion. In addition, there were restrictions on movement of wildlife and loss of access to Principal Findings wetland areas and key migration corridors and resources (Reid et al., 2004; Lamprey East Africa and Reid, 2004). • Impacts of restrictions on wildlife and live- ILRI hosted the Land Use Change, Impacts and stock movement: despite the gazetting of Dynamics (LUCID) network, beginning in 2000. national parks and reserves, dispersal of 406 P. Ericksen et al. animals into adjacent areas or for seasonal and forage resources, the research found that movements was still critical. Unfortunately, limitations on mobility were having serious the increased competition for land also impacts on rangeland ecology as well as on increased human–wildlife conflicts. livestock and people. Their thesis was that frag- mented landscapes are less productive than un- Extending the LUCID work while adding fragmented ones. They attributed fragmentation unique data on wildlife and livestock populations to modern systems of land tenure, which are not over time across Kenya, Joseph Ogutu and col- suited to arid and semi-arid rangelands (Hobbs leagues documented the impacts of pastoralism et al., 2008). and protected areas on wildlife numbers. The Reid et al. (2008) describe the Athi-Kaputiei first studies focused on the Mara ecosystem, Plains near Kitengela, Kenya, which they stud- where they documented the decline in a number ied intensively on the ground and in aerial sur- of wildlife species, which corresponded with veys during the period 1977–2002. The area is habitat deterioration and in some cases increased an unusual example of a fragmented pastoral numbers of people and settlements (Ogutu et al., ecosystem that still supports migration of large 2009). This trend was continuing by 2011, but wildlife despite its proximity to Nairobi. At pre- they noted the importance of ranches around sent, the fragmentation is exacerbated by fen- the park as wildlife dispersal areas, which led to a cing of land parcels. Fences had a particularly recommendation that land use and poaching negative impact on livestock movements and needed to be regulated (Ogutu et al., 2011). The grazing/foraging. This research informed the finding on the importance of pastoral areas to Wildlife Conservation Lease Program, which support protected areas emerged again, as these paid Kitengela residents to allow free movement areas provide corridors between seasonal grazing of wildlife on their lands, to avoid poaching and areas (Ogutu, 2013; Ogutu et al., 2017). The to avoid fencing of land. Initiated in 2000 and 2017 study estimated that Kenya’s communally expanded in 2007, the programme had the po- and privately protected pastoral areas sup- tential to be a promising innovation, but it fell port 65–70% of the country’s wildlife popula- short of success. Although Osano (2013) found tion. This is because the national parks and re- some positive impacts, for example in terms of serves are too small for many of the large wildlife more lions and payments to beneficiaries, the who must access areas outside the protected programme stopped in 2012 when the donor areas either seasonally or year-round. This re- funds closed. search influenced the support community con- servancies around the Maasai Mara National The conflict between wildlife and livestock Reserve, Nairobi National Park and Amboseli National Park through the Reto-o-Reto project. Reto-o-Reto sought solutions to land-use change ILRI researchers engaged in over a decade and wildlife conservation problems with a more of research on land-use change, land manage- action-oriented approach. It engaged pastoral ment, biodiversity and livestock across the arid communities in solutions to balance poverty and semi-arid regions of southern Kenya and alleviation and wildlife conservation in pastoral northern Tanzania. This collaboration produced systems. The research approach described in Reid a book in 2008, drawing on some of the LUCID et al. (2016) was one of ‘continual engagement’ studies and others, entitled Fragmentation in Semi- between scientific researchers and community Arid and Arid Landscapes (Galvin et al., 2008d). members which produced new knowledge and This edited volume is a comprehensive overview solutions. They intentionally used collaborative of why and how fragmentation has occurred research-facilitator teams to better engage with across a range of arid and semi-arid ecosystems communities and bridge across nested institu- and the implications of this fragmentation for tions from community up to the global level. people and animals. An increase in the ‘exclusivity This approach produced several key pieces of hy- of use’ across rangelands has restricted move- brid knowledge, from the Kitengela conservation ments across landscapes, limiting access to re- payments to the Amboseli modelling work on the sources. Given that mobility is key to managing impacts of subdivision and the Mara Community the temporal and spatial heterogeneity of water Conservation Planning Framework. Rangeland Ecology 407 Community-based conservation has emerged the subdivision of group ranches) and the growth as a promising approach for resolving conflicts of agriculture (Thornton et al., 2003, 2006; stemming from wildlife–livestock–pastoralist Boone et al., 2005, 2006) on livelihoods and interactions, and the Reto-o-Reto approach sup- ecosystem services. The coupled model was ported the establishment of a network of conser- subsequently used to explore the possibilities of vancies around the Maasai Mara National schemes of payment for ecosystem services in Reserve. This began with four in 2009, which areas of southern Kenya where agriculture is ex- expanded to 14 by 2013 when the Mara Wildlife panding rapidly, to compensate pastoralists for Conservancies Association was formed. At the losses arising from more wildlife-compatible forms heart of the conservancy model is that payments of land use (Bulte et al., 2008). A third set of stud- from tourists are used to compensate pastoral ies investigated the impacts of climate variability communities for adhering to rules about when on livelihoods and on the economic benefits of and where they are permitted to graze their ani- using weather forecasts (Galvin et al., 2004; mals, in order to ensure that wildlife have Thornton et al., 2004). This work was later enough area and forage. The community man- extended to more generalized studies of the rela- agement model offered by the conservancies was tionship between fragmentation and rainfall vari- strengthened by the 2013 Wildlife Conservation ability (Boone, 2007; Boone and Wang, 2007). and Management Act and the 2016 Community The CSU/ILRI collaboration on integrated Land Act. Today, all conservancies across Kenya assessment modelling was founded on the SA- are members of the Kenya Wildlife Conservan- VANNA model, originally developed by Coughe- cies Association and strive to offer a sustainable nour (1985, 1992) for Turkana District, Kenya, model that protects both wildlife and pastoral but further developed and applied in many other livestock production. settings in Africa, Asia and North America since. The final outcome for which the Reto-o- SAVANNA models primary ecosystem interactions, Reto work is known was the 2010 Greater simulating functional groups for plants and ani- Kitengela Land Use Master Plan, the first ever for mals over periods from 10 to 100 or more years a pastoral area. This plan evolved with support in a spatially explicit way. from research that showed, for example, that For household modelling a simple structure compensating pastoralists for loss of access to was built – the Pastoral Household and Economic grazing lands could increase their resilience. The Welfare Simulator (PHEWS). This tracks the flow multiple pressures on land in Kitengela, and the of cash and dietary energy in pastoralist house- cessation of funds for the compensation scheme, holds using a simple set of management rules have compromised the objectives of the plan. As that describe a reasonably realistic hierarchy of discussed at the end of the chapter, participatory goals at the household level. PHEWS was used to land-use planning with communities is still one model households of varying sizes and assets of the most promising solutions for improved and with different access to natural resources. management of rangeland ecosystems. PHEWS was built using survey data regarding household size, structure and income; species, Modelling numbers and sexes of livestock held; and culti- vated area (Thornton et al., 2003). A coupled The Colorado State University (CSU)/ILRI collab- SAVANNA–PHEWS model was calibrated for oration also led to innovations in modelling sites in northern Tanzania and southern Kenya rangelands and pastoral households. Research (Thornton et al., 2006, 2007; Galvin et al., spanning the 1990s and early 2000s sought to 2006) using ecosystem and household data quantify different alternatives in the rangelands collected in the previous decade. of sub-Saharan Africa resulting from the in- SAVANNA–PHEWS was innovative in at creasing pressures on natural resources owing least two ways. First, it is worth noting that to human population growth and the resultant SAVANNA itself is a complex and sophisticated conflicts among wildlife, cattle and agriculture model, and even now represents a summit of the (Galvin and Thornton, 2001). The major re- ecosystem modeller’s art. PHEWS, on the other search questions were the effects of landscape hand, occupies the other end of the spectrum: it fragmentation (in the case of southern Kenya, is simple, requires relatively few data to calibrate, 408 P. Ericksen et al. was developed very quickly and can easily be • A tool of moderate complexity – one that adapted to new situations. PHEWS was an early could be useful to a new user in a week or example of a ‘disposable’ model: because the less. builder invests only limited time and energy in • A monthly time step, with simulations that assembling it, the model can easily be thrown run for 5–100 years or more. away and a new start can be made, as there is • Representation of global vegetation at least little lost. The nature of the relationship between at the scale of herbaceous, shrubs and model complexity and model utility is probably trees. not as straightforward as is often imagined; this • The ability to include natural or manage- is borne out by recent trends in some quarters ment modifications to rangelands, such as towards model simplification. Second, despite fire and fertilization. the disparities in the detail and elegance of the • Programming structures and portable models, SAVANNA and PHEWS were tightly code, allowing the software to be run on dif- linked, and were run as part of the same simula- ferent platforms including multiprocessor tion. At each time step modelled, SAVANNA clusters or networks. passed information about livestock to PHEWS, • Output mostly as straightforward spatial and PHEWS passed information about the sale surfaces, without complex summary ana- or purchase of animals back to SAVANNA, where lyses that are more readily done in other herd sizes were adjusted accordingly. packages. PHEWS was essentially a population-based model, and it was used as the basis for develop- The idea was not to program G-Range from ing an agent-based approach to investigate simi- scratch but to use components from published lar questions. Like PHEWS, the DECUMA model models. The Century model (Parton et al., 1993) (Decisions under Conditions of Uncertainty by was used as the core of the soil modelling and Modelled Agents) was tightly linked with SA- physiological aspects of the G-Range model, given VANNA to represent livestock-owning and culti- Century’s wide use in rangelands over the past 20 vating households. As an agent-based model, years or so. Other aspects of G-Range were influ- DECUMA simulates in a relatively complex way enced by SAVANNA. G-Range does contain some individual households on a landscape in a spa- new contributions, notably in modelling plant tially explicit way (unlike PHEWS), adjusting livestock populations. distributions on a weekly basis and making There are several insights from the CSU/ other household decisions on a monthly basis ILRI collaboration on the modelling of coupled (Boone et al., 2011; Boone and Lesorogol, 2016). systems. An early synthesis of the coupled mod- One potentially important document from all this elling work supported the hypothesis that a work, unwritten as yet, is a comparison of SA- household’s capacity to stresses is governed by VANNA–PHEWS and SAVANNA–DECUMA simu- flexibility in livelihood options (Thornton et al., lations for the same situation: we have calibrated 2007). Households cope with stresses through applications of both coupled models for the same intensification, diversification and off-farm part of Kajiado district, Kenya. Such a compari- economic activities. Viable options depend on son would throw light on the added value of a household objectives and attitudes as well as on more complex household model (DECUMA) access to natural resources, inputs and output compared with a very simple model (PHEWS). markets. The study also highlighted the fact that Beginning in 2012, ILRI worked with CSU on generally it is the poorer households that can gain a simulation tool that could be used to project glo- the most from implementing risk-management bal rangeland changes in response to trend climate options. Furthermore, there are limits to the and climate variability. This led to a process-based adaptive capacity of households in the absence simulation model that is spatially explicit and of access to off-farm resources, and these limits of moderate complexity, called G-Range. Several likewise depend on local context. Much of the needs guided the design of G-Range: CSU/ILRI work around rangeland fragmenta- tion has been summarized in Hobbs et al. (2008) • A simulation tool for global rangelands that and Galvin et al. (2008b,c) as discussed above. captures main primary production and its A general insight from the modelling work dynamics. was the view of humans, cattle and wildlife as Rangeland Ecology 409 components of an integrated ecosystem (Gal- distribution as well as changing soil seed stocks vin et al., 2008a), if technical, advisory and (Hérault and Hiernaux, 2004). Analysis of aer- policy-related interventions are to be appro- ial photographs confirmed the match between priately targeted. A thread that remains to geomorphology and vegetation type (Hiernaux be unravelled is the appropriate use of inte- and Haywood, 1978). grated models to resolve conflicts between Additional observations of 169 sites across conservation and people in the rangelands the Macina floodplains from 1979 to 1983 (Galvin et al., 2006). (ODEM/CIPEA, 1983) demonstrated that biogeo- Simulations with G-Range summarized pro- graphic zoning only affected unflooded islands jected climate change impacts on livestock across or edges, along with land-use history including Africa, using a review of literature and model cropping and wood exploitation. In the flood results (Thornton et al., 2015; Boone et al., zones, the perennial herbaceous vegetation is 2015). While there are many options that can determined by the inter-annual variations in help livestock keepers adapt, there appear to be flooding regime (Hiernaux and Diarra, 1986). no widely applicable and unconstrained options Although poor in species diversity, they were (Boone et al., 2018; Thornton et al., 2019). highly productive. The patchy distribution required large-scale mapping, which was also a methodo- logical innovation (Marie, 2000). West Africa: assessment and monitoring Observations across the 25 sites in the of rangelands Gourma transect (Hiernaux et al., 2009b,c) following the 1983/84 drought confirmed the This was one of the major areas of research earlier Mali findings that vegetation is organized begun by ILCA in Mali, as a contribution to the by bioclimatic zones and then based on edaphic study of livestock production in a diversity of conditions within these zones. Land-use history pastoral and agro-pastoral systems (Wilson et al., had little impact. Soil type and rainfall distribu- 1983, 1988; ODEM/CIPEA, 1983) continued in tion determined where herbaceous and woody Niger by ILRI (Hiernaux and Ayantunde, 2004; plants were located and distributed. Analysis of Hiernaux and Turner, 2002). ILCA researchers satellite images for the same region also indicated invested large efforts to characterize, quantify that vegetation types can be mapped to edaphic and spatially assess forage resources across dif- zones and surface hydrology and soil type (Bre- ferent regions. Field observations of 331 sam- man and de Ridder, 1991; Gal et al., 2016). Field pled sites across central Mali from 1976 to 1980 observations supported the building up of a (Wilson et al., 1983) encompassed a large area in novel primary production model, STEP (Sahel- order to include most of the seasonal movements ian Transpiration, Evaporation and Productivity of herds managed by Macina Fulani pastoralists model) to simulate annual herbaceous growth as well as a diversity of livestock production sys- relying mainly on a soil water balance and me- tems. The observations defined bioclimatic zones teorological control of plant photosynthesis (Lo and noted how the forage species composition Seen et al., 1995). The model was validated with varied with these zones (Hiernaux and Le field monitoring data on the Ferlo and Gourma Houérou, 2006). Within the zones, the distribu- rangeland vegetation (Mougin et al., 1995; tion of woody plant species depends on soil tex- Tracol et al., 2006). The herbaceous vegetation ture first and then moisture regime and terrain decay and decomposition with or without graz- topography (Sankaran et al., 2005). Beyond ing is also simulated in STEP and has had a num- that, land-use history, especially clearing for ber of applications in assessing seasonal fodder cropping, was found to modify the composition resources (Diawara et al. 2018) and impact on of woody plant species (Cissé and Hiernaux, the environment (Delon et al., 2015; Pierre et al., 1984; Achard et al., 2001). The distribution of 2015). STEP was further adapted for particular herbaceous species, which are mostly annuals, perennial grasses growing in flood plains (Léau- is less consistently sensitive to these same factors thaud et al., 2018). as well as to the shadow of woody plants. This is Observations a decade later in the Fakara due to large inter-annual variations in species region of south-western Niger, a more densely composition in relation to variations in rainfall populated area with more sedentary populations, 410 P. Ericksen et al. found that vegetation distribution was explained spatial heterogeneity and patchy pattern of the by both the edaphic environment and land use vegetation. Consideration of the significant sea- (Turner and Hiernaux, 2015). The generally sonality of vegetation growth in arid to sub- poor soil fertility explained the poverty of the humid West Africa was a driving principle in vegetation floristic composition. These findings screening for more efficient metrics of forage were confirmed by analysis of high-resolution resources assessment by satellite remote sensing photographs and satellite images, which also such as normalized difference vegetation index were used to distinguish cropping land-use types (NDVI) metrics used to assess herbaceous and (Schlecht et al., 2006; Tong et al., 2020). crop production (Hiernaux, 1988; Bégué et al., The assessments were also combined with 2014) and also woody plant crown cover and foli- monitoring of herbaceous biomass and floristic age mass (Brandt et al., 2016a,b). composition at the same sites, observing vari- Indeed, the strength of the seasonality of ation in vegetation production over time (e.g. be- the Sahel ecosystems, including the complex fore, during and after droughts); across topog- situations created by the combination or rainfall raphy (lowlands versus uplands; sandy soils versus and flood seasons in the Macina flood plains clay soils); and in response to burning and grazing. (Hiernaux, 1983), forced the assessments to go The results confirmed the high spatial hetero- beyond static survey and to quantify, explain and geneity of Sahelian rangeland forage resources, attempt to predict the seasonal and inter-annual in response to rainfall, runoff and soil properties. variations in forage resources. Repeated meas- Land-use change was also found to be a factor in ures of vegetation in rangeland grazed or protected data collected from 1994 to 2006 in Fakara, from grazing together with rainfall records, soil Niger. Finally, woody plants were also assessed moisture and main nutrient (nitrogen, phos- and monitored during the same time and over phorus) contents were decisive to better under- the same sites to observe how their growth, stand what the limiting factors of rangeland phenology and distribution were affected by soils production were (Buerkert and Hiernaux, 1998; and rainfall as well as grazing and cropping Hiernaux and Diawara, 2014). They led to the (Hiernaux et al., 2009a, 2019). development of simulation models of the vegeta- Overall, this body of research was novel in tion growth (Lo Seen et al., 1995; Mougin et al., several respects. First was the quantification of 1995), grass tillering under grazing condition forage resources expressed in seasonal capacities (Hiernaux et al., 1994) and also of the straw and per landscape unit (Marie, 2000; Hiernaux, 2005; litter decomposition during the dry season de- Auda et al., 2012; Hiernaux et al., 2015). Second pending on grazing and trampling intensity were a number of methodological innovations. (Hiernaux et al., 2014). The assessment of the contribution of woody Stratified sampling along a linear transect plants to livestock nutrition led to new woody has been adopted in many rangeland resources as- plant population survey methods (Hiernaux, sessment and monitoring such as the monitoring 1980; Franklin and Hiernaux, 1991; Brandt by the Centre de Suivi Ecologique in Senegal (Di- et al., 2016a,b), development of allometric rela- ouf et al., 1998), the Institut de l’Environnement tionships between woody plant size and foliage et de Recherches Agricoles (INERA) national pas- and fruit masses (Cissé, 1980), and surveys to toral resource survey in Burkina Faso (Kiema, characterize the phenology of the main woody 2015), the carbon balance study in Widou Thien- plant species (Hiernaux et al., 1994b). It also led goly in Senegal (Assouma et al., 2018). Similarly, to a number of methodological advances in sys- the stratification of woody plants when assessing tematic field observations (Hiernaux, 1982, their density by an unbiased distance method 2016; de Leeuw and Hiernaux, 1990), aerial known as the ‘point-centred quadrant’ (Pollard, surveys (Milligan, 1982; Milligan et al., 1982) 1971) with points distributed at regular intervals and remote sensing using aerial photos and sat- along the same linear transect, associated with ellite images (Hiernaux, 1988; Hanan et al., classical dendrometry on woody plants sampled 1991). One of the principles put forward in field by the method and allometric relationships (Cissé, methodology is the stratified random sampling 1980; Henry et al., 2011), is increasingly used to along linear transects (Hiernaux, 2016), as a efficiently assess the contribution of woody plant more efficient strategy to account for the high population (Hoffmann et al., 2006). Rangeland Ecology 411 The seasonal curve integral of the NDVI concerns that grazing livestock was a driver of during either the growing season (herbaceous) environmental degradation. These experiments or the dry season (woody plants) is widely used in were aimed at measuring the impact of livestock satellite remote sensing the vegetation produc- grazing regimes on short (within season) and tion, as well as metrics based on seasonal NDVI medium (inter-annual) time frames. The ILRI maximum, mean or percentile values (Tucker experimental work clearly separated the impact et al., 1985; Hiernaux, 1988; Fensholt et al., during the short growing season from that dur- 2004; Dardel et al., 2014a), and based on STI in ing the long dry season (Hiernaux and Diarra, the dry season (Jacques et al., 2014; Kergoat et al., 1986; Hiernaux and Le Houérou, 2006). 2015). Accounting for the woody phenotypes Grazing experiments across three sites and improved the assessment of woody plant cover over two decades, in fenced paddocks, demon- and foliage masses (Brandt et al., 2016a,b, 2019). strated that herbaceous annuals responded to The simulation models developed and cali- grazing during the growing season with re- brated using field data collected by ILRI in Mali, Niger growth, but this decreased rapidly in response to and Senegal have been used to predict vegetation repeated grazing on the same plants. The short- production from rainfall, flood and grazing scenarios term impacts depend on the timing and intensity (Dardel et al., 2014a,b; Diouf et al., 2016; Diawara of the grazing (Hiernaux and Turner, 1996). et al., 2018), and also to assess the risks of soil Grazing during the dry season was found to ac- erosion by wind (Pierre et al., 2015), carbon seques- celerate the decomposition of straw and litter, so tration, carbon and nitrogen emissions to estimate that, at most, livestock intake reaches a third of (Le Dantec et al., 2009; Delon et al., 2015) and to cal- the herbaceous biomass at the onset of the sea- culate the main nutrient balance in a pastoral eco- son (Hiernaux et al., 2015). In the particular case system (Schlecht et al., 2004; Assouma et al., 2018). of the Macina grasslands, which are adapted to The quantitative assessments of grazing live- temporal floods, dry-season regrowth was found stock impact on the ecosystem in the short and to depend on grazing timing and intensity, with medium term, and the derived model tools have an optimum frequency specific to each grassland been used to advocate for protecting and ensuring type (Hiernaux, 1984). The longer-term effects livestock seasonal and regional mobility in pas- of grazing observed on species composition and toral systems (Hiernaux et al., 2015). They contrib- vegetation productivity are minor compared with ute to discussions on the desertification paradigm the effect of clearing a land to crop or for forestry in the Sahel and policies developed at national and exploitation (Hiernaux, 1998; Achard et al., international scales to combat alleged desertifi- 2001; Hiernaux and Turner, 2002). They de- cation (Dardel et al., 2015; Hiernaux et al., 2016). pend on the timing and intensity of grazing and Quantifying the carbon and nutrient recyc- involve forage intake but also trampling and live- ling role of livestock in pastoral and agro-pastoral stock excretions (Diarra et al., 1995; Schlecht et systems has been used to diagnose the corralling al., 1998). Another set of experiments assessed practices of agro-pastoralists in western Niger soil seed stocks and germination patterns; these and suggest improvements (Gandah et al., 2003; were found to be transient, with most seeds Powell et al., 2004; Djaby, 2010; Hiernaux and germinating during the wet season following Diawara, 2014; Coppock et al., 2017). More dispersion, particularly due to the capacity of generally, this knowledge was used to assess the annuals to produce large numbers of seeds, contribution of livestock to cropping intensifica- which were rapidly dispersed. tion with a perspective of ecological sustainabil- Another line of research observed the effects ity and improvement of agro-pastoralist welfare of grazing regimes on diet selection and animal (Hiernaux, 1996, 2013; La Rovere et al., 2005). performance. This demonstrated that grazing ruminants selected forages that were more nu- tritious and digestible than an overall pastural Effects of grazing regimes on animal evaluation would suggest. Night-time grazing and rangeland performance was found to increase forage intake and hence animal productivity, especially during the dry A series of experiments were carried out across season, as the quality of available forage decreases the various West African sites to try to address in the dry season. Supplementation is necessary 412 P. Ericksen et al. for animals to maintain their body weight in the Research from central Mali (Niono and late dry season. The foraging behaviour of free- Macina) in the late 1970s and early 1980s de- grazing livestock differs by species, with goats scribed the daily and regional grazing routes being the most selective and sheep the patchiest. used by pastoralists. The ILCA/ODEM project in A domain of the grazing livestock impact Macina mapped transhumance routes and docu- on which the ILRI team particularly focused mented the local governance and access rules. In their research, especially when studying agro- Niono, the researchers found that integrated pastoral systems, was the recycling of organic crop–livestock production was increasing. Con- material and minerals through faeces and urine tractual agreements and grazing rights and in- excretions. Faeces and urine excretions were stud- stitutions were declining in terms of authority/ ied in controlled conditions, in barns and meta- formal recognition. Similar research in Gourma, bolic cages at Sadoré research station (Fernández- Mali, also found diversities of pastoral and agro- Rivera et al., 2005), but also on grazing cattle on pastoral systems and a tradition that fixed graz- the ranch at Toukounouss (Ayantunde et al., ing rights through access to water points. Research 1999) and on cattle, sheep and goats in village from Fakara in the mid-1990s that was followed conditions (Schlecht et al., 1998). In the latter up in 2007 found that grazing management of case, the results indicated that grazing livestock most village livestock depended on movements through their local mobility with grazing–walk- outside the village territory, especially during ing and resting–rumination areas achieved a the rainy season to avoid damage to crops and to spatial transfer of organic matter and nutrient have access to natural pastures as there is always to the benefit of resting points such as paddocks restricted livestock mobility during the cropping and corralling spots (Hiernaux et al., 1998). These season. Second, the presence of extra-village nutrient concentrations are key to cropping in- movements of village livestock is higher in areas tensification and diversification (Hiernaux and of higher population density, which is expected Diawara, 2014). due to declining grazing areas as a result of demo- Turner (1999) and Turner et al. (2005) graphic pressure. Third, the perceived advantages looked at the issue of whether grazing livestock of herd mobility are to better provide livestock in mixed crop-grazing areas could possibly lead with pasture and water and, at least during the to grazing-induced degradation. Their results rainy season, to avoid crop damage. The perceived indicated this is possible if there is not enough disadvantages of herd mobility are losing access to labour/expertise to manage the animals or if milk and other livestock products, not finding dry- they are constrained in finding enough free-graz- season pasture or water outside the village territory ing areas. Turner and Hiernaux (2002) worked and spending more energy trekking the herd. with livestock herders to document how they Research on farmer–herder relationships moved their grazing animals across landscapes. and conflict management in agro-pastoral systems This approach resulted in better information on of Niger from 1995 onwards produced novel the spatiotemporal distribution of livestock findings (Turner et al., 2012; Turner 2017). across agro-pastoral landscapes, as this distribu- Community informants stated their preference tion reflects local land-use patterns, topography, to resolve conflicts without involving customary vegetation, settlements and water points. or government authorities, particularly at supra- village levels, as they did not believe that higher authorities could settle conflicts in a lasting fashion. Second, most of the farmer–herder dis- Mobility and access to grazing putes occurred during the cropping season and involved active mediation by customary author- Given the previously summarized evidence that ities at the local level, resulting in a crop damage livestock keeper management of grazing is critical fine. These fines were paid by the herding family to how much feed livestock are able to access as or livestock owner. These disputes are better well as the distribution of nutrients across land- managed when there is a convergence of product- scapes, the studies that researchers associated ive interests across herding and farmer social groups with ILCA and later ILRI conducted on mobility and a high perceived degree of interdepend- and access to grazing resources are relevant. ence. Later research in Mali (Turner et al., 2014) Rangeland Ecology 413 confirmed that increases in the number of con- works to make participatory and community- flicts due to unauthorized grazing of crop residues based rangeland management successful is is a reflection of the change in farmer–herder accumulating, but researchers have not yet relationships from that of mutual trust that char- systematically consolidated this evidence. For acterized manure and entrustment contracts to example, evidence is emerging to suggest that more inherently conflictual relationships based simple interventions by communities to establish on wage and tenancy contracts. (or re-establish) seasonal grazing patterns can have a quick and significant effect on rangeland condition. In order to convince various investors Conclusions and the Future of the feasibility of rangeland management, how- ever, we need more action research and trials of Rangelands research in arid and semi-arid sub- interventions such as seasonal planned grazing, Saharan Africa has been reinvigorated by renewed rangeland rehabilitation and management of government and donor interest in pastoral liveli- bush encroachment. This needs to be coupled with hoods. The challenges facing productive range- impact assessments of the social and biophysical lands remain competition over resources, which impacts of past and ongoing management initia- has been exacerbated by armed conflict; overuse tives, as well as cost–benefit analysis of the rela- of some rangelands as fragmentation continues; tive benefits of rangeland management. Finally, we and the failure of many technical and governance lack evidence of how community-level restor- interventions, as summarized by Reid (2012). ation interventions interact with socio-ecological These challenges cannot be met without a long- dynamics and larger landscape scales. term core-funded research investment, as has been achieved in the pastoral areas of the USA, Austra- lia, New Zealand and the Mediterranean Basin. Monitoring rangeland conditions Despite the recent revival of interest in pas- toral problems in the drylands, the Altmetric results highlight the serious underinvestment Quick successes in restoration often build the in African range sciences. Searches for the ex- confidence of community institutions to take on pressions ‘range ecology’, ‘pasture’, ‘pastoralism’ bigger challenges. However, scaling out these local and ‘grazing’ show that papers from African sites, successes often proves challenging, given the of all institutional affiliations, are only about 5% seemingly intractable challenge of securing of global papers and barely 4% of global citations. access to resources while also maintaining the The unresolved development challenges of flexibility needed to manage these resources. pastoralism in East and West Africa make it While rangeland management needs to be par- essential to renew long-term empirical research ticipatory, this participation may differ by social, to understand rangeland dynamics and to de- institutional and biophysical context. Support- velop appropriate public policies. The rangelands ing, validating and disseminating evidence for research agenda at ILRI focuses on: (i) govern- appropriate models of governance that build on ance for better rangeland management; (ii) community-based approaches, further enabling monitoring rangeland conditions to improve them with higher-level institutional arrange- development interventions; (iii) understanding ments, is thus another priority. While evidence the interactions between climate change and the suggests that community governance needs to rangelands; and (iv) improving rangelands prod- be enabled by higher-level arrangements, such uctivity for pastoral resilience. as national land-tenure policies or district land- use plans, again little systematic analysis of these nested models has been consolidated. Range governance Mitigating climate change As most rangelands are common pool resources and the rangelands that are shared by communities, development partners have been testing community-based Documentation of the ecosystem services provided approaches in different settings. Evidence of what by rangelands shows that they can generate a 414 P. Ericksen et al. competitive return on new investment. Much of can advance investigations of adaptation and ILRI’s previous work on ecosystem services fo- mitigation options there. Work is under way on cused on wildlife biodiversity. However, recently, coupling G-Range with household models to carbon sequestration has attracted more interest, allow integrated scenarios of ranges, animals as an option to offset greenhouse gas emissions and households. from other sectors. While the potential for carbon sequestration in rangelands is high, there is little empirical evidence about what is achievable in arid Rangeland productivity and semi-arid environments (Milne et al., 2016). With the increased interest in stimulating more market orientation and commercialization of Adapting to climate change pastoral and agro-pastoral systems, many devel- in the rangelands opment partners are forced to address the question of forage availability. While substantial new fod- Adapting rangelands to future climate change is der production is available from crop residues and an overarching priority. Climate change will potentially from the introduction of new plant- bring hotter temperatures and different rainfall ing materials in mixed systems, it is inevitable that patterns, with accompanying changes in the most rainy-season forage will be produced on composition, quantity and quality of forage. rangelands. With greater investment in tools such Models like G-Range are a major step forward in as land-use planning, rangeland management studies of climate change and rangelands and can be stimulated by the commercial opportunities. Note 1 ‘ILRI’ refers to ‘ILCA and ILRI’ unless a specific distinction is needed. References Achard, F., Hiernaux, P. and Banoin, M. (2001). Les jachères fourragères naturelles et améliorées en Afrique de l’Ouest. In: Floret, C. and Pontanier, R. (eds) La Jachère en Afrique Tropicale. De la Jachère Naturelle à la Jachère Améliorée. Le Point des Connaissances. John Libbey Eurotext, Paris, pp. 201–240. Angassa, A. and Oba, G. (2007) Relating long-term rainfall variability to cattle population dynamics in com- munal rangelands and a government ranch in southern Ethiopia. Agricultural Systems 94, 715–725. Angassa, A. and Oba, G. (2008) Herder perceptions on impacts of range enclosures, crop farming, fire ban and bush encroachment. Human Ecology 36, 201–215. Assouma, M.H., Lecomte, P., Hiernaux, P., Ickowicz, A., Corniaux, C., et al. (2018) How to better account for livestock diversity and fodder seasonality in assessing the fodder intake of livestock grazing semi-arid sub-Saharan Africa rangelands. Livestock Science 216, 16–23. Auda, Y., Mougin, E., Hiernaux, P., Diawara, M., Soumaguel, N., et al. (2012). Estimation of fodder produc- tion in Sahelian rangelands (Gourma, Mali) using ENVISAT/MERIS observations. ESA-2012 Frascati. Ayantunde, A.A., Hiernaux, P., Fernandez-Rivera, S., van Keulen, H. and Udo, H.M.J. (1999) Selective grazing by cattle on spatially and seasonally heterogeneous rangeland in Sahel. Journal of Arid Envir- onments 42, 261–279. Ayantunde, A.A., Fernandez-Rivera, S., Hiernaux, P. and Tabo, R. (2007) Implications of restricted access to grazing by cattle in wet season in the Sahel. Journal of Arid Environments 42, 261–279. Bégué, A., Vintrou, E., Saad, A. and Hiernaux, P. (2014) Differences between cropland and rangeland MODIS phenology (start-of-season) in Mali. International Journal of Applied Earth Observation and Geoinformation 31, 167–170. Behnke, R.H. and Scoones, I. (1993). Rethinking range ecology: implications for rangeland management in Africa. In: Behnke, R.H., Scoones, I. and Kerven, C. (eds) Rangeland Ecology at Disequilibrium: New Models of Natural Variability and Pastoral Adaptation in African Savannahs. Commonwealth Secretar- iat and ODI, London, pp. 1–30. Rangeland Ecology 415 Behnke, R.H., Scoones, I. and Kerven, C. (1993). Range Ecology at Disequilibrium: New Models of Natural Variability and Pastoral Adaptation in African Savannas. Overseas Development Institute, London. Boone, R.B. (2007) Effects of fragmentation on cattle in African savannas under variable precipitation. Landscape Ecology 22, 1355–1369. Boone, R.B. and Lesorogol, C.K. (2016). Modelling coupled human–natural systems of pastoralism in East Africa. In: Building Resilience of Human-Natural Systems of Pastoralism in the Developing World. Springer International Publishing, Cham, Switzerland, pp. 251–280. Boone, R.B. and Wang, G. (2007) Cattle dynamics in African grazing systems under variable climate. Journal of Arid Environments 70, 495–413. Boone, R.B., Coughenour, M.B., Galvin, K.A. and Ellis, J.E. (2002) Addressing management questions for Ngorongoro Conservation Area using the Savanna Modelling System. African Journal of Ecology 40, 138–150. Boone, R.B., BurnSilver, S.B., Thornton, P.K., Worden, J.S. and Galvin, K.A. (2005) Quantifying declines in livestock due to land subdivision in Kajiado District, Kenya. Range Ecology and Management 58, 523–532. Boone, R.B., Galvin, K.A., Thornton, P.K., Swift, D.M. and Coughenour, M.B. (2006) Cultivation and conservation in Ngorongoro Conservation Area, Tanzania. Human Ecology 34, 809–828. Boone, R.B., Galvin, K.A., BurnSilver, S.B., Thornton, P.K., Ojima, D.A. and Jawson, J.R. (2011) Using coupled simulation models to link pastoral decision making and ecosystem services. Ecology and Society 16, 6. Boone, R.B., Conant, R.T. and Sircely, J. (2015). Adjustment and sensitivity analyses of a beta global rangeland model. In: CGIAR Research Program on Climate Change, Agriculture and FoodSecurity (CCAFS). CCAFS, Copenhagen. Boone, R.B., Conant, R.T., Sircely, J., Thornton, P.K. and Herrero, M. (2018) Climate change impacts on global rangeland ecosystem services. Global Environmental Change 24, 1382–1393. Brandt, M., Hiernaux, P., Rasmussen, K., Mbow, C., Kergoat, L., et al. (2016a) Assessing woody vegetation trends in Sahelian drylands using MODIS based seasonal metrics. Remote Sensing of Environment 183, 215–225. Brandt, M., Hiernaux, P., Tagesson, T., Verger, A., Rasmussen, K., et al. (2016b) Woody plant cover esti- mation in drylands from Earth Observation based seasonal metrics. Remote Sensing of Environment 172, 28–38. Brandt, M., Hiernaux, P., Rasmussen, K., Tucker, C. J., Wigneron, J.-P., et al. (2019) Changes in rainfall distribution promote woody foliage production in the Sahel. Communications Biology 2, 133. Breman, H. and de Ridder, N. (1991). Manuel sur les Pâturages des Pays Sahéliens. Karthala, Paris. Breman, H. and de Wit, C.T. (1983) Rangeland productivity and exploitation in the Sahel. Science 221, 1341–1347. Buerkert, A. and Hiernaux, P. (1998) Nutrients in the West African Sudano-Sahelian zone: losses, transfers and role of external inputs. Journal of Plant Nutrition and Soil Science 161, 365–383. Bulte, E.H., Boone, R.B., Stringer, R. and Thornton, P.K. (2008) Elephants or onions? Paying for nature in Amboseli, Kenya. Environment and Development Economics 13, 395–414. Burnsilver, S.B., Worden, J. and Boone, R.B. (2008). Processes of fragmentation in the Amboseli ecosystem, Southern Kajiado District, Kenya. In: Galvin, K.A., Reid, R.S., Behnke, R.H. Jr. and Hobbs, N.T. (eds) Fragmentation in Semi-Arid and Arid Landscapes. Springer, Dordrecht, Netherlands, pp. 255–279. Cappelaere, B., Descroix, L., Lebel, T., Boulain, N., Ramier, D., et al. (2009) The AMMA-CATCH experi- ment in the cultivated Sahelian area of south-west Niger – investigating water cycle response to a fluctuating climate and changing environment. Journal of Hydrology 375, 34–51. Catley, A., Lind, J. and Scoones, I. (2013). Pastoralism and Development in Africa: Dynamic Change at the Margins. Earthscan/Routledge, London. Cissé, M.I. (1980) The browse production of some trees of the Sahel: relationships between maximum foli- age biomass and various physical parameters. In: Le Houérou, H.N. (ed.) Browse in Africa, The Cur- rent State of Knowledge, International Symposium on Browse in Africa, 8–12 April 1980, Addis Ababa. ILCA, Addis Ababa, pp. 205–210. Cissé, M.I. and Hiernaux, P. (1984). Impact de la mise en valeur agricole sur les ressources fourragères: le cas des jachères de Dalonguébougou. Doc. Prog. AZ 96. Bamako, CIPEA. Coppock, D.L. (1994) The Borana plateau of southern Ethiopia: synthesis of pastoral research, development, and change, 1980–91. Systems Study Number 5. ILCA, Addis Ababa. 416 P. Ericksen et al. Coppock, D.L., Fernandez-Gimenez, M., Hiernaux, P., Huber-Sanwald, E., Schloeder, C., et al. (2017). Rangeland systems in developing nations: conceptual advances and societal implications. In: Briske, D.D. (ed) Range-land Systems: Processes, Management and Challenges. Springer Series on Envir- onmental Management, pp. 569–641. Coppolillo, P. (2000) The landscape ecology of pastoral herding: spatial analysis of land use and livestock production. Human Ecology 28, 527–560. Cotula, L., Toulmin, C. and Hesse, C. (2004). Land tenure and administration in Africa: lessons of experience and emerging issues. IIED, London. Coughenour, M.B. (1985) Graminoid responses to grazing by large herbivores: adaptations, expectations and interacting processes. Annals of the Missouri Botanical Garden 72, 852–863. Coughenour, M.B. (1992). Spatial modelling and landscape characterization of an African pastoral ecosys- tem: a prototype model and its potential use for monitoring drought. In: McKenzie, D., Hyatt, D. and McDonald, V. (eds) Ecological Indicators, Vol. 1. Elsevier, New York, pp. 787–810. Dardel, C., Kergoat, L., Hiernaux, P., Grippa, M., Mougin, E., Ciais, P. and Nguyen, C.-C. (2014a) Rain-use efficiency: what it tells about conflicting Sahel greening and Sahelian paradox. Remote Sensing 6, 3446–3474. Dardel, C., Kergoat, L., Hiernaux, P., Mougin, E., Grippa, M., et al. (2014b) Re-greening Sahel: 30 years of remote sensing data and field observations (Mali, Niger). Remote Sensing of Environment 140, 350–364. Dardel, C., Kergoat, L., Hiernaux, P., Grippa, M. and Mougin, E. (2015). Entre desertification et reverdisse- ment du Sahel. Que se passe-t-il vraiment? In: Sultan, B., Lalou, R., Sanni, M.A., Oumarou, A. and Soumaré, M.A. (eds) Les Sociétés Rurales Face aux Changements Climatiques et Environnemen- taux en Afrique de l’Ouest. IRD éditions, Marseille, pp. 135–152. de Haan, C., Dubern, E., Garancher and B., Quintero, C. (2016) Pastoralism development in the Sahel. A road to stability? World Bank Group, Washington, USA. de Leeuw, P.N. and Hiernaux P. (1990) Range resource evaluation. Module 6 of the Livestock research manual, ILCA Working paper 1. ILCA, Addis Ababa. de Leeuw, P.N., Diarra, L. and Hiernaux, P. (1992) An analysis of feed demand and supply for pastoral livestock in the Gourma region of Mali. In: Behnke, R.H. Jr, Scoones, I. and Kerven, C. (eds) Range Ecology at Disequilibrium. Proceedings of a meeting on Savanna Development and Pasture Produc- tion, November 1990, Woburn, UK. ODI, London, pp. 136–152. Delon, C., Mougin, E., Serça, D., Grippa M., Hiernaux, P., et al. (2015) Modelling the effect of soil moisture and organic matter degradation on biogenic NO emissions from soils in Sahel rangeland (Mali). Biogeosciences 12, 3253–3272. Desta, S. and Coppock, D.L. (2002) Cattle population in the southern Ethiopian rangelands, 1980–1997. Journal of Range Management 55, 439–451. Desta, S. and Coppock D.L. (2004) Pastoralism under pressure: tracking system change in Southern Ethiopia. Human Ecology 32, 465–486. Diarra, L., Hiernaux, P. and de Leeuw, P.N. (1995) Foraging behaviour of cattle grazing semi-arid range- lands in the Sahel, in Mali. In: Powell, J.M., Fernandez- Rivera, S., Williams, T.O. and Renard, C. (eds), Livestock and Sustainable Nutrient Cycling in Mixed Farming Systems of Sub-Saharan Africa. Pro- ceedings of an International Conference. ILCA, Addis Ababa, pp. 99–113. Diawara, M.O., Hiernaux, P., Mougin, E., Grippa, M., Delon, C. and Diakité, H.S. (2018) Effets de la pâture sur la dynamique de la végétation herbacée au Sahel (Gourma, Mali): une approche par modélisation. Cahiers Agriculture 27, 15010. Diouf, A., Sall, M., Wélé, A. and Dramé, M. (1998). Méthode d’échantillonnage de la production primaire sur le terrain Dakar. Document Technique du Centre de Suivi Écologique. Diouf, A.A., Hiernaux, P., Brandt, M., Faye, G., Djaby, B., et al. (2016) Do agrometeorological data improve optical satellite-based estimations of the herbaceous yield in Sahelian semi-arid ecosystems? Remote Sensing 8, 668. Djaby, B. (2010) Modélisation spatiale des flux organiques et minéraux assurant la productivité durable des systèmes culture-élevage dans le sahel nigérien. Thesis, University of Liège, Liège, Belgium. Ellis, J. and Galvin, K. (1994) Climate patterns and land-use practices in the dry zones of Africa. BioScience 44, 340–349. Ellis, J.E. and Swift, D.M. (1988) Stability of African pastoral systems: alternative paradigms and implications for development. Journal of Range Management 41, 450–459. Fensholt, R., Sandholt, I. and Rasmussen, M.S. (2004) Evaluation of MODIS LAI, FAPAR and the relation between FAPAR and NDVI in a semi-arid environment using in situ measurements. Remote Sensing of the Environment 91, 490–507. Rangeland Ecology 417 Fernández-Rivera, S., Hiernaux, P., Williams, T.O., Turner, M.D., Schlecht, E., et al. (2005). Nutritional constraints to grazing ruminants in the millet–cowpea–livestock farming system of the Sahel. In: Ayantunde, A.A., Fernández-Rivera, S. and McCrabb, G. (eds) Coping with Feed Scarcity in Small- holder Livestock Systems in Developing Countries. ILRI, Nairobi, pp. 157–182. Franklin, J. and Hiernaux, P.H.Y. (1991) Estimating foliage and woody biomass in Sahelian and Suda- nian woodlands using a remote sensing model. International Journal of Remote Sensing 12, 1387–1404. Gal, L., Grippa, M., Hiernaux, P., Peugeot, C., Mougin, E. and Kergoat, L. (2016) Changes in lakes water volume and runoff over ungauged Sahelian watershed. Journal of Hydrology 540, 1176–1188. Galaty, J.G. (1994a) Ha(l)ving land in common: the subdivision of Maasai group ranches in Kenya. Nomadic Peoples 34/35, 109–122. Galaty, J.G. (1994b) The pastoralist’s dilemma: common property and enclosure in Kenya’s rangelands. In: Vernooy, R. and Kealey, K.M. (eds) Food Systems under Stress in Africa. African–Canadian Research Cooperation: Proceedings of a Workshop, 7–8 November 1993. International Development Research Centre, Ottawa, pp. 100–114. Galaty, J.G. (2013a). Land grabbing in the eastern African rangelands. In: Catley, A., Lind, J. and Scoones, I. (eds) Pastoralism and Development in Africa: Dynamic Change at the Margins. Earthscan/ Routledge, London, pp. 131–142. Galaty, J.G. (2013b) The collapsing platform for pastoralism: land sales and land loss in Kajiado County, Kenya. Nomadic Peoples 17, 20–39. Gallais, J. (1984). Hommes du Sahel. Espaces-temps et pouvoirs. Le Delta intérieur du Niger, 1960–1980. Flammarion, Paris. Galvin, K.A. and Thornton, P.K. (2001) Human ecology, economics and pastoral household modelling. In: Boone, R.B. and Coughenour, M.B. (eds) A System for Integrated Management and Assessment of East African Pastoral Lands: Balancing Food Security, Wildlife Conservation, and Ecosystem Integrity. Final Report to the Global Livestock Collaborative Research Support Program. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colorado, pp. 105–123. Galvin, K.A., Ellis, J., Boone, R.B., Magennis, A.L., Smith, N.M., et al. (2002). Compatibility of pastoralism and conservation? A test case using integrated assessment in the Ngorongoro Conservation Area, Tanzania. In: Chatty, D. and Colester, M. (eds) Displacement, Forced Settlement and Conservation. Berghahn Books, Oxford. Galvin, K.A., Thornton, P.K., Boone, R.B. and Sunderland, J. (2004) Climate variability and impacts on East African livestock herders. African Journal of Range and Forage Science 21, 183–189. Galvin, K.A., Thornton, P.K., de Pinho, J., Sutherland, J. and Boone, R.B. (2006) Integrated modelling and its potential for resolving conflicts between conservation and people in the rangelands of East Africa. Human Ecology 34, 155–183. Galvin, K.A., Boone, R.B., BurnSilver, S.B. and Thornton, P.K. (2008a). Humans and wildlife as ecosystem components in integrated assessments. In: Manfredo, M.J., Vaske, J.J., Brown, P., Decker, D.J. and Duke, E.A. (eds) Displacement, Forced Settlement and Conservations. Island Press, Washington, DC, pp. 129–142. Galvin, K.A., Boone, R.B., Thornton, P.K. and Knapp, L.M. (2008b) Northwest Province, South Africa: communal and commercial livestock systems in transition. In: Galvin, K.A., Reid, R.S., Behnke, R.H. Jr and Hobbs, N.T. (eds) Fragmentation in Semi-Arid and Arid Landscapes. Springer, Dordrecht, Netherlands, pp. 281–304. Galvin, K.A., Thornton, P.K. and Boone, R.B. (2008c) Ngorongoro Conservation Area, Tanzania: fragmen- tation of a unique region of the Greater Serengeti Ecosystem. In: Galvin, K.A., Reid, R.S., Behnke, R.H. Jr and Hobbs, N.T. (eds) Fragmentation in Semi-Arid and Arid Landscapes. Springer, Dordrecht, Netherlands, pp. 255–279. Galvin, K.A., Reid, R.S., Behnke, R.H. Jr and Hobbs, N.T. (2008d). Fragmentation in Semi-Arid and Arid Landscapes. Dordrecht, Netherlands, Springer. Gandah, M., Brouwer, J., Hiernaux P. and van Duivenbooden N. (2003) Fertility management and landscape position: farmer’s use of nutrients sources in western Niger and possible improvements. Nutrient Cycling in Agroecosystems 67, 55–66. Gaston, A. and Boerwinkel, E. (1982). Production Primaire Herbacée Aérienne. (PISCEPS), Dakar. FAO, Rome (accessed 25 February 2020). Hanan, N.P., Prevost, Y., Diouf, A. and Diallo, O. (1991) Assessment of desertification around deep wells in the Sahel using satellite imagery. Journal of Applied Ecology 28, 173–186. 418 P. Ericksen et al. Haywood, M. (1981). Evolution de l’utilisation des terres et de la végétation dans la zone soudano-sahélienne du projet CIPEA au Mali CIPEA-ODEM. ILCA, Addis-Ababa. Henry, M., Picard, N., Trotta, C., Manlay, R.J., Valentini, R., et al. (2011) Estimating tree biomass of sub-Saharan African forests: a review of available allometric equations. Silva Fennica 45, 477–569. Hérault, B. and Hiernaux, P. (2004) Soil seed bank and vegetation dynamics in Sahelian fallows; the impact of past cropping and current grazing treatments. Journal of Tropical Ecology 20, 683–691. Hiernaux, P. (1980). L’inventaire du potentiel fourrager des arbres et arbustes d’une région du Sahel malien. Méthodes et premiers résultats. In: Le Houérou, H.N. (ed) Les Fourrages Ligneux en Afrique. Etat Actuel des Connaissances. ILCA, Addis Ababa, pp. 195–202. Hiernaux, P. (1982) Méthode d’évaluation du potentiel fourrager des parcours sahéliens proposée par les chercheurs du CIPEA. In: Breman H. (ed.) La Capacité de Charge des Pâturages Sahéliens pour les Systèmes d’Elevage de la Région. Compte Rendu de la Table Ronde, 26–27 January. CABO, Wageningen, Netherlands, pp. 49–53. Hiernaux, P. (1983) Recherche d’une Solution aux problèmes de l’Elevage dans le Delta Intérieur du Niger au Mali. Vol. 1: Les Pâturages de la zone d’Etude. ODEM, Sévaré, and CIPEA, Bamako. Hiernaux, P. (1984) Bilan d’une saison des pluies 1984 très déficitaire dans le Gourma (Sahel malien). Première campagne de suivi et télédétection expérimentale. Doc. Prog. AZ 140. CIPEA, Bamako. Hiernaux, P. (1988). Vegetation monitoring by remote sensing: progress in calibrating a radiometric index and its application in the Gourma, Mali. ILCA Bulletin, 32, 14–21. Hiernaux, P. (1996). The crisis of Sahelian pastoralism: ecological or economic? Pastoral Development Network Paper 39a. ODI, London. Hiernaux, P. (1998) Effects of grazing on plant species composition and spatial distribution in rangelands of the Sahel. Plant Ecology 138, 191–202. Hiernaux, P. (2004). Equilibrium and non-equilibrium behaviours of range vegetation dynamics in the Sahel. Professional workshop 01, VII International Rangeland Congress. Hiernaux, P. (2005). Etude et cartographie des ressources végétales et évaluation de leurs utilisations pas- torales dans le Nord mali Contribution au Système d’Information Géographique monté par Action contre la Faim. CESBIO/IRD/ACF, Bamako. Hiernaux, P. (2013). Les facteurs d’insécurité lies à la dégradation des ressources pastorales: pratiques d’exploitation des terres et changement climatique. In: La Contribution de l’Elevage Pastoral à la Sécurité et au Développement des Espaces Saharo-Sahéliens. Colloque régional de N’Djamena, 27–29 May. AFD, N’Djaména, Tchad. Hiernaux, P. (2016). Note de synthèse sur les outils disponibles pour le suivi-évaluation des ressources fourragères et l’établissement de bilans fourragers. Projet Régional d’appui au Pastoralisme au Sahel (PRAPS). CILSS, Ouagadougou. Hiernaux, P. and Haywood, M. (1978). La carte de végétation et des pâturages du ranch de Niono au 1/25000. CIPEA, Bamako. Hiernaux, P., Cissé, M.I., Diarra, L. and de Leeuw, P.N. (1994). Fluctuations saisonnières de la feuillaison des arbres et des buissons sahéliens. Conséquences pour la quantification des ressources four- ragères. Revue d’Elevage et de Médecine Vétérinaire dans les Pays Tropicaux 47, 117–125. Hiernaux, P., and Diarra, L. (1986) Bilan de cinq années de recherches (1979–1984) sur la production végé- tale des parcours des plaines d'inondation du Fleuve Niger du Mali central. ILCA, Bamako, Mali. Hiernaux, P., Diarra L., De Leeuw, P.N. (1994b) Modelling tillering of annual grasses as a function of plant dens- ity: application to sahelian rangelands productivity and dynamics. Agricultural Systems 46, 121–139. Hiernaux, P. and Turner, M.D. (1996) The effect of clipping on growth and nutrient uptake of Sahelian annual rangelands. Journal of Applied Ecology 33, 387–399. Hiernaux, P. and Turner, M.D. (2002). The influence of farmer and pastoralist management practices on desertification processes in the Sahel. In: Reynolds, J.F. and Smith, M.S. (eds) Global Desertifica- tion. Do Humans Cause Deserts? Dahlem University Press, Berlin, pp. 135–148. Hiernaux, P. and Ayantunde, A. (2004). The Fakara: a semi-arid agro-ecosystem under stress. Report of research activities of International Livestock Research Institute (ILRI) in Fakara, South-western Niger, between 1994 and 2002, submitted to Desert Margins Program. ICRISAT, Niamey, Niger. Hiernaux, P. and Le Houérou, H.N. (2006) Les parcours du Sahel. Sècheresse 17, 51–71. Hiernaux, P., Ayantunde, A., Kalilou, A., Mougin, E., Gérard, B. et al. (2009a) Trends in productivity of crops, fallow and rangelands in Southwest Niger: impact of land use, management and variable rainfall. Journal of Hydrology 375, 65–77. Rangeland Ecology 419 Hiernaux, P., Diarra, L., Trichon V., Mougin, E. and Baup F. (2009b) Woody plant population dynamics in response to climate changes from 1984 to 2006 in Sahel (Gourma, Mali). Journal of Hydrology 375, 103–113. Hiernaux, P., Mougin, E., Diarra, L., Soumaguel, N., Lavenu, F., et al. (2009c) Sahelian rangeland response to changes in rainfall over two decades in the Gourma region, Mali. Journal of Hydrology 375, 114–127. Hiernaux, P. and Diawara, M.O. (2014) Livestock: recyclers that promote the sustainability of smallholder farms. Rural 21, 9–11. Hiernaux, P., Diawara, M.O. and Gangneron, F. (2014) Quelle accessibilité aux ressources pastorales du Sahel? L’élevage face aux variations climatiques et aux évolutions des sociétés sahéliennes. Afrique Contemporaine 249, 21–35. Hiernaux, P., Diawara, M.O., Kergoat, L. and Mougin, E. (2015). La contrainte fourragère des élevages pastoraux et agro-pastoraux du Sahel. Adaptations et perspectives. In: Sultan, B., Lalou, R., Sanni, M.A., Oumarou, A. and Soumaré, M.A. (eds) Les Sociétés Rurales Face aux Changements Clima- tiques et Environnementaux en Afrique de l’Ouest. IRD editions, Marseille, pp. 171–192. Hiernaux, P., Dardel, C., Kergoat, L. and Mougin, E. (2016). Desertification, adaptation and resilience in the Sahel: lessons from long term monitoring of agro-ecosystems. In: Behnke, R.H. and Mortimore, M. (eds) The End of Desertification. Springer-Verlag, Berlin/Heidelberg, pp. 147–178. Hobbs, N.T., Galvin, K.A., Stokes, C.J., Lackett, J.M., Ash, A.J., et al. (2008) Fragmentation of rangelands: implications for humans, animals, and landscapes. Global Environmental Change 18, 776–785. Hoffmann, I., Hof, A., Hiernaux, P. and Malami, B.S. (2006) Study on the Use Intensity of the Woody Layer in the Zamfara Reserve. International Conference on the Future of Transhumance Pastoralism in West and Central Africa, 21–25 November, Abuja, Nigeria. Homewood, K. (2008). Ecology of African Pastoralist Societies. James Curry, Oxford. Homewood, K., Lambin, E.F., Coast, E., Kariuki, A., Kikula, I., et al. (2001) Long-term changes in Serengeti-Mara wildebeest and land cover: pastoralism, population, or policies? Proceedings of the National Academy of Sciences USA 98, 12544–12549. Homewood, K., Kristjanson, P. and Pippa, T. (eds) (2009). Staying Maasai: Livelihoods, Conservation and Development in East African Rangelands. Springer, New York. ILCA (1975) Evaluation and Mapping of Tropical African Rangelands. Proceedings of the Seminar, 3–8 March, Bamako, Mali. ILCA, Addis Ababa. Jacques, D.C., Kergoat, L., Hiernaux, P., Mougin, E. and Defourny, P. (2014) Monitoring dry vegetation masses in semi-arid areas with MODIS SWIR bands. Remote Sensing of Environment 153, 40–49. Kergoat, L., Hiernaux, P., Dardel, C., Pierre, C., Guichard, F. and Kalilou, A. (2015) Dry-season vegetation mass and cover fraction from SWIR1.6 and SWIR2.1 band ratio: ground-radiometer and MODIS data in the Sahel. International Journal of Applied Earth Observation and Geoinformation 39, 56–64. Kiema, A. (2015). Etude d’évaluation de l’état général des ressources pastorales au Burkina Faso. Initiative mondiale pour un pastoralisme durable. Projet CPIG-DANIDA-WAP, Ouagadougou. La Rovere, R., Hiernaux, P., van Keulen, H., Schiere, J.B. and Szonyi, J.A. (2005) Co-evolutionary scen- arios of resource use change and intensification in rural communities of south-western Niger. Agricul- tural Systems 83, 251–576. Lamprey, R.H. and Reid, R.S. (2004) Expansion of human settlement in Kenya’s Maasai Mara: what future for pastoralism and wildlife? Journal of Biogeography 31, 997–1032. Le Dantec, V., Timouk, F., Hiernaux, P., Kergoat, L. and Mougin, E. (2009). Net CO2 and water exchanges in a semi-arid region (Gourma, Mali). In: AMMA International Conference, 20–24 July, Ouagadougou. Le Houérou, H.N. (1980). Browse in Africa: The Current State of Knowledge. ILCA, Addis Ababa. Le Houérou, H.N. (1989). The Grazing Land Ecosystems of the African Sahel. Springer, Berlin/Heidelberg. Léauthaud, C., Kergoat, L., Hiernaux, P., Grippa, M., Musila, W., et al. (2018) Modelling the growth of flood- plain grasslands to explore the impact of changing hydrological conditions on vegetation productivity. Ecological Modelling 387, 220–237. Lesnoff, M., Corniaux, C. and Hiernaux, P. (2012) Sensitivity analysis of the recovery dynamics of a cattle population following drought in the Sahel region. Ecological Modelling 232, 28–39. Lo Seen, D., Mougin, E., Rambal, S., Gaston, A. and Hiernaux, P. (1995) A regional Sahelian grassland model to be coupled with multispectral satellite data. 2: toward the control of its simulations by remotely sensed indices. Remote Sensing of Environment 52, 194–206. Maitima, J., Reid, R.S., Gachimbi, L.N., Majule, A., Lyaruu, H., et al. (2004) The linkages between land use change, land degradation and biodiversity across East Africa. LUCID Working Paper Series No. 42. ILRI, Nairobi. 420 P. Ericksen et al. Marie, J. (2000). DELMASIG: Hommes, Milieux, Enjeux Spatiaux et Fonciers dans le Delta Intérieur du Niger, HDR, University of Paris X, Paris. McCabe, J.T. (2003) Sustainability and Livelihood Diversification Among the Maasai of Northern Tanzania. Human Organization 62, 100–111. McCabe, J.T., Leslie, P.W. and DeLuca, L. (2010) Adopting cultivation to remain pastoralists: the diversification of Maasai livelihoods in northern Tanzania. Human Ecology 32, 321–334. Miehe, S., Kluge, J., von Wehrden, H. and Retzer, V. (2010). Long-term degradation of Sahelian rangeland detected by 27 years of field study in Senegal. Journal of Applied Ecology 47, 692–700. Milligan, K. (1982). Aerial survey of human, livestock and environmental conditions in a central region of the pastoral zone of Niger. Final report. USAID, Niamey. Milligan, K., Keita, M. and de Leeuw, P.N. (1982) Recensement aerien saisonnier du cheptel et types de paysage du delta central du Niger au Mali. ODEM/IBRD, Sevare, Mali. Milne, E., Aynekulu, E., Bationo, A., Batjes, N., Boone, R. et al. (2016) Grazing lands in Sub-Saharan Africa and their potential role in climate change mitigation: What we do and don’t know. Environmental Development 19, 70–74. https://doi.org/10.1016/j.envdev.2016.06.001. (http://www.sciencedirect.com/ science/article/pii/S2211464516300355) Monod, T. (ed)(1975). Les Sociétés Pastorales en Afrique Tropicale. Oxford University Press, London. Montgolfier-Kouèvi, C. and Le Houérou, H.N. (1980). Study on the economic viability of browse plantations in Africa. In: Le, Houérou, H. (ed) ILCA, Addis Ababa, pp. 7–24. Mougin, E., Hiernaux, P., Kergoat, L., Grippa, M., de Rosnay, P., et al. (2009) The AMMA Gourma observa- tory site in Mali: climatic variations to changes in vegetation, surface hydrology, fluxes and natural re- sources. Journal of Hydrology 375, 14–33. Mougin E., Lo Seen, D., Rambal, S., Gaston, A. and Hiernaux, P. (1995) A regional Sahelian grassland model to be coupled with multispectral satellite data. 1: validation. Remote Sensing of Environment 52, 181–193. Nguyen, C.C. (2015). Dynamique, structure et production de la végétation du Gourma (Sahel, Mali) en re- lation avec les sols, l’occupation des sols et les systèmes hydriques: étude de télédétection à haute et moyenne résolution. Thesis, Université Toulouse III Paul Sabatier, Toulouse, France. Niamir-Fuller, M. (ed) (1999). Managing Mobility in African Rangelands: The Legitimization of Transhu- mance. Practical Action Publishing, Rugby, UK. Nicholson, S.E. (2000) The nature of rainfall variability over Africa on time scales of decades to millennia. Global and Planetary Change 26, 137–158. Nicholson, S.E., Funk, C. and Fink, A.H. (2018) Rainfall over the African continent from the 10th through the 21st century. Global and Planetary Change 165, 114–127. Oba, G. (2013). The sustainability of pastoral production in Africa. In: Catley, A., Lind, J. and Scoones, I. (eds) Pastoralism and Development in Africa: Dynamic Change at the Margins. Earthscan/Routledge, London. Oba, G., Stenseth, N.C. and Lusigi, W.G. (2000) New perspectives on sustainable grazing management in arid zones of sub-Saharan Africa. BioScience 50, 35–51. ODEM/CIPEA (1983) Opération de développement de l’élevage dans la région de Mopti/Centre International pour l’élevage en Afrique. Recherche d’une solution aux problèmes de l’élevage dans le delta intérieur du Niger au Mali. ODEM/CIPEA, Addis Ababa. Ogutu, J.O. (2013) Changing wildlife populations in Nairobi National Park and adjoining Athi-Kaputiei plains: collapse of the migratory wildebeest. Open Conservation Biology Journal 7, 11–26. Ogutu, J.O., Piepho, H.-P., Dublin, H.T., Bhola, N. and Reid, R.S. (2009) Dynamics of Mara-Serengeti ungulates in relation to land use changes. Journal of Zoology 278, 1–14. Ogutu, J.O., Owen-Smith, N., Piepho, H.P. and Said, M.Y. (2011) Continuing wildlife population declines and range contraction in the Mara region of Kenya during 1977–2009. Journal of Zoology 285, 99–109. Ogutu, J.O., Kuloba, B., Piepho, H.P. and Kanga, E. (2017) Wildlife population dynamics in human-dominated landscapes under community-based conservation: the example of Nakuru Wildlife Conservancy, Kenya. PLoS One 12, e0169730. Olsen, J.M., Butt, B., Atieno, F., Maitima, J.M., Smucker, T.A., et al. (2004). Multi-Scale Analysis of Land Use and Management Change on the Eastern Slopes of Mt. In: Kenya. ILRI, Nairobi. Osano, P. (2013). A preliminary evaluation of the wildlife conservation lease demonstration project in the Athi-Kaputei Plains, Southern Kenya: key results, achievements and lessons learnt. Project Evaluation Report, The Wildlife Foundation, Nairobi. Rangeland Ecology 421 Parton, W.J., Scurlock, J.M.O, Ojima, D.S., Gilmanov, T.G., Scholes, R.J., et al. (1993) Observations and modelling of biomass and soil organic matter dynamics for the grassland biome worldwide. Global Biogeochemical Cycles 7, 785–809. Penning de Vries, F.W.T and Djiteye, M.A. (eds) (1991) La Productivité des Pâturages Sahéliens: Une Etude des Sols, des Végétations et de l’Exploitation de Cette Ressource Naturelle, 2nd edn. Centre for Agricultural Publishing and Documentation, Wageningen, Netherlands. Pierre, C., Kergoat, L., Bergametti, G., Mougin, E., Baron, C., et al. (2015) Modeling vegetation and wind erosion from a millet field and from a rangeland: two Sahelian case studies. Aeolian Research 19, 97–111. Pollard, J.H. (1971) On distance estimators of density in randomly distributed forests. Biometrics 27, 991–1002. Powell, M.J., Pearson, R.A. and Hiernaux, P. (2004) Crop-livestock interactions in the West African drylands. Agronomy Journal 96, 469–483. Pratt, D.J. and Gwynne, M.D. (eds) (1977). Rangeland Ecology and Management in East Africa. Hodder & Stoughton, London. Reid, R.S., Gachimbi, L.N., Worden, J., Wangui, E.E., Mathai, S., et al. (2004) Linkage between changes in land use, biodiversity and land degradation in the Loitokitok area of Kenya. LUCID Project Working Paper 49. ILRI, Nairobi. Reid, R.S., Gichohi, H., Said, M.Y., Nkedianye, D., Ogutu, J.O., et al. (2008) Fragmentation of a peri-urban savanna, Athi-Kaputiei Plans, Kenya. In: Galvin, K.A., Reid, R.S., Behnke, R.H. Jr and Hobbs, N.T. (eds) Fragmentation in Semi-Arid and Arid Landscapes. Springer, Dordrecht, Netherlands, pp. 195–224. Reid, R.S. (2012). Savannas of our Birth: People, Wildlife, and Change in East Africa. University of California Press, Berkeley, California. Reid, R.S., Nkedianye, D., Said, M.Y., Kaelo, D., Neselle, M., et al. (2016) Evolution of models to support community and policy action with science: balancing pastoral livelihoods and wildlife conservation in savannas of East Africa. Proceedings of the National Academy of Sciences USA 113, 4579–4584. Reij, C., Tappan, G. and Smale, M. (2009) Agroenvironmental transformation in the Sahel: another kind of ‘Green Revolution’. IFPRI Discussion Paper 00914. IFPRI, Washington, DC. Sandford, S. (1983). Management of Pastoral Development in the Third World. Wiley, Chichester, UK. Sankaran, M., Hanan, N.P., Scholes, R.J., Ratnam, J., Augustine, D.J., et al. (2005) Determinants of woody cover in African savannas. Nature 438, 846–849. Schlecht, E., Fernandez-Rivera, S. and Hiernaux, P. (1998) Timing, size and N-concentration of faecal and urinary excretions in cattle, sheep and goats - Can they be used for better manuring of cropland? In: Soil Fertility Management in West African Land Use Systems. Proceedings of the Regional Workshop University of Hohenheim, ICRISAT Sahelian Centre and INRAN, 4–8 March 1997, Niamey, Niger, pp. 361–368. Schlecht, E., Hiernaux, P. and Turner, M.D. (2001) Mobilité régionale du bétail: nécessité et alternatives? In: Tielkes, E., Schlecht E. and Hiernaux P. (eds) Elevage et Gestion de Parcours au Sahel, Implications Pour le développement. Verlag E. Grauer, Stuttgart, Germany, pp. 291–302. Schlecht, E., Hiernaux, P., Achard, F. and Turner M. (2004) Livestock related nutrient budgets within village territories in western Niger. Nutrient Cycling in Agroecosystems 68, 199–211. Schlecht, E., Hiernaux, P., Kadaouré, I., Hülsebusch C. and Mahler, F. (2006) A spatio-temporal analysis of forage availability and grazing and excretion behaviour of herded and free grazing cattle, sheep and goats in Western Niger. Agriculture, Ecosystems and Environment 113, 226–242. Scoones, I. (1995) Exploiting heterogeneity: habitat use by cattle in drylands Zimbabwe. Journal of Arid Environments 29, 221–237. Sendzimir, J., Reij, C.P. and Magnuszewski, P. (2011) Rebuilding resilience in the Sahel: Regreening in the Maradi and Zinder regions of Niger. Ecology and Society 16, 1. Solomon Bekure, S., de Leeuw, P., Grandin, B.E. and Neate, P.J.H. (eds) (1991) Maasai Herding: An Ana- lysis of the Livestock Production System of Maasai Pastoralists in Eastern Kajiado District, Kenya. ILCA Systems Study No. 4. ILCA, Addis Ababa. Thornton, P.K., Galvin, K.A. and Boone, R.B. (2003) An agro-pastoral household model for the rangelands of East Africa. Agricultural Systems 76, 601–622. Thornton, P.K., Fawcett, R.H., Galvin, K.A., Boone, R.B., Hudson, J.W. and Vogel, C.H. (2004) Evaluating management options that use climate forecasts: modelling livestock production systems in the semi-arid zone of South Africa. Climate Research 26, 33–42. Thornton, P.K., Burnsilver, S.B., Boone, R.B. and Galvin, K.A. (2006) Modelling the impacts of group ranch subdivision on households in Kajiado, Kenya. Agricultural Systems 87, 331–356. 422 P. Ericksen et al. Thornton, P.K., Boone, R.B., Galvin, K.A., BurnSilver, S.B., Waithaka, M.M., et al. (2007) Coping strategies in livestock-dependent households in East and southern Africa: a synthesis of four case studies. Human Ecology 35, 461–476. Thornton, P.K., Boone, R.B. and Ramirez-Villegas, J. (2015) Climate change impacts on livestock. UNFCCC SBSTA submission for the African Group of Negotiators, March 2015. Working Paper 120. CCAFS, Copenhagen. Thornton, P.K., Herrero, M. and Boone, R.B. (2019). Altered grazing systems: pastoralism to conventional agriculture. In: Gibson, D.J. and Newman, J. (eds) Grasslands and Climate Change. Cambridge Uni- versity Press, Cambridge, pp. 257–275. Tong, X., Brandt, M., Hiernaux, P., Herrmann, S.M., Tian, F., et al. (2017) Revisiting the coupling between NDVI trends and cropland changes in the Sahel drylands: a case study in western Niger. Remote Sensing of Environment 191, 286–296. Tong, X., Brandt, M., Hiernaux, P., Herrmann, S.,Vang Rasmussen, L., et al. (2020) The forgotten land use class: mapping of fallow fields across the Sahel using Sentinel-2. Remote Sensing of Environment 239, 111598. Toulmin, C. (1987) Drought and the farming sector: loss of animals and post-drought rehabilitation. Devel- opment Policy Review 5, 125–148. Toulmin, C., Delville, P.L. and Traore, S. (2002). The dynamics of resource tenure in West Africa. James Currey, Oxford. Touré, I., Ickowicz, A., Wane, A., Garba, I. and Gerber, P. (2012). Atlas of Trends in Pastoral Systems in Sahel. FAO/CIRAD, Montpellier, France. Tracol, Y., Mougin, E., Hiernaux, P. and Jarlan, L. (2006) Testing a Sahelian grassland functioning model against herbage mass measurements. Ecological Modeling 193, 437–446. Tucker, C.J., Vanpraet, C.L., Sharman, M.J. and Ittersum, G. (1985) Satellite remote sensing of total herbaceous biomass production in the Senegalese Sahel: 1980–1984. Remote Sensing of Environment 17, 233–249. Turner, M.D. (1992). Living on the edge: FulBe herding practices and the relationship between economy and ecology in the Inland Niger Delta of Mali. PhD thesis. University of California, Berkeley, California. Turner, M.D. (1999) Merging local and regional analyses of land-use change: the case of livestock in the Sahel. Annals of the Association of American Geographers 89, 192–219. Turner, M.D. (2017) Livestock mobility and the territorial state: south-western Niger (1890–1920). Africa 87, 578–606. Turner, M. and Hiernaux, P. (2002) The use of herders’ accounts to map livestock activities across agropastoral landscapes in semi-arid Africa. Landscape Ecology 17, 367–385. Turner, M.D., Hiernaux, P. and Schlecht, E. (2005) The distribution of grazing pressure in relation to vegetation resources in semi-arid West-Africa: the role of herding. Ecosystems 8, 668–681. Turner, M.D., Ayantunde, A.A., Patterson, K.P. and Patterson, E.D. (2012) Conflict management, decentralization and agropastoralism in Niger. World Development 40, 745–757. Turner, M.D., McPeak, J.G. and Ayantunde, A.A. (2014) The role of livestock mobility in the livelihood strategies of rural peoples in semi-arid West Africa. Human Ecology 42, 231–247. Turner, M. D. and Hiernaux, P. (2015) The effects of management history and landscape position on inter- field variation in soil fertility and millet yields in southwestern Niger. Agriculture, Ecosystem and Environment 211, 73–83. Turner, M. D., McPeak, J.G., Gillin, K., Kitchell, E. and Kimambo, N. (2016) Reconciling flexibility and tenure security for pastoral resources: the geography of transhumance networks in eastern Senegal. Human Ecology 44, 199–215. Valenza, J. (1984). Surveillance continue des pâturages naturels sahéliens sénégalais. Résultats de 10 années d’observation. Agrostologie 44. ISRA-LNERV Dakar. Vetter, S. (2005) Rangelands at equilibrium and non-equilibrium: recent developments in the debate. Journal of Arid Environments 62, 321–341. Wilson, R.T. (1986). Livestock production in central Mali: long-term studies on cattle and small ruminants in the agropastoral system. ILRI, Nairobi. Wilson, R.T., de Leeuw, P.N. and de Haan, C. (eds) (1983) Recherches sur les systèmes des zones arides du Mali: résultats préliminaires. Research Report No. 5. ILCA, Addis Ababa. Wilson, T., Hiernaux, P. and McIntire, J. (1988). Systems research in the West African arid and semi-arid zones: a synthesis of results from 1976 to 1986. ILCA, Addis Ababa. 12 Forage Diversity, Conservation and Use Jean Hanson1, Rainer Schultze-Kraft2, Michael Peters3, Peter Wenzl2, Ahmed Amri4, Ali Shehadeh5 and Mariana Yazbek5 1International Livestock Research Institute, Addis Ababa, Ethiopia; 2Alliance of Bioversity and CIAT, Cali, Colombia; 3Alliance of Bioversity and CIAT, Nairobi, Kenya; 4International Center for Agricultural Research in the Dry Areas, Rabat, Morocco; 5International Center for Agricultural Research in the Dry Areas, Terbol, Lebanon Contents Executive Summary 424 The problem 424 Scientific impacts 424 Collections 424 Characterization 424 Distribution 425 Scientific gaps 425 Development impacts 425 Uptake summary 425 Preservation of option value 425 Germplasm distribution as a proxy for use 426 Development gaps 426 Capacity building and partnerships as development impacts 426 The future 426 Introduction 427 Coverage of the International Forage Collections in CGIAR 427 Forage germplasm maintained in the ILRI gene bank 428 Forage germplasm maintained in the CIAT gene bank 428 Forage germplasm maintained in the ICARDA gene bank 429 Gaps in the germplasm collections 430 Describing Forage Traits in Germplasm to Support Use 431 Value of Forage Germplasm 433 Impact of the CGIAR Forage Gene Banks 433 Forage germplasm distribution 433 Estimating economic benefits of forage gene banks 435 Costing gene bank operations 435 Management of gene banks based on economics 436 ILRI example of a gene bank costing study 437 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 423 424 J. Hanson et al. Benefits from use of germplasm 438 Estimating scientific and field benefits of the CGIAR forage gene banks 438 Success stories in the use of forage germplasm 438 Example of impact from use of Napier grass germplasm from ILRI 440 Capacity building 441 Conclusions 443 Development impact 443 Potential option values 443 Scientific impact 443 Environmental benefits 443 Basic and applied research 444 The Future 444 References 446 Executive Summary Collections The three collections at ILRI, CIAT and ICARDA The problem comprise an estimated 72,000 distinct accessions of 1600 species of tropical and subtropical for- Managing biological diversity has long been age grasses and herbaceous and woody leg- r ecognized as an essential component of crop im- umes. With the large number of plants that can provement to ensure a pool of genetic material for be used as forage, this is likely to be less than selection and breeding for current and future 20% of global tropical forage biodiversity. The needs arising from agricultural development, regional origins of the collections are about population growth and growing food demand. 30% each from Latin America and the Carib- Tropical and subtropical forage genetic resources bean, West Asia and North Africa, and are particularly important because with few trop- sub-Saharan Africa, and the remaining 10% ical forage breeding programmes of limited spe- mostly from tropical Asia and South-east Asia. cies coverage, forage germplasm remains the basis Much of the forage diversity is found in the for selection and development of new feeds. With rangelands of sub-Saharan Africa, and conser- increasing demands for feeds for livestock intensi- vation efforts including protected areas for wild- fication, rapid rates of genetic erosion of forage life conservation and tourism will protect diversity in many regions and the need to select remaining forage diversity in situ for the foresee- new higher-yielding and better-adapted geno- able future. These collections constitute stocks types, the in-trust forage collections held in the of current and future value that can be used in International Livestock Research Institute (ILRI), crop and animal production. the International Center for Tropical Agriculture (CIAT) now part of the Alliance of Bioversity International and CIAT after the first mention of Characterization CIAT and the International Centre for Agricul- The CGIAR forage germplasm collection is con- tural Research in Dry Areas (ICARDA) are cru- tinuously characterized for morphological traits, cial. The knowledge generated from research on with about 60% already completed and species forage diversity allows scientists to identify geno- with high potential as feed (grasses: Brachiaria, types with higher potential and to support innova- Cenchrus, Chloris, Panicum and Pennisetum spp.; tive genotype s election and breeding programmes. herbaceous legumes: Arachis, Centrosema, Des- modium, Lablab, Medicago, Stylosanthes, Vicia and Vigna spp.; fodder shrubs and trees: Cratylia, Des- Scientific impacts manthus, Flemingia, Leucaena and Sesbania spp.) already evaluated for biomass production, nutri- The principal scientific impacts have been in tional value and some adaptive traits, often via germplasm collection, characterization and multilocational trials with national partners. distribution. Species characterization and evaluation have Forage Diversity, Conservation and Use 425 identified promising genotypes, some more pro- expected to achieve greater efficiencies and lead to ductive than current named cultivars, for a greater impact through increased use of the for- range of environments. The information gener- age genetic resources. The continued long-term ated has been used to identify genotypes for core safety and availability of the CGIAR forage collec- collections, for use in selecting parents for grass tions is being ensured by placing a safety dupli- breeding programmes and for users to better cate in an alternative location. A second safety select genotypes best suited to their needs. duplicate is also being placed in the Svalbard Technical advances in collection manage- Global Seed Vault and currently 50% of the ment have been possible as information has been collections are stored there. generated on environmental adaptation, seed production and seed longevity. Improvements in Scientific gaps procedures, such as breaking seed dormancy and seed production, and understanding the Despite substantial effort to characterize and longevity of forage seeds has led to more efficient evaluate materials in the collections, there is in- management of the large collections held in sufficient information on performance, disease CGIAR. Costing of gene bank operation method- resistance and drought resistance in many spe- ology d eveloped by the Systemwide Genetic Re- cies. The gene banks have assigned Digital Object- sources Programme and applied to the different ive Identifiers to the accessions to track germplasm areas of gene bank operations has shown the use while strengthening characterization work areas of highest cost and risk. This allows man- on the collections held in trust. agement to focus on these areas where efficiencies and savings can be made without compromising genetic integrity and long-term conservation of Development impacts the germplasm. Gene bank management of forages is a Uptake summary complex operation because of the large number The link from distribution from the gene bank to of wild species and the diversity of morphotypes farm use is often undocumented because the ac- and botanical species that are used as forages. cessions are first passed to research programmes Knowledge management tools such as the Crop and from there are further distributed to farm- Genebank Knowledge Base have been developed ers. Forages differ from crops in that most com- to share knowledge from the CGIAR centres and mon forages are selections from germplasm and assist gene bank managers both inside and out- not the result of breeding programmes. This dir- side CGIAR to better manage their collections. ect use should make it easier to track selections, Forage networks have made significant contri- but documentation is still scarce. Some success butions to forage selection and use, and sub- has been realized with large-scale adoption of regional networks continue to play a key role in Brachiaria and Panicum in Latin America and the studying the diversity in the collections and evalu- Caribbean, Napier grass in sub-Saharan Africa, ating genotypes for adaptation and utility as and medics and vetch in the dry areas (see livestock feed in a wide range of environments. Chapter 13, this volume). Distribution Preservation of option value An estimated 138,000 samples have been dis- Conservation of diversity in gene banks is a tributed from the CGIAR collection to 188 coun- long-term activity that needs to be resourced to tries from the gene banks, in addition to over ensure continued availability of germplasm for fu- 55,000 samples made available internally ture use. A return on the investment made can be within the centres for forage research and breed- quantified as the impact from actual and potential ing. Opportunities for better alignment of oper- use of the germplasm for livestock feed with re- ations and synergies between the CIAT and ILRI lated increases in on-farm income from sale of forage collections through coordinated curation livestock products. This can be directly related to and harmonization under the CGIAR Genebank the costs of conservation. However, much of the Platform were identified in 2018. This is value of the gene bank lies in the future. Option 426 J. Hanson et al. values are particularly important for germplasm technical capacities. Much of this training has in ex situ collections, which are natural resources, been done as part of scientific networks to and global public goods, which the world commu- strengthen the capacity of partners to evaluate nity relies on for their existence as a resource for and select forages for use in smallholder live- future needs. The option value of having the ger- stock systems. ILRI, CIAT and ICARDA have mplasm available to respond to as-yet-unidentified been working closely together under the System- future needs, such as changing feed needs due to wide Genetic Resources Programme and have climate change, and the existence value that soci- developed training materials and guidelines for ety derives from knowing that something exists gene bank management, including the sections and will be available for future needs, may be on forages in the Crop Genebank Knowledge more powerful economic drivers of conservation Base. ILRI and CIAT, along with the Common- of forage diversity than the cost–benefit analyses wealth Scientific and Industrial Research Organ- that can be done on its actual use. isation (CSIRO) and the Department of Primary Industry and Fisheries, Queensland Government, Germplasm distribution as a proxy for use were also founding partners in the development of the online Tropical Forages tool to support for- Using germplasm distribution data as a proxy for age selection. use is complex because many of the samples dis- tributed will not meet user needs or show super- ior performance in the intended location. Only a small percentage of the materials received will The future be taken forward from evaluation to adoption and use. Estimates from past variety develop- Over the past 30  years, the forage programmes ment in CIAT indicate that about 1% of the grass and gene banks in CGIAR have focused on build- collection and less than 0.2% of the herbaceous ing their germplasm collections to cover a broad legumes have been released as named varieties. range of genetic diversity of tropical and subtrop- ical forages to meet user demands, studying the Development gaps collections to characterize and evaluate d iversity and understand the traits in specific accessions, Barriers to development impacts include: (i) legal and improving their facilities to securely con- restrictions on new collections with new condi- serve the germplasm. Going forward, the focus tions for access to germplasm under the Nagoya will continue to be on the core operations that Protocol on Access and Benefit-sharing (Nagoya are essential to conserve and manage the diver- Protocol), which may make reaching agreement sity but with increased emphasis on efficiency on collection and sharing of germplasm a more and value for money in gene bank operations and lengthy process; (ii) costs of maintaining and on demonstrating the value and impact of that characterizing collections to make them more investment. This will include the following: accessible to users and prioritizing which species future users are likely to demand; (iii) high un- • Quantifying the diversity within the collec- certainty about options value, even for species tions to understand how much of the genetic that have been well characterized; and (iv) high diversity in key species is already held ex situ uncertainty about non-market benefits. and how much remains at threat of genetic erosion in the field where attention needs to be Capacity building and partnerships given. This will involve working with partners as development impacts in national forage gene banks to elucidate the geographical origin of populations, identify Improving the capacity of partners in gene bank gaps and determine priorities to fill them. management and forage research and develop- • Describing the diversity of important forage ment will have lasting impact. The CGIAR forage traits using genomic tools to understand gene bank programmes have provided training the relationships among traits and identify to over 1000 scientists and technicians in ger- genes responsible for particularly import- mplasm management, forage evaluation and ant plant traits that can be used in forage forage seed production to build scientific and selection and breeding. Forage Diversity, Conservation and Use 427 • Making more and better use of the collec- and they have limited species coverage (Sandhu tions by posting information about the col- et al., 2015). With increasing feed demands from lection and traits in specific accessions in livestock intensification, rapid genetic erosion of global genetic resources web databases forage diversity in many regions and the need to such as Genesys, ensuring sufficient disease- select new material, the forage collections held free seeds for distribution and sharing at ILRI (beginning in 1983), CIAT (beginning in knowledge on forages through the online 1972) and ICARDA (beginning in 1985) are key forage selection tool. to feed development. The knowledge generated • Improving efficiency and effective man- from research on forage diversity allows scien- agement of the germplasm by identifying tists to identify genotypes that have higher feed and eliminating genetic duplicates from potential and are adapted to changing environ- the collections, using genomic tools and mental conditions, and to support plant breed- identifying core collections for further ing programmes in the generation of new focus and study. material (particularly with tolerance to heat, • Improving facilities and procedures in la- droughts and flooding, and pests and diseases), boratories that support more extensive use which are becoming of increasing importance of genomic tools and increased use of eco- to respond to climate change. friendly energy. Solar and wind energy This chapter outlines the impact of forage can both reduce the environmental foot- diversity conservation, characterization and dis- print of gene banks and can also substan- tribution work under the international network tially reduce the costs of conservation and of forage collections in CGIAR. will be used more extensively in improved facilities. Coverage of the International • Continuing a gender-balanced capacity- Forage Collections in CGIAR building effort to support global efforts in forage conservation and sustainable use and make better use of the collections. This will There are three major international forage gen- include training, internships and graduate etic resources collections held in CGIAR at CIAT scholarships, mentoring, developing and (Cali, Colombia, established in 1972), ICARDA supporting communities of practice, online (Aleppo, Syria, established in 1985) and ILRI training and knowledge sharing. (Addis Ababa, Ethiopia, established in 1983) • Working closely with the CGIAR Research with an estimated total of 72,000 distinct acces- Programme (CRP) on Livestock, informa- sions of 1600 species. These collections are held tion will be collected on user demand for ‘in trust’ as international public goods. The lar- forages and traits, documenting the deliv- gest collection is held by ICARDA with 38,955 ery pathways from the gene bank to forages accessions of 447 species and includes subtrop- production and adoption of forages on farm ical and Mediterranean forage legumes adapted to better document the impact and use of to dry areas. The CIAT gene bank conserves the forage collections. These data will be more than 23,000 forage accessions of 737 spe- used to estimate current and option v alues. cies with a large collection of forage legume ger- mplasm adapted to low-fertility acid soils. ILRI conserves 18,672 accessions representing about Introduction1 1600 species, including the major collection of African grasses and tropical highland forages, as Identifying and managing biological diversity is well as a large collection of tropical herbaceous a component of crop improvement to ensure a forage legumes. The World Agroforestry Centre pool of genetic material for selection and breed- (ICRAF) holds a collection of fodder tree ger- ing for current and future needs arising from mplasm, which complements the collections land scarcity, population growth and the in- held in CIAT and ILRI. creasing demand for food (Hawkes, 1971). Trop- The germplasm of CGIAR is governed ical and subtropical forage genetic resources are under the Multilateral System of access and particularly important because few tropical benefit sharing of the International Treaty on forage breeding programmes remain in existence, Plant Genetic Resources for Food and Agri culture 428 J. Hanson et al. (ITPGRFA). Most of the accessions were acquired Forage germplasm maintained before the entry into force of the Convention on in the ILRI gene bank Biological Diversity (CBD) on 29 December 1993 and are in the public domain. Collections ILRI began assembling forage germplasm for use were made with the understanding that all the in African farming systems in 1983 as the Inter- materials could be made freely available for any national Livestock Centre for Africa (ILCA). Since agricultural research, breeding and training its establishment, the collection had grown to purposes. The germplasm was placed in trust as 18,672 accessions in early 2020. Tropical a global public good under agreements between grasses, herbaceous legumes and trees are con- the Centres and the Food and Agriculture Or- served, representing 426 genera. ILRI collected ganization of the United Nations (FAO) in Octo- forage germplasm in sub-Saharan Africa from ber 1994. In October 2006, the Centres then 1983 to 1993, although the majority of acces- signed an agreement to place the germplasm sions were donated by national partners. A major held in trust under the FAO agreement under donor has been CSIRO, Australia, which gave al- the Multilateral System of the ITPGRFA. The most 8000 accessions. Most of the collections CGIAR centres claim no ownership and seeks were made in collaboration with CIAT and Bio- no intellectual property rights over the ger- versity International (formerly the International mplasm and related information. To ensure Plant Genetic Resources Institute (IPGRI), in- continued free availability of this germplasm, cluding 542 accessions of Brachiaria spp. that are material is distributed with the Standard Mater- jointly held by ILRI and CIAT. Forage collections ial Transfer Agreement (SMTA) of ITPGRFA made by Bioversity International in sub-Saharan and requires all recipients of its germplasm to Africa were also donated to ILRI. ILRI studies the accept the same conditions. diversity in the collection and promotes know- Germplasm was collected by ILRI and CIAT ledge of forage crops with the goal of allowing until 1993 when the CBD came into force and users to make better-informed choices to select clarity over access and benefit sharing from the accessions and species best suited to their needs. use of crop diversity was questioned and collec- ILRI’s collection contains germplasm from tions were halted. Issues of access and benefit 160 countries. Much of the collection (43%) is sharing were resolved with the Nagoya Protocol from sub-Saharan Africa, with 17% collected in to the CBD in 2014, paving the way for future Ethiopia. In total, 97% of the accessions stored in collection and gap filling. Temperate forages the ILRI gene bank came from collections made continued to be collected by ICARDA because in the wild in rangelands and grasslands, with many of them are specifically listed under the only 3% represented by cultivars of common for- Multilateral System of ITPGRFA, making the ages that are already out of plant variety rights legal framework for collection clearer. Many and can be made freely available (Table 12.1). tropical forage species are not included in the Multilateral System, which has resulted in re- Forage germplasm maintained duced collection activities. ILRI, CIAT and ICAR- in the CIAT gene bank DA make the germplasm freely available as part of a global effort to promote genetic resources Forage genetic resources at CIAT have been a conservation and use. continuing activity since the inception of CIAT Table 12.1. Number of forage accessions conserved at ILRI, by region. (Data from ILRI gene bank, June 2018.) Australasia Latin America Sub- West Asia Unrecorded and the and the North Saharan and North regions and Forage type Asia Pacific Europe Caribbean America Africa Africa cultivars Browse 492 334 205 1,052 32 974 22 614 Grasses 259 44 20 2,778 159 2,707 53 883 Herbaceous 865 472 264 2,598 116 4,358 238 1,633 legumes Total 1,616 850 489 3,928 307 8,039 313 3,130 Forage Diversity, Conservation and Use 429 in 1972 (Schultze-Kraft et  al., 2020). The first 2006, a significant part of CSIRO’s former Aus- phase until 1993 focused on assembling and tralian Tropical Forages Collection. Over half of using the forage germplasm collection for forage- the grass accessions in the collection were ac- based livestock production on acid, low-fertility quired from donations, with 45% collected by soils in tropical American lowlands, particularly CIAT, mostly in collaboration with ILRI and na- savannahs. The objective was to create a ger- tional partners in sub-Saharan Africa. mplasm pool that was as diverse as possible for cultivar development via selection for direct use or, if natural variability fails to provide the de- Forage germplasm maintained sired combination of traits, via breeding. With a in the ICARDA gene bank total of approximately 23,000 accessions (about 21,500 legumes and 1500 grasses) from a total The ICARDA gene banks hold a large and highly of 75 origin countries, the CIAT collection is the diversified collection of temperate/Mediterra- largest tropical forage germplasm collection nean forages including globally important and worldwide (Table 12.2). Its value lies in its focus unique collections of Lathyrus, Medicago, Pisum, on plants from, and subsequently adapted to, Trifolium and Vicia spp. (Table 12.3). This collec- acid, low-fertility soils, and on legumes. The ma- tion is unique in terms of its geographical cover- jority of the legume accessions came from col- age, originating from 112 countries, its species lections made by CIAT, in collaboration with coverage with 631 taxa including many neg- national partners, in South and Central Amer- lected but potentially important species, and ica and South-east Asia, with the remainder with 20,831 accessions collected by ICARDA in acquired from donations from major tropical collaboration with partners. The bulk of the col- forage collections. More than 9000 accessions lection is still conserved in the gene bank in were received as donations, among them, in Syria. Since 2014, all gene bank core activities Table 12.2. Number of forage accessions conserved at CIAT, by region. (Data from CIAT gene bank database, June 2018.) Australasia Latin America Sub- West Asia Unrecorded and the and the North Saharan and North region and Forage type Asia Pacific Caribbean America Africa Africa cultivars Trees and shrubs 373 13 336 106 61 – 108 Grasses 12 1 150 1 1,103 3 381 Herbaceous 2,658 152 14,209 1,237 710 1 1,525 legumes Total 3,043 166 14,695 1,344 1,874 4 2,014 Table 12.3. Number of cultivated and wild accessions and taxa of forage germplasm conserved at ICARDA. (Data from the ICARDA germplasm database, March 2020.) Cultivated Wild and Uncertain Total Genus Accessions Taxa Accessions Taxa Accessions Taxa Lathyrus 2,703 17 1,748 51 4,451 55 Medicago 466 29 9,603 108 10,069 109 Pisum 1,941 11 4,188 13 6,129 15 Trifolium 185 23 5,555 95 5,740 97 Vicia 711 31 5,852 87 6,563 87 Other legume 131 15 4,287 193 4,418 198 forages Other non- 56 7 1,488 114 1,544 116 legume forages Total 6,193 133 32,721 661 38,914 677 430 J. Hanson et al. have been relocated to Lebanon and Morocco. wild, what additional diversity could be gained New facilities are being established, and regener- from gap filling, whether there are important ation and characterization are ongoing to recon- traits demanded by users that are missing from stitute the active and base collections and to existing germplasm, and whether the benefits of describe the diversity. ICARDA is also working adding additional collections outweigh the cost towards promoting in situ conservation and sus- of collection and conservation. Additionally, tainable use of dryland agrobiodiversity includ- with most of the important tropical forage spe- ing crop wild relatives and forage/range species. cies outside of the list of Annex 1 crops that are The activities include the identification of bio- key to food security under the ITPGRFA, reach- diversity hot spots for protection and recommen- ing bilateral agreements on access and benefit dation of management plans. sharing under the Nagoya Protocol will be com- plicated and, in some cases, may not be possible in the short term. These concerns, together with Gaps in the germplasm collections the limited current use of forage germplasm in breeding programmes (Sandhu et  al., 2015), Although the collections in CGIAR are diverse could lead to the conclusion that, until there is collections of grasses, legumes and trees in more widespread use of forage germplasm and terms of numbers of species and genera, there adoption of forages on farm, there is little need to are still important gaps in the collections, which fill the gaps. However, gene banks are also im- remain far from representative of the geograph- portant as sources of genetic diversity for the fu- ical diversity of tropical Poaceae and Legumi- ture, and future needs for traits are uncertain nosae. Geographically, there are few accessions and difficult to predict based on current use, so from Sudan, Somalia and Uganda in East Africa, gene banks remain the most important source of from Benin, Côte d’Ivoire and Ghana in West forage diversity that is being rapidly eroded in the Africa, and from Angola and Mozambique in wild due to overuse, land degradation and frag- southern Africa in the collections. All these mentation. For example, traits such as resistance areas are centres of diversity for tropical grasses, to a new disease are only sought once the disease and collections from these semi-arid regions spreads and its economic impact is realized. This may contribute traits of importance for selection was the case for stunt disease in Napier grass, of genotypes adapted to drylands. Germplasm of which has only been observed over the last species better adapted to areas with frost and 20 years, leading to research to identify resistant drought are needed for the tropical highlands genotypes (Asudi et  al., 2015; Wamalwa et  al., and subtropical regions. Some important species 2017) to protect dairy systems in East Africa that are still under-represented with relatively small rely on Napier grass as their most important feed numbers of accessions, such as Napier grass and (Muyekho et al., 2003). Guatemala grass. ICARDA and the School of The question of whether it is better to collect Biological Sciences at the University of Birming- and store in gene banks or to look for diversity on ham, UK, undertook ecogeographical surveys farm or, in the case of forages, in the extensive and gap analysis for temperate forage genera grasslands where they originated and continue (Lathyrus, Medicago, Pisum, Trifolium and Vicia) to evolve and adapt must also be considered. As- and identified areas for further collecting, sessments of amounts of genetic erosion, threat mainly to target threatened wild species with levels and the amount of diversity present in the adaptive traits that are needed for developing ecosystem are needed to make such decisions. new varieties that are better adapted to climate Such assessments have focused on systems and change. A complementary analysis showed that ecosystem scales and not on intraspecific and the Fertile Crescent region has the highest spe- genotypic variation (West, 1993) or general cies richness with the Syrian–Lebanese border rangeland degradation (Engler and von Wehrden, deserving efforts for ensuring in situ conserva- 2018). The lack of evidence on the threat of tion and management. genetic erosion for forage genotypes has led to Important considerations in expanding the scientists assuming the worst-case scenario and germplasm collection include what additional di- making ex situ collections of forages for gene versity is under threat from genetic erosion in the banks as a precaution to conserve diversity that Forage Diversity, Conservation and Use 431 could well be lost through overgrazing and poor Diversity can be measured through allele land management practices were they to be left frequency in accessions of the same species and in  situ. Given the levels of protection offered to genetic distance between accessions, often ex- rangelands in national parks in sub-Saharan pressed using Nei’s genetic distance (Nei, 1987). A frica (Reid et  al., 2005), ex situ may not have Genomic diversity and heterozygosity are esti- been the best choice for some species that are mated over several loci to obtain an estimate of well protected in the wild. However, having di- genetic diversity. These studies support func- versity readily available in the gene bank en- tional genomics and association mapping of hances access to a wide range of genotypes for phenotype and genotype and allow identifica- diversity studies and distribution, thereby in- tion of genes or markers influencing specific creasing their use and allowing an assessment of traits. Information generated from this research their potential value. In the face of lack of evi- is used to develop core collections and subsets for dence on whether collection is essential, deci- specific use cases and to identify superior acces- sions are based on economics and the perceived sions for evaluation, distribution and ultimately value of the resources (Evenson et al., 1998). use in farming systems. Understanding the diversity and expression of specific traits in each accession allows users to select genotypes to match their environment Describing Forage Traits and production systems and supports increased in Germplasm to Support Use use of the germplasm. Characterization and se- lection of traits are very context and species spe- The value of a gene bank is realized through the cific, and different traits may be selected when use of the genotypes or accessions that are ac- screening germplasm for different purposes. A cessed from the collection. The more informa- range of traits have been used to describe and tion available about the diversity and traits in cluster grasses and legumes based on standard the collections, the higher the probability that descriptors for forage grasses (IBPGR/CEC, 1985) accessions will be used, and therefore consider- and legumes (IBPGR/CEC, 1984). Standardized able effort is placed on research to characterize observation strategies were tested for morpho- the accessions. ILRI, CIAT and ICARDA’s re- logical characterization of forages at ILRI to search on forage genetic resources has focused ensure that methods used for crops were appro- on diversity within the genotypes for their use as priate for forages (van de Wouw et  al., 1999). feed and for natural resource management. Characterization of the germplasm is ongoing Such research includes assessing variations in for minimum morphological descriptors in the phenotype, agronomic traits and nutritional forage collections of CGIAR. More extensive traits, and in resistance to diseases and insects. characterization has been undertaken to describe Diversity within and between populations re- the diversity among accessions of species with sults from differences in genes and alleles that high economic importance. In ILRI, the focus make up the genotypes represented within the has been on forage grasses to estimate the diver- accessions. Each accession can be considered as sity in the collections and to select promising ac- a representative of diversity within the species. cessions for further evaluation of productivity Genetic diversity can be considered as a and feed value. In CIAT, the main focus was on proxy for future value of an accession. From the the evaluation and introduction of forage leg- 1990s, calculation of intraspecific diversity has umes until the mid-1990s, after which both been considered to determine what to conserve legumes and grasses were studied to select prom- and how much to conserve. Diversity is meas- ising germplasm to meet livestock producer ured on cardinal distances between individuals demands in a range of farming systems. Given based on similarity and dissimilarity (Weitzman, their potential for scaling through public– 1993). This concept is frequently used in the de- private partnerships, CIAT has concentrated its velopment of hierarchical diversity trees or breeding activities on forage grasses. ILRI has phylogenetic trees (dendrograms) that show focused mainly on selection of promising geno- how one accession relates to another (van Hin- types of Napier grass because of its importance tum, 1995; Osawaru et al., 2015). in dairy systems in East Africa (Negawo et  al., 432 J. Hanson et al. 2017, 2018), while CIAT since the 1990s has c lassify accessions into groups but that they focused on Brachiaria spp. because of the poten- were not sufficiently unique to reliably identify tial for grazing systems on acid soils in Central accessions (Heering et  al., 1996; Plumb et  al., and South America (Labarta et  al., 2017), and 1996). research has been expanded to Panicum spp. Data from morphological, nutritional and evaluation to select and breed for specific pro- genomic characterization can be combined to duction niches throughout the tropics. cluster accessions and to develop core collec- Genomics has wide application for describ- tions from the clusters. Core collections have ing diversity and promoting the use of acces- been described as a limited set of accessions that sions in gene banks (Kilian and Graner, 2012; represents the breadth of genetic diversity con- Wambugu et  al., 2018). Molecular approaches tained in the whole collection (Hodgkin et  al., provide useful selection tools to complement 1995; van Hintum et al., 2000). They are useful morphological characterization. Molecular marker to rapidly screen a species for potential use from and DNA sequence analysis can identify regions a few accessions in a cost-effective manner, or of the genome responsible for the expression of mini-cores can be created by selecting a subset economically important traits, such as improved of the core based on specific characteristics or digestibility and insect and disease resistance, traits. Clusters were identified in Lablab spp., that can be used in marker-assisted breeding. allowing identification of dissimilar accessions Molecular markers can also be used to confirm to cover maximum diversity for designation as the parentage of individuals in a breeding popu- a core collection (Pengelly and Maass, 2001). lation, for genetic diversity analyses and finger- Users can select either all accessions of a cluster printing individual accessions, for taxonomic of a specific type of direct interest or take one or identification, for finding duplicates within gene two accessions from each cluster to screen a bank collections, and for understanding system- wider range of morphological variation within a atic and evolutionary relationships within species. limited number of accessions. Diversity studies are also important for dif- Much of the evaluation of forage germplasm ferentiating among species and botanical types has been done by partners through research for taxonomic identification and for understand- networks using common protocols to identify ing the probable geographical origin and evolu- broadly adapted accessions as well as genotypes tion of crops and their related species. Molecular with specific trait adaptation. In Latin America approaches have been applied to elucidate taxo- and the Caribbean and in West Africa, forages nomic relationships and genetic distinctness of were screened for adaptation to acid soils, and accessions. While application of genomics is still promising accessions were identified for further in the early days for forages, these techniques promotion as ‘best bets’. In East Africa, the focus have been applied to studying diversity among was on identification of species adapted to the genotypes from the forage gene banks for species tropical highlands and on disease-resistant Na- in the genera Lablab (Maass et al., 2005; Roboth- pier grass, while in West Asia and North Africa, am and Chapman, 2017), Sesbania (Jamnadass the focus was on drought-tolerant forages for the et  al., 2005), Stylosanthes (Huang et  al., 2017), dry areas and cold-tolerant medics for use in Pennisetum (Lowe et  al., 2003; Wanjala et  al., Iran. These best bets were later elevated for larg- 2013; Negawo et  al., 2017, 2018), Brachiaria/ er-scale seed production (see Chapter 13, this Urochloa (Torres, 2005), Flemingia (Andersson volume) to provide sufficient seed supply to sup- et al., 2006) and Cratylia (Andersson et al., 2007). port forage adoption in the ILRI programme in Nutritional traits such as protein, fibre, lig- Kaduna, Nigeria, and the seed units of ILRI, nin and other chemical components that can CIAT and ICARDA. limit digestibility are also used to characterize Although emphasis has been placed on forages and to predict animal performance characterization and evaluation of the genetic (Cherney, 2000). Preliminary research on the resources, there is still insufficient information use of polyphenolic profiles determined by use of on specific traits such as disease and drought re- high-performance liquid chromatography on a sistance in large parts of the collection that limited number of accessions indicated that limits use. This gap could be partly addressed if there were some distinct profiles that could users provided more feedback on their research Forage Diversity, Conservation and Use 433 and on the use of germplasm as forage and feed. as a source of genes for adaptive traits. Non-use The CGIAR gene banks are assigning Digital Ob- components are even more difficult to quantify jective Identifiers to the accessions to better and value because of the difficulty in estimating track use of germplasm in publications and for- the existence or bequest value of an accession. age release in future. As a result of the germplasm evaluation and selection work at the centres within their re- spective mandate areas, many accessions from Value of Forage Germplasm the centres’ collections were released by national partners. centre accessions have also been The primary source of value of germplasm lies in adopted by end users without a formal and the actual or potential use of the collection. The documented release, and the real number of use value may depend on the extent of unique- such informal distributions is probably much ness of a collection and the number of unique higher than documented. In most cases, there is traits or traits that are present at low frequencies. a lack of accurate information about where re- This raises the issue of how large a collection is leased cultivars from centre or other sources needed and what the chance is of finding rare al- have been used, in which areas and with what leles. The probability of finding a specific trait productivity effects. can be modelled mathematically. For example, if a trait is present with a 0.01 frequency in the population, then there is an 87% chance of find- ing the trait within 200 accessions. Screening a Impact of the CGIAR Forage larger number of accessions increases the prob- Gene Banks ability of finding the trait, but as the number of accessions increases, the likelihood that the trait Forage germplasm distribution has already been detected is high and adding more accessions will not greatly increase the A major focus of any gene bank is to make ger- chance of finding the trait for added cost (Zohra- mplasm available for production use through bian et  al., 2003). This theory would indicate further evaluation, selection and plant breeding that there is little need for large germplasm col- by partners. The Centres’ collections supply for- lections if the focus is only looking for traits. age seed to collaborators in addition to their However, selection and breeding are done on work in collecting, evaluating, conserving and whole genotypes and the trait may be in an un- regenerating forage germplasm. Small experi- suitable genetic background for the production mental quantities of seeds or cuttings are pro- system. Currently, genotypes provide a short, vided from the in-trust collection under the cost-effective path to cultivar release, but as tech- terms of the SMTA as part of the CGIAR open- niques and costs of gene editing decline and more access policy to maximize utilization of ger- genomic information is known about accessions, mplasm for research, breeding and training. large collections may become redundant. The CGIAR centres have supported the use of The value of an individual accession may forage germplasm through production and dis- only be noticed when a new gene or allele that tribution of seed or plant materials. Distribu- addresses a major production problem, such as tion figures have been used as a proxy for use of disease or climate adaptation, is found in the ac- the collections. While they certainly provide in- cession. Information on the traits in an acces- formation on the demand for different species sion increases the chance of use and therefore and accessions, it is very difficult to link forage the value of an accession. This makes use com- germplasm to actual use in livestock systems ponents easier to quantify for the accessions that and to attribute forage use and impact back to have been characterized for yield, nutritive the gene bank accessions unless the materials value, or insect and disease tolerance. Use value are used in variety development. Many of the is much more difficult to estimate for the re- samples distributed will not meet user needs or maining accessions that lack detailed data on show superior performance in the location. their potential for future use but which might Only a small percentage of the materials re- have potential for as-yet-undocumented traits or ceived will be taken forward from evaluation to 434 J. Hanson et al. adoption and use. Estimates from past variety de- gene banks were physically located, and part- velopment in CIAT indicate that about 1% of the ners were aware of and had easy access to the grass collection and less than 0.2% of the herb- genetic resources available. aceous legumes have been released as named var- In accordance with their role, the gene ieties (see Table 5 in Schultze-Kraft et al., 2020). banks mainly cater to national and international Since their establishment, the CGIAR gene requests to supply a large variety of accessions banks have provided 137,816 accessions for for- in small quantities (Fig. 12.2). Most of the sam- age research and development activities in 158 ples from the gene bank are supplied to scientists countries. Over 30 countries with a strong live- in government agencies, non-governmental or- stock sector and demand for feed have received ganizations, educational institutions, and pri- more than 1000 samples each from the gene vate farmers and seed producers, with the banks over this period (Fig. 12.1). On average, general purpose of providing germplasm for for- the Centres freely distribute more than 5000 age production. In addition, large numbers of samples of forage germplasm globally per year samples of forage germplasm have been pro- for evaluation and selection by partners and for vided to the Centre’s own research and breeding further development and use by smallholder programmes for evaluation and forage develop- farmers. The m ajority of samples have been pro- ment, especially to the breeding programmes of vided to Colombia, E thiopia and Syria, where the CIAT and ICARDA. 20,000 18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0 Recipient countries (FAO country codes) Fig. 12.1. Countries receiving more than 1000 samples from CGIAR forage gene banks. Browse Grass Legume 60,000 50,000 40,000 30,000 20,000 10,000 – CIAT CIAT ICARDA ICARDA ILRI ILRI External Internal External Internal External Internal Fig. 12.2. Distribution of samples from CGIAR forage gene banks. (Data from CGIAR gene banks, February, 2019.) Number of samples distributed Number of samples distributed AUS BRA CAN CHN COL CRI DEU DZA ECU EGY ETH GBR IND IRN ITA KEN LBN MAR MEX NGA PAK PER PHL RUS SYR TUN TUR TZA USA YEN ZWE Forage Diversity, Conservation and Use 435 User demand has focused on a limited num- knowing whether or not they will ever be used. ber of forages that have been perceived to be of Non-use values include existence value – the high potential for further development as live- satisfaction that individuals or societies may stock feed. Different species are in demand in the derive from knowing that something exists and different regions and have been supplied by the will be available for future needs (Smale and CGIAR gene banks based on their respective Hanson, 2010). Option value is particularly im- mandates and locations (Table 12.4) in a com- portant for germplasm in ex situ collections, plementary way. which are natural resources and global public goods, and the world community relies on their existence as a resource for future needs. Estimating economic benefits The discussion of the economic benefits of of forage gene banks the forage gene banks is largely limited to show- ing the number of lines (accessions and hybrids) To estimate the impact of the forage gene bank developed from germplasm maintained in the germplasm, its sources of value must be under- various collections and subsequently released as stood. Forages have some unique features in that cultivars. To estimate economic benefits, field they are not directly used as human food but as studies are necessary, involving resource econo- feed for livestock, which convert the fibre that is mists, to assess areas planted to the new cultivars, inedible for humans into milk, meat and power. crop and livestock production gains, increases in Valuation frameworks consider the economic farmers’ incomes and non-market benefits (e.g. value of plant genetic resources to include both soil conservation and improvement). use and non-use components (Smale and Han- Some attention has been given to estimat- son, 2010), both of which can have current and ing the costs and benefits from gene banks (Koo future values. Use values may be derived from et al., 2003). The costs of conserving accessions the direct use of the product, which for forages in gene banks are easy to calculate, while the could be biomass and feed value of the forage or e xpected benefits are quite difficult to estimate. amount of milk and meat produced when fed to livestock. Use values can also be indirect, includ- ing the value of ecosystem or environmental Costing gene bank operations benefits from incorporation of forages into the system. Use values include the option value of The analytical framework used for the cost stud- having the genetic resources available without ies carried out on the CGIAR gene banks in 2008 Table 12.4. Most distributed forage species from the CGIAR gene banks. (Data from CGIAR gene bank databases, February 2019.) Organization Species Class of forage Number of samples CIAT Desmodium heterocarpon Legume 2801 Centrosema macrocarpum Legume 2639 Stylosanthes guianensis Legume 2455 Leucaena leucocephala Browse 2096 Brachiaria humidicola Grass 1797 ICARDA Pisum sativum Legume 13433 Vicia sativa Legume 5875 Medicago polymorpha Legume 2597 Lathyrus sativus Legume 2225 Vicia narbonensis Legume 1464 ILRI Sesbania sesban Browse 2555 Lablab purpureus Legume 1982 Vigna unguiculata Legume 1457 Cajanus cajan Browse 1379 Stylosanthes guianensis Legume 1085 436 J. Hanson et al. and 2009 was the microeconomic theory of pro- This approach enabled the estimation of possible duction (Pardey et al., 2001). For a gene bank, economies of scale or reduced costs per accession outputs were considered as numbers of accessions that would come from increases in the size of the characterized, stored, monitored and regener- collection. The average and marginal costs ated, and inputs were capital, labour, supplies change with the number of accessions and aver- and services. Production decisions involve age fixed or quasi-fixed costs per accession re- choosing which outputs to produce in which duce as the number of accessions increases. amounts, with which mix of inputs and input Marginal costs are the addition to total costs for quantities. Optimal resource allocation can be every accession added to the collection and in- achieved either by minimizing the costs of oper- crease as the number of accessions increases. ation given fixed physical resources and existing When inputs are used efficiently and are com- technology or by maximizing production subject plementary, the cost of conservation of each to a fixed budget and existing technology. a ccession should decrease. The approach selected by Koo et al. (2004) was cost minimization. The costs of running a gene bank vary with target species, type of con- Management of gene banks based servation, location, size of the collection and on economics amount of risk that has to be mitigated in the ac- tivities. Most of the benefits of gene bank collec- Management of large collections is based on the tions are public goods whose values are both science of conservation and economics of differ- expensive and difficult to estimate and are likely ent management options. Management deci- to be unreliably estimated, while the costs of sions based on changes in understanding of seed gene bank operations are relatively easy to esti- conservation science, genetic integrity, technol- mate with a reasonable degree of precision. ogy, facilities or staffing affect the costs of inputs. Pardey et al. (2001) reasoned that if the costs of Staff costs, supplies and services increase annu- conserving an accession are lower than the low- ally due to inflation, but better use of technology er-bound estimate of the corresponding benefits, and opportunities for adopting improved operat- it may not be necessary to estimate the actual ing procedures from better understanding of seed benefits of each accession. conservation science can reduce the time and Koo et al. (2003) considered gene bank op- supplies needed for routine operations and there- erations within this production economics fore significantly reduce costs and increase effi- framework, looking at the cost of inputs to de- ciency and cost:benefit ratios (Koo et al., 2004). liver outputs in terms of stored germplasm and These considerations were applied in gene information. Total costs were partitioned into bank costing studies carried out previously on components and each category was then sum- the CGIAR collection in 2009 (Horna and marized in terms of average and marginal costs. Smale, 2010). Data were compiled by gene bank They included variable inputs that are deter- managers on input use and expenditures and mined by the size of the collection, capital costs used to estimate average and marginal cost per that are independent of the size of the collection unit for routine gene bank activities. Data were (up to some limits) and quasi-fixed inputs that analysed across all CGIAR gene banks. Gene are neither fixed nor variable but indivisible bank operations are made up of a set of inter- large expenses, such as the cost of the gene bank related component activities that occur over the manager and scientific expertise, that are needed life of the accession. These include acquisition, independent of the size of the collection (Koo characterization, safety duplication, medium- et al., 2003). In this study, costs were calculated and long-term storage, germination and seed for the different gene bank operations and con- health monitoring, regeneration, seed processing, servation type based on actual annual costs and information management, distribution and numbers of accessions for a range of species and general management. Gene bank operations are locations, taking care to ensure consistency in dynamic; some such as storage are incurred data collection. This allowed comparisons to be annually, while others such as monitoring, regen- made on operational costs per accession be- eration and safety duplication are incurred peri- tween crops and locations (Koo et  al., 2004). odically. In order to estimate annual expenditure, Forage Diversity, Conservation and Use 437 the overall annual expenditure for the whole op- fodder tree species) to better understand the cost- eration was used and allocated back over the ing structure of operational costs (Horna and number of accessions involved in the operation Smale, 2010). This showed that the quasi-fixed in that year. The study found that the reproduct- costs accounted for the highest costs in gene ive biology of the species, which determines the bank operations at ILRI (Fig. 12.3). The cost of type of conservation and regeneration proced- maintaining an accession varied from US$125 ures, and level of quality management attained, to US$242 per accession per year (Table 12.5), contributed most to the costs of gene bank oper- depending on the category of the forage, which ations. Seed crops had the lowest costs of conser- was largely determined by the length of life cycle vation, while vegetatively propagated crops, and longevity of the seeds during storage. The trees and wild relatives, including forages, had study concluded that it was important to take higher costs. into consideration that the majority of the forages maintained at ILRI are wild species that ILRI example of a gene bank costing study require special management and have little pub- lished i nformation about their breeding systems, The study on the ILRI data looked at the costs of seed germination and storage behaviour in gene the different broad categories of forages (annual banks. Regeneration and multiplication of these and perennial grasses, annual and perennial materials are particularly challenging as the dif- herbaceous legumes, and short- and long-lived ferent species have very different behaviours in 60,000 50,000 40,000 30,000 20,000 10,000 0 Grasses Grasses Legumes Legumes Fodder Fodder Other Other annual perennial annual perennial trees <3 trees >3 annual perennial years years Capital cost Quasi-fixed cost Labour costs Non-labour costs Fig. 12.3. Costs by component and type of forage at ILRI in US$. (From Horna and Smale, 2010.) Table 12.5. Annual total cost per accession for four forage types, in US$ (c.2009). (From Horna and Smale, 2010.) Forage Classification Annual cost per accession Grasses Annual 149 Perennial 171 Herbaceous legumes Annual 242 perennial 191 Fodder trees Short-lived 197 Long-lived 205 Other forages Annual 204 Perennial 125 Cost (US$) 438 J. Hanson et al. the field. Therefore, although a small gene bank derived from plant collection and characteriza- in terms of number of accessions, the manipula- tion that increases its use value. A second scien- tion of a large diversity of materials requires con- tific benefit includes the tools for germplasm siderable investment in equipment and human management in gene banks and in the field. capital. Another important cost in using gene The field benefits accruing from the use of bank accessions is the lag between germplasm forages are: collection or introduction and cultivar release, • flows of economic benefits, defined as gains which reflects the intensity, and related costs, of in output or reductions in cost of feed in- research at each step in forage development. puts to animal production; • stocks of economic benefits, defined as gains in asset values, such as soil fertility; and Benefits from use of germplasm • environmental benefits, such as carbon s equestration (a stock), soil stabilization (a The economic benefits of a specific trait or a spe- stock) and water-holding capacity (a stock) cific accession are more difficult to estimate than or reduced greenhouse gas emissions the value of a species. When different accessions (a flow). are used in plant breeding the contribution of each accession to the new variety can be esti- Estimating these benefits requires data on mated from the breeding procedure. Most forage materials distributed, their use in production cultivars are selections made directly from ac- and their yield and cost effects. cessions, so in forages it is much simpler to esti- mate the contribution of a specific accession to a Success stories in the use of forage cultivar than for crops, where crossing, selection g ermplasm and backcrossing are used in cultivar develop- ment. The value of germplasm in use can be Forage cultivars have mostly come from selec- measured by the area planted, the yield per unit tion of more productive types rather than from area and the market price of the crop (or its breeding (Jank et al., 2011). This selection would equivalent when converted into animal prod- not have been possible without the sources of ucts). These data are in short supply for forage germplasm held in the world’s forage gene production in many countries, making it diffi- banks. A recent meta-review of global impacts cult to come up with good estimates of benefits. of forage cultivation revealed many data gaps in The question remains on how to value the use of improved forages owing to irregular re- potential benefits from germplasm that is not in porting of forage adoption, with few studies re- use and may not have been fully characterized. porting on economic benefits and costs (White Smale and Hanson (2010) concluded that use et al., 2013). value is a relatively small component of total Most historical success in selection of for- value because they have public goods attributes age cultivars has been in Australia, which relies and may not be reaching full market potential in heavily on pastures to sustain its lamb and beef transitional agricultural economies. In some industries (Oram, 1990). Some of these selec- cases, forecasts of future benefits can be esti- tions have been made on germplasm from the mated based on past benefits and patterns of CGIAR forage gene banks. For example, in 1994, f orage use. The potential benefits from use of an the University of Queensland registered Mount additional accession are usually greater than Cotton, a cultivar of Sesbania sesban selected the marginal cost of conserving it for the long from accession 15036 that is maintained in the term as seeds in a well-managed gene bank. ILRI gene bank (Gutteridge and Shelton, 1995). However, despite its identification as very pro- Estimating scientific and field benefits ductive and leafy, there is little evidence to sug- of the CGIAR forage gene banks gest that it is widely grown or adopted. Another species of interest currently in Australia is Lablab The scientific benefits of the gene bank are de- purpureus, because of its fast-growing nature, fined as advances in knowledge. The principal good protein concentration and the possibility of form of scientific benefit is the information its use as both food and feed (Pengelly and Maass, Forage Diversity, Conservation and Use 439 2001). Lablab is already widely grown in the species are benefiting from hybridization and se- tropics and recently selections have been made lection efforts for cultivar development, includ- from three lablab accessions (13685, 14428 ing lucerne (Medicago sativa) and a few temperate and 14437) provided from the in-trust collec- legumes and grasses such as Lolium perenne. The tion at ILRI to CSIRO in 2004. Five new cultivars breeding efforts are mainly by the private sector were selected from the three accessions and are and their success is tightly linked to efficient seed in the Australian release process (D.S. Loch, per- production, certification and marketing. sonal communication). The original germplasm The importance of Mediterranean forages will continue to be freely available from the ILRI is highlighted by several examples: in-trust collection and has also been deposited in the Australian Pastures Genebank. • Lucerne is among the most important per- Documentation of large economic benefits ennial legume forages in the world. Despite from use of improved forages is rare; a particu- extensive breeding efforts, little improve- larly important documented case is in Brazil ment in potential biomass production has where improved Brachiaria spp. grasses are grown been achieved during the last 60 years, but on an estimated 100 million ha (Jank et  al., the use of wild relatives has contributed to 2014). Except in rare cases, it is even more diffi- enhance resistance to diseases and pests cult to attribute any of these benefits to use of such as leafhopper to maintain high prod- accessions from the forage gene banks owing to uctivity. Most of the perennial Medicago spp. lack of information. Since 1980, a total of 29 can be used for further improvement of M. legume and 39 grass cultivars have been re- sativa but can also be domesticated as new leased in 17 countries, mostly in Latin America forage resources. While lucerne breeding is and the Caribbean (Schultze-Kraft et al., 2020). benefiting from genetic transformation ef- These materials are from the CIAT collection forts for tolerance to herbicides, genetic re- and can be directly attributed to accessions in sources are crucial for further improvement the CIAT gene bank. However, with some excep- of lucerne and other forage legumes as ani- tions, the extent of planting and economic im- mal feed, nutritional food and even for pact from use of these cultivars in tropical pharmaceutical use. livestock systems is not well documented. • Annual medics, mainly the self-regenerat- Although only a few of the registered culti- ing species such as Medicago truncatula, vars came from accessions maintained at the M.  polymorpha and M.  littoralis, have con- ILRI forage gene bank, ILRI has had a broader tributed significantly to improving and sus- impact through the distribution of forage leg- taining cereal and livestock production umes globally to national programmes. The lar- through their introduction in rotation with gest number of accessions distributed was in cereals within the ley-farming systems de- Ethiopia, where a large part of the forage ger- veloped by Australia and promoted in other mplasm being evaluated and the old cultivars in regions of the world with Mediterranean use on farm originated from the ILRI gene bank climates (Cocks and Bennett, 1996). In (Hanson and Tedla, 2010; Jorge et al., 2012). Australia, medics occupied 40 million ha, The domestication and evaluation of intro- but their importance has decreased since duced forage germplasm has allowed some 2005, mostly due to the effects of herbi- M editerranean species to play a major role in cides used on cereals. However, herbicide diversifying the livestock feeding calendars and tolerance is found in some accessions of M. sustaining the crop–livestock–pasture-based littoralis and M. truncatula. farming systems in Australia, New Zealand, the • White clover (Trifolium repens) originating USA, South Africa and several other regions from the Mediterranean region forms the with Mediterranean-type climates (Cocks and basis of the sown pastures covering 9 mil- Benett, 1996; Suttie et  al., 2005). In Australia lion ha in New Zealand, 5 million ha in alone, more than 50 cultivars of various forage Australia and 5 million ha in USA and species were released from direct selections from spreading to other regions in Europe and collected genetic resources. Forage breeding Latin America (Mather et  al., 1996). New started in the 1950s and only a few forage cultivars are being developed in the UK to 440 J. Hanson et al. replace the old cultivar ‘Grassland Huia’, and selling Napier grass as feed has good poten- released in New Zealand. tial for improving smallholder livelihoods (ILRI, • The Dryland Agricultural Research Insti- 2013). tute (DARI) in Iran, in collaboration with Dairying in Kenya depends on a few Napier ICARDA, has tested large numbers of ac- grass genotypes, many of them susceptible to cessions of forage legumes obtained from two major diseases – head smut, caused by Usti- the ICARDA gene bank for use in the wheat/ lago kamerunensis, and stunt, caused by a 16SrX1 barley-based livestock systems in the high- group phytoplasma, both of which are spread- lands of Iran. Preliminary results showed ing in the country (Jorge et al., 2014). Although the good adaptation of some accessions of other grasses can be introduced into the system, Vicia panonica and V. ervilia. Two of the ac- none is as productive, as broadly accepted or as cessions were released in 2010 and in well suited to the system as Napier grass. Losses 2013, and now almost 40,000  ha are in feed supply caused by these diseases are ser- planted. However, seed multiplication re- ious for smallholders (Mwendia et  al., 2006, mains a challenge for larger adoption of 2007). these species. Head smut is spreading across Africa from • Grass pea (Lathyrus sativus) is of economic west to east and in Kenya, where yield losses in and ecological significance in South Asia Napier grass of up to 46% due to head-smut in- and sub-Saharan Africa and is one of the fection have been reported (Farrell, 1998; neglected crops with good tolerance to Mwendia et  al., 2007). The diversity of Napier harsh conditions and with multiple uses. grass lines held in the forage collection at ILRI Although it is rich in proteins, when over- was an obvious target to look for tolerance. In consumed, the oxalyldiaminopropionic acid 1993, the Kenya Agricultural and Livestock Re- (ODAP) toxin may cause paralysis of the search Organization (KALRO) evaluated ten ac- lower limbs. Several accessions of L. sativus cessions from the collection of Napier grass from and L. cicera with low ODAP content from the ILRI gene bank for tolerance to head-smut ICARDA were selected and used in breeding. disease and identified two accessions (ILRI More than 20 lines were released by na- 16791 and 16798) that showed tolerance. tional agricultural research programmes These were released as Kakamega I and II, and in including Wasie in Ethiopia in 2005, Ali- 2007 were reported as being used by 19% of Bar in Kazakhstan in 2004, Tarman for the farmers surveyed in smut-affected areas (Mwen- highlands of Turkey in 2002, and more re- dia et  al., 2007; ILRI/KALRO, 2013). More re- cently, two varieties (BARI Khesari 3 and 4) cently, more germplasm from the ILRI gene bank with low ODAP content were released in has been screened, and one additional acces- Bangladesh (Kumar et al., 2013). sion, 16806, was found to be resistant, while three others showed tolerance to smut (Omayio Example of impact from use of Napier et  al., 2015; Kariuki et  al., 2016). Accession grass germplasm from ILRI 16806 was more productive than Kakamega I in the high potential area of Muguga in Kenya pro- ILRI has focused attention on Napier grass be- ducing 51 t dry matter/ha/year compared with cause of its importance as the major forage for Kakamega I, which produced 42  t dry matter/ East African cut-and-carry dairy systems due to ha/year. This is equivalent to the best-yielding its wide adaptation, high yield, and ease of cultivars of Napier grass under farm conditions propagation and management. Napier grass (Munyasi et al., 2015). provides more than 50% of the feed for more Based on figures from Mwendia et al. (2007), than 0.6 million smallholder dairy farms in production losses due to smut on smallholder Kenya (Orodho, 2006). More than 80% of milk farms would be about 0.2 t/ha/year for zero graz- produced and sold in Kenya comes from small- ing systems, giving an annual loss to a small- holder farmers with less than 2 ha of arable land holder farmer equivalent to 22 days of feed for a (Thorpe et  al., 2000), who feed one or two cow, a loss in income of 220–330 L of milk. Con- cross-bred dairy cows on small plots of Napier sidering the cost of Napier grass at US$15/t and grass. With fodder in high demand, producing an estimated yield under low-input smallholder Forage Diversity, Conservation and Use 441 conditions of 18 t/ha/year, a reduction of 40% accessions could prevent loss of income of over of the yield due to smut would reduce the amount U$11 million/year. Similar benefits can be pro- of fodder available for sale and cost a farmer jected in Tanzania and Uganda where the diseases US$108 per year in lost income from Napier are also spreading. grass sales (ILRI/KALRO, 2013). A recent eco- nomic analysis of the costs of production and Capacity building gross margin analysis per year from growing Napier grass in 2017, showed an expected profit Scientific and technical capacities are a major of US$245 per acre per year (NAFIS, 2017). This constraint to forage development. To build this is equivalent to US$605/ha, so a reduction of capacity, ILRI, CIAT and ICARDA have provided 40% of the yield would result in losses of about training for scientists, technicians and farmers US$242/ha/year at current costs for Napier grass. in germplasm management and forage seed pro- Stunt disease is even more devastating in duction. At ILRI, 50% of the 60 beneficiaries East Africa than smut and is spreading rapidly in taking individual academic training since 1983 Napier grass-growing areas (Khan et al., 2014), were from Ethiopia, plus a wide range of other causing significant yield loss for smallholder African, European, Asian and Western coun- dairy farmers in Kenya, Uganda and Tanzania. tries (Fig. 12.4) with a 50:50 split between males Yield losses of 40–90% have been reported due and females. Short courses and workshops were to stunt in western Kenya (Khan et  al., 2014), held on demand, and a large number of partners causing reductions of up to 65% in milk yield have been trained in both short courses and aca- over the year. Again, the national partners demic fellowships in ILRI. Earlier courses from turned to the ILRI forage germplasm collection 1983 to 1990 focused on forage evaluation and to look for resistance, and as a result, two differ- gene bank management; most recent courses ent accessions 16789 and 16807, have been have focused on forage seed production and identified that have tolerance to stunt (Wamalwa quality management systems for gene bank et al., 2017). With yield losses of 40–90% from operations. stunt, a conservative estimate of production At CIAT, the capacity-building-related im- loss due to the disease would be similar to head pact of its forage genetic resources was sought smut for lost feed and lost income from sales via its entire Tropical Pastures/Forages Pro- of  milk or fodder for many smallholders in gram including activities in different subre- stunt-affected regions. However, in areas where gions of Latin America and the Caribbean, production is reduced by 90%, losses would be annual training courses, and network-based very severe with little production of feed for multilocational germplasm testing within the dairy and potential losses of almost US$550/ Red Internacional de Evaluación de Pastos ha/year. Tropicales (RIEPT, International Tropical Pas- Use of the two head-smut resistant acces- tures Evaluation Network). During the period sions and two stunt-tolerant accessions from the 1978–1990, the CIAT Tropical Pastures Pro- ILRI gene bank can reduce the economic loss gram, in its coordinating role in the RIEPT net- and benefit smallholder farmers in disease- work, held a yearly course, ‘Programa de affected areas. Estimating the economic loss Capacitación Científica en Investigación para la requires data on areas cultivated and areas of the Producción de Pastos Tropicales’. The course resistant accessions that have been planted. was aimed at researchers from Latin America Data are scarce because many farmers are not and the Caribbean and consisted of an intensive sure of the varieties that they grow with differ- multidisciplinary phase, in which all partici- ent genotypes being grown under the same pants were exposed to lectures and practical name and with the same genotype being grown training in all disciplines represented in the pro- under different names (Jorge et al., 2014). Using gramme – thus including the field of genetic a conservative estimate of 600,000 smallholder resources of forage plants and germplasm dairy farmers in Kenya on farms of about 2 ha handling – and a specialization phase. With an and using an average of 4% of their land for average of 20 participants per course, a total of growing Napier grass, there are 48,000  ha about 250 researchers were trained during the of Napier in Kenya where use of the tolerant 13-year period, of which ten specialized in gen- 442 J. Hanson et al. USA, 6 China, 1 Uganda, 1 UK, 4 South Africa, 1 Sri Lanka, 1 Nigeria, 1 Netherlands, 3 Mali, 1 Ethiopia, 32 Kenya, 3 Germany, 11 Fig. 12.4. Number of associates working on forage genetic resources by nationality at ILRI from 1986 to 2020. (Data from ILRI databases, March 2020.) etic resources. In addition, a number of post- (with national research institution partners in graduate students from both Colombian and Latin America and the Caribbean); Reseau de foreign universities conducted, under the super- Recherche en Alimentation du Bétail en Afri- vision of the programme scientists, field and/or que Occidentale et Centrale (RABAOC; with na- laboratory research for their theses with a focus tional research institution partners in West and on genetic diversity. From 1990 onwards, train- Central Africa, in cooperation with ILRI); and ing activities in forage germplasm were mainly the South East Asian Forage and Feed Resources in the area of germplasm management and in Network (SEAFRAD; with national research in- the form of ‘field days’ for Colombian university stitution partners in South-east Asia; this net- students and technicians, with demonstrations work developed in the 1990s into the Australian at field, greenhouse and laboratory levels. Sev- Centre for International Agricultural Research eral hundred students, technicians and re- (ACIAR)-funded Forages for Smallholders pro- searchers participated in this scheme. ject). At ILRI, the networks carried out similar At ICARDA, major training efforts were activities of forage evaluation, information provided during 1990–2005 to North African sharing and capacity building with a focus on and South and Central Asian young researchers sub-Saharan Africa: the Pasture Network for through joint projects with Australian part- Eastern and Southern Africa (PANESA; with ners. In recent years, training on the use of for- national research institution partners in East age germplasm has been provided to Iranian and Southern Africa); the African Feeds Re- scientists focusing on diversification of cere- search Network (AFRNET; which replaced al-based systems in the highlands. An estimated PANESA with national research institution 750 forage professionals have been trained in partners with an Africa-wide focus); and con- forage g ermplasm activities in the last 30 years, tributions to the Alley Farming Network for including about 35 carrying out field research Tropical Africa (AFNETA; with national re- for their theses. search institution partners in cooperation with A basis of gene bank work is the develop- the International Institute of Tropical Agricul- ment and spread of scientific networks. The ture (IITA). These networks worked closely with networks, in which the CIAT Tropical Pastures the regional and subregional agricultural re- Program participated, were decisive for the search organizations that were established in multilocational testing of elite forage ger- all regions in the early 2000s, supporting them mplasm and development of a number of grass in their efforts to improve feed resources and and legume cultivars that were eventually re- nutrition as part of sustainable livestock pro- leased (Schultze-Kraft et  al., 2020): RIEPT duction systems. Network activities included Forage Diversity, Conservation and Use 443 the development, publication and use of net- Conclusions work-wide common research methodologies, and workshops and meetings with publication Development impact of research results. ILRI, CIAT and ICARDA have collaborated Despite the recognized value of forages for live- since their establishment on forage genetic re- stock production and land management, there sources conservation and use, initially through has been limited use of germplasm in tropical the Systemwide Genetic Resources Programme forage development except on larger farms in and currently through the CGIAR Genebank Latin America (White et  al., 2013). The eco- Platform. Activities have included joint collect- nomic impact of the forage germplasm in the ing missions for Brachiaria spp. germplasm in the ILRI gene bank is relatively low in Africa, except early 1980s in East and southern Africa, ger- for Napier grass where four accessions with dis- mplasm exchange, forage evaluation and devel- ease tolerance have been selected and released to opment of the forage pages of the Crop Genebank farmers. The cost of conservation of perennial Knowledge Base. grasses in ILRI is estimated to be US$170/year/ ILRI has developed training materials and accession, while the benefit to farmers of guidelines for gene bank management (Rao US$242/ha/year of Napier grass alone, when et al., 2006a,b), and with CIAT on regeneration taking into account the extent of production in for grasses (Hanson and Schultze-Kraft, 2009) Kenya, Tanzania and Uganda, where head smut and with ICARDA for legumes (Hanson et  al., and stunt disease are spreading, translates into 2009) and grass pea (Hanson and Street, 2008). benefits of over US$15 million/year (or US$37.5 ILRI and Bioversity International oversaw the million/year when considering that annual development of the online Crop Genebank profit is expected to be US$605/ha/year). Knowledge Base (http://cropgenebank.sgrp. cgiar.org/; accessed 26 February 2020) which was developed to facilitate access to best prac- Potential option values tices for gene bank management of selected crops, including forages. ILRI, working closely with Much of the value of the gene bank lies in the CIAT and ICARDA gene bank staff, developed future. The option value of having the ger- the pages on procedures for gene bank manage- mplasm available to respond to as-yet-unidenti- ment, forage legumes and forage grasses to fied future needs, such as changing feed needs support more efficient and effective ex situ con- due to climate change, and the existence value servation and use of crop genetic r esources. that society derives from knowing that some- Another form of capacity development is thing exists and will be available for future the distribution of germplasm data. ILRI, other needs, may be more powerful economic drivers centres, and national and international partners of conservation of forage diversity than the compile and develop forage germplasm data and cost–benefit analyses that can be done on its distribute it online and in print. Examples are a ctual use. manuals for forage evaluation (Tarawali et  al., 1995) and forage seed production (ILCA, 1994). ILRI and CIAT, together with CSIRO and the De- Scientific impact partment of Primary Industry and Fisheries, Queensland Government, were founding part- Environmental benefits ners in the development of the online Tropical Forages tool (www.tropicalforages.info; accessed Over the last 20 years, there has been a comple- 26 February 2020), an interactive tool for select- mentary research focus at the Centres on both ing forage species suitable for local climate, soils, feed and environmental benefits (White et  al., production system and management conditions 2013). This is the study of the contribution of in the tropics and subtropics with comprehen- forage plants to ecosystem services such as sive information on adaptation, uses and man- mitigation of the emissions of nitrous oxide, a agement of these forage species (Cook et  al., potent greenhouse gas, via biological nitrifica- 2005). tion inhibition. 444 J. Hanson et al. Basic and applied research Africa (BecA)-ILRI Hub have done limited but promising work on Brachiaria spp. in With so many genera and species used as forage, sub-Saharan Africa. CIAT’s private-sector part- there remains a gap in botanical research to bet- ner is commercializing hybrids and a selection of ter understand which species have more poten- P. maximum globally in over 40 countries tial for use as livestock feed and for genetic throughout the tropics and subtropics, for ex- research to quantify diversity in these species to ample in sub-Saharan Africa, the Mediterra- support forage selection and variety development. nean, tropical Asia, South and Central America, Any further diversity research should focus on its southern USA and southern Europe. relevance for: (i) enhanced germplasm manage- ILRI, CIAT and ICARDA have strong re- ment and utilization (identification of duplicates, search programmes on forage genetic resources. establishment of core collections); (ii) relation- A promising area is using molecular markers to ships between traits and geographical origin of describe genetic (intraspecific) diversity, clarify populations; and (iii) identification of genes re- species relatedness and elucidate phylogenetics, sponsible for particularly important plant traits. and to improve gene bank curation by identify- Centres have focused on specific forages ing duplicate accessions within collections. that meet the demands of livestock production Studies on basic floral biology are being done on systems in their mandate areas. The ILRI focus understudied species of production interest, has been on forages for smallholder systems, es- with the objective of optimizing germplasm pecially dairy, which provides a steady income management and eventual breeding. As tropical for smallholders in East and Southern Africa. species with forage potential are essentially wild Research has focused on three grasses: Napier (undomesticated) and not well described botan- grass and Rhodes grass for cut-and-carry sys- ically, reference herbaria were established at tems in the subhumid areas, and buffel grass for CIAT and ILRI to assist species identification, grazing in dryland areas. Research of forage leg- with the aid of an images database. umes has focused on lablab and cowpea for sup- plementation of low-quality diets that are high in fibre. The CIAT breeding focus is currently on The Future Brachiaria spp. and Panicum maximum, the most important forage grasses in the subhumid and Over the past 30 years, the forage programmes humid tropics, with goals of developing pest re- in CGIAR have focused on building their ger- sistance and nutritive value, with appropriate mplasm collections to cover a broad range of tolerance to drought, excess water, low soil fertil- genetic diversity of tropical and subtropical for- ity and aluminium toxicity. Seed production, ages to meet user demands, studying the collec- which is key for commercial success, has been tions to characterize and evaluate diversity and added as a criterion. Since 1988, breeding work understand the traits in specific accessions, and at CIAT has been using Brachiaria germplasm. improving their facilities to securely conserve More than 300 elite hybrid lines had been de- the germplasm. The focus for the future will con- livered and seven cultivars have been developed tinue to be on the core operations that are essen- and are being globally commercialized from the tial to conserve and manage the diversity with start of the programme to date: cvv. Mulato, Mu- increased emphasis on those that will ensure ef- lato II, Cayman, Cobra, Camello, Mestizo (a syn- ficiency and value for money in gene bank oper- thetic variety based on mixtures) and Converse. ations. In addition to these essential activities, One hybrid cultivar, Talisman, is in a pre- there are opportunities to be more forward and commercialization stage. From the Brachiaria outward looking and to link with activities in the humidicola collection, intraspecific hybrid lines Livestock CRP and the Excellence in Breeding are developed and are on the path to commer- Platform, as well as to contribute to the global cialization in the next 5  years. CIAT’s Panicum system of plant genetic resources, support the maximum breeding programme is using material FAO Global Plan of Action and contribute to the from the 560 accessions held at the CIAT gene Sustainable Development Goals. These activities bank. ILRI and the Biosciences Eastern and central will build on the core expertise in the forage gene Forage Diversity, Conservation and Use 445 banks and will use cutting-edge technologies Traditional gene banks rely heavily on elec- to add value to ongoing routine activities by tricity for cold storage or cryopreservation of increasing the relevance of the collections to seeds or tissues. Advances in renewable sources climate change, contributing to understanding of energy such as solar and wind can both re- forage diversity and supporting capacity build- duce the environmental footprint of gene banks ing in collaboration with partners in the Live- and can also substantially reduce the costs of stock CRP and national gene banks. conservation. Water will become a major limit- Going forward, the CGIAR forage gene ing factor, and installation of water-harvesting banks need to look at how to respond to future methods and drip irrigation for regeneration and needs to continue to be relevant. Global predic- multiplication can contribute to more sustain- tions of climate change indicate that even with a able operations and mitigation of the effects of cap of a 2°C increase, new varieties of forage climate change. ILRI and ICARDA have estab- crops will be needed to adapt to climate change. lished new gene bank facilities in the last 2 years, The traits and genes found in the germplasm in and CIAT has embarked on an ambitious initia- the CGIAR collections are essential to realize the tive called Future Seeds to build a state-of-the-art genetic gains needed both to adapt to climate genetic resources centre that not only ensures change and to produce sufficient livestock feed the conservation but also encourages the pro- with limited resources and competition for land active use of the germplasm. The environmen- to feed a growing population. Additionally, sev- tally sustainable facility will use genomics, eral forages are also wild crop relatives and are digital phenotyping and information technolo- an important source of traits and an asset for gies to gradually build a ‘knowledge bank’ that crop breeding. Wild crop relatives are not well enables a more data-driven deployment of crop represented in ex situ collections, and there is a diversity and serves as a meeting platform for lack of information on the location of existing scientists promoting biodiversity as a driving collections, optimum storage conditions and effi- force for innovation in agriculture. cient management. Despite three decades of training in the New techniques offer many opportunities management and use of forage germplasm, to understand and make better use of the forage many national gene banks lack the capacity to genetic diversity held in the collections. Finger- meet expanding national mandates and to fully printing tools can be applied for improved man- participate in the global system. The next gener- agement, as well as to study diversity and ation of forage scientists needs to be able to apply identify traits to support sustainable use in new techniques such as DNA fingerprinting, these crops. Sequencing data and DNA markers flow cytometry, bioinformatics and spatial ana- are essential not only to facilitate use but also lysis for forage development but need mentoring for improved management through accession from experienced genetic resources scientists. identity and tracking genetic integrity of acces- The move to using molecular techniques has sions. Genomic information on gene flow, tax- also resulted in a gap of young scientists with onomy, apomictic status and ploidy levels of the traditional biology skills such as taxonomy, accessions can be used to improve operations phenology and seed biology. A major gender- and introduce considerable efficiencies in re- balanced capacity-building effort is needed to generation and multiplication. Research on re- support global efforts in conservation, make bet- generation methods, breeding systems, seed ter use of the collections and work efficiently in storage and longevity will provide information the genetic resources policy environment. This that can be used to develop gene bank stand- will include training, internships and graduate ards suitable for low-cost operations and con- scholarships, mentoring, developing and sup- servation of forages and other wild species. This porting communities of practice, online training activity will be done through collaboration and knowledge sharing. with national programmes, the Genebank Plat- Gene banks are long term and expensive to form, the Livestock CRP and the Excellence in maintain and have an essential function to sup- Breeding Platform for phenotyping and geno- port current and future use. Yet questions con- typing to support conservation management tinue to be asked on the value of the accessions, and use. the efficiency and costs of operations, and the 446 J. Hanson et al. impact derived from the investments in having Surveys will follow up with users to provide feed- three forage gene banks in CGIAR. Future activ- back on their research and the use of ger- ities will focus on answering some of these mplasm as forage and feed. The CGIAR gene q uestions on the value and use of the collec- banks are assigning Digital Objective Identifiers tions. Working closely with the Livestock CRP, to the accessions to better track use of ger- information will be collected on user demand for mplasm in publications and forage release in fu- forages and traits, documenting the delivery path- ture. These data will be used to estimate current ways from the gene bank to forage production and option values and will lead to a better under- and adoption of forage on farm to better document standing of the value of the CGIAR ex situ forage the impact and use of the forage collections. collections. Notes 1 Information for this chapter was compiled by the authors mainly from information accessible at the websites of ILRI, ICARDA and CIAT, from their respective annual reports, unpublished reports on ger- mplasm collecting missions, research papers of national and international partners, and other material published and unpublished. We further note that there have been changes in plant nomenclature for several species of particular interest. As this overview is concerned with the past, we are still referring to the earlier accepted names. It should be noted that according to the USDA Genetic Resources Informa- tion Network (GRIN) taxonomy Brachiaria brizantha, B. decumbens and B. humidicola are now ac- cepted as Urochloa brizantha, U. decumbens and U. humidicola, respectively, Panicum maximum as Megathyrsus maximus, Pennisetum purpureum as Cenchrus purpureus, and Desmodium heterocarpon as Grona heterocarpa. References Andersson, M.S., Peters, M., Schultze-Kraft, R., Gallego, G. and Duque, M.C. (2006) Molecular character- ization of a collection of the tropical multipurpose shrub legume Flemingia macrophylla. Agroforestry Systems 68, 231–245. Andersson, M.S., Schultze-Kraft, R., Peters, M., Duque, M.C. and Gallego, G. (2007) Extent and structure of genetic diversity in a collection of the tropical multipurpose shrub legume Cratylia argentea (Desv.) O. Kuntze as revealed by RAPD markers. Electronic Journal of Biotechnology 10, 386–399. Asudi, G.O. van den Berg, J., Midega, C.A.O., Pittchar, J., Pickett, J.A., et al. (2015) Napier grass stunt dis- ease in East Africa: farmers’ perspectives on disease management. Crop Protection 71, 116–124. Cherney, D.J.R. (2000) Characterization of forages by chemical analysis. In: Givens, D.I., Owens, E., Axford, R.F.E. and Omed, H.M. (eds) Forage Evaluation in Ruminant Nutrition. CAB International, Wallingford, UK, pp. 281–300. Cocks, P.S. and Bennett, S.J. (1996) Role of pasture and forage legumes in Mediterranean farming sys- tems. In: Bennett S.J. and Cocks P.S. (eds) Genetic Resources of Mediterranean Pasture and Forage Legumes. Current Plant Science and Biotechnology in Agriculture, Vol 33. Springer, Dordrecht, Netherlands, pp. 9–19. Cook, B.G., Pengelly, B.C., Brown, S.D., Donnelly, J.L., Eagles, D.A., et al. (2005) Tropical Forages: An Interactive Selection Tool. CSIRO, Brisbane, Australia: Department of Primary Industries and Fisheries, Queensland, Australia, CIAT, Cali, Colombia, and ILRI, Nairobi, Kenya. Engler, J.O. and von Wehrden, H. (2018) Global assessment of the non-equilibrium theory of rangelands: revisited and refined. Land Use Policy 70, 479–484. Evenson, R.E., Gollin, D. and Santaniello, V. (1998) Agricultural Values of Plant Genetic Resources. FAO, Rome, and CAB International, Wallingford, UK. Farrell, G. (1998) Towards the Management of Ustilago kamerunensis H Sydow and Sydow, a smut pathogen of napier grass (Pennisetum purpureum Schum.) in Kenya. PhD thesis, University of Greenwich, London. Gutteridge, R.C. and Shelton, H.M. (1995) New herbage plant cultivars Sesbania sesban (L.) Merrill (sesban) cv. Mount Cotton. Tropical Grasslands 29, 188–189. Forage Diversity, Conservation and Use 447 Hanson, J. and Schultze-Kraft, R. (2009) Regeneration guidelines: forage grasses. ILRI, Addis Ababa, and CIAT, Cali, Colombia. Hanson, J. and Street, K. (2008) Regeneration guidelines: grasspea. ILRI, Addis Ababa, and ICARDA, Aleppo, Syria. Hanson, J. and Tedla, A. (2010) Forage seed production in Ethiopia. Seed Info 39, 16–20. Hanson, J., Amri, A., Street, K., Shehadeh, A., Rukhkyan, N. and Snowball, R. (2009) Regeneration guide- lines: forage legumes. ILRI, Addis Ababa. Hawkes, J.G. (1971) Conservation of plant genetic resources. Outlook on Agriculture 6, 248–253. Heering, H., Reed, J.D. and Hanson, J. (1996) Differences in Sesbania sesban accessions in relation to their phenolic concentration and HPLC fingerprints. Journal of the Science of Food and Agriculture 71, 92–98. Hodgkin, T., Brown, A.H.D., van Hintum, T.J.L. and Morales, E.A.V. (1995) Core Collections of Plant Gen- etic Resources. John Wiley & Sons, Chichester, UK. Horna, D. and Smale, M. (2010) Evaluating cost-effectiveness of collection management: ex-situ conser- vation of plant genetic resources in the CG system. Final Report. IPGRI, Rome. Huang, C., Liu, G. and Bai, C. (2017) Polymorphism analysis in identification of genetic variation and rela- tionships among Stylosanthes species. 3 Biotech 7, 39. IBPGR/CEC (1984) Forage legume descriptors. IBPGR, Rome, and Commission of European Communi- ties (CEC), Brussels. IBPGR/CEC (1985) Forage grass descriptors. IBPGR, Rome, and Commission of European Communities (CEC), Brussels. ILCA (1994) Improving livestock production in Africa: evolution of ILCA’s programme, 1974–94. ILCA, Addis Ababa. ILRI/KALRO (2013) Getting superior Napier grass to dairy farmers in East Africa. ILRI, Addis Ababa, and KALRO, Nairobi. Jamnadass, R., Hanson, J., Poole, J., Hanotte, O., Simons, T.J. and Dawson, I.K. (2005) High differenti- ation among populations of the woody legume Sesbania sesban in sub-Saharan Africa: implications for conservation and cultivation during germplasm introduction into agroforestry systems. Forest Ecology and Management 210, 225–238. Jank, L., do Valle, C.B. and Resende, R.M.S. (2011) Breeding tropical forages. Crop Breeding and Ap- plied Biotechnology 1, 27–34. Jank, L., Barrios, S.C., do Valle, C.B., Simeāo, R.M. and Alves, G.F. (2014) The value of improved pas- tures to Brazilian beef production. Crop and Pasture Science 65, 1132–1137. Jorge, A., Hanson, J., Tedla, A. and Abdena, A. (2012) Promoting forages for livestock feed through germplasm and knowledge. In: Assefa, G., Dejene, M., Hanson, J., Anemut, G., Mengistu, S. and Mengistu, A. (eds) Forage Seed Research and Development in Ethiopia. EIAR, Addis Ababa, pp. 45–58. Jorge, A., Lukuyu, B., Marita, C., Mwangi, D.M., Kinuthia, E., Baltenweck, I. and Poole, J. (2014) Assess- ing the uptake and disease impact of Napier grass in Kenya. ILRI Project Report. ILRI, Nairobi. Kariuki, I.W., Mwendia, S., Muyekho, F.N., Ajanga, S.I. and Omayio, D.O. (2016) Biomass production and forage quality of head-smut disease resistant Napier grass accessions. African Crop Science Journal 24, 157–165. Khan, Z.R., Midega, C.A.O., Nyang’au, I.M., Murage, A., Pittchar, J., et al. (2014) Farmers’ knowledge and perceptions of the stunting disease of Napier grass in Western Kenya. Plant Pathology 63, 1426–1435. Kilian, B. and Graner, A. (2012) NGS technologies for analyzing germplasm diversity in gene banks. Brief- ings in Functional Genomics 11, 38–50. Koo, B., Pardey, P.G. and Wright, B.D. (2003) The economic costs of conserving genetic resources at the CGIAR centres. Agricultural Economics 29, 287–297. Koo, B., Pardey, P.G. and Wright, B.D. (2004) Saving Seeds: The Economics of Conserving Crop Genetic Resources Ex Situ in the Future Harvest Centres of the CGIAR. CAB International, Wallingford, UK. Kumar, S., Gupta, P., Barpete, S., Sarker, A., Amri, A., et al. (2013) Grass pea. In: Singh, M., Upadhyaya, H.D. and Bisht, I.S. (eds) Genetic and Genomic Resources for Grasspea Improvement. Elsevier, Oxford, pp. 269–292. Labarta, R., Andrade, R., Marin, D., Rivera, T., Orrego, M. and Pinillos, J. (2017) The impacts of CIAT’s collaborative research. CIAT, Cali, Colombia. 448 J. Hanson et al. Lowe, A.J., Thorpe, W., Teale, A. and Hanson, J. (2003) Characterisation of germplasm accessions of Napier grass (Pennisetum purpureum and P. purpureum × P. glaucum hybrids) and comparison with farm clones using RAPD. Genetic Resources and Crop Evolution 50, 121–132. Maass, B.L., Jamnadass, R.H., Hanson, J. and Pengelly, B.C. (2005) Determining sources of diversity in cultivated and wild Lablab purpureus related to provenance of germplasm by using amplified fragment length polymorphism. Genetic Resources and Crop Evolution 52, 683–695. Mather, R.D.J., Melhuish, D.T. and Herlihy, M. (1996) Trends in the global marketing of white clover culti- vars. Agronomy Society of New Zealand Special Publication 11, 7–14. Munyasi, J.W., Auma, E.O., Ngode, L. and Muyekho, F.N. (2015) Evaluation of biomass yield of selected fodder species in relation to frequency of harvest alongside defoliation heights in Western Kenya. Journal of Agricultural Extension and Rural Development 7, 257–262. Muyekho, F.N., Mose, L. and Cheruiyot, D.T. (2003) Development and transfer of forage production tech- nologies for smallholder dairying: case studies of participatory evaluation of species and methods of establishment in western Kenya. Tropical Grasslands 37, 251–256. Mwendia, S., Wanyoike, M., Wahome, R. and Mwangi, D. (2006) Farmers’ perceptions on importance and constraints facing Napier grass production in Central Kenya. Livestock Research for Rural Develop- ment 18, 160. Mwendia, S., Wanyoike, M., Wahome, R. and Mwangi, D. (2007) Effect of napier head smut disease on Napier yields and the disease coping strategies in farming systems in central Kenya. Livestock Research for Rural Development 19, 109. NAFIS (2017) Napier grass economic analysis. National Farmers Information Service (NAFIS), Nairobi, Kenya. Available at: www.nafis.go.ke/fodders/napier-grass/economic-analysis/ (accessed 18 March 2020). Negawo, A.T., Teshome, A., Kumar, A., Hanson, J. and Jones, C.S. (2017) Opportunities for Napier grass (Pennisetum purpureum) improvement using molecular genetics. Agronomy 7, 28. Negawo, A.T., Jorge, A., Hanson, J., Teshome, A., Muktar, M.S., et al. (2018) Molecular markers as a tool for germplasm acquisition to enhance the genetic diversity of a Napier grass (Cenchrus purpureus syn. Pennisetum purpureum) collection. Tropical Grasslands-Forrajes Tropicales 6, 58–69. Nei, M. (1987) Molecular Evolutionary Genetics. Columbia University Press, New York. Omayio, D.O., Ajanga, S.I., Muoma, J.V., Muyekho, F.N., Yamame, M.K. and Kariuki, I.M.S. (2015) Using Napier grass accessions’ origins, neighbour joining groups and their response to Ustilago kamerun- ensis to predict a probable co-evolutionary scenario. International Journal of Recent Scientific Research 6, 2639–2645. Oram, R.N. (1990) Register of Australian Herbage Plant Cultivars. CSIRO, Canberra. Orodho, A.B. (2006) The role and importance of Napier grass in the smallholder dairy industry in Kenya. World Agricultural Information Centre Portal, FAO, Rome. Osawaru, M.E., Ogwu, M.C. and Aiwansoba, R.O. (2015) Hierarchical approaches to the analysis of gen- etic diversity in plants: a systematic overview. University of Mauritius Research Journal 21, 92–127. Pardey, P.G., Koo, B., Wright, B.D., van Dusen, M.E., Skovmand, B. and Taba, S. (2001) Costing the con- servation of genetic resources: CIMMYT’s ex situ maize and wheat collection. Crop Science 41, 1286–1299. Pengelly, B.C. and Maass, B.L. (2001) Lablab purpureus (L.) Sweet – diversity, potential use and deter- mination of a core collection of this multi-purpose tropical legume. Genetic Resources and Crop Evolution 48, 261–272. Plumb, V.E., Wood, C.D., Gay, C.H.J. and Heering, J.H. (1996) The use of HPLC-derived phenolic profiles as a means of classifying Sesbania accessions. Journal of the Science of Food and Agriculture 70, 481–486. Rao, N.K., Hanson, J., Dulloo, M.E., Ghosh, K., Nowell, D. and Larinde, M. (2006a) Manual of Seed- Handling in Gene Banks. Handbook for Gene Banks No. 8. Bioversity International. Rome. Rao, N.K., Hanson, J., Dulloo, M.E. and Goldberg, E. (2006b) Seed-handling in Gene Banks: Self Learning Module. Bioversity International, Rome. Reid, R.S., Serneels, S., Nyabenge, M. and Hanson, J. (2005) The changing face of pastoral systems in grass dominated ecosystems of Eastern Africa. In: Suttie, J.M., Reynolds, S.G. and Batello, C. (eds) Grasslands of the World. FAO, Rome, pp. 19–76. Robotham, O. and Chapman, M. (2017) Population genetic analysis of hyacinth bean (Lablab purpureus (L.) Sweet, Leguminosae) indicates an East African origin and variation in drought tolerance. Genetic Resources and Crop Evolution 64, 139–148. Forage Diversity, Conservation and Use 449 Sandhu, J.S., Kumar, D., Yadav, V.K., Singh, T., Sah, R.P. and Radhakrishna, A. (2015) Recent trends in breeding of tropical grass and forage species. In: Roy, M.M., Malaviya, D.R., Yadav, V.K., Singh, T., Sah, R.P., et al. (eds) Proceedings of the 23rd International Grassland Congress 2015, 20–24 No- vember, New Delhi, pp. 337–348. Schultze-Kraft, R., Peters, M. and Wenzl, P. (2020) A historical appraisal of the tropical forage collection conserved at CIAT. Manuscript submitted to Genetic Resources. Smale, M. and Hanson, J. (2010) Assessing the impact of CGIAR investments in germplasm collection, conservation, characterization and evaluation: the scope of the literature. Available at: https://cas. cgiar.org/sites/default/files/images/SmaleHansen2011.pdf (accessed 17 February 2020). Suttie, J.M., Reynolds, S.G. and Batello, C. (eds) (2005) Grasslands of the World. FAO, Rome. Tarawali, S.A., Tarawali, G., Larbi, A. and Hanson, J. (1995) Methods for the evaluation of forage legumes, grasses and fodder trees for use as livestock feed. ILRI Manual, no, 1. ILRI, Nairobi. Thorpe, W., Muriuki, H.G., Omore, A., Owango, M.O. and Staal, S. (2000) Dairy development in Kenya: the past, the present and the future. Paper presented at the Annual Symposium of the Animal Production Society of Kenya (APSK). Theme: Challenges to Animal Production in this Millennium, KARI Head- quarters, 22–23 March, Nairobi. ILRI, Nairobi. Torres, G. (2005) Molecular and morphological phylogenetic analysis of Brachiaria and Urochloa (Poaceae). Molecular Phylogenetics and Evolution 37, 36–44. van de Wouw, M., Hanson, J. and Nokoe, S. (1999) Observation strategies for morphological characterisation of forages. Genetic Resources and Crop Evolution 46, 63–71. van Hintum, T.J.L. (1995) Hierarchical approaches to the analysis of genetic diversity in crop plants. University of Mauritius Research Journal 21, 23–34. van Hintum, T.J.L., Brown, A.H.D., Spillane, C. and Hodgkin, T. (2000) Core collections of plant genetic resources. IPGRI Technical Bulletin No. 3. IPGRI, Rome. Wamalwa, N.I.E., Midega, C.A.O., Ajanga, S., Omukunda, N.E., Muyekho, F.N., et al. (2017) Screening Napier grass accessions for resistance to Napier grass stunt disease using the loop-mediated isothermal amplification of DNA (LAMP). Crop Protection 98, 61–69. Wambugu, P.W., Ndjiondjop, M. and Henry, R.J. (2018) Role of genomics in promoting the utilization of plant genetic resources in gene banks. Briefings in Functional Genomics 17, 198–206. Wanjala, B.W., Obonyo, M., Wachira, F.N., Muchugi, A., Mulaa, M., et al. (2013) Genetic diversity in Napier grass (Pennisetum purpureum) cultivars: implications for breeding and conservation. AoB Plants 5, plt022. Weitzman, M.L. (1993) What to preserve? An application of diversity theory to crane conservation. Quarterly Journal of Economics 1, 157–183. West, N.E. (1993) Biodiversity of rangelands. Journal of Range Management 10, 2–13. White, D.S., Peters, M. and Horne, P. (2013) Global impacts from improved tropical forages: a meta-analysis revealing overlooked benefits and costs, evolving values and new priorities. Tropical Grasslands- Forrajes Tropicales 1, 12–24. Zohrabian, A., Traxler, G., Caudill, S. and Smale, M. (2003) The marginal value of an accession. Biotech- nology and Genetic Resources Policies, Brief 9. IFPRI, Washington, DC. 13 The Impact of CGIAR Centre Research on Use of Planted Forages by Tropical Smallholders Alan J. Duncan1, Michael Peters2, Rainer Schultze-Kraft3, Philip K. Thornton4, Nils Teufel5, Jean Hanson6 and John McIntire7 1International Livestock Research Institute, Addis Ababa, Ethiopia and Global Academy of Agriculture and Food Security, University of Edinburgh, UK; 2Alliance of Bioversity and CIAT, Nairobi, Kenya; 3Alliance of Bioversity and CIAT, Cali, Colom- bia; 4CGIAR Research Programme on Climate Change, Agriculture and Food Security (CCAFS), International Livestock Research Institute, Nairobi, Kenya; 5International Livestock Research Institute, Nairobi, Kenya; 6International Livestock Research Institute, Addis Ababa, Ethiopia; 7Santa Barbara, CA, USA Contents Executive Summary 451 The problem 451 Potential of planted forages 451 Research spending at ILRI 451 Impacts 452 Scientific impacts 452 Development impacts 452 Capacity building and partnerships as development impacts 453 Future directions 453 Introduction 453 The Role of Planted Forages in Tropical Farming Systems 454 The extent of planted forages in tropical farming systems 456 Adoption of planted forages in sub-Saharan Africa 457 Fodder banks in West Africa 457 Multi-purpose trees 458 Napier grass in East Africa and beyond 459 Push–pull technology 461 Adoption of planted forages in Asia 462 Adoption of planted forages in Latin America and the Caribbean 462 Knowledge Products on Planted Forages 463 Tropical Forages tool 463 Tropical Grasslands-Forrajes Tropicales 464 Feed Assessment Tool (FEAST) 464 © International Livestock Research Institute 2020. The Impact of the International 450 Livestock Research Institute (eds J. McIntire and D. Grace) The impact of CGIAR research on use of planted forages by tropical smallholders 451 Forage Seed Technical Support and Distribution 464 Environmental Benefits of Planted Forages 466 What has Limited the Impact of Planted Forages in Tropical Africa and Asia? 467 Global Impacts from Planted Tropical Forages 468 New Adoption Potential 469 The Future 472 References 473 Executive Summary including reducing the labour burden associated with feed sourcing, especially on women. F orages The problem can also deliver economic benefits through i mproved livestock productivity, which often ex- Livestock are an integral component of the ceeds the opportunity costs of utilizing land for smallholder mixed systems that dominate sub- staple crops. Finally, forages can play a strong Saharan Africa and South Asia. Availability of role in delivering environmental impact through sufficient high-quality feed is a major constraint soil stabilization, carbon sequestration and main- to productivity; livestock are often fed opportun- tenance of habitats which support biodiversity. istically and on poor-quality feed resources. Information on the use of planted forages in A decline in grazing resources in response to the the tropics is scanty, although there have been expansion of cultivated land and poor control recent attempts to synthesize available informa- over grazing rights means ever-increasing pro- tion. These reviews point to large areas of improved portions of variable-quality cereal and legume crop planted forages in Latin America with a few residues in ruminant livestock diets. Cultivation documented successes in high-potential sites in of green forages specifically for feeding livestock China, India, South-east Asia and East Africa. is an important potential means of addressing the feed gap. The prominence of planted forages in smallholder farming systems varies hugely, and Research spending at ILRI the extent of their cultivation in sub-Saharan Africa and to some extent Asia is lower than would The International Livestock Research Institute (ILRI) be expected given their potential to alleviate the and its predecessor, the International Livestock chronic feed gap. This chapter explores the poten- Centre for Africa (ILCA), have made major invest- tial and actual impact of planted forages and reviews ments in livestock feed research. ILCA scientists success cases emerging from CGIAR research. (1974–1994) working on feed research of all types – range ecology, forage diversity, planted forages, multi- dimensional crops and the nutritional quality of feeds Potential of planted forages – were about 18.8% of the 1974–1994 number of ILCA staff. These figures imply real spending on vari- Planted forages offer a range of benefits, which ous aspects of livestock feeding of some US$70 mil- is at odds with their apparent underuse in many lion (in 2015 US$) in the 20 years of ILCA’s existence. smallholder farming systems. Well-managed ILRI (1995–2017) spending was some US$105 cultivated forages provide substantial yields million, for a 1975–2018 commitment of US$175 of nutrient-rich biomass. Grasses provide large million, or roughly 10% of the total. Planted forages amounts of moderate-quality feed. Legumes accounted for the majority of ILRI spending in the generally yield less, but the vegetative material area of livestock feeding. We have not been able to is of exceptionally high quality and provides an estimate research spending on forages at the Inter- excellent high-quality feed to complement the national Centre for Agricultural Research in Dry- basal resource of cereal crop residues that often land Areas (ICARDA) or at the Centro Internacional dominate livestock diets across the tropics. de Agricultura Tropical (CIAT, or International Planted forages have several potential im- Center for Tropical Agriculture now part of the pacts. Use of forages can deliver social benefits Alliance of Bioversity International and CIAT). 452 A.J. Duncan et al. Impacts multiple production benefits including improved livestock production, and the development of Research by the various CGIAR centres with intensive grass production on small plots in interests in planted forages has yielded a range South-east Asia in support of beef fattening. of success cases where research has led to Second, the development of improved plant- widespread uptake of specific interventions in- ed forage cultivars, especially in Latin America, volving planted forages. For example, the World involving the private sector, the national research Agroforestry Centre (ICRAF) has been working system and CIAT has clearly led to extensive up- on the use of multi-purpose trees as livestock fod- take by relatively large-scale farmers with clear der in East Africa for the past three decades. This livestock production dividends. research effort led to the development of feeding A third area of scientific impact has been practices especially in support of dairy production. the developing of understanding and methods In partnership with a range of development part- for prioritization of planted forage options for ners, multi-purpose trees were adopted by up to particular localities. Examples of research out- 230,000 farmers according to some estimates. puts that codify the extensive knowledge within Similar experiences apply to other forage-based the global forage/feed network including CGIAR technologies including the push–pull system, in- scientists have been the widely used Tropical tensive small-scale grass production for beef Forages tool (www.tropicalforages.info/; accessed fattening in South-east Asia and recent uptake 27 February 2020) and recently the Feed As- of planted forages – often based on selections by sessment Tool (FEAST). Linked to this, CGIAR CIAT and ILRI – in South-east Asia and eastern research has been important in understanding Africa. In Latin America and the Caribbean, planted the systemic constraints on planted forage use in forages, by spontaneous adoption and selection smallholder systems across the tropics. by the private sector, national agricultural re- search systems and CIAT have transformed live- Development impacts stock production over the last 60 years and are the main source of animal feed albeit mostly on There has been considerable development impact large farms. More recently, increased attention from planted forage cultivars notably Brachiaria has been given to bred grass cultivars, with early spp. in Latin America, especially in Brazil. In Brazil uptake estimated by extrapolating from seed sales the strong private sector and national research of more than 900,000 ha of CIAT bred hybrids. system (Brazilian Agricultural Research Corpor- Despite these successes, the uptake of plant- ation or EMBRAPA) have been instrumental along ed forages has been much less in sub-Saharan with CIAT. In other countries in Latin America and Africa and Asia than, for example, in Latin Amer- the Caribbean, the contributions of CIAT have ica. There are different analyses on the basis for been more central. Other examples noted in this this lack of uptake ranging from those who point chapter include Stylosanthes spp. in China and to lack of awareness of the potential merits of Thailand and fodder hedgerows in East Africa. planted forages among farmers through to those Several other development impacts can be who attribute lack of uptake to inherent system identified from CGIAR research on planted forages. characteristics that make them uncompetitive The forage gene banks of ILRI, ICARDA and CIAT with alternative use of land and labour. have been the principal sources of germplasm for development as described in Chapter 12 (this vol- Scientific impacts ume). The same centres have acted as a know- ledge repository on planted forages for example Several scientific impacts are clear from CGIAR through collation of the Tropical Forages tool. research on planted forages over the past four Studies of the economic impact of planted decades. First, CGIAR centres and national and forages related directly to ILRI’s work are rare. international partners have been successful in de- An impact assessment in the 1990s in 15 West veloping intervention strategies involving plant- African countries showed some 27,000 adopt- ed forages. Examples include the use of ers of the fodder bank technology on 19,000 ha. multi-purpose trees in East Africa in support of The calculated economic internal rate of return dairy production, the push–pull system with its to the fodder bank technology was 38%. The impact of CGIAR research on use of planted forages by tropical smallholders 453 Capacity building and partnerships as livestock nutrition. This includes forage grasses, development impacts herbaceous forage legumes and forage shrubs/ trees, the latter often legumes; in addition, some Capacity development impacts of CGIAR re- non-legumes and non-grasses such as Tithonia di- search have been mainly through the develop- versifolia, Morus alba and Trichanthera gigantea are ment of knowledge products on forages, notably planted and used for animal feeding. the Tropical Forages tool, which is the main glo- Forage-based livestock production is signifi- bal knowledge repository on tropical forages and cant in temperate systems. In some cases, plant- is widely used. In addition, the development of ed forages are the main source of nutrition (e.g. technical materials and training manuals on for- Williams et al., 2007, for New Zealand). In the age development, notably in South-east Asia, is an tropics and subtropics, cultivation of forage dom- excellent example of the capacity development inates livestock production in Latin America and work of CGIAR centres. Finally, CGIAR centres the Caribbean (Jank et al., 2014) but is much have had a strong convening function, for ex- less prominent in Asia and Africa. In the latter ample in convening of international networks on regions, livestock are fed on crop residues and forages through the Tropical Forages Newsletter natural forage either grazed in situ or cut and initiative and the African Feed Research Network carried to confined animals (Renard, 1997). (AFRNET) convened by ILCA in the 1990s and In a recent study of 12 global locations in sub- the RIEPT network in Latin America and the Saharan Africa and South Asia, forage crops Caribbean convened by CIAT in the 1980s. contributed no more than 10% of cattle diets, with crop residues accounting for a much larger share (Valbuena et al., 2015). Research in India Future directions has found that crop residues were the single most important feed resource, supplying around Future work on planted forages by CGIAR centres 70% of intake, with planted forages contribut- must involve renewed effort in breeding to pro- ing 15% (Blummel et al., 2014). There are pock- duce lines that are resilient to stress, particularly ets of intensive forage use; for example, Napier in the face of climate change and land pressure grass (Pennisetum purpureum1) in East Africa is a and responding to emerging and specific market base feed of choice in dairy systems (Staal et al., demands. The promising work on prioritization 2002). Moreover, recent work and case studies of feed options including planted forages must be have identified the potential benefits of planted extended to allow better targeting of forage op- forages in eastern Africa (González et al., 2016; tions to reduce effort to promote options that have N. Teufel et al., unpublished data) and tropical little prospect of success. Finally, CGIAR research Asia (Stür et al., 2013) with an increasing – needs to continue the process of stronger part- a lthough as yet limited – uptake of tropical forages nering with private-sector actors at a range of beyond Latin America and the Caribbean, as scales in the feed sector. In the case of planted shown later in this chapter. forages, this could involve the incubation of small- The work of CGIAR on forages commenced scale seed producers through technical training. in the early 1970s with research in CIAT as part of the Beef Production Systems Program. This programme later shifted its focus entirely to tropical forages and became the Tropical Forages Introduction Program in 1979 (Lynam and Byerlee, 2017b). In ILCA (and later ILRI), planted forages research The purpose of this chapter is to answer the ques- began in the late 1970s with evaluations of grass, tion of what the scientific and economic value of re- forage legume and forage tree species in Nigeria search and development work by CGIAR centres and later in Ethiopia. The work of ICARDA and partners has been in planted forages for small- started at a similar time, focusing on testing holders in tropical and subtropical livestock sys- and disseminating cereal/legume rotations with tems in Africa, Asia and the Americas. By planted farmers in several countries in the West Asia and forages, we mean the whole range of plants used North Africa region, including Algeria, Iraq, for feeding livestock and specifically cultivated for Jordan, Lebanon, Morocco, Syria and Tunisia. 454 A.J. Duncan et al. The empirical basis of the impact of planted work on Brachiaria coordinated by ILRI and forages on tropical and subtropical livestock pro- CIAT in East Africa with scaling of up to 25,000 duction in Africa and Asia is not as strong as it households adopting Brachiaria germplasm should be. Some argue that there has been little selections (S. Ghimire, personal communica- to show for this effort in CGIAR institutions tion) and the adoption of P. purpureum in eastern (Squires et al., 1992; Thomas and Sumberg, and Central Africa (Negawo et al., 2017; Staal 1995) beyond Latin America and the Caribbean, et al., 2002). where there are documented large-scale impacts The work of Peters et al. (2001) is a forceful (Lynam and Byerlee, 2017b). argument for planted forages as instruments for In this chapter, we first consider the tech- higher productivity and better natural resource nical merits of planted forages. We then review management. They reviewed ‘…the role of for- what is known about the extent of cultivation age crops in improving the productivity of small- of planted forages in selected locations before holder farming systems and breaking the cycle summarizing various successes in different of poverty and resource degradation [by pre- tropical regions. Last, we reflect on lessons senting] the contributions of forage crops to in- from these experiences and present ideas for creasing farm incomes, intensifying farm pro- future research. duction, and contributing to better human nutrition.’ Planted forages offer strong technical bene- fits in mixed crop–livestock systems. For example, The Role of Planted Forages Peters et al. (2001) gave the agronomic argu- in Tropical Farming Systems ments for planted forages in the tropics: (i) they provide higher crop yields per unit of land com- Livestock are ubiquitous in the mixed crop– pared with alternatives, such as crop residues, livestock systems of the tropics. They are often natural pastures and browse; (ii) they provide poorly fed, subsisting on poor-quality crop re- higher forage quality in terms of nutrients per sidues, scavenged grasses and leaves, and a limited unit of dry matter (DM); (iii) there is improvement ration of agro-industrial by-products. Planted of soil quality by fixing nitrogen and retaining forages have long been promoted as a way of water; and (iv) they may fill seasonal feed short- improving feed supply. However, the place of plant- ages, in terms of providing feed when alterna- ed forages in smallholder livestock production tives are scarcest (e.g. natural pastures in the dry outside of perhaps East Africa, especially Kenya, season). Similar arguments were advanced by is often limited2. Planted forages are widely used Shelton et al. (2005). Rao et al. (2015) outlined in Latin America, although the definition of the importance of planted forages for ecosystem smallholder there does not compare with the services. much smaller scale in sub-Saharan Africa. Some White et al. (2013) give the technical bene- have argued that they are not used because of fits of forages in three domains. First, planted for- systemic and economic constraints (McIntire and ages can improve labour productivity given Debrah, 1986), although analysis often focuses on that they can reduce work to collect natural vege- forage legumes (Thomas and Sumberg, 1995; tation. Second, forages can allow savings in in- Sumberg, 2002). There are, however, recent suc- put use, such as water and fertilizer. Cultivation cess reports such as from Asia (Stür et al., 2013) of forages can lead to improved productivity in with more than 10,000 farmers adopting inten- terms of biomass yield, energy or protein per sive grass production, eastern Africa (Maass et al., unit area. Finally, forages can also have envir- 2015) with stated adoption of Brachiaria hybrids onmental benefits. Such benefits include im- by at least 20,000 farmers, eastern Africa (Franzel proved soil cover, reduced erosion and less weed and Wambugu, 2007) on the uptake of fodder infestation. shrubs (mostly Calliandra calothyrsus) by more Table 13.1 lists the top 15 forage species re- than 200,000 farmers, the adoption of the push– quested from the ILRI forage gene bank since the pull system of the International Centre of Insect early 1980s. Of the top ten grasses, P. purpureum Physiology and Ecology (ICIPE) by more than (commonly known as Napier or elephant grass) 30,000 farmers (Khan et al., 2011), the recent has been reported to produce a DM yield of The impact of CGIAR research on use of planted forages by tropical smallholders 455 Table 13.1. Indicative yield and nutritive value of the forage species most commonly requested from the ILRI gene bank (1984–2016). (Data on number of requests extracted from ILRI gene bank database. Yield and protein data from Cook et al., 2005; nutritive value data from Feedipedia, 2013.) Number of Indicative yield Crude protein Metabolizable Forage Type requests (t DM/ha/year) (% DM) energy (MJ/kg DM) Observations Chloris gayana Grass 833 10–25 3–17 8.5 Cenchrus ciliaris Grass 699 2–9 6–16 8.0 Pennisetum Grass 564 10–30 Leaf: 9.5–19.7 8.2 purpureum Lablab purpureus Legume 1,874 4 Leaf: 21–38; 9.2 Yield is per season stem: 7–20 Vigna unguiculata Legume 1,412 3–10 Green foliage: 14–21; 9.8 Yields refer to 8–12 crop residue: 6–8 weeks Cajanus cajan Legume 1,354 ~2 Leaf: 10–15 9.6 DM yields can be much higher under optimal conditions Stylosanthes Legume 1,210 5–10 12–20 8.0 guianensis Medicago sativa Legume 1,078 8–27a 18.3 8.5 No crude protein data in Cook et al. (2005) Neonotonia wightii Legume 828 3–8 17.1 9.1 No crude protein data in Cook et al. (2005) Stylosanthes hamata Legume 785 1–7 Leaf: 17–24; 8.8 stem: 6–12 Stylosanthes scabra Legume 763 1–10 Leaf: 10–20 n/a Trifolium tembense Legume 590 3–6 10–24 10.9 Sesbania sesban Tree 2,581 6–12 25–30 11.5 Leucaena leucocephala Tree 780 1–15 19–24 11.0 Gliricidia sepium Tree 604 2–20 18–30 11.5 aThe yield of M. sativa is under irrigation. 456 A.J. Duncan et al. between 12 and 90 t/ha/year and crude protein periods of scarcity (Tarawali and Pamo, 1992; concentrations of between 5 and 16% (Negawo Mekuria and Veldkamp, 2012) although prac- et al., 2017, and references therein). The data tical uptake of such practices has generally been overall show the very high yield potential of limited (Tarawali et al., 1999). planted forages, particularly forage grasses, when As well as biological nitrogen fixation by compared with natural pasture DM yields in, for forage legumes, some tropical grass roots have example, Ethiopia, which range from 1 t/ha/year recently been shown to exude chemical inhibi- in the lowlands to 4–6 t/ha/year on seasonally tors of biological nitrification, which suppress waterlogged fertile areas (Alemayehu, 1998). soil-nitrifying bacteria, reducing the rate of The data also illustrate the high nutritive value of leakage of nitrogen from the system, by blocking forages, especially forage legumes, many of which the conversion of ammonium to nitrate. Inhibi- have crude protein concentrations of around 20%. tors have been identified in root exudates from a Comparing these with the other major sources number of legume and grass species including of nutrition for livestock in developing-world sorghum and rice, but by far the most intense smallholder systems, crop residues, under nor- activity was detected in Brachiaria humidicola mal agronomic practices, cereal straw yields in (Subbarao et al., 2007). Subsequently, the cyclic sub-Saharan Africa would be in the order of diterpene ‘brachialactone’ was demonstrated to 1.5–7.0 t/ha with crude protein concentrations contribute between 60% and 90% of the biological of 5–10% and metabolizable energy concentra- nitrification inhibition activity in Brachiaria root tions of 6–8 MJ/kg DM (Zaidi et al., 2013). exudate (Subbarao et al., 2009). A further bene- Forage legumes, through their ability to form fit of these soil-nitrifying bacteria is reductions a symbiotic relationship with nitrogen-fixing in emissions of the GHG nitrous oxide. rhizobia hosted in their root nodules, also offer The paradox is that despite the positive ni- improvements in soil quality and reductions in trogen-fixing properties of legumes, which offer greenhouse gas (GHG) emissions and nitrogen benefits to plant production and soil health when fertilizer application. After harvesting the aerial grown in nitrogen-constrained environments, portion of the plants, the remains are degraded to in general the uptake of forage legumes in small- produce organic matter and the nitrogen com- holder systems has been limited (Sumberg, 2002; ponent is mineralized to form ammonia, which Shelton et al., 2005). is released into the soil and can be utilized by other plants in close proximity or planted subse- quently. In addition, unlike in the above-ground portion of the plant where nitrogen concentrations The extent of planted forages in tropical do not vary significantly, nitrogen concentra- farming systems tions in below-ground legume tissues have been shown to vary considerably (Carranca et al., 2015). Global estimates of the area planted to forages Results by Muhr et al. (1999) showed the are not readily available. There are, however, coun- positive effects of nitrogen contribution in the tries where forage data are readily available. Data order of 80 kg/ha to the succeeding crop, even from the Indian Council for Agricultural Research when removing the above-ground biomass, of (ICAR) indicate the area planted to ‘improved’ Stylosanthes guianensis. Depending on the way in forages in India is around 8 million ha with the which livestock are managed, some nitrogen dominant species being Egyptian clover (Trifoli- from legumes may be deposited on arable land as um alexandrinum) and forage sorghum (ICAR, excreta returns but usually only a fraction (Rufino 2011). In Brazil, another hotspot of planted for- et al., 2007). Finally, there is the case that some age production, EMBRAPA estimates that the forages provide feed at times of general feed scar- total area planted to forages is 115 million ha city. For example, some fodder trees provide green with Brachiaria spp. accounting for 80% of the feed during the dry season when other feeds are area and Panicum maximum accounting for 10% scarce (Franzel et al., 2014). Seasonal livestock (Sluszz, 2012). exclosures, where livestock are excluded for a Planted forages have been promoted for time to allow biomass accumulation, can also many years in eastern Africa (Abate et al., provide dry-season feed by reserving biomass for 1985). A survey of areas with commercial dairy The impact of CGIAR research on use of planted forages by tropical smallholders 457 production was carried out in Ethiopia and varieties of Napier grass also figure highly. In Kenya in 2015 to determine the levels of forage Kenya, local Napier varieties were the most com- adoption, production practices and importance mon forage and were grown by over 90% of house- as a feed source (N. Teufel et al., unpublished holds in the surveyed areas, although they ac- data), in which 180 communities were selected counted for only around 5% of total cultivated across the major regions in each country. In area. Improved varieties of Napier grass ranked Ethiopia, 20 woredas were purposively selected second in frequency of occurrence. Measuring from Tigray (three), Amhara (six), Oromiya adoption by area favours crops that are grown on (seven) and Southern Nations Nationalities and larger farms, such as fodder oats in Ethiopia or Peoples (four) for intensity of dairy production. Rhodes grass and the Kakamega Napier grass Within these woredas, nine kebeles3 were ran- varieties in Kenya. In summary, the area shares of domly selected. In Kenya, 12 counties were se- fodder crops are generally low in both countries, lected from Nyanza (three), western (two), Upper with Napier grass in Kenya being the most wide- Rift (three), central (three) and coast (one) re- spread, grown on around 6% of cultivated land gions, based on the density of dairy animals and (combining all varieties). Because many of these information about forages. Random selection crops are harvested multiple times per year, their identified 15 sublocations, the lowest formal ad- biomass shares are higher than their area shares. ministrative level, within each selected county and one village within each selected sublocation for the survey. During the survey, focus groups Adoption of planted forages responded to quantitative and qualitative ques- in sub-Saharan Africa tions on farming characteristics and forage pro- duction practices regarding the whole village, In this and the following sections we review cur- such as the total number of farming households rent knowledge on adoption of planted forages and cultivated area, the number of households related to CGIAR research efforts. We do this by growing a specific forage and the area of this for- focusing on a series of significant bodies of work age planted. Most forages were differentiated by in different regions that have led to at least mod- species only. Because of its importance, Napier erate success. grass (P. purpureum) was differentiated into four variety types at data collection: ‘Kakamega I’, Fodder banks in West Africa ‘Kakamega II’, ‘other improved varieties’ and ‘local varieties’. The share of households grow- The fodder bank concept was developed by ILCA ing forages among farming households and the (now ILRI) and partners in the late 1970s in forage share of cultivated land, where forage ex- Nigeria to help crop–livestock farmers in the dry tent was expressed in area units, were calculated to subhumid zone to: (i) overcome dry-season for comparison. feed constraints; and (ii) improve soil fertility. It Results of this study indicate that in Ethiopia, consisted of establishing small (typically 4 ha) forage grasses are grown by 10–35% of house- fenced paddocks with a prolifically seeding, self- holds, while in Kenya 10–85% of households regenerating forage legume (mainly Stylosanthes grow grasses. Forage legumes are grown by hamata) for strategic supplementation of live- fewer households overall; in Ethiopia, their oc- stock grazing natural pastures (Mohamed-Saleem currence (up to 12% of households) is higher and Suleiman, 1986). Such supplementation led than in Kenya (up to 2% of households). Forage to significant increases in livestock productivity trees occur reasonably widely with up to 40% of under experimental conditions. After 2–3 years, households growing trees in Ethiopia and up the legume fodder bank was replaced by an un- to 20% in Kenya. The areas devoted to planted fertilized sorghum or maize crop that provided forages are small. However, forage trees are often yields equivalent to fertilization with up to 45 kg planted singly or in rows, for which area meas- nitrogen/ha (Tarawali, 1991). After the crop ures were not recorded. phase, the area was reconverted to a fodder bank Sesbania was the most common planted for- via hard-seeded soil seed reserves. Research and age in Ethiopia, with 34% of households growing promotional activities related to this technology it and 0.5% of cultivated area allocated to it. Local came to an end in the early 1990s. 458 A.J. Duncan et al. An impact assessment, conducted by well as a synthesis of the various adoption stud- Elbasha et al. (1999) in 15 West African coun- ies. The best estimate was that milk yield in- tries showed that, until 1995, there were about creased by 0.6–0.8 kg/kg intake of Calliandra 27,000 adopters of the fodder bank technology leaf. Scaling this production impact up using es- covering about 19,000 ha. Given the estimated timates of adopters in East Africa, the authors US$7 million research expenditure to develop this estimated the economic impact of forage tree technology, an internal rate of return of 38% adoption in Kenya in 15 years to be in the order was calculated. Elbasha et al. (1999) stressed the of US$20–30 million. The report also assembled finding that a considerable time lag (at least 15 adoption data from various sources to arrive at a years) was necessary for diffusion of the tech- total of 206,000 farmers having adopted multi- nology. The second part of their statement, ‘... purpose trees in East Africa by 2005. By the au- the impact of adopted fodder banks has paid for thors’ own admission, the farmer numbers are the research that went into their development at ‘rough guesses’, often partly based on data from least three times over, and this will increase sub- reports of non-governmental organizations stantially in the next few years, given current promoting multi-purpose trees with associated adoption trends’ remains to be confirmed. uncertainties. Nevertheless, it is clear that the number of farmers growing multi-purpose trees Multi-purpose trees in East Africa is notable. The ‘project focus’ of the reported adoption ICRAF and partners have been active in develop- of multi-purpose trees was highlighted in a re- ing multi-purpose trees and shrubs as fodder in cent paper by Brockington et al. (2016) in which East Africa over the past three decades4. Multi- a previous agroforestry intervention was re- purpose trees have been defined as ‘all woody visited 5 years after the project end. While the perennials that are purposefully grown to pro- primary focus was on fruit trees, multi-purpose vide more than one significant contribution to trees also formed part of the study, and the issues the production and/or service functions of a land- around ‘adoption’ are generic. Assessment of use system. They are so classified according to fodder tree adoption must be done over several the attributes of the plant species as well as to years, whereas conventional project cycles are the plant’s functional role in the agroforestry usually too short to allow this. The result is that technology under consideration’ (Huxley, 1984). most assessments track the very early stages of Functions include livestock feed, construction, adoption focusing on counting numbers of farm- fuel, improved soil fertility, erosion control and ers receiving extension support and materials shelter, among others. while neglecting evaluation of longer-term Initial research in the early 1990s estimated diffusion of tree technologies. Abandonment of the potential of C. calothyrsus, Sesbania sesban and fodder tree interventions following the end of Leucaena leucocephala to provide feed for small- projects is relatively common (Francis and Atta- holder dairy cows in the central highlands of Kenya Krah, 1989; Mekoya et al., 2008). Not all farmers (Franzel and Wambugu, 2007). C. calothyrsus who test fodder trees go on to adopt them (Kiptot proved most suitable, and much of the research et al., 2007). However, there are also cases of effort has focused on understanding how it fits spontaneous diffusion of fodder tree technolo- into small farms. Multi-purpose trees and forage gies without researcher involvement (Kiptot et al., trees in general have the advantage that they re- 2006; Wambugu et al., 2011) so the story is not quire minimal inputs and can be planted along clear. What is needed for this and other planted field margins and on soil bunds. They have the forage initiatives are some independent evalu- disadvantage of slow growth (first harvests in ations involving medium-term revisits to project much of East Africa are 12 months after trans- sites to assess the extent of sustained use of fod- planting) and the need for stable land tenure to der trees beyond the project life. Further studies justify investment in a multi-season crop. like that of Brockington et al. (2016) would be ICRAF published a comprehensive review welcome. of the impact of multi-purpose trees (Place et al., Successful use of multi-purpose trees for 2009). This review gave data on the production fodder is ‘knowledge intensive’, and much of impact of intake of tree forage by livestock as ICRAF’s recent work has been around suitable The impact of CGIAR research on use of planted forages by tropical smallholders 459 dissemination pathways to enhance adoption internodes (i.e. increased leaf:stem ratio) and are (Kiptot et al., 2006; Lukuyu et al., 2012). At the considered to be of higher quality and to have current stage of smallholder system development, greater resilience to abiotic stresses (Sollenberger it may therefore be some while before we see et al., 1987). spontaneous uptake of multi-purpose trees with- Napier grass is most prevalent in the cut-and- out project intervention. carry systems that dominate livestock production systems, particularly dairy, in East and Central Napier grass in East Africa and beyond Africa. It is considered the most popular peren- nial fodder for the smallholder crop–livestock Napier grass (P. purpureum, sometimes called ele- farming systems in Kenya (Nyambati et al., 2010), phant grass) is a perennial species originating in and has been reported to represent approxi- sub-Saharan Africa that is known for its high mately 80% of planted forages in this country biomass production and rapid regeneration cap- (Staal et al., 1997). Recent estimates by authors abilities, good palatability and nutritional qual- of this chapter based on the survey reported in ities. These features have made it highly popular, Tables 13.2 and 13.3 indicated that Napier grass mainly as a ‘cut-and-carry’ feed in livestock is an important feed resource for at least 1.3 mil- production systems across the tropics and sub- lion households in Kenya and Ethiopia. The variety tropics. The popularity of Napier grass is not Bana is currently the most commonly grown in attributable to CGIAR research as it was intro- Kenya, although, as indicated elsewhere in this duced from southern Africa in the colonial, pre- chapter, other selected varieties are growing in CGIAR era. However, because of its importance, importance. This is mainly because, while con- especially in East Africa, it has featured strongly sidered resilient in the face of many pests and in CGIAR research as outlined below. Napier diseases, the impact of some specific diseases grass is highly adaptable to a broad range of pro- (namely head smut and stunt) have had a nega- duction systems and environments. Although it tive impact on the further growth and distribu- is known to grow best in regions where annual tion of this species. However, new material rainfall exceeds 750 mm and in locations below such as the smut-resistant Kakamega I and 2100 m above sea level, it has been reported to Kakamega II, which are accessions from the grow at up to 3000 m above sea level in tropical ILRI gene bank, are alternatives where smut is a regions and to contain similar concentrations of threat (Mwendia et al., 2008). More recently, crude protein to lucerne, and has been used to other gene bank accessions have been identified replace lucerne hay in livestock production sys- as a source of tolerance, or resistance, to stunt tems (Criscioni et al., 2016). It can be chopped and disease (Wamalwa et al., 2017). The potential of fed directly and can also be grazed or conserved Napier grass with respect to yield, disease resist- and made into hay, silage or pellets (Figueira et al., ance and ease of harvest has been assessed in 2015; Mapato and Wanapat, 2018). There are Ethiopia (Kebede et al., 2017) and Tanzania approximately 25 Napier grass varieties or culti- (Sikumba et al., 2015). vars, and an additional 16 sterile hybrids formed Napier grass has also had a significant im- with pearl millet (Pennisetum glaucum) currently pact in production systems outside Africa and being grown around the world (Cunha et al., has been readily adopted in South and South- 2011). In addition, the species is open pollinated east Asia. For example, Pakchong 1 (or ‘Super and therefore highly heterozygous with signifi- Napier’), a tall hybrid cultivar recently developed cant variability for both nutritional and agro- in Thailand and distributed in other countries nomic traits, which indicates that there are sig- including the Philippines, Malaysia and India, is nificant opportunities for the development of being widely promoted for cultivation in small- new and improved varieties (Negawo et al., 2017). holder systems (Halim et al., 2013; Wangchuk The current varieties come in two main forms: et al., 2015). In India, the standard Bajra-Napier there are the standard ‘tall’ varieties, which are hybrid (also known as king grass) variety has renowned for their biomass production and are been identified as an option for sustained fodder most popular in the cut-and-carry systems and yields, particularly for cattle and buffalo produc- there are the dwarf or ‘short’ varieties, which do tion systems (Kadam et al., 2017). Napier grass is not produce as much biomass but have shorter also a popular cut-and-carry forage in the trop- 460 A.J. Duncan et al. Table 13.2. Distribution of fodder grasses, legumes and trees in areas with dairy production in Ethiopia and Kenya. (Data from N. Teufel. et al., unpublished survey.) Forage grasses Forage legumes Forage trees Country Region No. villages No. villagesa Hh (%)b Area (%)c No. villagesa Hh (%)b Area (%)c No. villagesa Hh (%)b Area (%)c Ethiopia Amhara 54 40 10.9 0.6 25 5.6 0.4 53 40.6 2.4 Oromiya 63 54 12.2 0.6 30 3.3 0.1 40 15.2 3.1 SNNP 36 34 34.5 1.3 23 12 0.4 21 6.1 0.3 Tigray 27 25 14.8 1.3 23 8.5 1.1 27 24.3 3.4 Kenya Central 45 45 82.7 12.9 13 2 0.2 11 0.7 0.1 Coast 15 6 9.2 0.5 1 0.1 0 Nyanza 45 39 67.1 6.7 3 0.8 0.1 16 20.3 0.7 Upper Rift 45 45 42.2 7.6 3 0.3 0.2 11 6.3 0.1 Western 30 30 68.3 6.8 4 0.8 0 12 4.2 0.1 SNNP, Southern Nations, Nationalities, and Peoples’ Region. aNumber of villages in which fodder was grown. bShare of households (Hh) growing forages among farming households. cArea share of cultivated area devoted to forage, considering only forages for which extent of adoption was recorded by area units. The impact of CGIAR research on use of planted forages by tropical smallholders 461 Table 13.3. The most important fodder crops by households in Ethiopia and Kenya. (Data from N. Teufel et al., unpublished survey.) Country Rank Species No. villagesa Hh (%)b Area (%)c Ethiopia 1 Sesbania (Sesbania sesban) 136 34 0.5 Napier, local (Pennisetum 2 141 16 0.3 purpureum) 3 Rhodes grass (Chloris gayana) 50 7 0.2 Desho grass (Pennisetum 4 44 6 0.1 pedicellatum) 5 Lucerne (Medicago sativa) 43 5 0.1 Napier, local (Pennisetum Kenya 1 162 93 5.2 purpureum) Napier, other improved 2 11 10 0.2 (Pennisetum purpureum) 3 Rhodes grass (Chloris gayana) 41 8 0.8 4 Calliandra (Calliandra calothyrsus) 43 7 0 5 Fodder sorghum (Sorghum spp.) 3 6 0 aNumber of villages in which fodder was grown. bShare of households (Hh) growing forages among farming households. cArea share of cultivated area devoted to forage, considering only forages for which extent of adoption was recorded by area units. ical southern states of China where king grass is a biological intervention to control maize stem also an option. Here, new varieties that have been borer (Khan et al., 2006, 2008a, 2014; ICIPE, developed include Guimu and Guimin Yin, which 2015). The basis of the technology is that stem are mainly considered for use as feed for cattle and borer moths are repelled by phytochemicals in the other livestock (Shilin et al., 2007). In the grazed legume Desmodium spp. (D. intortum, D. uncinatum) systems of Japan, the dwarf varieties, commonly but are attracted by the green biomass provided grown in combination with Italian ryegrass (Lolium by Napier grass. Planting patterns for Desmodi- multiflorum), are growing in popularity (Ishii um, Napier grass and maize that take account of et al., 2005; Fukagawa and Ishii, 2018). these effects can lead to reduced stem borer in- In South America, one of the few active festation. The planting pattern usually involves breeding programmes for Napier grass exists at maize intercropped with Desmodium as a repel- EMBRAPA in Brazil (Gomide et al., 2015). There ling plant (push) with the intercropped area is a history of exchange of material between the s urrounded by Napier as an attractant for pests ILRI gene bank and EMBRAPA, which enhanced (pull). A further benefit is reduced infestation the diversity in each of the collections in support by Striga spp. (S. hermonthica and S. asiatica), an of the development of new cultivars (Negawo et al., obligate parasitic weed that can greatly reduce 2018). Some of the most recent cultivars from host crop yields. The technology has potential EMBRAPA produced include Pioneiro, a standard environmental and economic benefits in that it variety for cut-and-carry systems (Figueira et al., does not require expensive and damaging pesti- 2016), and BRS Kurumi, a dwarf variety mainly cides. The technology also provides feed in the targeted for grazing but also considered suitable form of Napier grass and Desmodium fodder, as for cut-and-carry systems (Gomide et al., 2015; well as reducing pests. Pereira et al., 2017). Assessments by the originating institutions suggest that tens of thousands of farmers have adopted the practice in East Africa. For example, Push–pull technology a 2011 study put the number of adopters at 30,000 in East Africa, with the technology cover- The so-called push–pull system developed by ing 15,000 ha (Khan et al., 2011). More recent ICIPE and the Rothamsted Research Institute is promotional materials estimate the number of 462 A.J. Duncan et al. adopters to be closer to 100,000 (ICIPE, 2015). The planted forage initiative in South-east In common with fodder trees, push–pull is a Asia evolved over two decades through multiple knowledge-intensive technology that has been projects with estimated numbers of farmers heavily promoted by the originating institu- adopting it in excess of 10,000 in Vietnam alone tions. Definitions of ‘adoption’ in the literature among those directly involved in project areas around push–pull tend to be vague and often (Stür et al., 2013). Although assessments of the relate to the numbers of farmers ‘reached’ extent of adoption are conducted by the origin- with the technology rather than those accept- ating institutions and are thus not independent, ing and adopting the practice over the long the counts are based on surveys of random sam- term. Adoption estimates are usually presented ples of households in the study districts and thus without a clear definition of adoption and have some credibility. The technology has now without supporting evidence against which to been taken up by local government structures judge their reliability. As with fodder trees, (Millar and Connell, 2010), and the reach is likely there is a need for independent evaluations to to be much larger although, again, independent assess the true extent of adoption of the push– assessments are needed. pull technology. Other examples are the adoption of the leg- The assumed strength of push–pull is its ume Stylosanthes guianensis, based on accession multiple benefits, which include reduced pest CIAT 184, as a cover crop and for leaf meal pro- damage, fewer weeds, more feed and improved duction in tropical China, reaching about 300,000 soil fertility using legumes (Desmodium spp.). farmers to feed poultry and pigs (Guodao and However, in the context of assessing forage suc- Chakraborty, 2005) and the use of various cesses, push–pull is not necessarily the strongest Stylosanthes spp. in India by 250,000 farmers example because the potential benefits accrue (Shelton et al., 2005). from both improved cereal yields and improved forage production. In some economic assessments of the benefits of push–pull, the extra milk from improved livestock feeding does not feature in Adoption of planted forages in Latin cost–benefit calculations (Khan et al., 2008b). America and the Caribbean Other studies have factored in economic benefits from forages based on prevailing market prices In terms of area, adoption of tropical forages in for forages and found them to be important in the Latin America and the Caribbean is widespread overall profitability of the technology (de Groote and the role of the CGIAR centres has been import- et al., 2010). ant (Lynam and Byerlee, 2017a) although the beneficiaries have generally been larger farmers. In 2002, a Brachiaria decumbens × brizantha × ruziziensis cultivar coming out of CIAT’s breeding Adoption of planted forages in Asia programme (Miles, 2001, 2007) was released (Lynam and Byerlee, 2017b) as the first bred CIAT has stimulated adoption of intensive grass Brachiaria cultivar to be documented. Brachiaria production linked to beef cattle fattening and hybrids have since been commercialized through cow–calf breeding systems in South-east Asia interaction with the private sector, namely the (Stür et al., 2007). The technology involves Papalotla Group and Dow AgroSciences. Uptake cut-and-carry grass plots with the species being based on documented commercial seed sales to mainly Panicum maximum, B. humidicola and the end of 2018 was estimated to be some Brachiaria hybrids. The areas planted to grasses 960,000 ha mostly sown in the past decade. are small, averaging around 0.25 ha, but plots According to CGIAR reports, more than as small as 0.1 ha are viable because production 100,000 ha are planted on an annual basis, is intensive and grass yields can be very high. with numbers still below the potential as seed The success of the planted forages initiative production and commercialization channels are depended on the participatory nature of the still evolving (CGIAR, 2018). Most adopters ap- intervention coupled with improved planting pear to be small and medium-sized livestock pro- materials (Ayele et al., 2012). ducers (Papalotla, personal communication), The impact of CGIAR research on use of planted forages by tropical smallholders 463 although not equivalent to ‘smallholders’ as (Holmann et al., 2004). Surveys in Colombia’s used in the African context. Recently, Papalotla Eastern Plains carried out in 2017 suggested registered advanced cultivars in Kenya and is that about one-third (about 3 million ha) of im- commercializing these through licensing agree- proved pastures are sown using Brachiaria cul- ments (Government of Kenya, 2016). The ori- tivars selected by CORPOICA and CIAT or bred ginal cultivar Mulato had limited commercial by CIAT; across a set of five countries, an esti- success due to low seed production and was mated 59.2% of all pastures were found to be quickly replaced by Mulato 2 when higher seed planted with Brachiaria grasses, with about production was included as an additional breed- half being CIAT- selected Brachiaria (ISPC, ing objective. In subsequent years, a series of 2018). Through the work of EMBRAPA, with Brachiaria decumbens × brizantha × ruziziensis hy- contributions from CIAT, Rivas (2002) esti- brid cultivars and a synthetic mixture were re- mated that 1.5 million ha had been sown to An- leased, namely Cayman (tolerant to water log- dropogon gayanus by 2000. ging), Cobra (more erect growth habit), Camello The documented success of forage legumes (better drought tolerance), Mestizo (synthetic so far is less visible. However, for Arachis pintoi, mixture of three hybrids for better establishment cv. Amarillo developed in Australia and selected and pasture utilization) (Papalotla: www.grupo- by EMBRAPA and CIAT for tropical America, papalotla.com/productos.html and Tropical around 65,000 ha have been reported to be Seeds: www.tropseeds.com/varieties/, both ac- adopted in Acre in Brazil (Valentim and de An- cessed 27 March 2020) and Converse (Dow drade, 2005). A Stylosanthes spp. mixture (‘Esti- Agrosciences). Additional materials for in- losantes Campo Grande’), co-developed by B. creased tolerance to drought and shade (e.g. for Grof (EMBRAPA, ex-CIAT), has been sown on silvopastoral systems) and additional synthetic 150,000 ha in the southern Cerrados of Brazil mixtures are to be commercialized in the next (Fernandes et al., 2005). 2–4 years (Papalotla, personal communication). These South American examples point to CIAT is also advancing the development and possible future growth in use of planted forages in commercialization of B. humidicola and Panicum Africa once livestock production becomes more maximum breeding lines. commercial, farm sizes increase, private-sector Some of the most widespread adoption of actors have stronger engagement and the in- planted forages is in Brazil where it is estimated stitutional environment is more conducive to that about 120 million ha have been planted, of growth of the forage sector. which 99 million are Brachiaria spp. and about 17 million ha are P. maximum (Jank et al., 2014). This has largely been driven by the private sec- tor and EMBRAPA, which are valuable part- Knowledge products ners, and involves large farms that are only per- on planted forages ipherally within the mandate of the CGIAR system. As well as research on planted forages, CGIAR Similar production increases have been centres have generated a range of ‘knowledge achieved in the Eastern Plains of Colombia as products’, which collect internal and published part of the collaboration of the Corporacion knowledge and present it in a form that is useful Colombiana de Investigacion Agropecuaria for scientists and the wider livestock develop- (CORPOICA) and CIAT, with inclusion of ment community. Examples of such knowledge Brachiaria spp. in crop/pasture rotations leading products are the Tropical Forages tool, the Trop- to a doubling of carrying capacity compared ical Grasslands-Forrajes Tropicales journal and the with degraded pasture and a tenfold increase Feed Assessment Tool (FEAST). over native savannah (Rincón and Ligarreto, 2008; Rincón et al., 2010). Tropical Forages tool Better documented is the uptake of Brachiaria grasses in Mexico and Central America where A wide array of plant species is used as feed for by the early 2000s over 3 million ha were re- livestock and these differ in both their suitability ported based on extrapolation from seed sales as livestock feed and in the biophysical conditions 464 A.J. Duncan et al. that support their growth. In 2005, the inter- Feed Assessment Tool (FEAST) national planted forages community initiated the Tropical Forages tool to collate the tacit FEAST (Duncan et al., 2012) is a systematic, knowledge of forage experts as well as published participatory approach to supporting design of data on forage characteristics to develop an livestock feed interventions at the village/com- online tool to support selection of appropriate munity level. It was originally developed in 2008 forages for specific purposes and locations. The by ILRI in collaboration with CIAT. FEAST in- Tropical Forages tool (www.tropicalforages. volves a structured conversation with farmers info; accessed 28 February 2020) has been at the village level to characterize the local available since 2005 and is currently receiving farming system, the role of livestock in the farm- approximately 500,000 annual hits with visitors ing system and the way in which livestock are coming from universities, development agencies, currently fed. This is followed by application of a seed companies and (informed) farmers. It is open short household survey among selected farm- access and easy to use, providing detailed infor- ers. Data from the survey are used to develop a mation on more than 170 major forage species series of standardized graphical outputs, which and the environments they are adapted to. The support decision making on appropriate feed tool was developed by the Commonwealth Sci- interventions. A  further feature is an interven- entific and Industrial Research Organisation tion-ranking module, which generates a priori- (CSIRO), the Queensland Department of Agricul- tized list of candidate interventions derived from ture and Fisheries, CIAT and ILRI, and capitalized automatic analysis of survey data. The FEAST on the inputs on more than 50 forage experts, data application itself has been downloaded with widespread knowledge and experience in by 1400 individuals and has been applied in tropical and subtropical forages. An updated ver- over a dozen countries. Over 70 FEAST reports sion of the tool was launched in 2021. have been published online. Published outputs in the FEAST collection in the CGIAR publica- tion repository have had over 15,000 views and Tropical Grasslands-Forrajes Tropicales downloads per year in the past 5 years. The online journal Tropical Grasslands-Forrajes Tropicales was established in 2012 as a successor to the former journals, Tropical Grasslands, pub- Forage seed technical support lished during 1967–2010 by the Tropical Grass- and distribution land Society of Australia, and Pasturas Tropicales, published during 1979–2007 by CIAT. CGIAR has been active in the provision of advice The main features of Tropical Grasslands- and technical training in forage production as Forrajes Tropicales are that the journal is inter- well as production and sale of seeds for estab- national, published online only, open access lishment of tropical forages to address these con- (no charges for subscription or publication fees), straints. The seed production units of CIAT and bilingual (English and Spanish), peer reviewed ILRI provide a source of tropical forage seeds and and guided by a 23-member Editorial Board, planting material of selected best-bet species at which is composed of the world’s leading trop- cost for use in establishing national forage seed ical pasture scientists. Further information on production. CIAT provides seeds of 25 herb- the journal is available at its website (www.trop- aceous legumes, nine grasses and seven fodder icalgrasslands.info; accessed 28 February 2020). trees. ILRI currently can supply seeds of 33 spe- Back issues of the predecessor journals, Tropical cies of herbaceous legumes, ten species of grass Grasslands and Pasturas Tropicales, can also be and five species of fodder trees. Provision of seeds accessed at this site. The journal is indexed in the from the two sources is complementary, mostly major abstract and citation databases of peer- handling different species and distributing in reviewed literature and is widely used; as of different regions, primarily within the regions December 2019, there had been more than where the centres are located (Table 13.4). CIAT 492,000 abstract views and more than 669,000 has shipped seeds of over 20,000 samples to 88 PDF/eBook downloads; currently the journal different countries, while ILRI has provided over receives 229,000 annual visits. 8500 samples in response to over 1500 orders The impact of CGIAR research on use of planted forages by tropical smallholders 465 since the establishment of their forage seed ac- by CIAT that will be distributed by Papalotla. Both tivities. Seed distribution has supported the for- CIAT and ILRI have supported entrepreneur- age evaluation networks of CIAT and ILRI in or farmer-led seed supply systems to comple- Asia, Latin America and the Caribbean, and ment large-scale private seed production. Col- sub-Saharan Africa, and has strengthened for- laboration with government institutions and age research and development activities globally. non-governmental organizations has ensured Tropical forage seed production is a special- training of farmers in seed production, seed quality ist market because many species are perennial control and certification. A farmer-led seed and annual reseeding is not required. This leads enterprise, PRASEFOR (Artisanal Forage Seed to uncertain demand, which causes a high de- Production), was formed in Honduras as an gree of risk to both seed producers and sellers association of 12 smallholder farmers to produce and has reduced investment in a more formal grass seeds. In contrast to many other farmer-led distribution seed system for forage seeds in many seed enterprises, this was business oriented and developing countries (Hanson and Peters, 2003). formed with very limited financial support, mak- This has also contributed to fewer new varieties ing the approach easily replicable (Hanson and and species being released in recent years and Peters, 2003). Farmers were able to obtain a promotion of an informal integrated community- return of more than US$600/ha on forage seed based seed supply system to fill the gap. A recent production compared with an estimated US$60/ survey in Ethiopia showed that many small- ha for maize production. A similar approach was holders are willing to pay for forage seeds and to recently piloted in Ethiopia through the Deutsche use land for planted forages where alternative Gesellschaft für Internationale Zusammenarbeit feed is scarce and where market opportunities GmbH (GIZ)-funded FeedSeed project by mentor- for milk and meat exist (Negassa et al., 2016). ing individuals and forage seed businesses Both centres have supported alternative representing smallholder farmers, cooperatives suppliers in the tropical forage seed agribusi- and commercial seed companies, thereby help- ness. CIAT has partnered with the Papalotla ing to stimulate the availability of more and Group from the private sector to achieve a wide better- quality forage seed and forage use by small- dissemination of hybrid pasture seeds developed holder livestock producers. The pilot project Table 13.4. Major countries of recipients of forage seeds from CIAT and ILRI. (Data from ILRI and CIAT databases, February 2019.) CGIAR centre Country Number of seed samples Amount distributed (kg) CIAT Bolivia 492 306 Brazil 837 535 Colombia 16,202 55,912 Costa Rica 570 993 Ecuador 618 189 Honduras 476 95 Mexico 597 110 Nicaragua 522 162 Peru 1,024 1,295 Venezuela 605 1,723 ILRI Burundi 42 54 Cameroon 98 74 Ethiopia 6,989 34,591 Ghana 81 33 Kenya 159 199 Pakistan 51 44 Rwanda 47 29 Tanzania 130 136 Uganda 69 77 Zambia 58 43 466 A.J. Duncan et al. incubated 30 profitable private companies 2013). Moreover, the high-risk sites, or ‘hotspots’, whose annual seed sales reportedly started at for GHG emissions are located in areas of low US$20,000 and increased to US$400,000 by livestock productivity – extensively managed the end of the 2-year project. With seed prices at areas in eastern Africa, Latin America and the US$5–20/kg in early 2019, forage seed produc- Caribbean, and South Asia (Gerber et al., 2013; tion does seem to be economically attractive. Herrero et al., 2013). These are also the areas with the greatest potential for increased forage use, although as argued elsewhere in this chap- ter increased forage use would depend heavily Environmental benefits of planted forages on the economics of land and labour use. As examples of the potential importance of The potential impact of tropical forages on GHG forages in climate-change scenarios, the govern- emissions has been well documented. Mitigation ments of Brazil and Colombia have identified the approaches include direct reduction of emissions, intensification of livestock production, using enhancing of carbon sequestration, higher pro- planted forages and proper management, as key ductivity per land area and/or livestock unit and strategies to mitigate GHG emissions in agricul- of feed efficiency (Peters et al., 2013; Searchinger ture. The strategies are outlined in the Plano ABC et al., 2013; Rao et al., 2015) and avoiding detri- (Ministério da Agricultura, Pecuária e Abasteci- mental land conversion. By applying suitable mento, 2012) and the NINO (Ministerio de Am- practices, land can be freed for reforestation or biente y Desarrollo Sostenible, 2015) for Brazil for landscape restoration while livestock output and Colombia, respectively, as a combination of grows. Genetic solutions to curb nitrous gas emis- increased productivity per livestock unit and sions and nitrate leaching – while reducing the land area and favouring environmentally sound amount of nitrogen fertilizers in crop–livestock land-use changes. Thornton and Herrero (2010) rotations – have been demonstrated as a proof of has illustrated the possibility of specific technolo- concept for biological nitrification inhibition gies such as improved Brachiaria and Leucaena (Subbarao et al., 2017). The environmental role spp. to reduce GHG emissions through increased of symbiotic nitrogen fixation by tropical legumes per-animal productivity and assumed reductions is recognized, as is the potential of tannin-rich in livestock numbers. Lal (2010) has stated that legumes for reduction of methane emissions by 29% of the overall carbon mitigation potential livestock (Schultze-Kraft et al., 2018). Less infor- will be from pastureland. The biggest potentials mation is currently available on the water foot- for carbon sequestration (i.e. through restoring print of forage-based feed production, with fodder degraded grasslands) are in South America and production being a major driver of water use in Africa (Conant, 2002; Conant et al., 2011). livestock systems (Herrero et al., 2012). In a review of the potential of forages to The main potential environmental impact mitigate emissions, Peters et al. (2013) concluded of tropical and subtropical planted forages will be that forage-based systems have a lower ecological through intensification of livestock production and footprint than feedlots. Better management of accompanying reductions in emissions assum- crops and grasslands, and restoration of degraded ing livestock numbers decrease as production lands, can result in a mitigation potential as intensifies. The situation may change when the high as 3.5 billion t CO 2-eq/year or 75% of the role of tropical forages in providing ecosystem biophysical mitigation potential stated by Smith services is further recognized and economic et al. (2008). Thornton and Herrero (2010) cal- incentives for increased adoption of planted for- culated that a modest 30% adoption rate of im- ages are in place (such as Payment for Ecosystem proved deep-rooted Brachiaria pastures could Services schemes). Because pasture-based systems, yield a mitigation potential of 29.8 million t which are the largest single land-use category CO2-eq/year in the Cerrados of Brazil alone, an globally (Erb et al., 2007), are changing from ex- amount equivalent to 2% of the total mitigation tensive grasslands to more intensive mixed crop– potential of agriculture. According to Fisher livestock farming, there is the opportunity that this et al. (2007) and Blanfort et al. (2012), the miti- transition could reduce GHG emissions without gation potential of planted forages to accumu- negative effects on food security (Havlík et al., late carbon under adequate pasture and animal The impact of CGIAR research on use of planted forages by tropical smallholders 467 management is second only to forests. To refine issue. In such systems, land is scarce and use of these estimates, further research in a variety of land for production of staple and cash crops contexts is required. takes priority over the production of livestock feeds for reasons of economics. The barrier of system properties is also ap- parent in the arguments of Ruthenberg (1980) What has limited the impact of planted in his consideration of the scarcity of medium- forages in tropical Africa and Asia? to long-term grass cultivation in the tropics (‘ley farming’). Ruthenberg pointed out various char- Despite the technical potential of planted forages, acteristics of tropical farming systems that make a common research finding is that they are rarely ley farming unattractive. Among the reasons adopted beyond project-led initiatives in Africa offered by Ruthenberg were the poor nutritive and Asia. Evidence for spontaneous widespread quality of tropical grasses, the advanced animal adoption is rare, except, for example, Australia husbandry needed to make such systems work and Brazil where the context is very different and the relative (un)profitability of use of land and smallholders are not the main beneficiary. for livestock production compared with arable The reasons for lack of adoption are complex. options such as maize, sorghum and cassava. Promoters of planted forages are convinced of A more recent line of argument in the de- their technical merits and advocate greater bate about adoption success of planted forages ‘promotion and dissemination’ to convince frames the problem as one of ‘innovation cap- smallholders of their advantages. In the termin- acity failure’ (Hall et al., 2007). The applica- ology of Sumberg (2002) forage legumes have tion of innovation systems theory to the ques- taken on a ‘mantle of absolute goodness’. The tion of feed development follows a trend across realists view lack of adoption as being related to the CGIAR system to think about technical inherent system properties, which make them change in the context of innovation systems – unattractive for farmers given the prevailing the network of agents and their interactions economic environment. To quote Sumberg that are necessary to foster change. In ILRI (2002): ‘Particular attention is placed on the and CIAT, projects around the time, including idea that the biophysical and socio-economic the Fodder Adoption Project and Fodder Innov- factors that have previously been constraints to ation Project, attempted to move beyond tech- legume adoption should now be viewed as sys- nical feed research to investigate institutional tem properties and incorporated into the design barriers to technical change in the feed sector specification of technology.’ (Ayele et al., 2012; Reddy et al., 2013). In the This idea that system properties are what logic of these projects, there was a recognition constrain the uptake of planted legumes and that Sumberg’s (2002) ‘system properties’ planted forages in general is compelling and was were indeed impediments to progress but that also proposed by McIntire and Debrah (1986) as focusing on the institutional issues could over- long ago as the mid-1980s. McIntire and Debrah come some of these barriers including difficul- (1986) laid out a series of propositions regarding ties of seed supply, access to markets and the the suitability of forage legumes for smallholder need to deal with policy constraints to bring systems. Among their propositions were that at about improvements in livestock feeding. different population densities competition for While these projects did not fully achieve their land and labour will tend to disadvantage forage ambitions, they did bring in new thinking legumes as a viable part of the farm enterprise. about innovation in the livestock feed sector At low population densities (e.g. pastoral sys- and helped to broaden research perspectives tems), competition for labour is an issue. In such beyond a narrow focus on forage management. systems, crops and animals are managed as sep- In South-east Asia in particular, application of arate enterprises, as mobility of livestock is a ne- such thinking did lead to adoption of forages cessary property of the system. Devoting labour at a reasonable scale (Stür et al., 2013). Such to tending of forage crops becomes impractical. innovation systems thinking is now routinely At high population densities (e.g. mixed inten- embedded in CGIAR-led projects and pro- sive systems), competition for land becomes the grammes focused on feed and forage 468 A.J. Duncan et al. improvement in the CGIAR Research Pro- rural communities, and long-term involvement gramme on Livestock and Fish (Puskur et al., of champions (Shelton et al., 2005). 2011). The study by White et al. (2013) sought to quantify the impact of planted tropical forages in general (and not just legumes, as in the study Global impacts from planted tropical by Shelton et al., 2005). They defined the impacts forages in three main domains: economic, social and environmental. Positive impacts can include Early research on tropical planted forages focused improved soil cover and hence reduced erosion, on collection and characterization of forage spe- nitrogen fixation by legumes leading to improved cies with economic potential to improve livestock soil health, and deeper rooting leading to im- productivity. This involved collections in sub- proved water-use efficiency and soil carbon stor- Saharan Africa organized mainly from Australia. age. Negative impacts can include the introduc- Promising accessions are now stored in the vari- tion of invasive characteristics and loss of local ous gene banks of CGIAR, notably those at ILRI, biodiversity. CIAT and ICARDA. Despite extensive forage The authors set out a series of nine meth- research activity in the CGIAR and other centres, odological shortcomings of the studies reviewed forage adoption among smallholders has gener- as a caution to the conclusions about economic ally been disappointing. Pengelly et al. (2003) impact. Despite shortcomings in the methods concluded that ‘despite 50 years of investment and data in the evidence reviewed, the analysis in forage research in the tropics, forage adoption yielded some interesting insights. White et al. has been relatively poor across all tropical farm- (2013) identified a total of 118 million ha plant- ing systems’. Sumberg (2002) argued that for- ed to ‘improved’ tropical forages from the studies age legumes had not achieved their potential in they compiled. They estimated that Brazil ac- sub-Saharan Africa, despite 70 years of research counted for 86% of the known planted area of to promote them. Shelton et al. (2005) in their improved forages. The area under Brachiaria spp. review of forage legume successes found that dominated. The farm-size characteristics of this none of the 14 legume cultivars released in adoption were not fully defined and included Latin America and the Caribbean between 1980 large commercial operations. and 2000 was well adopted. Emphasizing the methodological short- In reaction to these disappointing findings comings of the sample studies, fewer than 20% about the impact of tropical forages research, of the studies directly attempted to quantify there have been attempts to re-evaluate the im- economic impacts. Furthermore, the lack of a pact. The most recent example is the meta-analysis common methodology to quantify economic of White et al. (2013), which built upon the impacts (mixture of net present value ap- study by Shelton et al. (2005). The Shelton study proaches and annual estimates) meant that it presented 19 case studies and estimated areas was difficult to come up with an estimate of planted to various tropical legumes around the total economic benefit. Environmental and so- world, along with numbers of farmers and gross cial impacts were even less frequently quanti- economic benefits. Notable successes were cow- fied, with 7% and 2% of the studies quoting pea in West Africa accounting for 1.4 million ha environment and social impacts, respectively. (ex ante impact estimate), fodder trees in East The authors recognized a trend for greater em- Africa totalling 4 million m of hedgerows, phasis on longer-term impacts in later studies, Stylosanthes spp. in southern China (more than suggesting that the research for development 200,000 ha), Thailand (more than 300,000 ha) community is increasingly aware of the need and India (less than 250,000 ha), Pueraria to quantify large-scale impacts. Of all the stud- p haseoloides as grazed pastures in Brazil (480,000 ies analysed, fewer than 15% were conducted ha) and A. pintoi also in Brazil (more than independently of the personnel affiliated with 65,000 ha). The authors identified factors fa- the programme or project. This may have led vouring adoption as meeting the needs of farm- to exaggeration bias, as those closely involved ers, building partnerships, understanding the would be under pressure to justify their re- resources and skills of farmers, engagement with search through impact. As previously noted, The impact of CGIAR research on use of planted forages by tropical smallholders 469 this lack of independent assessment is also an several climate models in relation to a high GHG issue with the various success cases outlined emissions scenario for periods centred on 2050 earlier in this chapter. and 2090 (Jones and Thornton, 2015). Aggregate differences in both projected temperature and rainfall for the climate models selected are shown New adoption potential in Table 13.5. To identify areas with ‘moderate’ popula- Given the advances in the quality and coverage tion densities, following Kruska et al. (2003) we of databases, scientists can target forage research used a lower limit of 20 people/km2, above which to sites of highest probable adoption. Several fac- increasing proportions of land are cultivated tors affect adoption potential. One is population (Reid et al., 1995, 2000). Goddard et al. (1975) density. Forage cultivation can be expected to and McIntire et al. (1992) showed that fallowing occur at intermediate population densities where disappears (and agricultural fields coalesce) at there is sufficient land but where labour does not densities above 85 people/km2 across the semi- become a limiting constraint, or in areas of high arid to humid zones in Africa. We thus set the population density that are highly productive. moderate population density to be between 20 A second factor is profitability in the situations and 85 people/km2, and we used the dataset where high demand and prices can provide the GPW v4 (CIESIN, 2016). incentive to invest in forages (e.g. instead of or as As a proxy for market pull, we used the ac- well as vegetable production). A third factor is cessibility dataset of Nelson (2008), which gives rainfall distribution: given that forage is a year- the travel time in minutes to the nearest city round enterprise especially for dairy production, with a population in excess of 50,000, based on forages are more likely to succeed where rainfall is the estimated travel time to cross each pixel in well distributed. If this is not the case, forages can relation to land cover, slope, elevation, the roads work but the capacity to conserve, especially network, and any railways, rivers and water through hay-making, may be crucial for adoption. bodies. As a threshold, following Jones and We outline an example for Napier grass Thornton (2009), we selected a value of 200 (P. purpureum) and its potential as a cut-and-carry min, allowing the possibility for a smallholder to forage. We utilized the model EcoCrop as imple- take produce to market and return home on the mented in DIVA-GIS (Hijmans et al., 2012). To same day. generate suitability surfaces for Napier grass, The results are shown in Table 13.6 and Fig. EcoCrop estimates a suitability index from 0 to 13.1. These ‘high- potential’ sites are defined as the 100 for a crop, based on monthly precipitation areas having at least moderate suitability for Na- and temperature surfaces and on a small number pier grass and having a moderate human popula- of crop-specific parameters. We evaluated Napier tion density (20–85 people/km2) and are within grass suitability for current climatic conditions 200 min travel time of large population centres. In using the WorldClim dataset (distributed as part all regions, suitability increases to the 2050s but of DIVA-GIS). For possible future climatic condi- then declines to the 2090s, presumably in response tions, we downscaled spatially coarse output from to ever-higher temperatures and increased plant Table 13.5. General Circulation Model (GCM) responses for the land areas between latitudes 30°N and 30°S. Values shown are for ‘future minus current.’ (From Jones and Thornton, 2015.) 2050s 2090s Annual Average annual Annual Average annual GCM rainfall (mm) temperature (°C) rainfall (mm) temperature (°C) GISS-E2-R (NASA Goddard –53 +1.96 –56 +3.63 Institute for Space Studies) Ensemble mean (17 GCMs) +12 +2.36 +32 +4.68 HadGEM2-ES (UK Met –2 +2.93 –10 +5.83 Office, Hadley Centre) 470 A.J. Duncan et al. Table 13.6. High-potential sites for Napier grass as a cut-and-carry forage by region, period and climate model (km2). (Unpublished data from K. Kekae, G. Brychkova, P.C. Mckeown, J. Hanson, C.S. Jones, P.K. Thornton and C. Spillane, 2018.) Current conditions Ensemble mean Ensemble mean (2000) (17 GCMs) (2050s) (17 GCMs) (2090s) Moderate Excellent Moderate Excellent Moderate Excellent Region suitability suitability suitability suitability suitability suitability Global 600,847 1,355,312 516,552 2,169,758 652,472 1,668,437 Latin America 174,039 363,101 97,149 560,998 162,458 394,257 Sub-Saharan 201,548 712,057 109,829 920,105 186,875 725,251 Africa Asia 149,343 257,897 89,741 436,231 125,500 304,854 Current conditions (2000) GIS2 GCM, 2050s GIS2 GCM, 2090s Moderate Excellent Moderate Excellent Moderate Excellent suitability suitability suitability suitability suitability suitability Global 600,847 1,355,312 546,007 1,981,659 524,104 2,044,916 Latin America 174,039 363,101 115,716 522,540 88,633 504,335 Sub-Saharan 201,548 712,057 117,360 766,692 120,844 689,778 Africa Asia 149,343 257,897 94,726 467,864 82,912 483,741 Current conditions (2000) HADG GCM (2050s) HADG GCM (2090s) Moderate Excellent Moderate Excellent Moderate Excellent suitability suitability suitability suitability suitability suitability Global 600,847 1,355,312 557,462 2,212,979 571,497 2,365,266 Latin America 174,039 363,101 114,498 557,242 131,724 491,588 Sub-Saharan 201,548 712,057 116,960 876,141 109,760 848,881 Africa Asia 149,343 257,897 87,816 446,145 94,043 441,943 Top, mean of 17 General Circulation Models (GCMs); middle, a ‘cool’ GCM (GISS-E2-R, see Table 13.5); bottom, a ‘warm’ GCM (HadGEM2-ES, see Table 13.5). evapotranspiration. Nevertheless, the proportion particularly in peri-urban areas. These effects of area that is of high suitability increases over are difficult to estimate, however. For example, time compared with the area of moderate suitabil- increasing population densities to the middle ity, and the proportion of highly suitable hotspot of the current century are likely to decrease the land outside the tropics is increasing over time. The range of tsetse-transmitted trypanosomiasis trends for the cooler General Circulation Model in parts of Africa through habitat destruction (GCM; Table 13.6, middle panel) are similar, al- (McDermott et al., 2002). However, evaluating though Asia sees an increasing area of highly suit- possible future changes in land use and land able high-potential sites up to the 2090s. The hot- competition due to population growth and ter GCM results (Table 13.6, lower panel) indicate urban migration is more challenging, mostly be- that the tropics are in general becoming rather cause of the difficulty in projecting future devel- warm for Napier grass, and there are large areas in opment of road and other transport infrastruc- the northern temperate zones that are projected to ture, which is often the forerunner of highly be highly suitable high-potential sites. localized increases in population density Changing human population dynamics in This simple illustrative analysis suggests that the coming decades can be expected to affect the potential high-potential sites for Napier grass suitability by increasing pressure on land, adoption as a cut-and-carry forage are relatively The impact of CGIAR research on use of planted forages by tropical smallholders 471 Current Moderate/good suitability Very good/excellent suitability 2050s (all GCMs) Moderate/good suitability Very good/excellent suitability 2090s (all GCMs) Moderate/good suitability Very good/excellent suitability Fig. 13.1. High-potential Napier adoption sites in sub-Saharan Africa. 472 A.J. Duncan et al. limited in the global tropics. A more robust ana- east Asia about the systemic conditions needed lysis could be done that takes rainfall distribution for successful planted forage development into account, as well as by using a more sophisti- among smallholders. These South American cated niche and suitability model than EcoCrop, examples point to possible future growth in such as MaxEnt (Warren and Seifert, 2011). An- use of planted forages in Africa once livestock other missing element is information on possible production becomes more commercial, farm changes in forage quality under a changing cli- sizes increase, private-sector actors have mate. Nevertheless, there is considerable poten- stronger engagement and the institutional tial to exploit new datasets for more appropriate environment is more conducive to growth targeting. Local evaluation, coupled with informa- of the forage sector. tion on where particular forages are actually • To exploit the full potential of environmental being utilized, has considerable potential to guide benefits, it is evident that work on forages (and future forage targeting and adoption. other feeds) needs to be integrated with ani- mal health and animal genetics. • We need to better define system constraints The Future to forage use and to target solutions for sys- tem constraints along with a continuing Planted forages have the potential to fuel growth focus on farm-level experimentation and on in the smallholder crop and livestock sector forage breeding. Understanding system in Africa and Asia by providing high-quality constraints will facilitate better forage tar- feed and by filling seasonal gaps. In addition to geting using spatial forecasting of global p roductivity benefits, planted forages can reduce high-potential sites. the environmental footprint of livestock. • Create markets for public and private agents Adoption of planted forages has been below by: (i) building germplasm supply arrange- its potential in nearly all of sub-Saharan Africa ments (public or private, depending on lo- and South Asia. Adoption in Latin America has cation); and (ii) identifying value chain been better, notably in Brazil and in parts of Cen- constraints and resolving them through tral America. Such success is rare in Asia and novel business arrangements. sub-Saharan Africa (with the possible exception of Napier grass in East Africa), but as mixed crop–livestock systems continue to expand, so Acknowledgements will planted forage use. To conclude, we offer some lessons for the Thanks to Werner Stür Chris Jones and Bruce future: Pengelly for commenting on a draft of this chapter, • Apply lessons from Latin America and the and to Steve Franzel and Jim Sumberg for critical Caribbean to sub-Saharan Africa and South- reviews. Notes 1 The nomenclature of several tropical forage species has changed recently. Because this review considers past research, we use the earlier accepted names throughout. It should be noted, however, that according to the taxonomy proposed by the USDA Germplasm Resources Information Network (GRIN), Brachiaria brizantha, B. decumbens and B. humidicola are now accepted as Urochloa brizantha, U. decum- bens and U. humidicola, respectively; Pennisetum glaucum, P. pedicellatum and P. purpureum as Cenchrus americanus, C. pedicellatus and C. purpureus, respectively; and Panicum maximum as Megathyrsus max- imus and Pueraria phaseoloides as Neustanthus phaseoloides. 2 Older accounts from East Africa show Napier grass in the early 20th century (Boonman, 1993). 3 A kebele is the smallest administrative unit in Ethiopia. A ‘woreda’ is the second smallest admininstrative unit in Ethiopia. 4 Fodder tree research has been pursued in East Africa for decades, as shown in the archives of the East African Agricultural and Forest Journal: (www.tandfonline.com/toc/teaf20/current). The impact of CGIAR research on use of planted forages by tropical smallholders 473 References Abate, A., Kayongo-Male, D. and Wanyoike, M. (1985) Fodder for high-potential areas in Kenya. In: Animal Feed Resources for Small-scale Livestock Producers. Proceedings of the second PANESA workshop, 11–15 November, Nairobi. Available at: https://cgspace.cgiar.org/bitstream/handle/10568/49993/x5548e. pdf?sequence=1 (accessed 27 March 2020). Alemayehu, M. (1998) Natural pasture improvement study around smallholder dairy areas. Small Dairy Development Project (SDDP), Addis Ababa. Ayele, S., Duncan, A.J., Larbi, A. and Khanh, T.T. (2012) Enhancing innovation in livestock value chains through networks: Lessons from fodder innovation case studies in developing countries. Science and Public Policy 2012, 333–346. Blanfort, V., Ponchant, L., Dezecache, C., Stahl, C., Freycon, V., et al. (2012) Dynamique du carbone dans les sols de prairies issues de la déforestation de la forêt amazonienne: étude d’une chronoséquence en Guyane française. [Study of carbon dynamics in grassland soils derived from the deforestation of the Amazonian forest: chronosequence study in French Guiana.] In: 19ème Rencontres Autour Des Recherches Sur Les Ruminants, 5–6 December, Paris. Blummel, M., Haileslassie, A., Samireddypalle, A., Vadez, V. and Notenbaert, A. (2014) Livestock water productivity: feed resourcing, feeding and coupled feed-water resource data bases. Animal Production Science 54, 1584–1593. Boonman, J.G. (1993) Elephant grass as fodder crop. In: East Africa’s Grasses and Fodders: Their Ecology and Husbandry. Springer Netherlands, Dordrecht, Netherlands, pp. 217–234. Brockington, J.D., Harris, I.M. and Brook, R.M. (2016) Beyond the project cycle: a medium-term evaluation of agroforestry adoption and diffusion in a south Indian village. Agroforestry Systems 90, 489–508. Carranca, C., Torres, M.O. and Madeira, M. (2015) Underestimated role of legume roots for soil N fertility. Agronomy for Sustainable Development 35, 1095–1102. CGIAR (2018) CGIAR Research Program on Livestock Agri-Food Systems 2018 Annual Report. CGIAR, Nairobi. CIESIN (2016) Gridded population of the world version 4 (GPWv4): population density adjusted to match 2015 revision UN WPP country totals. Center for International Earth Science Information Network (CIESIN), Palisades, New York. Conant, R.T. (2002) Potential soil carbon sequestration in overgrazed grassland ecosystems. Global Biogeochemical Cycles 16, 1–9. Conant, R.T., Ogle, S.M., Paul, E.A. and Paustian, K. (2011) Measuring and monitoring soil organic carbon stocks in agricultural lands for climate mitigation. Frontiers in Ecology and the Environment 9, 169–173. Cook, B.G., Pengelly, B.C., Brown, S.D., Donnelly, J.L., Eagles, D.A., et al. (2005) Tropical forages: an inter- active selection tool. Web Tool. CSIRO, Brisbane, Australia, DPI&F, Queensland, Australia, CIAT, Cali, Colombia, and ILRI, Addis Ababa. Criscioni, P., Marti, J.V., Pérez-Baena, I., Palomares, J.L., Larsen, T. and Fernández, C. (2016) Replace- ment of alfalfa hay (Medicago sativa) with maralfalfa hay (Pennisetum sp.) in diets of lactating dairy goats. Animal Feed Science and Technology 219, 1–12. Cunha, M.V. da, Lira, A.M., Santos, M.V.F. dos, Freitas, E.V. de, Dubeux Junior, J.C.B. et al. (2011) Associ- ation between the morphological and productive characteristics in the selection of elephant grass clones. Revista Brasileira de Zootecnia 40, 482–488. de Groote, H., Vanlauwe, B., Rutto, E., Odhiambo, G.D., Kanampiu, F. and Khan, Z.R. (2010) Economic analysis of different options in integrated pest and soil fertility management in maize systems of West- ern Kenya. Agricultural Economics 41, 471–482. Duncan, A., York, L., Lukuyu, B., Samaddar, A. and Stür, W. (2012) Feed Assessment Tool (FEAST): a system- atic method for assessing local feed resource availability and use with a view to designing intervention strategies aimed at optimizing feed utilization. Questionnaire for Facilitators, Version 5.3; updated: 15 June 2012. ILRI, Addis Ababa. Elbasha, E., Thornton, P.K. and Tarawali, G. (1999) An ex post economic impact assessment of planted forages in West Africa. ILRI Impact Assessment Series No. 2. ILRI, Nairobi. Erb, K.H., Gaube, V., Krausmann, F., Plutzar, C., Bondeau, A. and Haberl, H. (2007) A comprehensive glo- bal 5 min resolution land-use data set for the year 2000 consistent with national census data. Journal of Land Use Science 2, 191–224. Feedipedia (2013) Animal feed resources information system. INRA, Paris, CIRAD, Montpellier, France, AFZ, Paris, and FAO, Rome. Available at: www.feedipedia.org/node (accessed 18 February 2020). 474 A.J. Duncan et al. Fernandes, C.D., Grof, B., Chakraborty, S. and Verzignassi, J.R. (2005) Estilosantes Campo Grande in Bra- zil: a tropical forage legume success story. Tropical Grasslands 39, 223. Figueira, D.N., Neumann, M., Ueno, R.K., Muller, M.M.L. and Faria, M.V. (2015) Produção e composição química do capim elefante cv. pioneiro em diferentes alturas de resíduo – revisão de literatura. [Pro- duction and composition of chemical elephant grass cv. pioneer in different heights of waste – literature review.] Brazilian Journal of Applied Technology for Agricultural Science, Guarapuava-PR 8, 103–110. Figueira, D.N., Neumann, M., Ueno, R., Galbeiro, S. and Bueno, A.V.I. (2016) Forage yield and quality in elephant grass cv. Pioneiro harvested at different cutting heights and times. Semina: Ciências Agrári- as 37, 1017–1028. Fisher, M.J., Braz, S.P., Dos Santos, R.S.M., Urquiaga, S., Alves, B.J.R. and Boddey, A.R.M. (2007) Another dimension to grazing systems: soil carbon. Tropical Grasslands 41, 65–83. Francis, P.A. and Atta-Krah, A.N. (1989) Sociological and ecological factors in technology adoption: fodder trees in southeast Nigeria. Experimental Agriculture 25, 1–10. Franzel, S., Carsan, S., Lukuyu, B., Sinja, J. and Wambugu, C. (2014) Fodder trees for improving livestock productivity and smallholder livelihoods in Africa. Current Opinion in Environmental Sustainability 6, 98–103. Franzel, S. and Wambugu, C. (2007) The uptake of fodder shrubs among smallholders in East Africa: key elements that facilitate widespread adoption. In: Hare, M.D. and Wongpichet, K. (eds) Forages: A Pathway to Prosperity for Smallholder Farmers. Proceedings of an International Symposium, Faculty of Agriculture, Ubon Ratchathani University, Thailand, pp. 203–222. Fukagawa, S. and Ishii, Y. (2018) Grassland establishment of dwarf Napiergrass (Pennisetum purpureum Schumach) by planting of cuttings in the winter season. Agronomy 8, 12. Gerber, P.J., Steinfeld, H., Henderson, B., Mottet, A., Opio, C., et al. (2013) Tackling climate change through livestock: a global assessment of emissions and mitigation opportunities. FAO, Rome. Goddard, A., Mortimore, M. and Norman, D. (1975) Some social and economic implications of popu- lation growth in rural Hausaland. In: Caldwell, J.C. (ed.) Population Growth and Socioeconomic Change in West Africa. Columbia University Press for the Population Council, New York, pp. 321–336. Gomide, C.A.M., Chaves, C.S., Ribeiro, K.G., Sollenberger, L.E., et al. (2015) Structural traits of elephant grass (Pennisetum purpureum Schum.) genotypes under rotational stocking strategies. African Jour- nal of Range & Forage Science 32, 51–57. González, C., Schiek, B., Mwendia, S.W. and Prager, S. (2016) Improved forages and milk production in East Africa. A case study in the series: economic foresight for understanding the role of investments in agriculture for the global food system. CIAT Publication No. 422. CIAT, Cali, Colombia. Government of Kenya (2016) The Seeds and Plant Varieties Act – crop varieties. The Kenya Gazette 74, 2696–2701. Guodao, L. and Chakraborty, S. (2005) Stylo in China: a tropical forage legume success story. In: O’Mara, F.P., Wilkins, R.J., ’t Mannetje, L., Lovett, D.K., Rodgers, P.A.M. and Boland, T.M. (eds) Proceedings of the XX International Grassland Congress: Offered Papers, Dublin. Wageningen Academic Publishers, Wageningen, Netherlands, p. 322. Halim, R.A., Shampazuraini, S. and Idris, A.B. (2013) Yield and nutritive quality of nine Napier grass varieties in Malaysia. Malaysian Journal of Animal Science 16, 37–44. Hall, A., Sulaiman, R. and Bezkorowajnyj, P. (2007) Reframing technical change: Livestock Fodder Scarcity Revisited as Innovation Capacity Scarcity. ILRI, UNU-MERIT, ICRISAT and IITA on behalf of the System-wide Livestock Programme, CGIAR Systemwide Livestock Programme, Addis Ababa. Hanson J. and Peters, M. (2003) Meeting the need for herbage seeds in developing countries. In: Loch, D.S. (ed.) Proceedings of the 5th International Herbage Seed Conference. Herbage Seeds in the New Millennium – New Markets, New Products, New Opportunities, 23–26 November, Gatton, Australia. Queensland Department of Primary Industries, Cleveland, Australia, pp. 26–37. Havlík, P., Valin, H., Mosnier, A., Obersteiner, M., Baker, J.S., et al. (2013) Crop productivity and the global livestock sector: implications for land use change and greenhouse gas emissions. American Journal of Agricultural Economics 95, 442–448. Herrero, M., Grace, D., Njuki, J., Johnson, N., Enahoro, D., et al. (2012) The roles of livestock in developing countries. 8th International Symposium on the Nutrition of Herbivores 7 (s1), 3–18. Herrero, M., Havlik, P., Valin, H., Notenbaert, A., Rufino, M.C., et al. (2013) Biomass use, production, feed efficiencies, and greenhouse gas emissions from global livestock systems. Proceedings of the National Academy of Sciences USA 110, 20888–20893. The impact of CGIAR research on use of planted forages by tropical smallholders 475 Hijmans, R.J., Guarino, L. and Mathur, P. (2012) DIVA-GIS, Version 7.5. Available at: www.diva-gis.org (accessed 27 February 2020). Holmann, F., Rivas, L., Argel, P.J. and Pérez, E. (2004) Impact of the adoption of Brachiaria grasses: Central America and Mexico. Livestock Research for Rural Development 16, 98. Huxley, P.A. (1984) The basis of selection, management and evaluation of multi-purpose trees: an overview. International Council for Research in Agroforestry, Nairobi. ICAR (2011) Forage crops and grasses. In: Handbook of Agriculture. Indian Council of Agricultural Research (ICAR), New Delhi, pp. 1353–1417. ICIPE (2015) The ‘push–pull’ farming system: climate-smart, sustainable agriculture for Africa. ICIPE, Nairobi. Ishii, Y., Yamaguchi, N. and Idota, S. (2005) Dry matter production and in vitro dry matter digestibility of tillers among napiergrass (Pennisetum purpureum Schumach) varieties. Grassland Science 51, 153–163. ISPC (2018) Tropical forages and the diffusion of Brachiaria cultivars in Latin America. Standing Panel on Impact Assessment (SPIA) Brief No. 70. CGIAR Independent Science and Partnership Council, Rome. Jank, L., Barrios, S.C., Do Valle, C.B., Simeao, R.M. and Alves, G.F. (2014) The value of improved pastures to Brazilian beef production. Crop and Pasture Science 65, 1132–1137. Jones, P.G. and Thornton, P.K. (2009) Croppers to livestock keepers: livelihood transitions to 2050 in Africa due to climate change. Environmental Science & Policy 12, 427–437. Jones, P.G. and Thornton, P.K. (2015) Representative soil profiles for the Harmonized World Soil Data- base at different spatial resolutions for agricultural modelling applications. Agricultural Systems 139, 93–99. Kadam, S.S., Kumar, A. and Arif, M. (2017) Hybrid Napier for round the year quality fodder supply to the dairy industry – a review. International Journal of Current Microbiology and Applied Sciences 6, 4778–4783. Kebede, G., Feyissa, F., Assefa, G., Alemayehu, M., Mengistu, A., et al. (2017) Agronomic performance, dry matter yield stability and herbage quality of Napier grass (Pennisetum purpureum (L.) Schumach) accessions in different agro-ecological zones of Ethiopia. Journal of Agricultural and Crop Research 5, 49–65. Khan, Z.R., Midega, C.A.O., Hutter, N.J., Wilkins, R.M. and Wadhams, L.J. (2006) Assessment of the potential of Napier grass (Pennisetum purpureum) varieties as trap plants for management of Chilo partellus. Entomologia Experimentalis et Applicata 119, 15–22. Khan, Z.R., Midega, C.A.O., Amudavi, D.M., Hassanali, A. and Pickett, J.A. (2008a) On-farm evaluation of the push pull technology for the control of stemborers and striga weed on maize in western Kenya. Field Crops Research 106, 224–233. Khan, Z.R., Midega, C.A.O., Njuguna, E.M., Amudavi, D.M., Wanyama, J.M. and Pickett, J.A. (2008b) Economic performance of the ‘push–pull’ technology for stemborer and Striga control in smallholder farming systems in western Kenya. Crop Protection 27, 1084–1097. Khan, Z., Midega, C., Pittchar, J., Pickett, J. and Bruce, T. (2011) Push–pull technology: a conservation agriculture approach for integrated management of insect pests, weeds and soil health in Africa. Inter- national Journal of Agricultural Sustainability 9, 162–170. Khan, Z.R., Midega, C.A.O., Pittchar, J.O., Murage, A.W., Birkett, M.A., et al. (2014) Achieving food security for one million sub-Saharan African poor through push–pull innovation by 2020. Philosophical Transac- tions of the Royal Society B: Biological Sciences 369, 20120284. Kiptot, E., Franzel, S., Hebinck, P. and Richards, P. (2006) Sharing seed and knowledge: farmer to farmer dissemination of agroforestry technologies in western Kenya. Agroforestry Systems 68, 167–179. Kiptot, E., Hebinck, P., Franzel, S. and Richards, P. (2007) Adopters, testers or pseudo-adopters? Dy- namics of the use of improved tree fallows by farmers in western Kenya. Agricultural Systems 94, 509–519. Kruska, R.L., Reid, R.S., Thornton, P.K., Henninger, N. and Kristjanson, P.M (2003) Mapping livestock oriented agricultural production systems for the developing world. Agricultural Systems 77, 39–63. Lal, R. (2010) Managing soils and ecosystems for mitigating anthropogenic carbon emissions and advan- cing global food security. Bioscience 60, 708–721. Lukuyu, B., Place, F., Franzel, S. and Kiptot, E. (2012) Disseminating improved practices: are volunteer farmer trainers effective? Journal of Agricultural Education and Extension 18, 525–540. Lynam, J. and Byerlee, D. (2017a) Forever Pioneers – CIAT: 50 Years Contributing to a Sustainable Food Future… and Counting. CIAT, Cali, Colombia. 476 A.J. Duncan et al. Lynam, J. and Byerlee, D. (2017b) The Tropical Forages Program. In: Forever Pioneers – CIAT: 50 Years Contributing to a Sustainable Food Future… and Counting. CIAT, Cali, Colombia, pp. 80–91. Maass, B.L., Midega, C.A.O., Mutimura, M., Rahetlah, V.B., Salgado, P., et  al. (2015) Homecoming of Brachiaria: improved hybrids prove useful for African animal agriculture. East African Agricultural and Forestry Journal 81, 71–78. Mapato, C. and Wanapat, M. (2018) Comparison of silage and hay of dwarf Napier grass (Pennisetum purpureum) fed to Thai native beef bulls. Tropical Animal Health and Production 50, 1473–1477. McDermott, J.J., Kristjanson, P.M., Kruska, R.L., Reid, R.S., Robinson, T.P., et al. (2002) Effects of climate, human population and socio-economic changes on tsetse-transmitted trypanosomiasis to 2050. In: The African Trypanosomes. World Class Parasites, Vol 1. Springer, Boston, MA, pp. 25–38. McIntire, J. and Debrah, S. (1986) Forage research in smallholder and pastoral production systems. In: Util- ization of Agricultural By-Products as Livestock Feeds in Africa. Proceedings of a workshop held at Ryall’s Hotel, Blantyre, Malawi, September 1986. ILCA, Addis Ababa, pp. 118–126. McIntire, J., Bourzat, D. and Pingali, P. (1992) Crop–livestock interaction in sub-Saharan Africa. World Bank Regional and Sectoral Studies. World Bank, Washington, D.C. Mekoya, A., Oosting, S.J., Fernandez-Rivera, S. and van der Zijpp, A.J. (2008) Farmers’ perceptions about exotic multipurpose fodder trees and constraints to their adoption. Agroforestry Systems 73, 141–153. Mekuria, W. and Veldkamp, E. (2012) Restoration of native vegetation following exclosure establishment on communal grazing lands in Tigray, Ethiopia. Applied Vegetation Science 15, 71–83. Miles, J.W. (2001) Achievements and perspectives in the breeding of tropical grasses and legumes. In: Grassland Ecosystems: An Outlook into the 21st Century. Proceedings of the XIX International Grassland Congress, 11–21 February, Sao Paulo, Brazil. Brazilian Society of Animal Husbandry, Brazil, pp. 509–515. Miles, J.W. (2007) Apomixis for cultivar development in tropical forage grasses. Crop Science, 47 (Suppl. 3), S238–S249. Millar, J. and Connell, J. (2010) Strategies for scaling out impacts from agricultural systems change: the case of forages and livestock production in Laos. Agriculture and Human Values 27, 213–225. Ministério da Agricultura, Pecuária e Abastecimento (2012) Plano setorial de mitigação e de adaptação às mudanças climáticas para a consolidação de uma economia de baixa emissão de carbono na agri- cultura: plano ABC (Agricultura de Baixa Emissão de Carbono). Ministério da Agricultura, Pecuária e Abastecimento (MAPA), Ministério do Desenvolvimento Agrário, Coordenação da Casa Civil da Presidência da República, Brasília. Ministerio de Ambiente y Desarrollo Sostenible (2015) NINO Ganadería Bovina Sostenible: Densificación productiva, reconversión de pasturas y devolución a la naturaleza. Ministerio de Ambiente y Desar- rollo Sostenible (MADS), Bogotá, Colombia. Mohamed-Saleem, M.A. and Suleiman, H. (1986) Nigeria and West Africa: fodder banks. Dry-season feed supplementation for traditionally managed cattle in the subhumid zone. World Animal Review 59, 11–17. Muhr, L., Tarawali, S.A., Peters, M. and Schultze-Kraft, R. (1999) Forage legumes for improved fallows in agropastoral systems of subhumid West Africa. III. Nutrient import and export by forage legumes and their rotational effects on subsequent maize. Tropical Grasslands 33, 245–256. Mwendia, S.W., Mwangi, D.M., Wahome, R.G. and Wanyoike, M. (2008) Assessment of growth rate and yields of three Napier grass varieties in Central Highlands of Kenya. East African Agricultural and For- estry Journal 74, 211–217. Negassa, A., Shapiro, B., Kidane, T., Abdena, A. and Hanson, J. (2016) Ex-ante assessment of demand for improved forage seed and planting materials among Ethiopian smallholder farmers: a contingent valu- ation analysis. ILRI Project Report. ILRI, Nairobi. Negawo, A., Teshome, A., Kumar, A., Hanson, J. and Jones, C. (2017) Opportunities for Napier grass (Pen- nisetum purpureum) improvement using molecular genetics. Agronomy 7, 28. Negawo, A.T., Jorge, A., Hanson, J., Teshome, A., Muktar, M.S., et al. (2018) Molecular markers as a tool for germplasm acquisition to enhance the genetic diversity of a Napier grass (Cenchrus purpureus syn. Pennisetum purpureum) collection. Tropical Grasslands-Forrajes Tropicales 6, 58–69. Nelson, A. (2008) Estimated travel time to the nearest city of 50,000 or more people in year 2000. Global Environment Monitoring Unit, Joint Research Centre of the European Commission, Ispra, Italy. Nyambati, E.M., Muyekho, F.N., Onginjo, E. and Lusweti, C.M. (2010) Production, characterization and nu- tritional quality of Napier grass [Pennisetum purpureum (Schum)] cultivars in western Kenya. African Journal of Plant Science 4, 496–502. The impact of CGIAR research on use of planted forages by tropical smallholders 477 Pengelly, B.C., Whitbread, A., Mazaiwana, P.R. and Mukombe, N. (2003) Tropical forage research for the future – better use of research resources to deliver adoption and benefits to farmers. Tropical Grass- lands 37, 207–216. Pereira, A.V., Lédo, F.J. da S. and Machado, J.C. (2017) BRS Kurumi and BRS Capiaçu – new elephant grass cultivars for grazing and cut-and-carry system. Crop Breeding and Applied Biotechnology 17, 59–62. Peters, M., Horne, P., Schmidt, A., Holmann, F., Kerridge, P.C., et al. (2001) The role of forages in redu- cing poverty and degradation of natural resources in tropical production systems. ODI Agricultural Re- search & Extension Network (AgREN), London. Peters, M., Herrero, M., Fisher, M., Erb, K.-H., Rao, I., et al. (2013) Challenges and opportunities for improv- ing eco-efficiency of tropical forage-based systems to mitigate greenhouse gas emissions. Tropical Grasslands-Forrajes Tropicales 1, 156–167. Place, F., Roothaert, R., Maina, L., Franzel, S., Sinja, J. and Wanjiku, J. (2009) The impact of fodder trees on milk production and income among smallholder dairy farmers in East Africa and the role of research. ICRAF Occasional Paper. World Agroforestry Centre, Nairobi. Puskur, R., Baker, D., Omore, A.O. and Staal, S.J. (2011) Changing approaches to pro-poor livestock market development: Innovation and upgrading in the value chain. Livestock Exchange Issue Brief 17. ILRI, Nairobi. Rao, I., Peters, M., Castro, A., Schultze-Kraft, R., White, D., et al. (2015) LivestockPlus - The sustainable intensification of forage-based agricultural systems to improve livelihoods and ecosystem services in the tropics. Tropical Grasslands-Forrajes Tropicales 3, 59–82. Reddy, T.S.V., Puskur, R., Hall, A. and Sulaiman, R. (2013) Applying innovation system principles to fodder scarcity: experiences from the Fodder Innovation Project. CRISP Working Paper 2013-001. Centre for Research on Innovation and Science Policy (CRISP), Hyderabad, India. Reid, R., Kruska, R., Perry, B., Ellis, J. and Wilson, C. (1995) Environmental impacts of trypanosomiasis control: conceptual model, approach and preliminary results of studying indirect effects through changes in land-use. In: Proceedings of the 22nd Meeting of the International Scientific Council for Trypanosomiasis Research and Control, Nairobi, Kenya. OAU/ISCTRC, Nairobi, pp. 235–243. Reid, R.S., Kruska, R.L., Deichmann, U., Thornton, P.K. and Leak, S.G.A. (2000) Human population growth and the extinction of the tsetse fly. Agriculture, Ecosystems and Environment 77, 227–236. Renard, C. (ed.) (1997) Crop Residues in Sustainable Mixed Crop/Livestock Farming Systems. CAB Inter- national, Wallingford, UK. Rincón, A., Bueno, G.A., Álvarez, M., Pardo, Ó., Pérez, O. and Caicedo, S. (2010) Establecimiento, manejo y utilización de recursos forrajeros en sistemas ganaderos de suelos ácidos. Corporación Colombiana de Investigación Agropecuaria (Corpoica), Villavicencio, Colombia. Rincón, Á. and Ligarreto, G. (2008) Productividad de la asociación de maíz-pastos en suelos ácidos del Piedemonte Llanero colombiano. Corporación Colombiana de Investigación Agropecuaria (Corpoica), Villavicencio, Colombia. Rivas, L. (2002) Impacto económico de la adopción de pastos mejorados en América Latina tropical. CIAT, Cali, Colombia. Rufino, M.C., Tittonell, P., van Wijk, M.T., Castellanos-Navarrete, A., Delve, R.J., et al. (2007) Manure as a key resource within smallholder farming systems: analysing farm-scale nutrient cycling efficiencies with the NUANCES framework. Livestock Science 112, 273–287. Ruthenberg, H. (1980) Farming Systems in the Tropics. Clarendon Press, Oxford. Schultze-Kraft, R., Rao, I.M., Peters, M., Clements, R.J., Bai, C. and Liu, G. (2018) Tropical forage legumes for environmental benefits: an overview. Tropical Grasslands-Forrajes Tropicales 16, 1–14. Searchinger, T., Waite, R., Hanson, C., Ranganathan, J., Dumas, P. and Matthews, E. (2013) Creating a sustainable food future: a menu of solutions to sustainably feed more than 9 billion people by 2050. World Resources Institute, Washington, D.C. Shelton, H.M., Franzel, S. and Peters, M. (2005) Adoption of tropical legume technology around the world: analysis of success. In: McGilloway, D.A. (ed.) Grassland: A Global Resource. Wageningen Academic Publishers, Wageningen, Netherlands, pp. 149–166. Shilin, W., Jones, R.M., Minggang, X. and Pingna, H. (2007) Quality and seasonal yields of promising for- age species in the red soils region of southern China. Australian Journal of Experimental Agriculture 47, 942–948. Sikumba, G.N., Mangesho, W., Lukuyu, B.A., Ngulu, F. and Bekunda, M. (2015) Participatory evaluation of farmer preferences and productivity of selected Napier grass (Pennisetum purpureum) accessions in 478 A.J. Duncan et al. northern Tanzania. Poster prepared for the International Conference on Integrated Systems Research, 3–6 March 2015, Ibadan, Nigeria. ILRI, Nairobi. Sluszz, T. (2012) Monitoramento tecnologico de cultivares de forrageiras tropicais. Cadernos de Prospecção 5, 1–13. Smith, P., Martino, D., Cai, Z., Gwary, D., Janzen, H., et al. (2008) Greenhouse gas mitigation in agriculture. Philosophical Transactions of the Royal Society B: Biological Sciences 363, 789–813. Sollenberger, L.E., Prine, G.M., Ocumpaugh, W.R., Schank, S.C., Kalmbacher, R.S. and Jones, C.S. (1987) Dwarf elephant grass: a high quality forage with potential in Florida and the tropics. Proceedings of the Soil and Crop Science Society of Florida 46, 42–46. Squires, V.R., Mann, T.L. and Andrew, M.H. (1992) Problems in implementing improved range management on common lands in Africa: an Australian perspective. Journal of the Grassland Society of Southern Africa 9, 1–7. Staal, S.J., Baltenweck, I., Waithaka, M.M., de Wolff, T. and Njoroge, L. (2002) Location and uptake: integrated household and GIS analysis of technology adoption and land use, with application to smallholder dairy farms in Kenya. Agricultural Economics 27, 295–315. Staal, S.J., Chege, L., Kenyanjui, M., Kimari, A., Lukuyu, B., et al. (1997) Characterisation of dairy systems supplying the Nairobi milk market: a pilot survey in Kiambu District for the identification of target groups of producers. ILRI, Nairobi. Stür, W.W., Horne, P.M., Phengsavanh, P., Gabunada, F., Khanh, T.T. and Connell, J. (2007) Planted forages – the key for making money from smallholder livestock production: experiences from CIAT’s forage R&D in Southeast Asia. In: Forages – A Pathway to Prosperity for Smallholder Farmers. Proceedings of a Forage Symposium, Ubon Ratchathani University, Thailand, pp. 313–331. Stür, W.W., Khanh, T.T. and Duncan, A.J. (2013) Transformation of smallholder beef cattle production in Vietnam – technology, innovation and markets. International Journal of Agricultural Sustainability 11, 363–381. Subbarao, G.V., Wang, H.Y., Ito, O., Nakahara, K. and Berry, W.L. (2007) NH +4 triggers the synthesis and release of biological nitrification inhibition compounds in Brachiaria humidicola roots. Plant and Soil 290, 245–257. Subbarao, G.V., Nakahara, K., Hurtado, M.P., Ono, H., Moreta, D.E., et al. (2009) Evidence for biological nitrification inhibition in Brachiaria pastures. Proceedings of the National Academy of Sciences USA 106, 17302–17307. Subbarao, G.V., Arango, J., Masahiro, K., Hooper, A.M., Yoshihashi, T., et al. (2017) Genetic mitigation strategies to tackle agricultural GHG emissions: the case for biological nitrification inhibition technol- ogy. Plant Science 262, 165–168. Sumberg, J. (2002) The logic of fodder legumes in Africa. Food Policy 27, 285–300. Tarawali, G. (1991) The residual effect of Stylosanthes fodder banks on maize yield at several locations in Nigeria. Tropical Grasslands 25, 26–31. Tarawali, G. and Pamo, E.T. (1992) A case for on-farm trials of fodder banks on the Adamawa Plateau in Cameroon. Experimental Agriculture 28, 229–235. Tarawali, G., Manyong, V.M., Carsky, R.J., Vissoh, P.V., Osei-Bonsu, P. and Galiba, M. (1999) Adoption of improved fallows in West Africa: lessons from mucuna and stylo case studies. Agroforestry Systems 47, 93–122. Thomas, D. and Sumberg, J.E. (1995) A review of the evaluation and use of tropical forage legumes in sub-Saharan Africa. Agriculture, Ecosystems and Environment 54, 151–163. Thornton, P.K. and Herrero, M. (2010) Potential for reduced methane and carbon dioxide emissions from livestock and pasture management in the tropics. Proceedings of the National Academy of Sciences USA 107, 19667–19672. Valbuena, D., Tui, S.H.K., Erenstein, O., Teufel, N., Duncan, A., et al. (2015) Identifying determinants, pres- sures and trade-offs of crop residue use in mixed smallholder farms in Sub-Saharan Africa and South Asia. Agricultural Systems 134, 107–118. Valentim, J.F. and de Andrade, C.M.S. (2005) Forage peanut (Arachis pintoi): a high yielding and high quality tropical legume for sustainable cattle production systems in the western Brazilian Amazon. In: O’Mara, F.P., Wilkins, R.J., ’t Mannetje, L., Lovett, D.K., Rogers P.A.M. and Boland T.M. (eds) Proceedings of the XX International Grassland Congress: Offered Papers. Wageningen Academic Publishers, Wage- ningen, Netherlands, p. 329. The impact of CGIAR research on use of planted forages by tropical smallholders 479 Wamalwa, N.I.E., Midega, C.A.O., Ajanga, S., Omukunda, N.E., Muyekho, F.N., et al. (2017) Screening Napier grass accessions for resistance to Napier grass stunt disease using the loop-mediated iso- thermal amplification of DNA (LAMP). Crop Protection 98, 61–69. Wambugu, C., Place, F. and Franzel, S. (2011) Research, development and scaling-up the adoption of fodder shrub innovations in East Africa. International Journal of Agricultural Sustainability 9, 100–109. Wangchuk, K., Rai, K., Nirola, H., Thukten, T., Dendup, C. and Mongar, D. (2015) Forage growth, yield and quality responses of Napier hybrid grass cultivars to three cutting intervals in the Himalayan foothills. Tropical Grasslands-Forrajes Tropicales 3, 142–150. Warren, D.L. and Seifert, S.N. (2011) Ecological niche modeling in Maxent: the importance of model complexity and the performance of model selection criteria. Ecological Applications 21, 335–342. White, D.S., Peters, M. and Horne, P. (2013) Global impacts from improved tropical forages: a meta-analysis revealing overlooked benefits and costs, evolving values and new priorities. Tropical Grasslands- Forrajes Tropicales 1, 12–24. Williams, W.M., Easton, H.S. and Jones, C.S. (2007) Future options and targets for pasture plant breeding in New Zealand. New Zealand Journal of Agricultural Research 50, 223–248. Zaidi, P.H., Vinayan, M.T. and Blummel, M. (2013) Genetic variability of tropical maize stover quality and the potential for genetic improvement of food-feed value in India. Field Crops Research 153, 94–101. 14 Multidimensional Crop Improvement by ILRI and Partners: Drivers, Approaches, Achievements and Impact the late Michael Blümmel1, Anandan Samireddypalle2, Perez Haider Zaidi3, Vincent Vadez4, Ramana Reddy5 and Pasupuleti Janila4 1International Livestock Research Institute, Addis Ababa, Ethiopia; 2Indian Council of Agricultural Research, Bangalore, India; 3Centro Internacional de Mejoramiento de Maíz y Trigo, Patancheru, India; 4International Crops Research Institute for the Semi-Arid Tropics, Patancheru, India; 5International Livestock Research Institute, Patancheru, India Contents Executive Summary 481 The problem 481 Scientific impacts 481 Development impacts 482 Policy impacts through the provision of information 482 Capacity building and partnerships on multidimensional crop improvement as outcomes 483 Introduction 483 Fodder markets 484 CR fodder quality and livestock productivity 487 Trait identification and tools for affordable phenotyping for crop residue fodder quality 488 Validation of laboratory fodder quality traits 488 Calibration and validation of NIRS tools 489 Exploitation of Existing Cultivar-dependent Variation in CR Fodder Quality 490 Phenotype pipeline and releases for variations in CR fodder quality 490 Targeted genetic enhancement towards food–feed crop cultivars 492 Recurrent selection 492 Hybrid breeding for dual-purpose maize 493 QTL identification and backcrossing 494 GWAS and GS 494 Trade-offs Among Crop Residue Fodder Traits and Primary Traits 495 Primary and secondary traits 495 CR N content and grain and CR yield 496 CR digestibility and grain and CR yield 496 © International Livestock Research Institute 2020. The Impact of the International 480 Livestock Research Institute (eds J. McIntire and D. Grace) Multidimensional Crop Improvement by ILRI and Partners 481 Outcomes and Aspects of Impacts of Multidimensional Crop Improvement 499 Economic impact of multidimensional crop improvement 500 The Future 500 References 501 Executive Summary However, little uptake of chemical treatments was observed, despite efforts by the international re- The problem search and development communities, and invest- ments into chemical straw treatments have Livestock provides food and income for almost declined since (Owen and Jayasuriya, 1989). 1.3 billion people across the world. Grazing has The lack of adoption of postharvest treat- long been a principal source of feed in much of ments of CRs gave way to a new model of improv- South Asia and in sub-Saharan Africa. Due to ing the fodder value of CRs by selection and plant population pressure, land degradation and con- breeding (Reed et  al., 1988a) and by identifying version from grazing to arable land, grazing areas anti-nutritive factors in crop biomass (Reed et al., have contracted, resulting in feed shortages. The 1988b, 1990). In the mid-1990s, the International conversion of grazing land is likely to be aggra- Livestock Research Institute (ILRI) and Inter- vated by climate change (Blümmel et al., 2015b). national Crops Research Institute for the Semi-Arid The increasing demand for animal-sourced food Tropics (ICRISAT) began a joint programme on im- is another factor in putting pressure on feed from provement of grain and CR traits, focusing on sor- all sources (Blümmel et al., 2017). ghum (Sorghum bicolor) and pearl millet (Pennisetum Feed supply and demand scenarios for glaucum (L.) R. Br.) in the semi-arid tropics of India. South Asia and sub-Saharan Africa have shown Ex ante estimates of potential productivity gains that crop residues (CRs) such as straws, stover from genetic improvement of the digestibility of and haulms commonly provide 50–70% of the multidimensional food and fodder crops would pro- feed resources in smallholder systems (Blümmel duce high rates of economic return in the form of et al., 2014b; Duncan et al., 2016). In the high- incremental meat, milk and draught power (e.g. lands of Ethiopia, cereal CRs have emerged as Kristjanson and Zerbini, 1999, for pearl millet and the main components of the livestock diet but sorghum in semi-arid India). Similar work started are generally poor in their nutritive value with a in West Africa in the 1990s among the Inter- low crude protein content (4%) and digestible national Institute of Tropical Agriculture (IITA), organic matter (less than 50%). ICRISAT and ILRI, targeting cowpea. Lignocellulosic biomass from forest, agricul- This chapter therefore addresses the fol- tural waste and CRs is the most abundant renew- lowing questions. What is the extent of cultivar- able biomass on earth with a total production dependent variation in CR fodder quality? Can estimated to range from about 10 billion to 50 these variations be exploited without detriment billion t (Sanchez and Cardena, 2008). About to grain yield? Have quality improvements in 3.8 billion t are contributed by CRs, with cereals CRs from plant selection and breeding been contributing 74%, sugar crops 10%, legumes achieved? Have such improvements made a field 8%, tubers 5% and oil crops 3% (Lal, 2005). Con- impact on crop and animal productivity? sidering the quantities of CRs available and the high nutritive quality of its basic constituents – hexose and pentose sugars – attempts to improve Scientific impacts CR biomass for fodder began a century ago (Fin- gerling and Schmidt, 1919; Beckmann, 1921). The principal scientific achievement was to force These and later attempts to improve CR bio- a reconsideration of the single-trait (i.e. grain) mass included chemical, physical and biological model in favour of the multi-trait and whole-plant treatments. Chemical treatments, particularly the (i.e. food and fodder) model. While there are as yet use of hydrolytic agents such as sodium hydroxide few public-sector decisions to include stover traits and ammonia (Jackson, 1977; Owen and Jayasur- as cultivar release criteria – sorghum and pearl iya, 1989), received significant r esearch attention. millet are recent examples – public and private 482 M. Blümmel et al. crop-improvement programmes have reoriented a ttributable to CR quality in India and West their efforts towards whole-plant improvement. Africa. This information is valuable to crop Crop-improvement paradigms are changing to e xtension programmes. whole-plant optimization, as, for example, re- Adoption studies have shown that materials flected in the new CGIAR Research Program with higher straw digestibility improve livestock (CRP) on Grain Legumes and Dryland Cereals. productivity, which is again valuable to exten- Under this principal achievement, scientific sion work. impacts are the findings that: (i) there is significant Plant breeding and selection have led to the variation in CR quality; (ii) such variation does availability of crop cultivars with higher-quality not compromise grain yield; (iii) near-infrared CRs in sorghum, pearl millet, groundnut, rice spectroscopy (NIRS) methods are accurate for and maize in India, and in cowpea in West Africa. rapid screening of quality traits; and (iv) recent Higher productivity and income come from molecular analyses can detect variations in fod- sales of CRs and from livestock production. der quality early in breeding material. S alient examples are as follows: NIRS equations were also developed for grains of key crops, including routine quality • An ILRI-CIMMYT collaboration identified a traits such as protein, starch and fat but also multidimensional maize hybrid (NK 6240), amino and fatty acids. which is now a very popular hybrid in India The fodder quality of CRs can be increased by (Anandan et al., 2013). ILRI, CIMMYT and targeted genetic enhancement using conventional Syngenta are now exploring branding for or molecular crop-improvement approaches such CR fodder quality traits. as marker-assisted breeding, use of quantitative • Adoption of improved multidimensional cul- trait loci (QTLs) or genome-wide association stud- tivars based on seed production has been dif- ies (GWAS). Nepolean et  al. (2009) used QTL to ficult and at times contradictory to estimate. map the genomic regions controlling stover qual- Randomized adoption studies by household ity and yield traits in pearl millet, while Blümmel surveys show generally less adoption than et al. (2015a), used stay-green QTLs in sorghum. estimates based on seed production. GWAS was used to unravel favourable native gen- • Adoption of hybrids is much faster because etic variations for traits of agronomic and eco- seed availability is less of a problem than nomic importance across many cereal crops with open-pollinated varieties. Thus, a new (Vinayan et al., 2013). dual-purpose maize hybrid (MHM4070 or Genomic selection (GS) or marker-enabled Lall-454) specifically bred by CIMMYT and predictions can predict untested phenotypes ILRI for high temperatures in India reached from whole-genome information. Blümmel et al. more than 23,000 ha within 3 years. (2014b) developed a GS model of fodder quality • Concomitant increases of about 10% each traits to predict superior lines from a collection of pod yield, haulm yield and haulm fodder of doubled-haploid lines from the maize work of quality in some new cultivars has provided the International Maize and Wheat Improve- sufficient incentives for their fast and large- ment Centre (CIMMYT) in Asia. scale adoption. Apparently, small differences in CR fodder quality result in substantial differences in live- Policy impacts through the provision stock productivity because of the additive effects of information of higher diet quality and higher feed intake. Mapping recommendation domains has al- Fodder market studies in South Asia and West lowed spatial stratification of farming systems to Africa have shown that: (i) market prices reflect better assess the potential of multidimensional fodder quality differences within and between cereals (Kristjanson and Zerbini, 1999). crops; (ii) customers are willing to pay price pre- miums for apparently small differences in fodder quality traits; (iii) the price of CRs relative to grain Development impacts has increased during recent decades (Kelley et al., 1993; Sharma et al., 2010); and (iv) in some Market studies in India and West Africa have Indian markets, income from CR sales exceeded identified significant differences in CR prices that from grain sales (Samireddypalle et al., 2017). Multidimensional Crop Improvement by ILRI and Partners 483 Capacity building and partnerships methane production) and indirect (land use and on multidimensional crop conversion) greenhouse gas emissions (Steinfeld improvement as outcomes et al., 2006). Previous work by the ILRI and partners ILRI has established scientific partnerships in has identified feed shortage as a major con- the CGIAR system with ICRISAT, CIMMYT, IITA, straint to higher livestock yields; this feed con- national agricultural research and extension straint will worsen with the increasing demand systems (NARES) in Ethiopia (Ethiopian Insti- for animal-s ourced food (Blümmel et al., 2017). tute of Agricultural Research, EIAR) and India Opportunities for improving feed resources (National Research Centre for Sorghum, now are constrained by shortages of arable land the Indian Institute for Millet Research, IIMR), and, increasingly, water, and these constraints and the private sector (SeedCo, Syngenta and are likely to become aggravated by climate Advanta). change (Blümmel et  al., 2015b). Feed supply– Affordable and comprehensive phenotyp- demand scenarios for South Asia and East and ing for food–feed and fodder traits in all key West Africa have shown that CRs such as cereal and legume crops is feasible. The ILRI- straws, stover and haulms are already the most crop-centre collaboration developed and valid- important feed r esources, commonly providing ated NIRS equations for nitrogen (N), neutral 50–70% of the feed resources in smallholder detergent fibre (NDF), acid detergent fibre mixed crop–l ivestock systems (Blümmel et  al., (ADF), acid detergent lignin (ADL), in vitro 2014b; Duncan et al., 2016). organic matter digestibility (IVOMD) and metab- Generally, lignocellulosic biomass from olizable energy (ME) of CRs of sorghum, pearl forest, agricultural waste and CRs is the most m illet, groundnut, pigeon pea, chickpea, cow- abundant renewable biomass on earth, with a pea, rice, wheat and maize. ILRI NIRS specialists total production estimated ranging from about have trained hundreds of laboratory technicians 10 billion to 50 billion t (Sanchez and Cardena, from public and private sectors in South Asia 2008). About 3.8 billion t are contributed by and East and West Africa on NIRS operations. CRs, with cereals contributing 74%, sugar crops NIRS hubs exist in India and Ethiopia, and NIRS 10%, legumes 8%, tubers 5% and oil crops 3% hubs in Nigeria, Mali and Burkina Faso are being (Lal, 2005). Considering the huge quantities of established. These hubs are based on NIRS equa- CRs available from agricultural production tions developed by ILRI and partners and on ex- and the high nutritive quality of their basic tensive training given by ILRI NIRS specialists. c onstituents – hexose and pentose sugars – it comes as no surprise that attempts to upgrade CR biomass for livestock fodder reach back to the Introduction beginning of the 20th century (Fingerling and Schmidt, 1919; Beckmann, 1921). These and The growth in demand for animal-source food later attempts included chemical, physical and in low- and middle-income countries provides biological treatments, but chemical treatments challenges and opportunities. A principal chal- received the maximum attention of researchers, lenge is to raise fodder and animal yields per unit particularly the use of hydrolytic agents such of land in a situation where the shrinking nat- as sodium hydroxide and ammonia (reviewed by ural resource base in terms of land and water Jackson, 1977; Owen and Jayasuriya, 1989). makes feed production harder. However, comparatively little uptake of these In addition, feed resourcing and feeding technologies was observed, even though consid- are at the very interface where positive and erable effort was made by the international negative effects from livestock occur (Blümmel research and development community (Owen et  al., 2013b). Feeds are the single most im- and Jayasuriya, 1989). For example, Owen and portant input cost into livestock production and Jayasuriya (1989) listed and reviewed 12 major largely determine its profitability (Swanepoel international conferences addressing the im- et  al., 2010). Feed production accounts for the proved use of CR biomass for livestock feed from bulk of water required in livestock production 1981 to 1988 and concluded that large-scale (Singh et  al., 2004), as well as direct (enteric adoption of treatment interventions was very 484 M. Blümmel et al. rare and did not continue once project activities mid-1990s that ILRI, a successor of ILCA, and ceased, despite efforts to simplify treatment tech- ICRISAT concluded a memorandum of under- nologies and to use local inputs. standing to jointly attempt concomitant im- The lack of adoption of postharvest ap- provement of grain and CR traits. proaches to improving CRs gave way to a new The present chapter reviews the findings, research paradigm of targeted improvement of outputs and outcomes of research on multidi- CR fodder value by plant breeding and selection. mensional crops in the tropics, focusing mainly This was discussed at an international confer- on cereals and grain legumes. Specifically, the ence by the International Livestock Centre for chapter addresses the following: Africa (ILCA) in 1987 (Reed et  al., 1988a). At that time, research on improving CR fodder • Establishment of CRs as traded commod- value at source was largely restricted to barley ities and their changing valuation as because of the importance of green barley in the the  impetus for multidimensional crop mixed systems of the eastern Mediterranean i mprovement. (Capper et  al., 1988). The ILCA proceedings • Trait identification and development of in- (Reed et al., 1988a) contained 12 papers: three frastructure for quick and affordable phe- addressed the use of CRs as livestock feed in notyping for CR fodder quality. smallholder crop–livestock farming systems • Exploitation of existing cultivar-dependent (globally: McDowell, 1988, and Kossila, 1988; variations in CR fodder quality. West Asia and North Africa: Nordblom, 1988) • Targeted genetic enhancement for multi- and three focused on the limited nutritive qual- trait food–feed–fodder cultivars. ity and characteristics of CRs but exclusively on • Trade-offs between CR fodder traits and pri- cereal CRs (Mueller-Harvey et  al., 1988; Owen mary traits, notably grain and pod or straw and Aboud, 1988; van Soest, 1988). The excel- yields. lent fodder quality of many of the legume res- • Outcomes of multidimensional crop im- idues was not addressed. Crop and cultivar provement and future work. variations in CR fodder traits were explored by Future work on multi-trait crop improvement. Ørskov (1988) and Capper et al. (1988) in some depth with regard to the number of cultivars in- vestigated, while the remaining papers focused Fodder markets more on types of cultivars, such as bird-resistant versus non-bird-resistant cultivars (McIntire et al., Increasing the feeding value of CRs by multidi- 1988; Reed et al., 1988b), or on very few culti- mensional crop improvement depends on the in- vars (Khush et  al., 1988; Pearce et  al., 1988). herent variation among cultivars of the same Both Ørskov (1988) and Capper et  al. (1988) crop in the nutritive value of their residues fed to reported highly significant cultivar- dependent livestock. Practical evidence of such variation has variations in CR fodder quality traits with limited been observed in fodder markets in India for many trade-offs with grain yields. years, as reviewed by Kelley et al. (1993, 1996). Kelley et  al. (1993) at ICRISAT surveyed While the fodder quality of CRs was largely fodder trading of cereal straws and farmer per- ignored in historical crop-improvement pro- ceptions of grain and straw value in India from a grammes, farmers and fodder traders long rec- more demand-side perspective. These authors ognized differences in the fodder quality of CRs, found that farmers paid attention to stover even within the same species. At the farm level, quantity and fodder quality in new sorghum new pearl millet cultivars that had been im- cultivars and that new cultivars could be re- proved only for grain yields had sometimes been jected if found lacking in these traits. The au- rejected by farmers because of low CR quantity thors furthermore reported that the monetary and quality (Kelley et  al., 1996), and similar value of sorghum grain relative to stover de- findings were reported by Traxler and Byerlee creased from about 6:1 to 3:1 within two decades (1993) for wheat. Kelley et  al. (1993) reported (1970–1990) and concluded and recommended from surveys of sorghum stover trading from that crop improvement consider CR fodder traits 1985 to 1989 in four districts of Maharashtra, in future crop improvement work. It was in the India, that stover from landraces realized on Multidimensional Crop Improvement by ILRI and Partners 485 average 41% (range 24–61%) higher prices than ILRI-ICRISAT work on fodder trading in modern cultivars. These surveys provided early India was followed by research in Mali, Niger and evidence that CR fodder quality differences are Nigeria by ILRI, ICRISAT and IITA. Price pre- reflected in livestock production responses of miums related to fodder quality differences were some magnitude. In addition, the collaboration also observed in West African markets (Jarial between ILRI and ICRISAT starting in the mid- et  al., 2016a,b). Livestock producer preferences 1990s was preceded by an ex ante assessment of for haulms from groundnut or cowpea varied the impact of improving the quality of sorghum with haulm quality between groundnut and and pearl millet stover on livestock performance cowpea. Thus, cowpea haulms were costlier than (Kristjanson and Zerbini, 1999). These authors groundnut haulms in fodder markets in Mali and calculated that a 1% increase in digestibility in Niger, but they also had superior N content and sorghum and pearl millet stover would increase IVOMD than groundnut haulms, while the re- milk, meat and draught power outputs by 6–8%. verse was true in Nigeria (Table 14.1). These estimates appeared very high and were Price differences between cowpea haulm questioned by Thornton et al. (2003), who argued and groundnut haulm reflected quality differ- that a mere increase in only digestible energy, for ences. There was also consistency in pricing of example, without regard for protein would not cowpea haulm, groundnut haulm, sorghum result in a significant improvement in livestock stover and pearl millet stover over 2  years at productivity. four fodder markets in Niger (Table 14.2). The One support for a higher productivity im- average price per kg of legume haulms was pact is market prices of sorghum stover where a about five times that of the cereal stover; the difference in digestibility of 5% was associated average price per unit of N was about 2.7 times with price premiums of 25% and higher. Blüm- as high. Sorghum stover received about 30% mel and Rao (2006) surveyed six major sorghum higher prices than pearl millet stover, probably stover traders in Hyderabad, India, monthly from because of a 5% unit difference in IVOMD. 2004 to 2005 and observed that six different Across the four CRs, N accounted for 98% stover types were usually traded. Customers usu- (p = 0.008) of the variation in price and IVO- ally had the choice of two or three sorghum stov- MD for 91% (p = 0.04), respectively. While Jar- er types offered by the same trader. The poorest ial et al. (2016b) did not report price differences and best-quality stover (perceived in terms of col- for CRs within crops related to cultivar differ- our, softness, sweetness, etc.) were sold on aver- ences, observations at a fodder market in Kano age for INR3 and INR4 per kg of dry matter, in September 2016 found cultivar differences respectively. Blümmel and Rao (2006) investi- in price in sorghum stover and in groundnut gated these traded stovers for laboratory fodder haulms (M. Blümmel, personal observation, quality traits, such as crude protein and IVOMD, September 2016). and related these laboratory traits to stover A further point is the relative monetary prices. While stover crude protein content was value of grains and CRs. In legume haulms, the not related to stover prices, IVOMD accounted for monetary value of grain and CRs can reach par- 75% of the price variation. In rice straw, trading ity (Samireddypalle et al., 2017), and grains can differences in IVOMD as low as 2–3 percentage occasionally (e.g. when there is high demand for points were associated with similar price pre- mutton during Muslim festivals) even be cheaper miums (Teufel et  al., 2010). Incidentally, these than haulms (Ayantunde et  al., 2014). In sor- findings were in accord with the above-reported ghum stover trading in India during the past observations of Kelley et  al. (1993) that stover decade, stover prices were about 50–60% that of from sorghum landraces achieved on average sorghum grain value (Sharma et al., 2010). mean prices 41% higher than modern cultivars. In summary, CR fodder market studies in Customers would not pay such price premiums if South Asia and West Africa showed that: (i) traders feeding of stover from landraces would not result and customers are aware of CR fodder quality in significantly higher livestock productivity. differences within and across crops; (ii) customers Findings from the surveys of sorghum stover are willing to pay considerable price premiums (Blümmel and Rao, 2006) and rice straw (Teufel for apparently small differences in fodder quality et al., 2010) trading are combined in Fig. 14.1. traits; and (iii) the monetary value of CRs relative 486 M. Blümmel et al. 4.2 4.0 Best rice straw 3.8 High-quality sorghum stover ‘Raichur’ 3.6 3.4 Good rice straw 3.2 3.0 Medium/low Low-quality rice straw sorghum stover ‘Local Yellow’ 2.8 35.0 37.5 40.0 42.5 45.0 47.5 50.0 52.5 55.0 Crop residue in vitro digestibility (%) Fig. 14.1. Relationship between cost of sorghum stover and cost of rice straw and their in vivo digestibility. DM, dry matter. (Data from Blümmel and Rao, 2006, and Teufel et al., 2010.) Table 14.1. Prices, relative value and fodder quality traits of cowpea haulm (CPH) and groundnut haulm (GNH) traded in Niger, Mali and Nigeria. Nigera Malib Nigeriac Variable CPH GNH CPH GNH CPH GNH Price (US cents/kg) 28 20 95 86 19 72 Price grain:haulm 2.4:1 4:1 1:1.7 1:1.5 3.5:1 1.1:1 Haulm nitrogen (%) 2.2 1.7 2.9 2.4 2.0 2.4 Haulm IVOMD (%) 61.3 58.4 65.4 64.2 55.6 57.1 aJarial et al. (2016a) for four fodder markets, over 2 years, bimonthly. bAyantunde et al. (2014) for five fodder markets, over 1 year, monthly. cSamireddypalle et al. (2017) for five fodder markets, over 2 years, monthly. Table 14.2. Average N content, IVOMD and prices (CFA franc) of cowpea haulm, groundnut haulm, sorghum stover and pearl millet stover traded over 2 years at four fodder markets in Niger. (Modified from Jarial et al., 2016b.) Fodder N (%) IVOMD Price (CFA franc/kg) Price (CFA franc/N) Cowpea haulm 2.22 61.3 164 73.9 Groundnut haulm 1.60 58.4 119 74.4 Sorghum stover 1.03 52.2 31 30.1 Pearl millet stover 0.98 47.2 24 24.5 Crop residue price (INR/kg DM) Multidimensional Crop Improvement by ILRI and Partners 487 to grain values is considerable and has been in- invest in higher-quality stover. However, only creasing over recent decades (Kelley et al., 1993; part of the incremental increase in milk poten- Sharma et al., 2010). In fact, depending on har- tial was due to the higher-quality stover, as this vest indices and/or CR fodder quality, more group also consumed more concentrate money can be earned from CRs than from the (0.85 kg/day), which contributed about half to primary product (Samireddypalle et  al., 2017). the DTMR. The increased milk potential attrib- The findings from fodder markets as far apart as utable to higher stover quality is estimated to be West Africa and South Asia send strong signals 2.4  kg/day (increase from 4.4 to 6.8  kg/day; that both fodder quantity and fodder quality of Table 14.3). This would be an increase of about straws, stovers and haulms do matter. 24% relative to the milk potential of the DTMR with the lower-quality stover of 9.9 kg/day. This increase appears to agree with the price pre- miums paid for the higher-quality sorghum CR fodder quality and livestock stover at the fodder markets in India. It also productivity seems to align with the price differences ob- served between sorghum and pearl millet stover Livestock productivity trials conducted with the traded at fodder markets in Niger (Table 14.2). private sector confirmed information from fod- The effect of CR quality on livestock prod- der market studies. In India, Miracle Fodder and uctivity is clearer in cases where the residues are Feeds Pvt Ltd designed so-called densified total fed as sole diets rather than as basal diets, as is mixed ration (DTMR) feed blocks that consist generally the case with cereal CRs. Table 14.4 largely of by-products such as sorghum stover summarizes work where legume haulms were (about 50%), bran, oilcakes and husks (about fed as sole diets to small ruminants. Cultivar- 36%), with the rest contributed by molasses dependent variations in haulm fodder quality (8%), maize grain, urea, minerals, vitamins, etc. were considerable. In the case of groundnut (Shah, 2007). In a series of experiments with haulms harvested from six different cultivars in Miracle Fodder and Feeds Pvt Ltd, the authors Nigeria, sheep could lose weight on haulms from tested these feed blocks with two objectives: (i) to one cultivar while gaining 46 g/day on haulms estimate probable maximum productivity levels from another cultivar. In India, weight gains in on cereal CR-based diets; and (ii) to estimate the sheep could differ by more than twofold (from 65 importance of the quality of the basic CR going to 137 g/day) depending on haulm fodder qual- into the blocks on overall livestock performance. ity difference among groundnut cultivars. Simi- In an experiment with a large private Indian lar proportional genotypic variations have been buffalo dairy (Anandan et al., 2010), two experi- reported for faba bean haulms in Ethiopia mental DTMR feed blocks were produced from (Table 14.4). For unsupplemented barley straw low-quality (47% IVOMD) and premium-quality from eight different cultivars, Capper et al. (1988) (52% IVOMD) sorghum stover traded in the fod- reported daily weight changes from 150 g to lit- der markets (Blümmel and Rao, 2006). tle above live weight maintenance. Protein sup- The results from these trials are reported in plementation resulted in cultivar-dependent Table 14.3. Using premium sorghum stover variations in weight gain from 100 to 250 g/day. (‘Raichur’ in Fig. 14.1) resulted in more than These examples show that the effect of cul- 5 kg higher daily milk potential than using the tivar variations on fodder quality of CRs on live- lower-quality stover (‘Local Yellow’ in Fig. 14.1). stock productivity can be substantial. The high This differential yield potential was due to higher response in livestock performance to apparently ME content/kg DTMR and also higher feed in- small differences in CR fodder quality is the re- take in the ration containing the premium stover. sult of two cumulative effects: higher diet qual- These accumulating effects of higher ME con- ity and higher feed intake. However, this effect tent and higher feed intake are the reason that can only be effective where feed is offered ad lib- apparently small difference in feed quality can itum, which is not always the case, and often CRs have considerable effects on animal performance. are in short supply and fed in a restricted fashion The increase in milk potential of 5 kg compared (Mayberry et al., 2017). It is also worth pointing with the ration containing the lower-quality out that higher productivity can be achieved on stover explains the decisions of customers to mostly, or even completely, by-product-based 488 M. Blümmel et al. Table 14.3. Milk potential in Indian dairy buffalo fed two DTM feed blocks based on premium-quality (52% digestibility, 7.39 MJ ME/kg) and low-quality (47% digestibility, 6.52 MJ ME/kg) sorghum stover with total by-product proportion of feed blocks greater than 90%. (Data from Blümmel et al., 2017, based on the actual milk fat contents of buffalo milk.) Low-quality stover Premium-quality stover Protein (%) 17.1 17.2 ME (MJ/kg) 7.37 8.46 Voluntary intake of feed block (kg/day) 18.0 19.7 Voluntary intake of feed block (% /kg LWa) 3.6 3.8 ME intake (MJ/day) 132.7 166.7 ME intake stover (MJ/day) 58.7 72.7 Milk fat (%)b 7.4 7.6 Milk potential (kg/day) 9.9 15.5 Milk potential from stoverc 4.4 6.8 Milk potential from cross-bred cattle (kg/day) 14.0 21.0 Milk potential from stoverc 6.2 9.2 aLive weight (LW) of buffalo was calculated by body measurements and estimated to be on average 506 and 525 kg in the low-quality and premium-quality feed block, respectively. bMilk fat in cattle was assumed to be 4% with a cross-energy content of 3.13 MJ/kg. cEstimated based on ME contribution of stover to ME of DTMR as: ME stover/ME DTMR × milk potential. Table 14.4. Effect of groundnut and faba bean cultivars on live-weight changes in sheep fed exclusively with haulms offered ad libitum. Experiment Average (g/day) Lowest (g/day) Highest (g/day) Haulms of ten groundnut cultivars 94.1 65 137 fed to Indian Deccan sheepa Haulms of six groundnut cultivars fed 26.5 –6 46 to West African Dwarf sheepb Haulms of five faba bean cultivars fed to 49.2 37.5 64.6 Ethiopian Arsi sheepc aPrasad et al. (2010). bEtela and Dung (2011). cWegi (2016). feeding systems. In the case of DTMR, milk in a later section that the fodder quality of CRs yields in cross-bred cattle of more than 20 kg/ can be increased further by targeted genetic day seem achievable (Table 14.3) and these enhancement using conventional or molecular DTMR consist of more than 90% by-products. breeding crop-improvement approaches. Feeding legume haulms as the sole feed to sheep can result in daily weight gains of well over 100 g/day (Table 14.4). These are productivity Trait Identification and Tools for levels more commonly associated with concen- trates than with CR diets. Findings from the live- Affordable Phenotyping for Crop stock productivity trials are consistent with price Residue Fodder Quality premiums paid for fodder quality differences (Fig. 14.1, Tables 14.1 and 14.2). It is important Validation of laboratory fodder to point out that the variations seen in the fodder quality traits markets and livestock productivity trials came about largely by chance and that those differ- Fodder quality is ultimately determined only by ences in fodder quality were not the intentional livestock production and productivity, but livestock results of crop breeding or selection. We will see performance trials are unsuitable for routine Multidimensional Crop Improvement by ILRI and Partners 489 feed and fodder quality analysis. This is particu- traits using stepwise multiple regressions im- larly valid in crop improvement, where many proved predictions of in vivo measurements in samples must be analysed and where initially most cases in pearl millet, sorghum stover and the biomass availability is low. Simple laboratory groundnut haulms. In all three cases, laboratory fodder quality traits are needed, but these traits traits related to available feed energy (in vitro di- must be well correlated with actual livestock gestibility, ME and fibre constituents) were found performance measurements. ‘Simple’ here refers to exhibit more consistent relationships with not only to logistical and economical laboratory in vivo measurements than CR N content (Pra- demand but also to the need for the traits to be sad et al., 2010; Ramakrishna et al., 2010). comprehensible to, and usable by, crop scien- tists, seed producers, fodder traders and develop- ment practitioners with no or little training in livestock nutrition. When the ILRI-ICRISAT col- Calibration and validation of NIRS tools laboration on multidimensional crop improve- ment started, a wide range of potential Conventional laboratory analysis cannot effi- morphological, chemical and in vitro traits were ciently cope with the large set of sample entries investigated and related to livestock perform- from multidimensional crop-improvement pro- ance measurements usually obtained with sheep grammes. NIRS is a non-invasive technique rou- (Sharma et al., 2010). tinely used since the 1960s in the food industry, Ravi et  al. (2010) investigated morpho- forage breeding and pharmaceutical industry. logical, chemical and in vitro traits in pearl millet Most instruments used are manufactured by stover and related these traits to organic matter FOSS (Forage Analyser 500 and 6500), which digestibility, organic matter intake, digestible has the advantage that NIRS equations developed organic matter and N balances in sheep. Gener- in one laboratory can be transferred to other ally, fibre components and in vitro laboratory l aboratories using FOSS. The ILRI-crop-centre traits were more closely related to in vivo meas- collaboration developed and validated NIRS equa- urements than morphological traits, even tions for N, NDF, ADF, ADL, IVOMD and ME of though plant height and stem diameter were CRs of sorghum, pearl millet, groundnut, pigeon both consistently and statistically significantly pea, chickpea, cowpea, rice, wheat and maize. We inversely related to the in vivo measurements of generally expected an R2 value of at least 0.90 be- 40 pearl millet stovers. In contrast, traits such as tween conventionally analysed laboratory traits leafiness, including estimates of residual green and blind predictions by NIRS (see also Sharma leaf area, which are often employed for sensory et al., 2010). With new global interest in mono- phenotyping by crop-improvement programmes gastric and fish feed, NIRS equations were also de- and farmers, were less well related to in vivo veloped for grains of key crops, including routine measurements. It is important to realize that all quality traits such as protein, starch and fat stovers were offered chopped (which is increas- (Choudhary et al., 2010) but also amino and fatty ingly the practice or the trend, at least for stover acids (Prasad et al., 2015), which still mostly rely utilization in India and elsewhere), which might on costly high-performance liquid chromatog- reduce the importance of leafiness and other raphy analysis. morphological traits on intake responses. NIRS equations can be transferred across Bearing in mind the above considerations FOSS-type instruments with little spectra about the simplicity and meaningfulness of fod- standardization to account for instrument-to- der quality traits, NDF (a cell wall estimate), ADF instrument variation. Over the past one and a (an estimate of cellulose) and in vitro digestibility half decades, ILRI NIRS specialists have trained seem to be good indicators for ranking fodder hundreds of laboratory technicians from CGIAR quality in pearl millet, sorghum (Ramakrishna and the national public and private sectors in et al., 2010) and maize (Ravi et al., 2013) stover, South Asia and East and West Africa on NIRS while ADL (an estimate of lignin) seems to pre- operations, including NIRS networking and the dict fodder quality in groundnut haulms better generation of NIRS equations. Fully functioning than any of the aforementioned traits (Prasad NIRS hubs exist now in India and Ethiopia, and et  al., 2010). Combining different laboratory NIRS hubs in Nigeria and Mali are being set up. 490 M. Blümmel et al. Thus, quick, affordable and comprehensive Asia, and its producer, Syngenta, has recently phenotyping for food–feed and fodder traits in approached ILRI and CIMMYT for ways of ad- all key cereal and legume crops is feasible, but vertising the high fodder quality on the seed sample processing (drying, grinding and ship- packets of the hybrid. ILRI, CIMMYT and Syn- ping) limit experimental efficiency. Mobile NIRS genta are now exploring processes to bring applications can potentially overcome this con- about such branding and seed bag labelling for straint. Two new mobile hand-held systems CR fodder quality traits. Work is ongoing in the manufactured by Phazir and Brimstone have use of check cultivars analogous to current been explored during the past 2 years to remove, methods of comparing grain yields of yet-to- or at least mitigate, the sample processing con- be-released cultivars with yields of selected straint (Prasad et  al., 2015). Phazir and Brim- check cultivars; in addition to grain yields, CR stone currently cost about US$40,000 each, but quality traits could also be compared. Another recently an extremely cheap (about US$450–500) option is comparing CR quality of yet-to-be- and small pocket NIRS system called Scio came released cultivars with longer-term average on the market and is currently being tested at qualities of CRs traded at fodder markets or ILRI in India and Ethiopia. with the average values of CR qualities given in nutritional textbook/feeding tables for a given country. In any event, getting the private sec- Exploitation of Existing tor interested in dual-purpose traits is of great strategic importance for mainstreaming Multi- Cultivar-dependent Variation dimensional crop improvement and for scaling in CR Fodder Quality of new cultivars, as public-sector crop im- provement groups are watching the private Phenotype pipeline and releases sector closely. for variations in CR fodder quality Phenotyping pipeline hybrids that are close to release is also cost-effective and has A widespread misconception about how super- short delivery pathways. This was imple- ior CRs can be generated is that targeted crop mented with a private-sector maize pro- breeding is invariably required. However, phe- gramme in India; examples of this work from notyping for fodder quality to detect genetic 2014 onwards are presented in Fig. 14.2, differences in food–feed–fodder traits in ad- where 24 pipeline hybrids were tested at four vanced cultivars and exploiting them often suf- locations in India. The hybrids with the highest fices. Exploiting existing variations in traits average grain yield also had highest stover N and targeting genetic enhancement towards and second highest stover IVOMD (Fig. 14.2a,b). specific traits are separate approaches, and the The variation in stover IVOMD among the top first is possible without the second. The first grain yielders of 9–10 t/ha was like the vari- approach does not require much investment ation between the best and poorest sorghum besides phenotyping for CR traits and has short stover in fodder trading in India (Fig. 14.1) or delivery pathways. The second approach re- between the average sorghum and pearl mil- quires more investment and time but promises let stover traded in Niger (Table 14.1), which greater impact. resulted in appreciable price premiums for the This timespan of crop improvement can better-q uality stover in both cases. The impli- be shortened by phenotyping CRs of released cation for promoting the maize hybrid with cultivars for fodder quality and by promoting the highest IVOMD rather than the lowest superior dual-purpose cultivars with farmers, IVOMD among the top yielders in dairy prod- traders and processors. This approach is par- uctivity can be extrapolated from the findings ticularly promising where the private sector is in Table 14.3. However, while combination of involved, usually in the promotion and mar- highest grain yields and highest stover traits keting of hybrids. A collaboration between such as N and IVOMD are entirely feasible, ILRI and CIMMYT identified such a superior these trait combinations seem to be associ- dual-purpose maize hybrid (Anandan et  al., ated only with intermediate stover yields 2013), which is now a very popular hybrid in (Fig. 14.2c,d). Multidimensional Crop Improvement by ILRI and Partners 491 (a) (b) 11,000 11,000 p = 0.22 p = 0.09 10,000 10,000 9,000 9,000 8,000 8,000 7,000 7,000 6,000 6,000 5,000 5,000 1.00 1.05 1.10 1.15 1.20 1.25 1.30 47 48 49 50 51 52 53 54 55 56 57 Stover N content (%) IVOMD (%) (c) (d) 11,000 11,000 p = 0.6 p = 0.44 10,000 10,000 9,000 9,000 8,000 8,000 7,000 7,000 6,000 6,000 1.00 1.05 1.10 1.15 1.20 1.25 1.30 46 48 50 52 54 56 58 Stover N content (%) IVOMD (%) Fig. 14.2. Relationship between stover N and grain yield (a), between stover IVOMD and grain yield (b), between stover N and stover yield (c) and between stover IVOMD and stover yield (d) in 24 pipeline maize hybrids grown at four locations in India. S t o v e r y i e l d ( k g / h a ) G r a i n y i e l d ( k g / h a ) S t o v e r y i e l d ( k g / h a ) G r a i n y i e l d ( k g / h a ) 492 M. Blümmel et al. Institutionalized Multidimensional crop im- matter intake of pearl millet stover measured in provement has advanced only slowly. In 2002, sheep increased from 12.9 to 15.1  g/kg live the N ational Research Centre for Sorghum weight (LW), an increase of 17%, and the N bal- (NRCS) decided to include sorghum stover traits ance changed from negative (−0.016 g/kg LW/ as release criteria for new sorghum cultivars. day) to positive (0.05  g/kg LW/day). The im- Interestingly, this was influenced by a visit of the provement in stover quality did not come at any then Director of the NRCS to the sorghum fodder penalty for grain or stover yield. Choudhary markets in Hyderabad described earlier. This in- et  al. (2012) investigated the mode of inherit- volved seconding NRCS technicians to the ILRI ance of stover N and IVOMD. From a full-sibling NIRS Hub hosted by ICRISAT to analyse stover of (FS) base population of pearl millet variety ICMV all new sorghum cultivars submitted for release 221, three high and low N and three high and under the All-India Coordinated Research Pro- low IVOMD FSs were selected. Crosses were ject (AICRP) on Sorghum (Venkatesh et  al., made for high × high (H × H), low × low (L × L), 2006). This work continues and is now being ex- and high × low (H × L) FS trait contrasts and plored for minor millets by IIMR. Stover traits evaluated at Patancheru, in India, in the rainy have now also been included as release criteria seasons of 2007 and 2008. The high- and low-N for pearl millet, although this crop is, paradoxic- (HN and LN, respectively) FS parents were 0.85% ally, currently not under the mandate of IIMR. and 0.72% N, respectively. In the crosses, stover Less formalized pilot studies have been under- N contents were: HN × HN = 0.85%, LN × LN = taken with the Indian Directorate for Maize, 0.73% and HN ×LN = 0.80% (p  <  0.05). The where the modification of cultivar release cri- high- and low-digestibility (HD and LD, respect- teria to include maize stover traits was discussed ively) FS parents were 43.3% and 40.3% IVO- during recent annual maize meetings, although MD, respectively. In the crosses, stover IVOMD without a formal decision yet being taken. The were: HD × HD = 43.7%, LD × LD = 40.3% and situation is similar in Ethiopia, where the Inter- HD × LD = 42.2% (p < 0.05). The intermediate national Centre for Agricultural Research in the results of H × L crosses strongly indicated the Dry Areas (ICARDA) prepared the ground with additive nature of the stover quality traits of N EIAR by phenotyping lentils, chickpeas and faba and IVOMD and suggest the application of cyclic beans for haulm fodder quality traits during re- breeding methods for increasing stover N con- lease processes (Alkhtib et al., 2016, 2017). tent and IVOMD in pearl millet. A further pilot study was conducted to increase the key fodder quality traits of N Targeted genetic enhancement towards c ontent and IVOMD through two cycles of FS food–feed crop cultivars recurrent selection of open-pollinated pearl millet cultivar ICMV 221 (base population, C 0). Six experimental varieties were selected from The targeted concomitant improvement of grain the first cycle (C ) and second cycle (C ) of se- and CR traits requires more investment and time 1 2lection for: (i) high grain yield; (ii) high grain than the mere detection and exploitation of al- and stover yield; (iii)  high stover IVOMD, ready existing variations but promises greater (iv)  low stover IVOMD; (v) high stover N con- impact. In ILRI-ICRISAT-CIMMYT collabor- tent; and (vi) low stover N content. Stover N ations, both conventional and molecular breed- and IVOMD increased by 9.5% and 2%, re- ing approaches were applied for targeted genetic spectively, in the C bulk, and by 21% and 5%, enhancement of CR fodder traits within the 1respectively, in the C bulk over the base popu- paradigm of simultaneous improvement of 2lation C . The high-N experimental varieties grain and fodder traits. 0showed the highest N percentage and stover N yield, while the high-digestibility experimental varieties showed the highest ME and IVOMD Recurrent selection values from both selection cycles. The findings suggest that stover N and IVOMD can be im- Bidinger et  al. (2010) showed that within two proved without significant detriment to grain recurrent selection cycles, digestible organic and stover yield. Multidimensional Crop Improvement by ILRI and Partners 493 Hybrid breeding for dual-purpose maize commercial hybrids within India also led to identification of promising hybrids such as In South Asia, dual-purpose maize breeding was NK6240 (Syngenta) with high digestibility (over supported by the CRP on Maize through a com- 50%) (Anandan et  al., 2013) and high grain petitive grant scheme to ILRI. Zaidi et al. (2013) yields (over 9.0 t/ha) during the rainy season. reported substantial variability for stover quality Maize is fast replacing some of the major in maize working with germplasm available cereal crops grown widely in these regions and from CIMMYT-Asia with no negative effect of currently ranks first followed by rice and wheat the stover quality traits (IVOMD and ME) on in terms of production and growth. One of the grain yield, indicating the possibility for simul- emerging seasons for maize cultivation in India taneous improvement of both stover quality and is spring, particularly in South India (usually a grain yield. In addition, substantial progress has rice–fallow system), where adverse weather con- been made in identifying trait-specific genomic ditions prevail (high temperature and low rain- regions for use in targeted breeding programmes fall). Several pipeline hybrids and breeding lines to improve stover quality and grain yield have been tested to suit this environment, and (Vinayan et  al., 2013). This breeding initiative preliminary investigations led to identification for improving stover quality has led to the devel- of potential hybrids that have good grain yield opment of advanced lines with high digestibility and high stover quality. The progress of this (over 50%) and energy (greater than 8.0 MJ/g) maize hybridization programme to simultan- for use as parents of new hybrid combinations. eously improve food and fodder traits is exempli- Results from evaluation of these experimental fied in Fig. 14.3 using data from sorghum stover hybrids under optimal growing conditions have trading as reference values; the perceptions of shown promise in terms of their yield perform- farmers and traders in India are that sorghum ance (roughly 8.0 t/ha) and in vitro digestibility stover is nutritionally superior to maize stover (over 50%). Studies of the performance of (Blümmel et al., 2014b). Mean IVOMD (range 53.6–56.0%) of 11 experimental heat-tolerant dual- purpose maize hybrids generated 4.2 4.0 High-quality sorghum stover 3.8 3.6 Mean IVOMD (range 53.7– 59.8%) of 11 advanced dual- purpose maize breeding 3.4 lines generated 3.2 Low-quality 3.0 sorghum stover 2.8 45.0 47.5 50.0 52.5 55.0 57.5 60.0 Stover in vitro digestibility (%) Fig. 14.3. Breeding advances in dual-purpose maize stover quality relative to different sorghum stovers traded in rainfed India in the past decade. (Data from Blummel et al., 2014b). Stover price (INR/kg DM) 494 M. Blümmel et al. Fig. 14.3 shows that maize stover is not in- multi-location field trials. The results from the ferior to sorghum stover, which was also con- laboratory analysis of stover samples showed firmed in trials with dairy animals (Blümmel that one of the improved hybrids was at least et al., 2014b). Furthermore, the average IVOMD 8.5% higher in ME and 6.3% higher in IVOMD (54.4%) of the new hybrids targeting areas with than the control hybrid. The new hybrid also adverse weather conditions is about 2.5% higher produced a 10% increase in grain yield and a 4% than that of the highest-quality traded sorghum increase in stover yield. These results suggest stover (Fig. 14.3). This was one of the sorghum that new hybrids can be developed, concomi- stovers used for the dairy experiments described tantly improving grain and stover traits using in Table 14.3. It is very likely that dairy product- QTLs (Nepolean et al., 2009). ivity would be substantially further enhanced if Blümmel et al. (2015a) introgressed stay- the sorghum stover were replaced by a maize green QTLs into the sorghum genetic back- stover with 56% IVOMD as available in the new grounds S-35 and R-16, generating 52 and hybrids, and even more so by a maize stover with 39 lines, respectively, to investigate the effects an IVOMD of close to 60% now available in the of stay-green introgression on stover traits new dual-purpose breeding lines (Fig. 14.3). and grain–stover relationships. The stover qual- Similar findings were reported from CIM- ity traits analysed were N, IVOMD, ADF, ADL MYT-ILRI dual-purpose maize breeding research and neutral detergent solubles (=  100  –  NDF) in East Africa (Ethiopia, Tanzania and Kenya). using a combination of conventional nutritional Ertiro et  al. (2013) produced 60 experimental laboratory analysis with NIRS. Field trials dual-purpose hybrids from 16 parental lines, were conducted under treatments of unlimited yielding 10 t of grain with an IVOMD of up to (control) and limited water supply. Significant 62% (range 53.1–62.3%). Mid-parental key (p < 0.0001) differences were found among lines stover traits such as IVOMD were well related for grain and stover yield and all stover quality (r = 0.78; p < 0.0001) to the IVOMD of the hy- traits under both water treatments. Water treat- brids produced from them, also strongly suggest- ment had greater effects on grain and stover ing the opportunity for dual-purpose hybrid yields, which decreased by between 20% and breeding. 32% under water stress, than on stover quality traits, which varied at most by 8% between treatments. Year had the greatest effect among QTL identification and backcrossing treatments, followed by water treatment and cultivar. Trade-offs between stover quality traits Nepolean et  al. (2009) used QTLs to map the and grain yields were largely absent in both genomic regions controlling stover quality and backgrounds. However, the effect of QTLs on se- yield traits in pearl millet. Marker-assisted breed- lected stover quality traits was background de- ing would be an effective tool to exploit these pendent. In S-35, one stay-green QTL (stgB) genomic regions and to choose breeding lines significantly increased stover IVOMD and grain having combinations of better stover quality and stover yield, while no concomitant trait im- and high grain yield without linkage drag be- provement was observed in the background tween these traits. With these objectives in mind, R-16. The QTL in S-35 also increased the water- QTLs for stover IVOMD and ME content were use efficiency of the whole plant in terms of identified and introgressed into four parental grain yield, stover yield and stover ME (Blümmel lines of existing hybrids showing good agro- et al., 2014a). nomic performance. Three generations of mark- er-assisted backcrossing and subsequent selfing of backcrossed progenies having target QTLs GWAS and GS was carried out with the help of QTL-flanking microsatellite simple-sequence-repeat markers. GWAS have the potential to unravel favourable Single QTL introgression lines that were homo- native genetic variations for traits of agronomic zygous for target regions were identified. Im- and economic importance across a wide range proved hybrids were synthesized from these of cereal crops. Vinayan et al. (2013) studied a QTL homozygous lines and were evaluated in panel of 276 inbred lines from CIMMYT’s Multidimensional Crop Improvement by ILRI and Partners 495 Drought Tolerant Maize for Africa (DTMA) pro- and (iv) higher livestock productivity with CRs ject using their test-cross hybrids with the maize with a higher feed quality. Evidence for cultivar line CML312, and the single crosses were evalu- preferences based on feed traits comes from ated for grain and stover yields, plant height, farmer rejection of new sorghum and pearl mil- days to 50% anthesis and silking, stover N, NDF, let cultivars that had been improved only for ADF, ADL, IVOMD and ME content. GWAS grain yields and had low stover quantity and analysis was carried out using genotyping by quality (Kelley et  al., 1996). Recently, farmers sequencing, and 55K single-nucleotide poly- ranked maize stover traits highly when assessing morphism arrays revealed several regions of cultivars in East Africa (de Groote et al., 2013). significant association for N, ADF and IVOMD, Trading of CRs is expanding in volume and dis- each explaining from 3% to 9% of the pheno- tances, and CR:grain price ratios during the typic variance for these fodder quality traits. past two decades have decreased (Kelley et  al., GWAS was helpful in uncovering genomic re- 1993; Blümmel and Rao, 2006; Berhanu et al., gions of interest for target traits. 2009). Nevertheless, grain yields remain the GS or marker-enabled predictions can pre- primary trait that most crop-improvement pro- dict untested phenotypes from whole-genome grammes focus on. When multidimensional information. In one study, GS models were devel- crop-improvement programmes target CR traits, oped for fodder quality traits to predict superior they need to address potential trade-offs between lines from the collection of doubled-haploid lines grain and CR traits. generated by the Global Maize Program of It is important to understand what causes C IMMYT in Asia. Using high-density genotypic trade-offs between grain and CR traits. In its sim- information as well as fodder quality phenotypes plest form, a nutrient limited by soil fertility and/ of approximately 700 lines from two association or fertilizer application, such as N, is partitioned panels – DTMA and the CIMMYT-Asia Associ- between grain and the CR. A more complex ex- ation Panel (CAAM) – marker effects were ob- ample is in the partitioning of photosynthetic tained for fodder quality traits using GS models. products (which are not finite quantities such as The results indicated significant relationships soil and fertilizer N), notably soluble carbo- between genotyping-by-sequencing-derived val- hydrates, which contribute significantly to CR ues and the phenotypes, with r values ranging digestibility and therefore to fodder quality. from r = 0.44 to r = 0.45 across IVOMD and ME, Trade-offs can also arise from more indirect respectively (Blümmel et al., 2014b). These pre- mechanisms of ensuring grain yields and effi- dictions of fodder quality phenotypes in biparen- cient harvest, such as lodging resistance, which tal populations indicated that genomic selection can affect fodder quality of CRs through in- can be used to: (i) improve fodder quality in creased stem lignification. maize breeding populations; and (ii) select par- On the most basic level of trade-offs, grain ents in breeding for fodder quality from maize and CR yields were only moderately correlated repositories without phenotyping the lines. in sorghum (Blümmel et al., 2010), groundnut (Nigam and Blümmel, 2010), pearl millet (Bidinger and Blümmel, 2007), cowpea (Samireddypalle et al., 2017), maize (Blümmel et al., 2013a) and Trade-offs Among Crop Residue wheat (Blümmel et  al., 2012a). Grain yields Fodder Traits and Primary Traits rarely accounted for more than 50% of the vari- ation in CR yields. In other words, variation in Primary and secondary traits harvest indices were considerable and grain yield is an insufficient predictor of CR yield. The increasing importance and demand for Breeding for increases in grain yield was often CRs as fodder is reflected in four major trends: accompanied by shortening of stems to prevent (i)  increasing labour investment in collecting lodging, resulting in the longer term in increas- and storing CRs in more extensive systems ing harvest indices (Hay, 1995). While this rela- (Valbuena et  al., 2015); (ii) farmer preferences tionship has been shown in temperate cereals, for dual-purpose crop varieties; (iii) higher mar- it is less clear in pulses and tropical cereals ket price for CRs with a higher feed quality; such as rice and maize (Hay, 1995). In recent 496 M. Blümmel et al. years, investments in second-generation biofuel to provide minimum microbial N requirements technologies have resulted in renewed interest in the rumen (van Soest, 1994). Sorghum stover in variations in harvest indices, as CRs provide protein and stover yield were significantly posi- valuable feedstock for ethanol production tively correlated, but the correlation coefficients (e.g. Dai et  al., 2016). These authors also re- were low (Fig. 14.4b). ported considerable cultivar- and management- Bidinger and Blümmel (2007) and Blüm- dependent variations in harvest indices, mel et  al. (2007a) imposed N restrictions by suggesting that CR yields cannot be satisfactor- limiting fertilizer application while increasing ily calculated from grain yields. Grain yield and pressure on partitioning of N and adjusting total biomass yield should therefore be recorded planting densities on different cultivars (land- in Multidimensional crop-improvement efforts. races, open-pollinated varieties (OPVs) and hy- These considerations are also relevant for con- brids) of pearl millet. Even under these imposed servation agriculture, as higher biomass yield restrictions, the authors found no inverse rela- would make the partitioning of CRs between tionship between the stover N of pearl millet livestock feeding and soil improvement perhaps and grain yields (Fig. 14.5a). However, stover less contentious (Baudron et al., 2014). N and straw yield could be significantly inversely associated under low fertility and high popula- CR N content and grain and CR yield tion density (Fig. 14.5b). Water restriction reinforces trade-offs Relationships between the N content of CRs and under normal management and growing condi- grain and CR yields vary. Under balanced crop tions. For example, in chickpea cultivars, haulm, management, when no restrictions were im- N and grain yield were inversely correlated posed on fertilizer or water, trade-offs between (r  =  −0.41) under normal growing conditions the N content of CRs and grain and CR yield but this association became closer (r  =  −0.62) were largely absent (Fig. 14.4). (The data in under water restriction (Fig. 14.6a). Associ- Fig. 14.4 were derived from a collaboration be- ations were positive between haulm N and grain tween the National Research Center for Sorghum, yield and again the association was stronger later renamed the Directorate for Sorghum under water restriction (Fig. 14.6b). Similar re- Research, and Indian Institute for Millet Re- lationships have been observed for groundnut search). No relationship was observed between (Blümmel et al., 2012b). the protein content of sorghum stover (which is calculated as stover N × 6.25) and grain yield CR digestibility and grain and CR yield (Fig. 14.4a). Under high Kharif (sorghum grown in the rainy season in semi-arid India) grain As with CR N content, relationships between CR yielders of 5 t/ha, stover protein content could digestibility and grain and CR yield are affected vary from 4% to 7%, the latter being adequate by water stress. No relationship was observed (a) 7,000 Kharif: p = 0.60 (b) 20,000 Kharif: y = 7782 + 724x, r = 0.20; p = 0.02 Rabi: p = 0.33 Rabi: y = 4558 + 295x; r =0.21; p = 0.006 6,000 17,500 15,000 5,000 12,500 4,000 10,000 3,000 7,500 2,000 5,000 1,000 2,500 0 0 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 Stover crude protein content (%) Stover crude protein content (%) Fig. 14.4. Relationships between mean stover crude protein and grain yield (a) and between mean stover crude protein and stover yield (b) in Kharif and Rabi sorghums submitted for cultivar release from 2002 to 2008. (Unpublished data, Michael Blümmel). Grain yield (kg/ha) Fodder yield (kg/ha) Multidimensional Crop Improvement by ILRI and Partners 497 (a) 500 HF HP: p = 0.52 (b) 600 HF HP: p = 0.24 HF LP: p = 0.32 HF LP: p = 0.87 LF HP: p = 0.32 400 500 LF HP: r = –0.67; p = 0.005 LF LP: p = 0.92 LF LP: p = 0.16 400 300 300 200 200 100 100 0 0 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 N content (%) of stovers N content (%) of stover Fig. 14.5. Relationships between N content of pearl millet stover and grain yields (a) and between N content of pearl millet stover and stover yields (b) under high (HF) and low (LF) fertility and high (HP) and low (LP) population density. (Data from Bidinger and Blümmel, 2007). (a)3,500 (b) 6,000 3,000 5,000 2,500 4,000 2,000 3,000 1,500 2,000 1,000 500 1,000 Irrigation: r = –0.41; p < 0.0001 Irrigation: r = 0.14; p = 0.016 No irrigation: r = –0.62; p < 0.0001 No irrigation: r = 0.54; p < 0.0001 0 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 N content (%) N content (%) Fig. 14.6. Relationship between haulm N content and grain yield (a) and between haulm N content and haulm yield (b) in 280 chickpea cultivars. (Data from Blümmel et al., 2012b). between stover digestibility and grain yield in trade-offs became more pronounced (r = −0.50, Kharif sorghum, while this relationship was p < 0.0001) under water restriction (Fig. 14.9a). significantly inverse in Rabi sorghum (grown in In all three crops, stover and haulm digestibility the dry season) (Fig. 14.7a). The variation in and stover and haulm yields were significantly stover digestibility in high Kharif grain yielders positively associated (Figs 14.7b and 14.8). of about 5  t/ha was close to 10% (Fig. 14.7a), The relationships between stover and which is twice the difference observed in sor- haulm digestibility and grain yield would be af- ghum stover trading situations (Fig. 14.1). fected, for example, by arrested translocation of Even in Rabi sorghum, with the overall nega- soluble carbohydrate from the stem to the grain tive association between stover digestibility and or from lignification of stems to prevent or coun- grain yield, stover digestibility among Rabi high teract lodging. While these mechanisms might grain yields of about 3.5  t/ha could vary by a be real, they were expressed only mildly in the similar magnitude. relationships of CR digestibility and grain yields In pearl millets, stover digestibility and in rice (Blümmel et  al., 2007b), groundnut grain yield were unrelated, regardless of N fertil- (Nigam and Blümmel, 2010), cowpea (Samired- izer level and population density (Fig. 14.8a). In dypalle et  al., 2017), maize (Blümmel et  al., chickpea haulm, digestibility and grain yield 2013a) and wheat (Blümmel et al., 2012a). were weakly although significantly (r = −0.13, Considerable elasticity exists between bio- p  <  0.03) associated under irrigation, but the mass yield (grain and CR) and CR fodder quality. Grain yield (kg/ha) 2 Grain yield (g/m ) Haulm yield (kg/ha) 2 Stover yield (g/m ) 498 M. Blümmel et al. (a) (b) 20,000 Kharif y = 1308 + 219x; r = 0.28; p = 0.003 7,000 Rabi y = –2845 + 163x; r = 0.32; p < 0.0001 17,500 6,000 15,000 5,000 12,500 4,000 10,000 3,000 7,500 2,000 5,000 1,000 2,500 Kharif: y = 321 + 70x; r = 0.2; p = 0.04 Rabi: y = 8176 – 115x; r = –0.55; p < 0.0001 0 0 35 38 41 44 47 50 53 56 59 62 35 38 41 44 47 50 53 56 59 62 Stover in vitro digestibility (%) Stover in vitro digestibility (%) Fig. 14.7. (a) Relationships between mean stover in vitro digestibility and grain yield in Kharif and Rabi sorghum cultivars submitted for release from 2002 to 2008. (b) Relationships between mean stover in vitro digestibility and stover yield in Kharif and Rabi sorghum cultivars submitted for release from 2002 to 2008. (Data from Blümmel et al., 2010). (a) (b) 500 600 HF HP: r = 0.65; p = 0.007 HF LP: r = 0.58; p = 0.02 LF HP: r = 0.60; p = 0.02 400 500 LF LP: r = 0.51; p = 0.04 400 300 300 200 200 100 HF HP: p = 0.86 100 HF LP: p = 0.99 LF HP: p = 0.88 LF LP: p = 0.86 0 0 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 In vitro stover digestibility (%) of stover In vitro digestibility (%) of stover Fig. 14.8. (a) Relationships between in vitro digestibility of pearl millet stover and grain yield (a) and between in vitro digestibility of pearl millet stover and stover yield under high (HF) and low (LF) fertility and high (HP) and low (LP) population density. (Data from Bidinger and Blümmel, 2007). (a) 3,500 (b) 6,000 3,000 5,000 2,500 4,000 2,000 3,000 1,500 1,000 2,000 500 1,000 Irrigation: r = –0.13; p = 0.03 Irrigation: r = 0.18; p = 0.002 Noirrigation: r = –0.50; p < 0.0001 No irrigation: r = 0.43; p < 0.0001 0 0 42.5 45.0 47.5 50.0 52.5 55.0 57.5 60.0 42.5 45.0 47.5 50.0 52.5 55.0 57.5 60.0 In vitro organic matter digestibility (%) In vitro organic matter digestibility (%) Fig. 14.9. (a) Relationship between haulm digestibility and grain yield in 280 chickpea cultivars and (b) Relationship between haulm digestibility and haulm yield in 280 chickpea cultivars. (Data from Blümmel et al., 2012b). Grain yield (kg/ha) 2Grain yield (g/m ) Grain yield (kg/ha) Stover yield (kg/ha) 2 Stover yield (g/m ) Stover yield (kg/ha) Multidimensional Crop Improvement by ILRI and Partners 499 Evidence comes from the water production func- Much of described work was conducted tion for groundnut components. Water stress within the framework of CGIAR and its national had a substantial negative effect on biomass partners. While drafting proposals for the se- yield in groundnut, while fodder quality traits cond phase of the CRPs (2017–2022), several of such as N and IVOMD were much less affected the former crop commodity programmes, such (Table 14.5). as the CRPs on Grain Legumes and Dryland Cer- eals and on Maize, specifically devoted flagships to work simultaneously for grain and CR im- provement, suggesting further mainstreaming Outcomes and Aspects of Impacts of of a paradigm shift in crop-improvement efforts. Multidimensional Crop Improvement The CGIAR nomenclature chosen for food and fodder improved cultivars was ‘full-purpose crops’. Outcomes are commonly defined by behavioural These CRPs have considerable reach as they changes and changes in mindsets by secondary work in global consortia comprising a wide beneficiaries. The work presented in this chapter range of national and international public and has contributed to such changes in both public private research organizations, development prac- and private crop improvement. The principal titioners and private-sector companies. outcome of research on multi-trait crop im- A milestone is reached when cultivar re- provement was the reconsideration of the single lease agencies start to amend release criteria that trait (i.e. grain) model in favour of the multi-trait include CR fodder traits, as has happened with and whole-plant (i.e. food and fodder) model. the AICRPs on Sorghum and recently on Pearl While there are as yet few formal decisions such Millet. Co-option and buy-in of the private sector as the decision of the NRCS (now IIMR) to in- will also be crucial. It is encouraging to see the clude stover traits as new cultivar release criteria increasing interest of the seed sector in exploring in sorghum (and now pearl millet, although marketing of CR fodder traits. The discovery, under a different mandate), there are strong in- proof-of-concept, pilot and, to a lesser degree, dications that public and private crop-improve- scale phases described above have helped to build ment programmes have reoriented their efforts a community of practice of experts and practi- towards whole-plant improvement. In the de- tioners from animal nutrition, crop improve- sign of the second phase of the CRPs, most crop ment, socio-economics and private-sector seed, commodity institutes targeted whole-plant feed and dairy companies, and from non-govern- improvement for which the expression ‘full-pur- mental organizations and NARES. This commu- pose crop’ established itself. Syngenta was nity of practice is the core around which further joined by other private breeders such as Seed Co multi-trait crop-improvement efforts need to take targeting dual-purpose maize in East and south- place. CGIAR crop institutes have well- established ern Africa, exploring branding and seed bag la- relationships and collaborations with NARES belling for CR fodder traits in their hybrids. mandated to work on specific crops. Table 14.5. Means of grain yields, CR yields, and CR N content and in vitro digestibility in groundnut and sorghum cultivars grown under water and control condition at Patancheru, India, in 2009 and 2010. (Data from Blümmel et al., 2012b, 2015a.) Characteristic Water management Groundnut Sorghum Grain yield (kg/ha) Stress 988 2542 Control 1753 3526 CR yield (kg/ha) Stress 2916 2970 Control 3840 3788 CR nitrogen (%) Stress 2.41 0.71 Control 2.23 0.72 CR digestibility (%) Stress 60.9 47.3 Control 61.6 47.5 500 M. Blümmel et al. Economic impact of multidimensional under new cultivars has obvious logistical chal- crop improvement lenges; the approaches currently being explored are around genotypic fingerprinting of new cul- Describing the adoption of new cultivars is a key tivars (Kosmowski et al., 2016). variable for estimating the impacts of crop im- The new cultivars benefit farms, fodder mar- provement. Assessing levels of adoption of new kets and livestock production. A general conclu- cultivars is usually done indirectly through the sion of our India work on dual-purpose crops is monitoring of seed production and sales and that adoption is faster and broader where the pri- crop-specific seed rates to estimate the areas vate sector is engaged. This conclusion usually planted under new cultivars (Teufel et al., 2011). applies to hybrids rather than to OPVs, where An example of the problem of measuring and seed multiplication is public. Work on multi-trait evaluating adoption is that of an early-maturing, crop improvement with OPVs identified promis- high-yielding and drought-tolerant dual-purpose ing new cultivars to scale (e.g. to more than groundnut variety (ICGV 91114) introduced in 100,000  ha) or at least to pilot (more than the Anantapur district of semi-arid India. ICGV 1000  ha), but this work was frustrated by a 91114 produced 15% higher pod yields, 17% dearth of seed (for one recent trial, just 100 g of more haulm and 11% better-quality fodder than seed of a dual-purpose legume was provided). The the locally grown variety in on-farm trials in reason for this lack of seeds in new public-sector three villages in the Anantapur district of India. OPVs might be related to misplaced incentives in Farmers who fed their dairy cows and buffaloes public-sector crop improvement, where the re- the improved fodder saw daily milk production lease and registration of new cultivars is recog- increase by about 10% per animal (Pande et al., nized rather than their adoption. Often, it would 2006, p. 23). have been necessary, even before piloting, to An impact study of 376 farmers estimated multiply seed for several years – a challenging that adopters of ICGV 91114 earned 34% add- proposition. In contrast, where private-sector hy- itional net revenue compared with traditional brids are concerned, as they are in maize, seed varieties, including a 29% gain in haulm value, availability has rarely been a constraint. while incurring unit costs that were 6% lower (Birthal et al., 2011, p. 22). A non-governmental organization, the Rural Development Trust/ Accion Fraterna, promoting the new cultivar es- The Future timated, based on seed production and sales, that by 2005 about 10,000–12,000  ha had been The traditional large-scale seed sector can bring planted with ICGV 91114. However, when Teufel hybrid crop cultivars to scale and collaborate in et  al. (2011) tried to trace this adoption using their ‘branding’ and seed labelling processes. randomly selected villages in the district, they Small- and medium-sized seed enterprises can reported only a ‘handful’ of adopters and con- move new cultivars from proof-of-concept stage cluded that the previous estimates of adoption to pilot stage by multiplying basic/foundation were dramatic overestimates. ICRISAT staff have seeds of OPVs/niche crops, often obtained from since maintained that: (i) ICGV 91114 is the NARES. Once a threshold in supply of OPV seeds third most popular cultivar in what is called is passed, farmer-to-farmer seed exchange be- ‘Breeder Seed Indented’, providing about 13% of comes significant. Small- and medium-sized feed all the groundnut seeds produced in this nation- enterprises can provide decentralized feed pro- wide scheme in India; (ii) groundnut breeders cessing and value addition to improved CRs, can e stimated a lower figure of 4% of area coverage; provide income and employment opportunities and (iii) 4% of area coverage equals about to disadvantaged rural people, and can act as a 185,000  ha under ICGV 91114 (P. Janila, Hy- ‘pull factor’ for the adoption of new cultivars. derabad, personal communication, 2016). While Large dairy enterprises using smallholder milk the estimates based on seed production and area suppliers can serve as mediators and conveyors planted are in considerable disagreement, they of new cultivars, feed intervention packages and are strongly suggestive of more than a ‘handful’ customers for existing small- and medium-sized of adopters. Making direct assessments of areas enterprises, and as stimulators of new ones. Multidimensional Crop Improvement by ILRI and Partners 501 Acknowledgements V. Padmakumar and Y. Asmare; from ICRISAT: C.T. Hash, the late F. Bidinger, S. Belum Reddy, The authors acknowledge the important contri- K.N. Rai, N. Nigam, A.A. Kumar and S. Jarial; bution of many collaborators: from ILCA/ILRI: from CIMMYT: M.T. Vinayan, O. Erenstein; and E. Zerbini, S. Fernandez Rivera, A. Ayantunde, from NRSC/DSR/IIMR: N. Seetharama, V.A. Tonapi E. Grings, K.V.S.V. Prasad, D. Ravi, A.A. Khan, and V. Bhat. References Alkhtib, A.S., Wamatu, J.A., Wegi, T. and Rischkowsky, B.A. (2016) Variation in the straw traits of morpho- logical fractions of faba bean (Vicia faba L.) and implications for selecting for food-feed varieties. Animal Feed Science and Technology 222, 122–131. Alkhtib, A.S., Wamatu, J., Ejeta, T.T. and Rischkowsky, B. (2017) Integrating straw yield and quality into multi-dimensional improvement of lentil (Lens culinaris). Journal of the Science of Food and Agricul- ture 97, 4135–4141. Anandan, S., Khan, A.A., Ravi, D., Jeethander, R. and Blümmel, M. (2010) A comparison of sorghum stover based complete feed blocks with a conventional feeding practice in a peri urban dairy. Animal Nutrition and Feed Technology 10, 23–28. Anandan, S., Khan, A., Ravi, D., Rao, M.S.B., Reddy, Y.R. and Blümmel, M. (2013) Identification of a super- ior dual purpose maize hybrid among widely grown hybrids in South Asia and value addition to its stover through feed supplementation and feed processing. Field Crops Research 153, 52–57. Ayantunde, A.A., Blummel, M., Grings, E. and Duncan, A.J. (2014) Price and quality of livestock feeds in suburban markets of West Africa’s Sahel: case study from Bamako, Mali. Revue d’Elévage et de M édecine Vétérinaire des Pays Tropicaux 67, 13–21. Baudron, F., Jaleta, M., Okitoi, O. and Tegegne, A. (2014) Conservation agriculture in African mixed crop– livestock systems: expanding the niche. Agriculture, Ecosystems and Environment 187, 171–182. Beckmann, E. (1921) Conversion of grain straw and lupins into feeds of high nutrient value. Festschrift der Kaiser Wilhelm-Gesellschaft zur Förderung der Wissenschaften 16, 18–26. Berhanu, G., Adane, H. and Kahsay, B. (2009) Feed marketing in Ethiopia: results of a rapid market appraisal. IPMS Working Paper 15. ILRI, Addis Ababa. Bidinger, F.R. and Blümmel, M. (2007) Effects of ruminant nutritional quality of pearl millet [Pennisetum glaucum (L.) R. Br.] stover: I. Effects of management alternatives on stover quality and productivity. Fields Crops Research 103, 129–138. Bidinger, F.R., Blümmel, M., Hash, C.T. and Choudhary, S. (2010) Genetic enhancement for superior food- feed traits in a pearl millet (Pennisetum glaucum (L.) R. Br.) variety by recurrent selection. Animal Nutrition and Feed Technology 10, 61–68. Birthal, P.S., Nigam, S.N., Narayanan, A.V. and Kareem, K. (2011) An economic assessment of the potential benefits of breeding for drought tolerance in crops: a case of groundnut in India. Research Bulletin No. 25. ICRISAT, Andhra Pradesh, India. Blümmel, M. and Rao, P.P. (2006) Economic value of sorghum stover traded as fodder for urban and peri- urban dairy production in Hyderabad. India International Sorghum and Millet Newsletter 47, 97–100. Blümmel, M., Bidinger, F.R. and Hash, C.T. (2007a) Management and cultivar effect on ruminant nutritional quality of pearl millet [Pennisetum glaucum (L.) R. Br.] stover. Effects of cultivar choice on stover qual- ity and productivity. Fields Crops Research 103, 119–128. Blümmel, M., Virk, P. and Xianglin, L. (2007b) Opportunities for improving the fodder value of rice straw by multidimensional crop improvement. In: Wright, I.A (ed.) Proceedings of the International Symposium on Production and Utilization of Paddy Rice as Feed, 9–12 October, Japan, China, Korea. pp. 17–21. Blümmel, M., Vishala, A., Ravi, D., Prasad, K.V.S.V., Ramakrishna Reddy, C. and Seetharama, N. (2010) Multi-environment investigations of food–feed trait relationships in Kharif and Rabi sorghum [Sorghum bicolor (L.) Moench] over several years of cultivars testing in India. Animal Nutrition and Feed Technology 10, 11–21. Blümmel, M., Joshi, A.K., Teufel, N. and Wright, I.A. (2012a) Looking beyond grain for overall benefit from wheat in mixed crop livestock systems. Presented at the Wheat for Food Security in Africa Conference, 8–12 October, Addis Ababa. ILRI, Nairobi. 502 M. Blümmel et al. Blümmel, M., Ratnakumar, P. and Vadez, V. (2012b) Opportunities for exploiting variations in haulm fodder traits of intermittent drought tolerant lines in a reference collection of groundnut (Arachis hypogaea L). Field Crops Research 126, 200–206. Blümmel, M., Grings, E.E. and Erenstein, O. (2013a) Potential for dual-purpose maize varieties to meet changing maize demands: synthesis. Field Crops Research 153, 107–112. Blümmel, M., Homann-Kee, T., Valbuana, S., Diego, V., Duncan, A.J. and Herrero, M. (2013b) Biomass in crop–livestock systems in the context of the livestock revolution. Sécheresse 24, 330–339. Blümmel, M., Haileslassie, A., Samireddypalle, A., Vadez, V. and Notenbaert, A. (2014a) Livestock water productivity: feed resourcing, feeding and coupled feed–water resource databases. Animal Production Science 54, 1584–1593. Blümmel, M., Zaidi, P., Vinayan, M., Babu, R. and Yadav, O. (2014b) The fodder value of maize stover vis-à-vis other cereals residues and opportunities for improvement through crop breeding and value addition. In: 12th Asian Maize Conference and Expert Consultation on Maize for Food, Feed, Nutrition and Environmental Security, 30 October–1 November, Bangkok, Thailand. CIMMYT, Mexico, and APAARI, Bangkok, pp. 82–90. Blümmel, M., Deshpande, S., Kholova, J. and Vadez, V. (2015a) Introgression of staygreen QTLs for concomi- tant improvement of food and fodder traits in Sorghum bicolor. Field Crops Research 180, 228–237. Blümmel, M., Haileslassie, A., Herrero, M., Beveridge, M., Phillips, M. and Havlik, P. (2015b) Feed re- sources vis-a-vis livestock and fish productivity in a changing climate. In: Malik, P.K., Bhatta, R. Taka- hashi, J., Kohn, R. and Prasad, C.S. (eds) Livestock Production and Climate Change. CAB Inter- national, Wallingford, UK, pp. 8–24. Blümmel, M., Muller, A., Schader, C., Herrero, M. and Garg, M.R. (2017) The use and abuse of cereals, leg- umes and crop residues in rations for dairy cattle. In: Webster, J. (ed.) Achieving Sustainable Production of Milk, Vol. 3: Dairy Herd Management and Welfare. Burleigh Dodds Science Publishing, Cambridge. Capper, B.S., Thomson, E.F. and Herbert, F. (1988) Genetic variations in the feeding value of barley and wheat straw. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 177–193. Choudhary, S., Hash, C.T., Sagar Prem, Prasad, K.V.S.V. and Blümmel, M. (2010) Near infrared spectros- copy estimation of pearl millet grain composition and feed quality. Proceedings of the 14th International Conference on NIR Spectroscopy. Bangkok, Thailand. Choudhary, S., Bidinger, F.R., Hash, C.T., Vadez, V. and Blummel, M. (2012) Gene action governing pearl millet stover nitrogen and in vitro digestibility and opportunities for improvement. European Journal of Plant Science and Biotechnology 6, 129–131. Dai, J., Bean, B., Brown, B., Bruening, W., Edwards, J., et al. (2016) Harvest index and straw yield of five classes of wheat. Biomass and Bioenergy 85, 223–227. de Groote, H., Dema, G., Sonda, G. and Gitonga, Z. (2013) Maize for food and feed in East Africa – the farmer perspective. Field Crops Research 153, 22–36. Duncan, A.J., Bachewe, F., Mekonnen, K., Valbuena, D., Rachier, G., et al. (2016) Crop residue allocation to livestock feed, soil improvement and other uses along a productivity gradient in Eastern Africa. Agriculture, Ecosystems and Environment 228, 101–110. Ertiro, B.T., Zeleke, H., Friesen, D., Blümmel, M. and Twumasi-Afiyie, S. (2013) Relationships between the performance of parental inbred lines and hybrids for food-feed traits in maize (Zea mays L.) in Ethiopia. Field Crops Research 153, 86–93. Etela, I. and Dung, D.D. (2011) Utilization of stover from six improved dual-purpose groundnut (Arachis hypogaea L.) cultivars by West African dwarf sheep. African Journal of Food, Agriculture, Nutrition and Development 11, 4538–4545. Fingerling, G. and Schmidt, K. (1919) Die Strohaufschliessung nach dem Beckmannschen Verfahren. 2. Einfluss der Aufschliessungszeit auf den Umfang der Nährwerterschliessung. Landw. Versuchsst. Sonderdruck, 115–152. Hay, R.K.M. (1995) Harvest index: a review of its use in plant breeding and crop physiology. Annals of Applied Biology 126, 197–216. Jackson, M.G. (1977) The alkali treatment of straws. Animal Feed Science and Technology 2, 105–130. Jarial, S., Blümmel, M., Soumana, I., Prasad, K.V.S.V., Issa, S., Ravi, D.K. (2016a) Price-Quality Relation- ships in Fodder Trading in Niger with Special Regards to Comparisons of Cowpea and Groundnut Haulms with Concentrates and Collected Shrubs and Grasses. In: Pan-African Grain and World Cowpea Conference, 28 February–4 March, Livingstone, Zambia. ICRISAT, Niamey. Multidimensional Crop Improvement by ILRI and Partners 503 Jarial, S., Blümmel, M., Soumana, I., Ravi, D.K., Issa, S., Whitbread, A. and Tabo, R. (2016b) Comparison of Cowpea and Groundnut Haulm Trading in Urban and Rural Fodder Markets in Niger. In: Pan-African Grain and World Cowpea Conference, 28 February–4 March, Livingstone, Zambia. ICRISAT, Niamey. Kelley, T.G., Rao, P.P. and Walker, T.S. (1993) The relative value of cereal straw fodder in India: implications for cereal breeding programs at ICRISAT. In: Dvorak, K.A. (ed.) Social Science Research for Agricul- tural Technology Development. CAB International, Wallingford, UK. Kelley, T.G., Rao, R.P., Weltzien, R.E. and Purohit, M.L. (1996) Adoption of improved cultivars of pearl millet in arid environment: straw yield and quality considerations in western Rajasthan. Experimental Agricul- ture 32, 161–172. Khush, G.S., Bienvenido, B.O. and Roxas, D.B. (1988) Genetic selection for improved nutritional quality of rice straw – a plant breeder’s viewpoint. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 261–282. Kosmowski, F., Aragaw, A., Kilian, A., Ambel, A., Ilukor, J., et al. (2016) Varietal identification in household surveys: results from an experiment using DNA fingerprinting of sweet potato leaves in Southern Ethiopia. Policy Research Working Paper No. WPS 7812. World Bank Group, Washington, DC. Kossila, V. (1988) The availability of crop residues in developing countries in relation to livestock populations. ILCA, Addis Ababa. Kristjanson, P.M. and Zerbini, E. (1999) Genetic enhancement of sorghum and millet residues fed to ru- minants. An ex ante assessment of returns to research. ILRI Impact Assessment Series. No. 3. ILRI, Nairobi. Lal, R. (2005) World crop residues production and implications of its use as a biofuel. Environment Inter- national 31, 575–584. Mayberry, D., Ash, A., Prestwidge, D., Godde, C.M., Henderson, B., et al. (2017) Yield gap analyses to es- timate attainable bovine milk yields and evaluate options to increase production in Ethiopia and India. Agricultural Systems 155, 43–51. McDowell, R. (1988) Truth conditions, bivalence, and vericationism. In: Meaning, Knowledge, and Reality. Harvard University Press, Cambridge, Massachusetts, pp. 3–28. McIntire, J., Reed, J.D., Jutzi, S.C., Tedla, A. and Kebede, Y. (1988) Evaluating sorghum cultivars for grain and straw yield. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 283–300. Mueller-Harvey, I., McAllan, A.B., Theodorou, M.K. and Beever, D.E. (1988) Phenolics in fibrous crop residues and plants and their effect on the digestion and utilisation of carbohydrates and proteins in ruminants. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 97–132. Nepolean, T., Blum̈ mel, M. and Hash, C.T. (2009) Improving straw quality traits through QTL mapping and marker-assisted selection in pearl millet. In: Forage Symposium 2009: Imaging Trends in Forage Research and Livestock Production, 16–17 February, CAZRI RRS, Jaisalmer, Rajasthan, India, p. 12. Nigam, S.N. and Blümmel, M. (2010) Cultivar-dependent variation in food-feed-traits in groundnut (Arachis hypogaea L.) (2010) Animal Nutrition and Feed Technology 10, 39–48. Nordblom, T.L. (1988) The importance of crop residues as feed resources in West Asia and North Africa. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Res- idues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 41–64. Ørskov, E.R. (1988) Consistency of differences in nutritive value of straw from different varieties in different seasons. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 163–176. Owen, E. and Aboud, A.A.O. (1988) Practical problems of feeding crop residues. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 133–156. Owen, E. and Jayasuriya, M.C.N. (1989) Use of crop residues as animal feeds in developing countries. Research and Development in Agriculture 6, 129–138. Pande, S., Sreenivas, B., Rao, P., Rao, J. and Reddy, P. (2006) Farmers’ Participatory Management of Dis- eases for Higher Yield and Nutritive Value of Crop Residues of Groundnut, Deccan Plateau, India. 504 M. Blümmel et al. Proceedings of a Workshop, 3–4 January 2005, ARS, ANGRAU, Rekulakunta, Anantapur, Andhra Pradesh. ICRISAT, Andhra Pradesh, India. Pearce, G., Lee, J., Simpson, R. and Doyle, P. (1988) Sources of variation in the nutritive value of wheat and rice straws. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp.195–231. Prasad, K.V.S.V., Khan, A.A., Vellaikumar, S., Devulapalli, R., Ramakrishna, R.C., et al. (2010) Observa- tions on livestock productivity in sheep fed exclusively on haulms from ten different genotypes of groundnut. Animal Nutrition and Feed Technology 10, 121–126. Prasad, K.V.S.V., Rashid, M.M., Sharda, P., Ravi, D., Ramana Reddy, Y., et al. (2015) Handheld mobile near infrared spectroscopy – a rapid tool for quality evaluation of feeds. Abstract of a presentation at the 9th Biennial Conference of the Animal Nutrition Association, 22–24 January, Guwahati, India. Ramakrishna, R.C., Khan, A.A., Prasad, K.S.V.S., Ravi, D., Reddy, B.S.V., et al. (2010) Stover quality traits for improvement of dual-purpose sorghum. Animal Nutrition and Feed Technology 10, 113–119. Ravi, D., Anandan, S., Khan, A.A., Bidinger, F.R., Nepolean, T., et al. (2010) Morphological, chemical and in vitro traits for prediction of stover quality in pearl millet for use in multidimensional crop improve- ment. Animal Nutrition and Feed Technology 10, 49–59. Ravi, D., Khan, A.A., Rao, M.S. and Blümmel, M. (2013) A note on suitable laboratory stover quality traits for multidimensional maize improvement. Fields Crops Research 153, 58–62. Reed, J.D., Capper, B. and Neate, P.J.H. (eds) (1988a) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa. Reed, J.D., Kebede, Y. and Fussell, L.K. (1988b) Factors affecting the nutritive value of sorghum and millet crop residues. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, E thiopia. ILCA, Addis Ababa, pp. 233–251. Reed, J.D., Soller, H. and Woodward, A. (1990) Fodder tree and straw diets for sheep: intake, growth, digestibility and the effects of phenolics on nitrogen utilisation. Animal Feed Science and Technology 30, 42371–42371. Samireddypalle, A., Boukar, O., Grings, E., Fatokun, C., Prasad, K.V.S.V., et al. (2017) Cowpea and ground- nut haulms fodder trading and its lessons for multi-trait cowpea improvement for mixed crop livestock systems in West Africa. Frontiers of Plant Science 8, 30. Sanchez, O.J. and Cardona, C.A. (2008) Trends in biotechnological production of fuel ethanol from different feedstocks. Bioresource Technology 99, 5270–5295. Shah, L. (2007) Delivering nutrition. Power Point Presentation delivered at the CIGAR System Wide Live- stock Program Meeting, 17 September, ICRISAT, Patancheru. Sharma, K., Pattanaik, A.K., Anadan, S. and Blümmel, M. (2010) Food–feed crop research: a synthesis. Animal Nutrition and Feed Technology 10, 1–10. Singh, O.P., Sharma, A., Singh, R. and Shah, T. (2004) Virtual water trade in dairy economy. Economic and Political Weekly 39, 3492–3497. Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M. and de Haan, C. (2006) Livestock’s Long Shadow: Environmental Issues and Options. FAO, Rome. Swanepoel, F., Stroebel, A. and Moyo, S. (2010) The Role of Livestock in Developing Communities: Enhan- cing Multifunctionality. University of the Free State and CTA, Cape Town. Teufel, N., Samaddar, A., Blümmel, M. and Erenstein, O. (2010) Quality characteristics of wheat and rice straw traded in Indian urban centres. Presentation at Tropentag 2010, World Food System – A Contri- bution from Europe, 14–16 September, Zurich. Teufel, N., Johnson, N. and Singh, D.K. (2011) The adoption and impact of an improved drought-tolerant, dual-purpose groundnut variety in Southern India. Paper presented at the 7th ASAE International Conference ‘Meeting the Challenges Facing Asian Agriculture and Agricultural Economics Toward a Sustainable Future,’ 13–15 October, Hanoi, Vietnam. ILRI, Nairobi. Thornton, P.K., Kristjanson, P.M. and Thorne, P.J. (2003) Measuring the potential impacts of improved food–feed crops: methods for ex ante assessment. Field Crops Research 84, 199–212. Traxler, G. and Byerlee, D. (1993) A joint-product analysis of the adoption of modern cereal varieties in developing countries. American Journal of Agricultural Economics 75, 981–989. Valbuena, D., Tui, S.H.-K., Erenstein, O., Teufel, N., Duncan, A., et al. (2015) Identifying determinants, pres- sures and trade-offs of crop residue use in mixed smallholder farms in Sub-Saharan Africa and South Asia. Agricultural Systems 134, 107–118. Multidimensional Crop Improvement by ILRI and Partners 505 van Soest, P.J. (1988) Effect of environment and quality of fibre on the nutritive value of crop residues. In: Reed, J.D., Capper, B.S. and Neate, P.J.H. (eds) Plant Breeding and the Nutritive Value of Crop Res- idues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 71–96. van Soest, P.J. (1994) The Nutritional Ecology of the Ruminant, 2nd edn. Cornell University Press, New York. Venkatesh, B.B., Umkanth, A.V., Madhusudhana, R., Vishala, A.D., Ravi, D., et  al. (2006) Genetic enhancement of sorghum stover quality: collaborative research between NRCS and ILRI. Jowar Samacha 2, 6–8. Vinayan, M.T., Raman, B., Jyothsna, T., Zaidi, P.H. and Blümmel, M. (2013) A note on potential candidate genomic regions with implications for maize stover fodder quality. Fields Crops Research, 153, 102–106. Wegi, T. (2016) Effects of feeding different varieties of faba bean (Vicia faba L.) straws with concentrate on feed intake, digestibility, body weight gain and carcass characteristics of Arsi-Bale sheep. MSc the- sis, Haramaya University, Dire Dawa, Ethiopia. Zaidi, P.H., Vinayan, M.T. and Blümmel, M. (2013) Genetic variability of tropical maize stover quality and the potential for genetic improvement of food-feed value in India. Field Crops Research 153, 79–85. Preface to Part III: Research Spending and Publications on Tropical Livestock Systems This preface first shows the estimated spending production, 50%; (ii) livestock systems, 11%; in the principal domains corresponding to four (iii) economics and policy, 7.6%; (iv) primary chapters in Part III – livestock systems, climate production, 4.1%; and (v) management, tech- change and livestock in the tropics, economics nical support and capacity development, 27%. and policy research and gender impact (www. ILRAD investment in systems, economics and ilri.org/ImpactBook/Finance). policy was limited to its role in the trypanotoler- The preface then presents ‘scientific impact’ ance network with ILCA before 1987, when as a function of publications and citations extracted ILRAD’s veterinary epidemiology unit began. from the Scopus database using search keywords ILRAD spending on systems, economics and relevant to the four domains. p olicy was less than 3% of its 1975–1994 total. Total spending by ILRI was US$ 1.75 billion. Spending by domain over ILRI’s lifetime was: Research spending (i) animal genetics, production and health, 39%; (ii) livestock systems, 10%; (iii) economics and Data from the financial and annual reports of policy, 11%; (iv) primary production, 6.4%; and the International Livestock Centre for Africa (v) management, technical support and cap- (ILCA), International Laboratory for Research acity development, 34%. on Animal Diseases (ILRAD) and International Livestock Research Institute (ILRI) were used to compile a spending database for 1975–2018 Altmetrics (www.ilri.org/ImpactBook/Finance). Current spending for each year and institution was assigned An Altmetric search (www.altmetric.com/; ac- to scientific domains using spending detail by cessed 13 March 2020) was carried out on the project, by scientists’ fields of expertise and, keywords ‘livestock farming’, ‘farming system’, occasionally, from cost accounting by the insti- ‘mixed system’, ‘pastoralism’ and ‘grazing’ as in- tutions. Current annual spending in US$ was dicators of LSR content. ILRI work dominates converted to constant annual spending in 2015 the research (as measured by papers and cit- US$ using the global Manufacturers’ Unit Value ations) on sub-Saharan Africa but is typically Index (Fig. PIII.1). small as a share of global research (Table PIII.1). Total 1975–1994 spending by ILCA and An initial Altmetric analysis was done for the ILRAD was US$636 million. Mean spending by systems domain1 on the expressions ‘livestock domain at ILCA and ILRAD as a share of total farming’, ‘mixed farming’, ‘farming system’ and spending was: (i) animal genetics, health and ‘crop–livestock’ for papers relating to research in 507 508 Research Spending and Publications on Tropical Livestock Systems 400 200 0 1975–80 1981–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Period Livestock systems Animal health and genetics Domain Management, technical, capacity development Economics and policy Plant sciences Total spending Fig. PIII.1. ILRI economics and policy research share of total ILRI spending, 1975–2018. (Data from ILCA/ ILRAD/ILRI Annual and Financial Reports.) Table PIII.1. Results of the Altmetric search for ILRI livestock systems and related studies. (Data from: www.altmetrics.com/explorer, accessed 13 March 2020.) Keywords Region Share of papers (%) Share of citations (%) ‘Livestock farming’ Global (n=310) 8.7 3.4 Sub-Saharan Africa (n=27) 55.2 67.3 ‘Farming system’ Global (n=1,453) 8.1 6.3 Sub-Saharan Africa (n=118) 36.9 31.8 ‘Mixed farming’ Global (n=126) 15.1 18.9 Sub-Saharan Africa (n=19) 64.0 64.7 ‘Pastoralism’ + ‘grazing’ Global (n=3,878) 2.5 2.3 Sub-Saharan Africa (n=258) 20.5 15.5 sub-Saharan Africa. The Altmetric analysis does 2011. It does show substantial ILRI contribu- not generally allow analysis before the advent tions at the global level, where papers having at of the CGIAR research programmes (i.e. before least one ILRI author of any rank represented 2011); Altmetric therefore excludes older ILCA between 2.5% (‘pastoralism’ plus ‘grazing’) and and ILRAD papers and many ILRI papers before 18.5% (‘mixed farming’); citations from these Spending in million of 2015 US$ Research Spending and Publications on Tropical Livestock Systems 509 papers at the global level ranged from 2.3% this volume). The sum of livestock systems re- (‘pastoralism plus ‘grazing’) to 18.9% (‘mixed search, which often had a policy component, farming’). Notable work with an ILRI author in plus economics and policy research during the these areas was Herrero et al. (2010) and Thorn- ILCA era was approximately US$120 million or ton and Herrero (2015) on adaptation to cli- 19% of the 1975–1994 ILCA/ILRAD total. The mate among mixed smallholders in Africa, Giller scientific return on this investment was 26 cit- et al. (2011) on soil fertility management (‘farm- ations per paper; this return was skewed with a ing system’) and Silvestri et al. (2012) on perception median of 11 and the top ten papers had 49% of of climate change among Kenyan pastoralists all citations in a sample of 196 papers. Papers to (‘pastoralism + grazing’). which ILCA staff contributed generated about 46% of a global sample of livestock systems re- search papers before 1995 and about 53% of the Spending on livestock systems research citations, and clearly had a powerful influence in and the return in publications describing the rationale and operation of Afri- can livestock production (Fig. PIII.2). ILCA spending on livestock systems was some ILRI lifetime spending on livestock systems US$71 million from 1975 to 1994, or 11% of the has been about US$177 million since 1975, ILCA/ILRAD total of US$636 million (see Chapter 15, or 10% of the 1975–2018 total of US$1.75 Papers, before merger into ILRI Papers, after merger into ILRI 5,000 300 4,000 3,000 200 2,000 100 1,000 0 0 1977–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Citations, before merger into ILRI Citations, after merger into ILRI 60,000 7,500 40,000 5,000 2,500 20,000 0 0 1977–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Africa Global Fig. PIII.2. ILRI papers in systems research in Africa and globally, 1977–2018. ILRI papers before merger = 196 and after merger = 889. Africa papers = 207 and 2,475; global papers = 895 and 10,058. (Data from www.scopus.com/.) Counts by publication period 510 Research Spending and Publications on Tropical Livestock Systems billion. (The sum of economics and policy work with the exception of King (1983) on water, plus livestock systems research, which nearly al- have been published since 2003. Over the life- ways made some policy recommendations, was time of ILRI (1975–2018), scientists produced a about US$375 million, or 21% of the 1975–2018 sample of 568 papers on some aspect of climate total.) Mean citations per paper were 19, with a and agriculture, generating a mean of 36 citations median of 8; the top ten papers had 17% of all per paper, a median of 11 (Fig. PIII.4) and with citations in a sample of 1,085 papers. The major the top ten cited papers producing 24% of total systems papers tended to be older, including the ILRI citations in this domain. Climate change work of Sandford (1983) on pastoral develop- papers tend to involve new tools (such as the rain- ment, Wilson (1986) on central Mali, Solomon fall data generator MarkSim), or global mapping Bekure et al. (1991) on Kenyan Maasailand, Cop- of vulnerability and impact. Adaptation to climate pock (1994) on Borana in Ethiopia, McIntire change appears as a theme more frequently et al. (1992) on sub-Saharan Africa, and Hier- than mitigation of climate change effects, and naux and Ayantunde (2004) on the Fakara sub- ILRI has made a major global contribution in region of Niger. Livestock systems work evolved areas related to livestock and adaptation to after 1990 to include such landmark pieces as climate change. Norval et al. (1992) on the field epidemiology of Theileria spp., Reid et al. (2000) on land use in Ethiopia, Kruska et al. (2003) on mapping global livestock systems, Reid et al. (2008) on broader Economics and policy issues in pastoralism, including interactions with wildlife, and Robinson et al. (2011) who mapped ILCA/ILRAD spending on economics and policy poverty measures to global livestock systems. research was about US$48 million between Papers produced by ILCA/ILRAD/ILRI staff, or 1975 and 1994, or 8% of the total (see Chapter in close collaboration with them, contributed 17, this volume). From a sample of 86 papers in about 29% of Africa-wide livestock systems re- this domain, the mean of citations per paper in search and about 31% of the citations and have economics and policy research was 40, the me- clearly had a powerful influence on policy and dian was 13 and the top ten papers contributed development interventions since 1975. 70% of ILCA/ILRAD citations (Fig. PIII.3). ILRI lifetime spending on economics and policy research has been about US$198 million since 1975, or 11.3% of the 1975–2018 total of Climate change US$ 1.75 billion. From a sample of 947 ILRI pa- pers in this domain, the mean of citations per Climate change research has only become a paper was 34, the median was 10 and the top major part of ILRI research since about 2000 ten papers contributed 24% of ILRI citations in (see Chapter 16, this volume). It was not possible this domain (Fig. PIII.3). The scientific return on to estimate spending on climate research, given spending was 5 papers per US$ million and 161 that it is a transversal field, typically matched citations per US$ million over the long period of against several fields in the citations databases 1975–2018. and was not, until recently, accounted separ- The older papers on economics and policy ately. In the ILCA/ILRAD era, a sample of 48 pa- were typically microeconomic, notably in making pers classed as having some element of ‘climate benefit–cost estimates of technologies, evaluat- and agriculture’ research generated a mean of ing public spending on agriculture, or testing 44 citations per paper, a median of 15 and with efficiency in factor and product markets (McCarthy the top ten papers producing 77% of all citations et al., 1999, as discussed in Chapter 17, this vol- (Fig. PIII.4). ume). Major economics and policy papers in this After 1999, there was a burst of work on century took a more macro view, and are usu- climate, crop and livestock interactions led by ally of the characterization type, including the Philip K. Thornton and Mario Herrero. All of the work of Delgado et al. (1999) on the ‘Livestock major ILRI-affiliated papers on climate change, Revolution’, Thornton et al. (2002) on poverty Research Spending and Publications on Tropical Livestock Systems 511 Papers, before merger into ILRI Papers, after merger into ILRI 75 1,000 50 500 25 0 0 1977–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Citations, before merger into ILRI Citations, after merger into ILRI 1,500 15,000 1,000 10,000 500 5,000 0 0 1977–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Africa Global Fig. PIII.3. ILRI papers cited in economics and policy research on Africa and global problems related to livestock systems, 1977–2018. ILRI papers before merger = 86; after merger = 861. Africa papers = 45 and 644; global papers = 154 and 2,484. (Data from www.scopus.com/.) and livestock development, Perry et al. (2002) on has been quite productive relative to global animal health research and poverty alleviation, (Fig. PIII.5a) and African efforts (Fig. PIII.5b). and Herrero et al. (2010) on sustainable live- (This subsample was limited to papers having stock investments. Two ILRI projects – one leading at least ten citations.) One measure of major to beneficial reforms in Kenyan dairy policy ILRI scientific contributions is the share of (Kaitibie et al., 2010) and another that created a ILRI papers in the top 5% of citations in a given new insurance instrument (Chantarat et al., 2013) field, compared with the share of all papers – and Jensen et al. (2019) – had important devel- ILRI plus global – in the top 5%. This prod- opment impacts. uctivity is notable in the best cited papers for economics and policy, where ILRI-associated High citation papers work was 32% of all global citations and 27% of citations in the top 5%. For livestock systems, ILRI research in livestock systems, economics and the corresponding shares were 10% and 9%; for policy, and in climate change related to agriculture climate, they were 18% and 15%. Counts 512 Research Spending and Publications on Tropical Livestock Systems Papers, before merger into ILRI Papers, after merger into ILRI 2,500 30 2,000 1,500 20 1,000 10 500 0 0 1977–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Citations, before merger into ILRI Citations, after merger into ILRI 30,000 1,000 20,000 500 10,000 0 0 1977–85 1986–90 1991–94 1995–00 2001–05 2006–10 2011–18 Publication period Institution ILRI Africa Global Fig. PIII.4. ILRI and other publications in climate change research related to livestock systems have become widely cited since 2000, 1977–2018. ILRI papers before merger = 48; and after merger = 520. Africa papers = 13 and 1,010; global papers = 56 and 3,276. (Data from www.scopus.com/.) Counts by publication period Research Spending and Publications on Tropical Livestock Systems 513 (a) Climate, before merger into ILRI Climate, after merger into ILRI 300 10 200 3 100 0 0 first second third fourth top 5% first second third fourth top 5% Economics and policy, before merger into ILRI Economics and policy, after merger into ILRI 20 15 200 10 100 5 0 0 first second third fourth top 5% first second third fourth top 5% Livestock systems, before merger into ILRI Livestock systems, after merger into ILRI 1,200 90 900 60 600 30 300 0 0 first second third fourth top 5% first second third fourth top 5% Citation quantiles Institution ILRI Global (b) Climate, before merger into ILRI Climate, after merger into ILRI 10 7.5 100 5 50 2.5 0 0 first second third fourth top 5% first second third fourth top 5% Economics and policy, before merger into ILRI Economics and policy, after merger into ILRI 20 90 15 10 60 5 30 0 0 first second third fourth top 5% first second third fourth top 5% Livestock systems, before merger into ILRI Livestock systems, after merger into ILRI 40 300 30 200 20 100 10 0 0 first second third fourth top 5% first second third fourth top 5% Citation quantiles Institution ILRI Africa Fig. PIII.5 (a) Frequency of papers of ILRI and other institutions in global systems research, 1977–2018. ILRI climate papers before merger = 33; after merger = 289. Global climate papers, 31; 1,311. ILRI livestock systems papers = 107 and 407; global livestock systems papers = 398 and 4,478; ILRI economics and policy papers = 56 and 447; global economics and policy papers = 57 and 1,001. (b) Frequency of papers of ILRI and other institutions in African systems research, 1,977–2018. ILRI climate papers before merger = 33; after merger = 289. Global climate papers, 31, 1,311. ILRI livestock systems papers = 107 and 407; Africa livestock systems papers = 83 and 1,119; ILRI economics and policy papers = 56 and 447; Africa economics and policy papers = 16 and 269. (Data from www.scopus.com/.) Counts of papers > 9 citations Papers > 9 citations 514 Research Spending and Publications on Tropical Livestock Systems Note 1 The Altmetric analysis returned very few matches for ‘climate change’ and related terms, compared with the Scopus database, and is therefore not reported. References Chantarat, S., Mude, A.G., Barrett, C.B. and Carter, M.R. (2013) Designing index-based livestock insurance for managing asset risk in northern Kenya. Journal of Risk and Insurance 80, 205–237. Coppock, D.L. (1994) The Borana Plateau of Southern Ethiopia: Synthesis of Pastoral Research, Development and Change, 1980–91. Systems Study No. 5. ILCA, Addis Ababa. Delgado, C.L., Rosegrant, M., Steinfeld, H., Ehui, S. and Courbois, B. (1999) Livestock to 2020: the next food revolution. Food, Agriculture and the Environment Discussion Paper No, 28. IFPRI, Washington, DC, FAO, Rome, and ILRI, Nairobi. Giller, K.E., Tittonell, P., Rufino, M.C., van Wijk, M.T., Zingore, S., et al. (2011) Communicating complexity: integrated assessment of trade-offs concerning soil fertility management within African farming systems to support innovation and development. Agricultural Systems 104, 191–203. Hiernaux, P. and Ayantunde, A. (2004) The Fakara: a semi-arid agro-ecosystem under stress. Report of research activities of International Livestock Research Institute (ILRI) in Fakara, South-western Niger, between 1994 and 2002, submitted to Desert Margins Program. ICRISAT, Niamey, Niger. Herrero, M., Thornton, P.K., Notenbaert, A.M., Wood, S., Freeman, H.A., et al. (2010) Smart Investments in Sustainable Food Production: Revisiting Mixed Crop-Livestock Systems. Science 327, 822–825. Jensen, N., Stoeffler, Q., Fava, F., Vrieling, A., Atzberger, C., et al. (2019) Does the design matter? Comparing satellite-based indices for insuring pastoralists against drought. Ecological Economics 162, 59–73. Kaitibie, S., Omore, A., Rich, K. and Kristjanson, P. (2010) Kenya dairy policy change and economic importance. World Development 38, 1494–1505. King, J.M. (1983) Livestock water needs in pastoral Africa in relation to climate and forage. ILCA Research Report No. 7. ILCA, Addis Ababa. Kruska, R.L., Reid, R.S., Thornton, P.K., Henninger, N. and Kristjanson, P. (2003) Mapping livestock-oriented agricultural production systems for the developing world. Agricultural Systems 77, 39–63. McCarthy, N., Swallow, B., Kirk, M. and Hazell, P. (eds) (1999) Property Rights, Risk, and Livestock Development in Africa. ILRI, Nairobi, and IFPRI, Washington, D.C. McIntire, J., Bourzat, D. and Pingali, P. (1992) Crop–Livestock Interaction in Sub-Saharan Africa. World Bank, Washington, D.C. Norval, R.A.I., Perry, B.D. and Young, A.S. (1992) The Epidemiology of Theileriosis in Africa. Academic Press, London. Perry, B.D., Randolph, T.F., McDermott, J.J., Sones, K.R. and Thornton, P.K. (2002) Investing in animal health research to alleviate poverty. ILRI, Nairobi. Reid, R.S., Kruska, R.L., Muthui, N., Taye, A., Wotton, S., et al. (2000) Land-use and land-cover dynamics in response to changes in climatic, biological, and socio-political forces: The case of southwestern Ethiopia. Landscape Ecology 15, 339–355. Reid, R., Galvin, K. and Kruska, R. (2008) Global significance of extensive grazing lands and pastoral societies: an introduction. In: Galvin, K.A., Reid, R.S. Behnke, R.H. Jr and Thompson Hobbs, N. (eds) Fragmentation in Semi-arid and Arid Landscapes. Springer, New York, pp. 1–25. Robinson, T., Thornton, P., Franceschini, G., Kruska, R., Chiozza, F., et al. (2011) Global Livestock Production Systems. FAO, Rome, and ILRI, Nairobi. Sandford, S. (1983) Management of Pastoral Development in the Third World. Wiley, Chichester, UK. Silvestri, S., Bryan, E., Ringler, C., Herrero, M. and Okoba, B.O. (2012) Climate change perception and adaptation of agro-pastoral communities in Kenya. Regional Environmental Change 12, 791–802. Solomon Bekure, de Leeuw, P.N., Grandin, B.E. and Neate, P.J.H. (1991) Maasai Herding: An Analysis of the Livestock Production System of Maasai Pastoralists in Eastern Kajiado District, Kenya. ILCA, Addis Ababa. Thornton, P.K., Kruska, R.L., Henninger, N., Kristjanson, P.M., Reid, R.S., et al. (2002) Mapping Poverty and Livestock in the Developing World. ILRI, Nairobi. Thornton, P.K. and Herrero, M. (2015) Adapting to climate change in the mixed crop and livestock farming systems in sub-Saharan Africa. Nature Climate Change 5, 830–836. Wilson, R.T. (1986) Livestock production in central Mali: long-term studies on cattle and small ruminants in the agropastoral system. ILCA Research Report No. 14. ILCA, Addis Ababa. 15 African Livestock Systems Research, 1975–2018 John McIntire1, Tim Robinson2 and Caroline Bosire3 1Santa Barbara, USA; 2FAO, Rome, Italy; 3International Livestock Research Institute, Nairobi, Kenya Contents Executive Summary 516 The problem 516 Research spending 516 Scientific impact 517 Development Impact 517 Defence of pastoralists’ interests 517 Valuing land rights 518 Identifying conditions for successful technical change 518 The problem of translating scientific into development impact 518 Introduction 519 Objectives of LSR 519 Stages of systems research 520 Potential impacts of systems research 521 Classifying Livestock Systems 522 Climate and growing period 522 Modes of production 522 Ranching 522 Commercial dairying 522 Nomadic and semi-nomadic grazing 522 Mixed farming by smallholders 524 Trends in African Agricultural Systems 525 Human populations 525 Livestock numbers 527 Changes in species mix 527 Land use, fertilizer use and cereal yields 531 Inequality of livestock holdings 536 Climate to 2000 536 Poverty 536 Livestock Systems Research 537 Research impacts on livestock 537 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 515 516 J. McIntire, T. Robinson and C. Bosire The development problem of pastoralism 537 Productivity of subsistence and commercial grazing systems 543 The problem of animal mobility 546 ILCA’s programme 546 Borana, Ethiopia 548 Maasailand, Kenya 549 Niono, Mali and the Malian Delta 550 Kaduna, Nigeria 550 Fakara, Niger 551 What did the pastoral systems find? 551 Sedentarization 551 Grazing organization 553 Water 554 Feed systems 556 Modelling pastoralism 561 What did the mixed systems studies find? 564 Characteristics of mixed-systems in sub-Saharan Africa 564 Crop–livestock interactions in mixed systems 567 Feed systems 574 Conclusions 575 Scientific impact 575 Development impact 578 Mixed systems 580 Farm mechanization 580 What blocked the translation of scientific into development impact? 582 The Future 584 Pastoral systems 585 Mixed systems 585 References 589 Executive Summary • What have been the development impacts of LSR since the 1970s? The problem • Can the development impacts of LSR be dis- tinguished from long-term trends in Afri- Livestock systems research (LSR) at the Inter- can livestock systems? national Livestock Research Institute (ILRI) sought to answer two questions: • What are the major livestock systems in the Research spending sub-Saharan Africa tropics and subtropics (Ruthenberg, 1980)? ILRI1 and one of its predecessors, the Inter- • What technical and organizational changes national Livestock Centre for Africa (ILCA), can be introduced into these systems to spent about US$212 million (in 2015 US$) on make them productive? LSR from 1975 to 2018. This was roughly 22% of ILCA spending before 1995 and about 11% of This chapter reports the answers of decades ILRI spending from 1995 to 2018. The majority of of research at ILRI, its predecessors and its this spending was on ruminants in sub-Saharan principal partners to these questions. The chapter Africa, mainly in systems where cattle were the continues to ask: dominant stock. A crude estimate of expenditure • What have been the scientific impacts of by system was 80% on mixed systems and 20% LSR since the 1970s? on grazing/rangelands during the 1975–1994 African Livestock Systems Research, 1975–2018 517 era of ILCA and the International Laboratory for and Ethiopia, led by Michael Blümmel, has Research on Animal Diseases (ILRAD). No such done much to correct this neglect (see estimate by system is possible for the ILRI era Chapter 14, this volume) and has shown (1995–2018). the value of complementary field and sta- tion research. Scientific impact • Quantitative modelling of livestock sys- tems. Models of pastoralism often gener- ated scientifically reliable projections of the The scientific impact of LSR was substantial: outputs of policy experiments. They have • Mapping systems. The greatest scientific succeeded in establishing new research impact of LSR was in mapping systems and lines, notably ‘livestock as a pathway out of identifying their technical possibilities, po- poverty’ and the study of livestock and cli- tential for growth and poverty reduction, mate change. and research priorities. • Defining how mechanization with animals • Understanding transversal and system evolved. This work had an immense scien- studies. The second most important scientific tific impact but little positive development impact was in understanding the evolution impact in the sense of inducing new invest- of animal production systems, grazing and ments in mechanization. There was some mixed alike, to prepare appropriate technical additional scientific impact in elucidating interventions, to avoid repeating past fail- why some apparently promising technical ures in technology generation and transfer, changes had failed. and to identify promising sites for technical • Ley farming in the Mediterranean and in transfer from other regions. the highlands of East Africa. • Estimating production parameters. Esti- • Other efforts to introduce planted forages in mating input and output relationships from smallholder systems, such as pasture im- field data in the systems studies was feasible provement in arid areas, or into beef produc- for the first time, using the work started in tion in the humid and subhumid tropics. the 1970s by ILCA and continuing to the • Smallholder dairying models borrowed from present by ILRI. temperate areas2 and introduced into the • Creating and defending a new view of Afri- tropics. can grazing systems. • Station and farm studies of nutrition and • Understanding the value of mobility in other factors affecting output from working grazing systems. The various field studies in animals. sub-Saharan Africa showed the importance of mobility in grazing and revealed the costs of new policies and organizations that Development impact would compromise mobility. This includes understanding of seasonal and annual The development impact of LSR was limited. It variability and the opportunistic character consisted of: (i) defining the economic weight of pastoralism. and rationality of pastoralism3 as a means of • Understanding complementarities in mixed defending pastoralists’ livelihoods against the wildlife and domestic livestock systems. This incursion of crops and wildlife, corruption and was most important in East Africa. bad policy; (ii) valuing land rights of pastoral- • Understanding the roles of crop residues ists; (iii) defining the conditions in which exter- and manure in soil nutrient cycling. These nal technologies could be introduced into mixed included quantifying the benefits and trade- systems; and (iv) testing and validating im- offs among uses of crop residues and manure proved dairying models for smallholders. in mixed systems as functions of crop and live- stock potential in given environments. Defence of pastoralists’ interests • Correcting the neglect of the feed value of crop breeding programmes in Africa by the The systems studies of ILCA, ILRI and their crop research centres. ILRI work in India many collaborators demonstrated the economic 518 J. McIntire, T. Robinson and C. Bosire rationality of extensive grazing systems in sub- (i) a degree of intensification that was compat- Saharan Africa. Homewood’s contemporary ible with local population density; (ii) a recogni- book (Homewood, 2008) followed by that of tion that the introduction of animal power was Catley et  al. (2013) are thorough analyses of not sufficient and was in many instances not ne- how the new view has contributed to the de- cessary; and (iii) a recognition that flexibility in fence of pastoralists’ interests and avoided managing crops and livestock in the same farm economic losses to the vulnerable groups. At the should not be sacrificed to rigid external ideas of same time, such work called into question the an optimal enterprise mix. ‘inevitable overgrazing’ critique of pastoralism. The single most important development In so doing, this research thoroughly discredited, impact was the design and extension of the even if it did not eliminate, the ‘mainstream broad bed maker (BBM) tool for cultivation on view’ of pastoralism. The new view of pastoral- Vertisols in highland Ethiopia. This is one of ism, supported by research throughout three ILRI research programmes that has docu- sub-Saharan Africa, reduced some of the policy mented the costs and benefits of research and threats to those systems, including forced extension in relation to a specific product; the settlement, confinement into inviable grazing other two are ILRI’s contribution to Kenya dairy schemes and dispossession by elites. policy (see Chapter 17, this volume) and the vaccine against East Coast fever (ECF) (see Valuing land rights Chapter 6, this volume). Related to the ‘defence of pastoralists’ interests’ was the impact of work on valuing land rights and thereby permitting better land policies. All The problem of translating scientific of the systems studied – Maasailand, Borana, impact into development impact Kaduna, Niono, the Niger Delta and Fakara, plus the extensive work of independent re- It was difficult or impossible to translate scien- searchers in East and West Africa – showed the tific impact into development impact for a num- economic and environmental rationality of ber of reasons: pastoralism; these demonstrations strength- Improving primary productivity in mobile ened the case of the pastoralists against forced • grazing areas was generally unprofitable sedentarization and other policies that threat- because of: (i) competition from natural ened their livelihoods. The review of land rights pastures with introduced pastures; and research in Chapter 17 (this volume) on policy (ii) the high costs of water, fertilizer, rota- shows how a better understanding of land tional grazing, fencing and other invest- markets and land rights in Niger and Ethiopia ments in relation to the weak productivity supported policies that could lead to greater ef- effects of introduced pastures. ficiency and equity. • Introducing planted forages into grazing or mixed systems generally failed because of Identifying conditions for successful low incremental yields and low adoption technical change rates outside dairy production areas. • Introducing mixed farming only took place The systems studies of ILRI and partners, com- over many years and its utility as an inte- plemented with findings from the transversal grated model for smallholders was limited studies, made it possible to identify conditions where the conditions – population density, for success or failure of proposed technical market access and seasonality – were not change. For pastoral systems, conditions of suc- favourable. cess included: (i) components that used local • Introducing mechanization with animals knowledge of production; and (ii) components as a project component. International agri- that did not restrict animal mobility in search of cultural research centre (IARC) research water or pasture. For mixed systems with domin- had no evident impact on technical change ant cropping, conditions of success included: in animal production or on efficiency and African Livestock Systems Research, 1975–2018 519 equity related to the ownership and use of Agricultura Tropical (CIAT), International Cen- livestock as draught animals. The excep- ter for Agricultural Research in the Dry Areas tions to this generalization were found only (ICARDA), regional and national programmes, where profitable cash crops, such as cotton and others), given their important contributions and groundnut, received significant exten- in many areas. The chapter first defines the sion support. objectives, stages and potential impacts of LSR. • Application of randomized control trial It then summarizes tropical systems with ani- methods for technology evaluation and for mals as treated by Robinson et  al. (2011) and making policy recommendations to govern- earlier by Seré and Steinfeld (1996), Thornton ments is generally too costly because of sam- et  al. (2002), Steinfeld et  al. (2006)4, Kruska ple size problems in livestock research. et al. (2003), Reid et al. (2008a) and Robinson • Many models were inapplicable for policy et  al. (2011). Subsequent sections review the advice. Using models, quantitative or not, scientific and development impacts of ILRI re- failed to raise productivity. Many models search on pastoral and mixed systems in were developed and later not used at all, sub-Saharan Africa, with some reference to even for scientific purposes, or were never South Asia and Latin America, and explains used for investment analysis or policy simu- why the scientific impact has often been strong lations; some of the models were poorly while the development impact has been weak5. documented and their results could not be The concluding part of the chapter summar- replicated. izes the scientific and development impacts and the problems in translating scientific impacts into development impacts and indicates some future priorities. Introduction The chapter does not cover mapping methods as discussed in Robinson et al. (2011) and in the This chapter covers the scientific and develop- agricultural land cover literature, such as Global ment impact of international LSR in animal Land Cover (GLC) 2000 (Mayaux et  al., 2004; health and management, grazing management Bartholomé and Belward, 2005), GlobCover and plant production, economics and policy, soil (Bontemps et  al., 2011) and MODIS products fertility management, farm mechanization and (Morisette et al., 2002), nor does it treat more re- crop–livestock interactions. It reviews both ex- cent efforts on composite products (Fritz et  al., tensive systems, in which mobile herding is the 2015; See et al., 2015). principal activity with little or no arable agricul- ture, and mixed systems, in which arable farm- ing is the dominant enterprise, animals and crops are managed jointly, and where animals Objectives of LSR are much less mobile. The chapter answers the following questions: The broadest objective of system characteriza- • What are the livestock systems in the trop- tion is that of Robinson et al. (2011, p. 17): ‘…to ics of sub-Saharan Africa? predict how the production systems may change • How have these systems evolved since in the future…and to assess the potential impact around 1970 just before the founding of of changes in crop–livestock systems on agro- ILCA and ILRAD? ecosystem services’. While Robinson’s is a • What are the major findings of LSR in recent statement, it is representative of the pur-6 sub-Saharan Africa beginning around poses of system characterization , as they were 1970 and continuing to the present? expressed when international centre livestock • What are the scientific and development research began in the 1970s. The objectives impacts of LSR since the 1970s? were as follows: The focus of the chapter is on ILRI, with refer- • To define land and climate units by tempera- ence to partners (e.g. Centro Internacional de ture, rainfall, altitude and soils to express 520 J. McIntire, T. Robinson and C. Bosire output potential in systems with a major Crops Research Institute for the Semi-Arid Trop- livestock component. ics (ICRISAT) in the 1970s to 1990s was chiefly • To describe and map crop and livestock pat- in the first three stages, with national pro- terns from field observations of land use grammes leading in the fourth stage. and production within land and climate A major part of the descriptive and diag- units. nostic stage in international livestock research • To describe animal health risks by system was agroecological zoning, which classified ‘…to develop a good understanding of the agricultural systems by ‘climate, soil and ter- differences among production systems to…min- rain’ (Collinson, 2000, p. 51; Seré and Steinfeld, imize the risk of disease emergence and spread 1996) and later work in this century such as that (Robinson et al., 2011, p. ix).’ of Kruska et al. (2003) and Robinson et al. (2011), • To apply maps of livestock densities and which used data and tools that had not existed output (Robinson et al., 2014) to illustrate when the IARCs were expanding quickly in the disease risks and to project global and local 1970s and 1980s. Its purpose was to project car- environmental impacts of livestock (Gerber rying capacities, to identify domains for genetic et al., 2013). and management improvement, to describe field • To estimate factor and input productivities conditions for experiments8 and to show the scope by system and then to define constraints to for technology transfer among zones. development interventions that could be re- The basis of the design and planning stage leased by research and extension in the was to create average farm typologies within ‘diagnostic stage’ of the farming systems agro-ecologies. These typologies would be used research sequence. to make recommendations applicable to dairy • To apply systems maps – comprising cli- farms using cut forages, to smallholders prac- mate, soils, altitude, water, animal diseases tising rainfed cropping without animal traction, and vectors, and productivity – to plan or to systems such as watering practices for development interventions and to better transhumant herders. The ILCA systems studies target public investment in support of sec- in grazing and mixed systems are a mix of the toral goals. descriptive/diagnostic and design/planning • To propose technical and managerial changes stages9. The most similar CIAT study (Rivas Rios, to relieve productivity constraints in the 1974) is the second stage, in which beef ranches ‘design and testing’ stages of the farming in the same agroecology of Colombia were systems research sequence. grouped by farm size. The ICRISAT village-level • To provide information for process and agent studies in the semi-arid tropics of India drew models of technologies and policies. samples of farms based, inter alia, on agroecolog- • To improve risk management and assist ical zoning and ultimately constructed Stage recovery from shocks7. 2 farm typologies using irrigation, mechaniza- tion and the principal crop as variables (Walker and Ryan, 1990). The construction of farm typ- ologies was a principal goal of the French trad- Stages of systems research ition of agronomy in West Africa and this had some influence on the work of ILCA and ICRI- Norman and Collinson (1985) defined four stages SAT in that region from the 1970s. in farming systems research: (i) a ‘descriptive The third stage was to estimate productiv- and diagnostic stage’ to analysis of constraints to ity determinants and to test technologies using productivity; (ii) the ‘design or planning stage’ experiments in environments and farm types as in which strategies are identified for resolving d efined in the first two stages. This stage is to constraints; (iii) a stage in which strategies calculate the levels, rates of change and deter- (e.g. new varieties) are tested on-farm, with varying minants of enterprise and whole-farm product- degrees of researcher management; and (iv) an ivity to be used in identifying constraints and in extension stage in which recommendations are targeting extension. This estimation was done applied. The comparative advantage of the using farm data gathered in surveys, sometimes IARCs, such as ILCA, CIAT and the International combined with farmer- or researcher-managed African Livestock Systems Research, 1975–2018 521 experiments (an early summary of mainly crop Class II is contributing to technical change, work from sub-Saharan Africa and Latin America which lifts productivity. This can involve esti- is given by Matlon et  al., 1984). The third stage mating optimal levels of water, veterinary drug was prominent in the early systems research port- and feed use under controlled situations and folios; crop research tended to be more prominent, testing these levels on farms or among herds. with the early ILCA work in Ethiopia, Mali and Ni- Other examples would be testing innovations, geria and some CIAT work in Latin A merica being such as vaccines, tropical pastures and multidi- the main livestock exceptions. mensional crops, or improved processing. The fourth stage was ‘recommendation and A subset of Class II is to estimate the nega- diffusion’ of new technologies. This was to be tive externalities (costs) of farming, such as achieved by models of ‘entry points’ (as de- greenhouse gas emissions, land degradation scribed much later by van Wijk et al., 2009) at and water pollution from livestock effluents. which technical changes could be introduced Examples are the book Livestock’s Long Shadow based on an understanding of the complex inter- (Steinfeld et  al., 2006), which calculated the actions among system components. This stage local and global environmental effects of ani- was broader in LSR than in cropping, which fo- mal production, and research on antimicrobial cused on crop cultivars and associated input resistance. packages, because of the need to integrate more Class III is advising on policies to raise prod- components – animal species and breed, animal uctivity or to improve income distribution. This health, animal management (fencing, housing can involve adjusting terms of trade, devising and grazing rotations) and feed production. better institutions, eliminating distortions in in- Livestock-related examples include mechaniza- centives, identifying costs of bad policies and tion: animal traction, the use of cows for draught proposing measures to achieve higher growth. power and the BBM in Ethiopia; the introduction The potential scientific and development im- of trypanotolerant animals in combination with pacts of LSR would occur through the following: drug treatments and vector control (see Chap- ters 2 and 3, this volume); a vaccine against ECF, • Defining and mea suring land and climate using the infection-and-treatment method (see units with similar production potential Chapter 6, this volume); the use of trypanocidal (Class 1). drugs (see Chapter 3, this volume); feed improve- • Characterizing and mapping production ments with grasses and legumes (see Chapters systems – combinations of resources by 11 and 12, this volume); various forms of land season – across the land and climate units management, such as fodder banks with Sty- (Class 1). losanthes spp. in central Nigeria; and alley farm- • Estimating productivity relationships to ing in Nigeria and other humid countries in identify priority technologies in animal West Africa. health, breed, feed and management, or to propose changes in organization or policy (Classes 1 and 2). • Targeting these technologies (new methods Potential impacts of systems research of production or organization, or policy shifts) for controlled testing in relevant We set the potential scientific and development s ystems (Class 2). impacts of LSR into the three classes of policy • Simulating potential growth as functions of studies as defined by Zilberman and Heiman hypothesized changes in technologies, pol- (2004, pp. 278–279). icy, organization or species/breed between Class I is scientific understanding, ad- domesticated animals and wildlife (Class 2). vanced by estimating system scale, productivity • Providing information to guide extension of and sectoral accounts. Class I includes advances proposed technologies to farmers (Class 3). in research methods, such as using remote sens- • Writing ex post evaluations of the develop- ing to map production systems and carrying out ment and environmental impact of changes long-term field studies on production and envir- in technologies, policies and organizations onmental relations. (Classes 2 and 3). 522 J. McIntire, T. Robinson and C. Bosire • Proposing new research hypotheses after farming with irrigation or under rain-fed condi- the ex post evaluations of new methods tions. Ruthenberg used two variables – animal (Classes 2 and 3). mobility (1980, p. 18) and rainfall (1980, pp. • Revising methods of experimentation and 322–323) – in his classification. He grouped modelling to test the new hypotheses animal mobility into permanent nomadism, sea- (Classes 2 and 3). sonal (proximate or distant transhumance) or none, as with village grazing or stabulation. Classifying Livestock Systems Ranching Ranching is a commercial system on large, pri- Classifying livestock systems has three steps: vate, fenced holdings, usually for beef production (i) describing the biophysical conditions of pro- using grade animals, with no seasonal or annual duction in terms of climate and length of grow- mobility outside the ranch. The amount and sea- ing period, as determined by rainfall, tempera- sonality of rainfall would vary from arid Botswana ture and altitude; (ii) describing the modes of (Köppen climates BWh and BSh), Australia (Köp- livestock and crop production with respect to pen climates BSk, BWh, plus the humid subtrop- animal mobility, principal product and enterprise ical classes Cfa and Cwa) to subhumid and humid scale; and (iii) mapping the modes of crop and (Colombia and Brazil; Köppen climates Af, Am and livestock production by biophysical conditions Cwb). There is no reliable estimate of ranched cat- to define systems (see Maps 1–3. p. xvii-xix). tle in the African tropics, but that number would be small as a share of all cattle in sub-Saharan Af- Climate and growing period rica and certainly much less than the correspond- ing shares in the humid tropics of Latin America or in the arid tropics of Australia. The typologies of the Food and Agriculture Organization of the United Nations (FAO, 1975) Commercial dairying classified the physical conditions of animal pro- duction by climate and length of growing period Commercial dairying is milk production for the (LGP)10 as the following: market, with some form of sown pastures, little • seasonal mobility, no annual mobility, and both Arid climates, of the Köppen types BWh large and small holdings. Commercial dairying (arid, desert, hot arid) and BSh (arid, sa- in the tropics is typically in the cool subhumid vannah, hot arid), with an LGP of less than highlands (East Africa and parts of Latin Amer- 75 days11. • ica) or in the hotter subhumid lowlands, both of Semi-arid to subhumid climates, of the which have high rainfall and year-round pasture Köppen types Am (equatorial monsoon), production. In sub-Saharan Africa, commercial Aw (equatorial winter dry), BSh and BWh, dairying is restricted to highland countries in and having an LGP of 75–150 days. • East Africa with less risk of trypanosomiasis and Subhumid to humid climates, of the ECF, or to parts of southern Africa with cooler K öppen types Aw and Am, with an LGP of climates and where insect-borne diseases can be 181–270 days. • better managed. No reliable estimate of animals Humid climates, of the Köppen types Aw or land in commercial dairying is available for and Am, with rare areas in Af (‘fully humid’), sub-Saharan Africa in the 1970s, but both with an LGP of more than 270 days. would have been small shares of their respective continental totals and with respect to the num- bers of commercial dairy animals in Latin Amer- Modes of production ica and the Caribbean. Ruthenberg (1980) proposed four modes of Nomadic and semi-nomadic grazing tropical livestock production – ranching, com- mercial dairying, nomadic and semi-nomadic Nomadic and semi-nomadic grazing is milk subsistence dairying, and mixed crop–livestock production for subsistence, in which extensive African Livestock Systems Research, 1975–2018 523 seasonal and annual mobility are the defining fea- The east and central group extends from tures. The climate is arid12 of the Köppen types Bw central Sudan south and east into Eritrea, Ethi- and Bs with rainfall usually below 300 mm in a opia, Kenya, Somalia and parts of Uganda, Tan- single wet season. Grazing systems in sub-Saharan zania, Rwanda and Burundi. East African sys- Africa cross a wide band from Senegal to Eritrea, tems are more diverse, ranging from arid to into the northern parts of the West Africa coast semi-arid (LGA) to subhumid (LGT), with some (seasonally), into Sudan, Cameroon and the Cen- bimodal rainfall, and some LGH systems at alti- tral African Republic, and then south from Eritrea tudes above 1000  masl. The east and central and Ethiopia around the Rift Valley into Mozam- groups further differ from the western group in bique. Practically all of the African nomadic having: (i) longer growing periods (60–90 days and semi-nomadic modes are at altitudes below and sometimes 90–120  days); (ii) colder min- 1500 m above sea level (masl). These systems, al- imum temperatures; (iii) higher CP content in though sparsely populated and understocked rela- native pastures; (iv) more competition with wild- tive to their potential, were the most important life; and (v) because of (i), more diverse mobility in grazing areas and in ruminant numbers in patterns and more diverse human ecology sub-Saharan Africa at ILCA’s founding. (Homewood, 2008; Blench, 1999). Systematic area data exist for 1991–1993 grazing systems. We use ‘grazing systems’, ‘pas- (Fig. 15.1; Seré and Steinfeld, 1996), for 1990– toral systems’ and ‘pastoralism’ interchangeably 2000 (Kruska et al., 2003) and for 2000–2010 to refer to African farming systems in which ru- (Robinson et al., 2011). The data in Fig. 15.1 for minants are the main stock. These are livestock/ 1991–1993 are adapted from Seré and Steinfeld grazing/arid (LGA), livestock/grazing/humid (LGH) (1996) as the nearest global benchmark for the and livestock/grazing/tropical highlands (LGT) ILCA studies that began in the late 1970s. Ru- in the terminology of Sere and Stenfeld (1996) minant systems on grasslands (LGA, LGH and and Robinson et al. (2011). LGT) covered about 1463 million ha, of which The LGT system is livestock found only in 25% was in Latin America and the Caribbean, humid and subhumid zones. These are found in 40% in sub-Saharan Africa, 27% in Asia, and 8% Köppen climate Aw (tropical savannah), which in West Asia and North Africa. The largest single is generally hot and wet with the driest month grassland was LGA in sub-Saharan Africa, cover- having less than 60 mm rainfall. ing some 392 million ha, which comprised 27% The LGH system is livestock only in (tem- of the world’s grasslands and 67% of sub-Saharan perate and) tropical highland zones. The LGT African grasslands. The grassland systems in and LGH systems constitute perhaps 26% of the Latin America and the Caribbean are in the sub-Saharan Africa grazing area, a similar share humid and subhumid climates, while those in of the ruminant populations and perhaps one- Central Asia, China, Mongolia and Russia are in sixth of the sub-Saharan African population. the cold and arid climates. Shares of the global to- African rangelands fall into two groups. tals of ruminant livestock in agroecologies classi- The west and central group extends from Sene- fied as ‘grasslands’ are (Table 15.1): Latin Amer- gal to eastern Chad, nearly all of which is in the ica and the Caribbean, 55%; sub-Saharan Africa, LGA and LGH systems13. All West African grazing 33%; Asia, 12%; and West Asia and North Africa, systems have a single rainy season at elevations less than 1%. The respective population shares below 1000 masl. The shorter rainy seasons are Latin America and the Caribbean, 22%; in West Africa (60–90  days), uniformly lower sub-Saharan Africa, 28%; Asia, 24%; and West elevations, generally hotter temperatures and Asia and North Africa, 6%. Animal density – ani- strictly monomodal rainfall imply lower average mals per hectare or animals per human popula- primary productivity, more variable grazing and tion – is highest for the three grassland systems in therefore stronger reasons for extended trans- Latin America and the Caribbean and next high- humance (Pratt and Gwynne, 1977; Thornton est in sub-Saharan Africa. Reid et  al. (2005) re- et al., 2002). The West African group would tend viewed 40 years of research on extensive systems to have lower average pasture quality, as indi- in East African grasslands. cated by crude protein (CP) content per unit of LGA is defined as livestock only. The grazing dry matter. areas and animal numbers of the most arid 524 J. McIntire, T. Robinson and C. Bosire Population, Agropastoral and pastoral Population, Mixed extensive Population, Mixed intensifying 90 300 750 60 200 500 30 100 250 0 0 0 Area (ha), Agropastoral and pastoral Area (ha), Mixed extensive Area (ha), Mixed intensifying 2.5 5 4 2.0 10 3 1.5 2 5 1.0 1 0.5 0 0 0.0 Cattle (TLU), Agropastoral and pastoral Cattle (TLU), Mixed extensive Cattle (TLU), Mixed intensifying 60 60 90 40 40 60 20 20 30 0 0 0 CSA EA SA SEA SSA WANA CSA EA SA SEA SSA WANA CSA EA SA SEA SSAWANA Region Fig. 15.1. Characteristics of global livestock systems by region, 1991–1993. CSA, central and southern Africa; EA, East Africa; SA, South Africa; SEA, South-east Asia; SSA, sub-Saharan Africa; WANA, West Asia and North Africa. (Data from Robinson et al., 2011, p. 48.) system are by far the highest in sub-Saharan Af- less than 25 persons/km2; ruminant stocking rica. These facts indicate that sub-Saharan Afri- rates were generally from 5 to 20 ha per animal. ca’s most important growth potential in terms of animal numbers is also its highest cost area in Mixed farming by smallholders terms of grazing land used per animal unit. LGA types are typically in the arid and semi-arid Ruthenberg’s fourth mode of tropical livestock (Köppen climate BSh). They are generally hot production – mixed farming by smallholders – with a single rainy season and with no month has always been the focus of international agri- having a minimum temperature below 0°C. The cultural research in sub-Saharan Africa. Such LGA system (390 million ha in sub-Saharan Af- modes blend crop and livestock activities. Ani- rica) covered some 27% of global grazing and mals eat crop residues and provide draught power, some 53% of sub-Saharan Africa grazing. The dairy products, meat and manure. Animal mo- great majority of African production in LGA sys- bility is restricted to seasonal grazing with some tems is nomadic herding with milk as the subsist- transhumance as part of animal tenure relation- ence good. Population density is low, typically ships between crop farmers and pastoralists. Millions African Livestock Systems Research, 1975–2018 525 mixed systems. We use ‘mixed systems’ from the cash crop – cotton, groundnut, rice and classification of Seré and Steinfeld, 1996)14. This other cereals, roots and tubers – that did classification defines mixed systems as those in not exist in rangelands. In the Latin Ameri- which ‘…livestock contribute more than 10 per can and Middle Eastern mixed systems, cent to total farm output in value terms or where animals were the cash commodity. intermediate contributions such as animal trac- • Important shares in income and employment tion or manure represent more than 10 per cent of perennial cash crops – coffee, tea, cocoa, of the total value of purchased inputs’. ‘Mixed rubber and oil palm – were found in mixed sys- systems’ comprise the rain-fed types – mixed/ tems with poultry, swine and small ruminants, rain-fed/arid and semi-arid (MRA), mixed/rain-fed/ although much less often with cattle. humid and subhumid (MRH) or mixed/rain-fed/ • There was smallholder commercial dairy- temperate and tropical (MRT) – and the irrigated ing with zero grazing, planted forages, pur- types – mixed/irrigated/arid and semi-arid (MIA), chased feeds, and cultivation of annual and mixed/irrigated/humid and subhumid (MIH) and perennial crops on the same farm. mixed/irrigated/temperate and tropical (MIT). • There was a more unequal distribution of The mixed-farming types15 – MRA, MRT, cattle among households, although this MRH and some MIA– formed the majority of was compensated to some degree by more rain-fed cropped area and income in sub-Saharan equal distribution of small ruminants. Africa in the mid-1970s. These shares grew • There was commercial on-farm fattening of after the mid-1970s with the expansion of ani- ruminants, swine or poultry, in zero-grazing mal traction in much of West Africa and in parts or transhumant systems, using purchased of East Africa. MRA and MRH systems repre- feeds and crop residues from annual and sented some 121.9 million poor rural livestock perennial crops. keepers in sub-Saharan Africa in 2000 or three- quarters of the total of all poor rural livestock keepers (Robinson et al., 2011, pp. 145–152)16. Trends in African Agricultural The mixed systems are characteristically Systems more diverse in enterprise and spatial patterns than the pastoral types17. In this section, we discuss the trends in African African mixed-farming systems18 differed agricultural production systems in a sample of from pastoralism in the following aspects: countries where livestock production is promin- ent and where ILRI has had a significant re- • Most land use, income and employment search presence. were from cropping. • Crop residues displaced pastures as the main feed. Human populations • The shares of animal products in food con- sumption were less than among pastoralists. Data on the distribution of African populations • Animal mobility was limited to seasonal across agricultural systems have long been patchy. transhumance or to daily movements near The information compiled by Jahnke (1982), as permanent villages. In the Latin American presented in the Introduction to this volume, systems, with private and fenced grazing, show that perhaps 238 million rural people lived in there would be no seasonal or annual mobility. sub-Saharan Africa in the mid-1970s in all five • The functions of livestock were more diverse. agroecological zones. Of that mid-1970s total, They included recycling soil nutrients and nearly 40% lived in the arid and semi-arid zones that using draught animals for cultivation. These were the initial focus of ILCA. The focus of ILRAD’s functions of animals were less important or work can be mapped largely to the humid and sub- even absent in Latin America where min- humid agroecological zones given that trypanosom- eral fertilizers replaced manure and tractor iasis and Theileria are more present there. mechanization replaced animal power. The more detailed estimates of Seré and • There were important shares in income and Steinfeld (1996) for nine livestock production employment of annual and semi-perennial systems gave a 1991–1993 average of some 526 J. McIntire, T. Robinson and C. Bosire 519 million people who depended on those sys- Robinson et al. (2011, p. 48) described four tems. The latter estimated human populations broad classes of livestock systems from vintage as: (i) a total of 173 million people in the grass- 2000 data: (i) agro-pastoral and pastoral, corres- land systems, constituting about one-third of ponding to the grassland systems LGA, LGH, LGT; the sub-Saharan African population and imply- (ii) ‘mixed extensive’ systems, corresponding to ing a population density of approximately 30 MRA, MRH, MRT, MIA, MIH and MIT; (iii) ‘mixed persons/km2 on grazing plus arable lands; and intensifying’ systems as those parts of MRA, (ii) a total of 346 million in the mixed rain-fed MRH, MRT, MIA, MIH and MIT with higher pro- and irrigated systems, implying a population duction potential and better market access; and density of approximately 140 persons/km2 on (iv) ‘others’, which were mainly urban and areas grazing plus arable lands. Of this 1991–1993 having less than ‘10 percent of the total land area total, about one-third lived in the LGA/LGT/LGH covered by crops’ (Robinson et al., 2011, p. 46), as production systems, which is not greatly differ- shown globally in Fig. 15.1. Sub-Saharan Africa ent from Jahnke’s estimated share of the rural is dominated by agro-pastoral and pastoral sys- population living in the arid and semi-arid agro- tems, and most of the original work of ILCA and ecological zones. ILRAD focused there (Fig. 15.2). Population (m) Population density (n/km2) 200 150 150 100 100 50 50 0 0 Arable land (m ha) Grazing land (m ha) 50 400 40 300 30 200 20 10 100 0 0 Ruminants (n/ha) Meat (kg/ha) 0.6 6– 0.4 4– 0.2 2– 0.0 0– LGA LGH LGT MIA MIH MIT MRA MRH MRT LGA LGH LGT MIA MIH MIT MRAMRHMRT Agroecology Fig. 15.2. Characteristics of sub-Saharan Africa livestock systems by agroecology, 1991–1993. (Data from Seré and Steinfeld, 1996.) African Livestock Systems Research, 1975–2018 527 Livestock production A measure of ruminant stock numbers is the tropical livestock unit (TLU)19. Figure 15.4 We know generally that the evolution of Africa expresses ‘TLU density’ as the ratio of ruminant livestock systems after the colonial era has been TLU to arable land. Growth in TLU numbers ac- determined by: (i) rainfall and its effects on pas- celerated sharply in East and southern Africa ture and crop growth, as the latter is a partial and in West and Central Africa around the mid- determinant of fodder availability for rumin- 1990s. This ‘area-weighted TLU density’ grew at ants; (ii) control of human and animal diseases; an annual rate of 0.6% in East and southern Af- (iii) the growth of human populations and in- rica20 during the period 1970–1994, at a rate of come and the resulting demand for agricultural 2.0% for 1995–2016 and at 1% for the entire products; and (iv) the expansion of cropland. period of 1970–2016 (Fig.15.2a). These factors have been compounded by exogen- The estimated TLU density grew at an an- ous changes in farming methods, the introduc- nual rate of 1.2% in West and Central Africa tion of arable and permanent cash crops, and during the period 1970–1994 and at 3.6% for the development of irrigated farming (e.g. Inner 1995–2016. Over the period of 1970–2016, Delta of the Niger River in Mali, the Awash Val- TLU density in West and Central Africa rose at ley in Ethiopia, the Niger–Benue confluence in an annual rate of 2.3%. Nigeria, the Senegal River Valley and along the other major river basins of sub-Saharan Africa). Evidence about stock holdings – numbers, Changes in species mix species, breeds, productivity and locations – is important for understanding the development One factor affecting the relevance and impact of impact of livestock research. Data on African research is long-term changes in ruminant spe- livestock numbers and productivity vary erratically cies mix. This mix appeared to have shifted over time, agroecological zone, country, season slightly to cattle from sheep and goats from 1970 and mode of production. Estimates of animal to 2016 (Fig. 15.5). This continues the trend numbers are particularly unreliable because of from 1950 to 1980 as noted earlier by Le systematic undercounting. It is generally impos- Houérou (1989, p. 125). The ratio of cattle TLUs sible to relate stock numbers or their productiv- to sheep/goats TLUs was about 6.7 in 1960, ity to inputs of research, infrastructure or policy. about 4.1 in the mid-1980s and about 3.7 in With these warnings, the following summarizes 2009 (FAO/CIRAD, 2013, p. 6). De Haan (2016, information over the period 1970–2016 for aggre- p. 25) found similar patterns for stock numbers gate information from sub-Saharan Africa on and species composition in East Africa, exclud- livestock production, stocks of ruminants and ing Ethiopia, over the period 1960–2011. There the species composition of ruminant numbers. was little change in the ratio of sheep and goat We defined a productivity measure as numbers to cattle numbers in East and southern growth of the FAO livestock production index, in Africa (Fig. 15.5a) over the period 1970–2016; rural per capita terms, for the East and southern this ratio rose modestly in West and Central Af- Africa (Fig. 15.3a) for the period 1970–2016. rica (Fig. 15.5b), which is possibly related to the The total index grew at an annual rate of about rising scarcity of grazing in that subregion. The 1.9% from 1970 to 2016; a related index, nor- shift in species from cattle to small ruminants in- malized by rural population, grew at an annual dicates that some opportunity has been lost by rate of 2% in those countries. The corresponding not focusing more on diseases specific to small values for ten countries in West and Central Africa ruminants, such as peste des petits ruminants (Fig.15.3b) were 2.8% and 0.7% for the same (see Chapter 7, this volume). period. In sum, the great majority of livestock The ILCA systems studies (Ethiopia, Kenya, production growth in sub-Saharan Africa for Mali, Nigeria and Niger) were not conducted nearly 50 years is closely associated with popula- over a long enough period to say anything reli- tion growth in 23 countries. The principal direct able about species mix over time. One exception research/development influence would have was an analysis of a mixed sheep and goat flock been through the eradication of rinderpest, at Elangata Wuas Group Ranch in Kenya where which was achieved early in the 21st century. goats had become more numerous relative to 528 J. McIntire, T. Robinson and C. Bosire (a) 100 50 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year (b) 250 200 150 100 50 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year Fig. 15.3. Livestock production index for East and southern Africa (a) and West and Central Africa (b), 1970–2016. Black lines, selected countries; red dashed line, regional trend; green dashed line, regional trend per capita terms estimated by ln(Y) = 0.3599 + 0.0021 × year (a) or by ln(Y) = –9.1269 + 0.0068 (b). (Data from www.faostat.org; accessed 30 April 2020.) Index (2004–2006 = 100) Index (2004–2006 = 100) African Livestock Systems Research, 1975–2018 529 (a) 300 200 100 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year (b) 800 600 400 200 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year Fig. 15.4. TLU density for East and southern Africa (a) and West and Central Africa (b), 1970–2016. Black lines, selected countries; red dashed line, regional trend estimated by ln(Y) = –14.7 + 0.0097 × year (a) or by ln(Y) = –42.1 + 0.0233 × year (b). (Data from www.faostat.org; accessed 30 April 2020). TLu/km2 TLu/km2 530 J. McIntire, T. Robinson and C. Bosire (a) 6 4 2 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year (b) 10.0 7.5 5.0 2.5 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year Fig. 15.5. Sheep and goat numbers relative to cattle numbers for East and southern Africa (a) and West and Central Africa (b), 1970–2016. Black lines, selected countries; red dashed line, regional trend estimated by ln(Y) = 0.1 + 0.0001 × year (a) or ln(Y) = –33.2 + 0.0171 × year. (Data from www.faostat.org; accessed 30 April 2020.) Ratio of sleep and goat numbers to cattle numbers Ratio of sheep and goat numbers to cattle numbers African Livestock Systems Research, 1975–2018 531 sheep between 1978 and 1986 (Wilson and 12 kg/ha in 2001–2005 and had risen to about Maki, 1989). Coppock (1994, pp. 169–170) 17 kg/ha in 2011–2016 (Fig.15.7b). found that short-term droughts in Borana af- The regional mean cereal yield in East and fected cattle numbers more adversely than small- southern Africa was about 1.1  t/ha in 1970– ruminant numbers. 1975 and had risen to 1.9 t/ha in 2011–2016 One aspect of species composition is the use (Fig. 15.8a); the estimated average rate of of animals, mainly oxen, for power. Indeed, ani- growth for East and southern Africa was 0.9% mal traction research was an important part of from 1970 to 2016. The West and Central Africa ILCA research until the 1990s. Systematic and mean was 0.6  t/ha in 1970 and had risen to comprehensive information on draught animals roughly 1.4 t/ha by 2016 (Fig. 15.8b); the esti- is unavailable in sub-Saharan Africa. We do know mated average rate of growth for the ten West that animal traction has expanded widely in this and Central Africa nations was 1.3% from 1970 area from a very low base of use, while tractor to 2016, so this subregion’s levels and growth mechanization has displaced animal and human rates lag behind those of southern Asia and of power in the other dryland and subhumid regions Latin America. The mean cereal yield in West of the world, although lack of data make it im- and Central Africa was about 0.7 t/ha in 1970 possible to estimate changes in stocks of TLUs and had risen to 1.4 t/ha by 2016 (Fig. 15.8b). according to the use of animals for power. The index of labour productivity in cereals therefore fell systematically over the long period of 1970–2016; it had been roughly 0.45 t per capita in the 1970s and fell to an average of Land use, fertilizer use and cereal yields 0.38 in the period 2010–2016. There was some small improvement in an Evidence about land holdings and rural popula- index of food production in East and southern tion shows rising land pressure throughout Africa and in West and Central Africa over the sub-Saharan Africa. The stock of arable land per period of 1970–2016. Food production per rural population fell continuously over the long rural inhabitant among the 13 East and south- period of 1970–2016. In East and southern Af- ern African nations (Fig. 15.9a) rose at an aver- rica, there was about 0.3 ha of arable land per age annual rate of 1.4% from 1970 to 1994; it capita in 1970 and this had fallen to about 0.2 rose at a rate of 3.1% from 1995 to 2016. The by 2016 (Fig. 15.6a); the annual rate of decline corresponding values for West and Central is estimated to have been –1.1%. In West and Africa (Fig. 15.9b) were 0.5% and 1.7%. Central Africa, there was about 0.7 ha of arable The aggregate evidence, with all the quali- land per capita in 1970 and this had fallen to fications about potential internal errors, shows about 0.3 by 2016 (Fig. 15.6b); the annual rate the following: (i) 23 countries in sub-Saharan of decline is estimated to have been about –1.4%. Africa have much less land per capita than they The amount of arable land per rural population had 40–50 years ago; (ii) these countries have began to fall less sharply after the mid-1990s, re- increased cereal yields at a rate close to that of flecting the broad deceleration of population rural population growth, indicating that rural growth across Africa. labour productivity in the major crop group has Evidence about fertilizer use and cereal barely grown; (iii) livestock productivity has yields suggests a modest intensification of farm- similarly increased at an annual rate roughly ing systems in response to land pressure. Na- equal to that of population growth; (iv) most of tional data on fertilizer use (Fig. 15.7) are not the gains in livestock productivity are due to available before 2002. In both subregions, there more animals per unit of land, not to more had been a modest increase in fertilizer use per productivity per animal; (v) there have been hectare to 2016, although levels in Africa re- modest gains in fertilizer use per unit of land main much lower than in southern Asia or in and sub-Saharan Africa still lags regions of the Latin America. National data on fertilizer use for world where agriculture has grown more West and Central Africa are no more compre- quickly; and (vi) there have been modest gains hensive than for East and southern Africa. The in food production per capita in the two princi- average rate in West and Central Africa was pal sub-regions. 532 J. McIntire, T. Robinson and C. Bosire (a) 0.6 0.4 0.2 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year (b) 2.5 2.0 1.5 1.0 0.5 0.0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year Fig. 15.6. Arable land per capita for East and southern Africa (a) and West and Central Africa (b), 1970–2016. Black lines, selected countries; green dashed line in per capita terms, regional trend estimated by ln(Y) = 21.4 – 0.0115 × year (a) or ln(Y) = 27.5 – 0.0143 × year. (Data from www.faostat. org; accessed 30 April 2020.) Hectares per capita Hectares per capita African Livestock Systems Research, 1975–2018 533 (a) 75 50 25 0 2002 2005 2008 2011 2014 Year (b) 50 40 30 20 10 0 2002 2005 2008 2011 2014 Year Fig. 15.7. Fertilizer use per hectare for East and southern Africa (a) and West and Central Africa (b), 2002–2016. Black lines, selected countries; red dashed line, regional trend estimated by ln(Y) = –19.4 + 0.0112 × year (a) or ln(Y) = –93.8 + 0.0478 × year (b). (Data from www.faostat.org; accessed 30 April 2020.) Fertilizer (kg/ha) Fertilizer (kg/ha) 534 J. McIntire, T. Robinson and C. Bosire (a) 5000 4000 3000 2000 1000 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year (b) 2000 1500 1000 500 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year Fig. 15.8. Cereal yields per hectare for East and southern Africa (a) and West and Central Africa (b), 1970–2016. Black lines, selected countries; red dashed line, regional trend estimated by ln(Y) = –10.9 + 0.0091 × year (a) or ln(Y) = –18.6 + 0.0127 × year (b). (Data from www.faostat.org; accessed 30 April 2020.) Yield (kg/ha) Yield (kg/ha) African Livestock Systems Research, 1975–2018 535 (a) 200 150 100 50 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year (b) 150 100 50 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Year Fig. 15.9. Food production index for East and southern Africa (a) and West and Central Africa (b), 1970– 2016. Black lines, selected countries; red dashed line, regional trend estimated by ln(Y) = –59.3 + 0.0319 × year (a) or ln(Y) = –68.6 + 0.0365 × year (b); green dashed line, regional trend in per capita terms estimated by ln(Y) = –5.6 + 0.0051 × year (a) or ln(Y) = –19.2 + 0.0119 × year (b). (Data from www. faostat.org; accessed 30 April 2020.) Index (2004–2006 = 100) Index (2004–2006 = 100) 536 J. McIntire, T. Robinson and C. Bosire Inequality of livestock holdings temperatures and drier growing conditions. Glo- bal atmospheric studies (Hulme et  al., 2001) Inequality of livestock holdings would determine showed African temperatures to have risen from to some extent the distribution of benefits from the mid-1970s to 2000. The most regular in- research across households within given agro- creases occurred in the northern hemisphere ecologies. The African field studies carried out by rainy (June–August) and post-rainy (Septem- ILCA and later ILRI all showed substantial in- ber–November) months. There was less pro- equality among farm units in livestock holdings, nounced annual warming in East Africa and indicating that research would benefit fewer even a cooler period in the Borana and Turkana households than in a more even distribution of ranges. The mid-1970s to 2000 tended to be animals across production units. Inequalities in drier than the previous 75  years in the Sahel ownership of cattle were high within and be- (about 25%); there was a mix of wetter and tween Malian survey villages, as were those in dryer trends in East Africa, with more extremely ownership of small ruminants (Wilson, 1986, p. wet years. 23). High degrees of inequality in stock holdings Vegetation trends in the Sahel, despite the were found in the Borana study led by Coppock apparent warming and drying, tended to show (1994, Table D1). In the Maasailand study of recovery from earlier drought cycles. Herrmann Solomon Bekure et  al. (1991, p. 84), 29% of et al. (2005) contended that land degradation in households in three group ranches owned 57% the Sahel, caused by drought, had been partly of cattle in the sample while 24% of sample reversed by conservation efforts. The effects of households owned only 4% of the cattle. Hier- climate on sub-Saharan agricultural production, naux and Ayantunde (2004, pp. 35–27) found on the whole, were mixed in the final 30 years of significant differences in land and animal hold- the previous century and no continental general- ings per adult household member in a sample of ization can be made about climate as a determin- 492 households in western Niger studied in the ant of research success/failure in that context. 1990s and in the early part of this century. A comparison of global data from 12 coun- tries21 confirmed the general finding of livestock Poverty holding inequality across households. With household samples divided into expenditure Poverty was not a theme of ILCA/ILRAD/ILRI quintiles, measured by the presence or absence research before the mid-1990s and the words of animals and by TLUs per household, the com- ‘poor’ or ‘poverty’ as keywords in published parison showed that about 65% of rural house- work rarely appear before 2000. There was some holds in all expenditure groups had livestock analysis of wealth disparities in Maasailand in (1.9 TLUs on average across countries and the 1980s by King et  al. (1984) and Grandin expenditure quintiles); about 68% in the lowest (1988) but no systematic or even sporadic effort expenditure quintile had livestock (1.4 TLUs on to relate ILRI’s work to poverty in Africa, or any- average), while 58% in the highest expenditure where else, before the work of Thornton et  al. quintile had livestock (3.3 TLUs on average). (2002) and that of Perry et al. (2002). Surveys of pastoralists from 2003 to 2011 in 11 There were no reliable estimates of rural sub-Saharan African nations showed a median poverty in sub-Saharan Africa before those of Gini coefficient of 0.673 on TLU holdings (de Haan, Thornton et  al. (2002). Robinson et  al. (2011) 2016, p. 238). The older evidence is not directly estimated that LGA, LGH and LGT represented comparable to the more recent data from the 11 some 29.8 million poor livestock keepers in countries and hence we cannot say if concentra- sub-Saharan Africa, or about 18% of poor rural tion of ownership has changed in recent years22. people. The dry (LGA, MRA and MIA) areas were about 45% of poor livestock keepers in sub- Saharan Africa. The most recent compilation Climate to 2000 (de Haan, 2016) confirms the historical pattern of higher poverty among animal keepers in the The ILCA systems studies in West Africa, begin- drier climates of sub-Saharan Africa, suggesting ning in the 1970s, took place at a time of rising that little relative progress has been made in African Livestock Systems Research, 1975–2018 537 those areas. While there was no poverty focus, They estimated that annual TFP growth in Bra- as such, of ILCA/ILRAD work until this century, zilian livestock (1970–1985) was only 0.09, of the choice of agroecologies for the original sys- which 55% was due to public research and 40% tems studies imposed such a focus on research at to other, unidentified, factors. They made no es- those sites. timates for livestock in India, although their esti- mates of TFP growth in crops were from 1.27 (1956–1965) to 1.14 (1977–1989) and had Livestock Systems Research public research shares of 22% and 45%, respect- ively. Hazell (2008) found no post-Green Revolu- Research impacts on livestock tion research impact on livestock productivity in South Asia. He mentioned the substantial ex- Evaluations of the impact of LSR in sub-Saharan penditures of the Indian national programme Africa confront difficult empirical problems of (Hazell, 2008, pp. 5 and 16) but did not analyse context, measurement and attribution. The their productivity effects. availability of baseline estimates for other re- Alene (2010) calculated an annual rate of gions and production systems is limited. Pub- growth in TFP, for crops and livestock, of 1.8%23 lished estimates of the impact of research on across sub-Saharan Africa for 1970–2004 and long-term productivity changes in livestock are a slightly faster rate of 2.1% from 1991–2004, uncommon globally and rare in Africa. A global possibly reflecting better weather or economic review of returns to agricultural research in reforms. He estimated an elasticity of TFP with crops, livestock, fisheries and forestry found a respect to research and development expend- livestock focus in 14.4% of a sample of 292 pa- itures of 0.10–0.22, depending on the statistical pers, which included a total of 1886 estimates. model used. One specific estimate of research The great majority of papers were from the USA, impact on African livestock productivity is that Latin America and the Caribbean (Alston et al., of Thirtle et al. (1998) for South Africa, which 2000), not from sub-Saharan Africa. The au- found rates of return ranging from 11% to 16% thors found livestock research – including ‘beef, for livestock, pasture and range improvements, swine, poultry, sheep or goats, all livestock, dairy, and returns to animal health research exceeding other livestock, pasture, and “dairy and beef ”’ – 36%. The estimates from South Africa were to have a positive but highly variable, rate of re- made for that country’s national programme turn. A sample of 233 estimates gave a mean and hence include research failures as well as internal rate of return (IRR) of 120.7%, with a successes. coefficient of variation of 4.0, a mode of 14.0 We will discuss mainly African systems and a median of 53.0 (Alston et al., 2000, p. 58). with some reference to other regions whose A meta-regression across research programmes studies are relevant to sub-Saharan Africa, given (crops, animals and natural resources) with rate that most international livestock research has of return as the dependent variable had a live- been in sub-Saharan Africa (Table 15.1). We stock research coefficient of 12.1% (p < 0.90); focus first on grazing systems (LGA, LGT and field crops had a marginal return of 25.1% LGH), of both nomadic and seasonally transhu- (p > 0.99) and natural resources research (fish- mant types24, and later on mixed farming, both eries and forestry) had a negative return of dryland (MRA, MRT and MRH) and dryland/ir- 94.5% (p > 0.99). rigated (MIA, MIT and MIH), where crops and Other international comparisons of research livestock interact25. effects on livestock productivity do not allow firm generalizations. Evenson and Rosegrant (2003) reviewed published studies of research productivity from US, Brazilian and Indian agri- The development problem culture. They estimated that annual total factor of pastoralism productivity (TFP) growth in US livestock pro- duction was 0.55 from 1950 to 1982, of which The initial studies of pastoral and agro-pastoral 9% was due to public research, 17% to extension systems confronted a ‘mainstream view’26 which and 54% to private research and development. informed official and scientific opinions about Table 15.1. Selected pastoral systems studies, various years. Population Country/ density at Farming Scales of Livestock Cultivation Proposed Study Region/period sub-region/area Climate outset (n/km2) systems research Species density intensity technologies Dyson- East Africa, Uganda/ Aw, BSh <20 Subhumid Regional Cattle Variable, Very low Not mentioned Hudson 1966 Karamoja pastoral territory typically (1966) low Monod 1950s–1970s North Africa and Arid Variable Arid grazing Saharan, Camels, Variable Very low Not highlighted in (1975) rangelands of Sahel cattle, book sub-Saharan goats Africa ILCA (1975) 1960s–1975 Sub-Saharan Arid, semi-arid, Variable Arid grazing, Arid and Cattle, Varied across Not studied Grazing Africa BWh, BSh across arid semi-arid goats, continent management continent cropping sub-Saharan sheep, Africa camels Dahl and East Africa, East Africa, Arid, semi-arid, <20 Arid grazing, Semi-arid Cattle, Very low Not studied Herd Hjort 1976 Somalia BWh, BSh limited and goats, management, (1976) cropping subhumid sheep, milk offtake East Africa camels Fricke Nigeria, 1960s Nigeria Semi-arid, Variable from LGA, LGT, Cattle Herd (1979) and 1970s subhumid 20 to 250 MRA, MRT management, animal health, pasture production Jahnke Sub-Saharan Sub-Saharan Arid to humid Variable LGA, LGH, Nation, Ruminants Variable Variable (1982) Africa, 1970s Africa LGT, MRA, territory, MRH MRT farm Sandford 1960s–1970s Global tropics Many, mainly Variable; not LGA, LGH, Territory Cattle, Variable, Zero to low Herd (1983a) arid and highlighted LGT camels, typically management, semi-arid in book goats low water development, secure land rights de Leeuw Sub-Saharan Ethiopia, Mali, Arid, semi-arid >100 in LGA, LGH, Territory, aerial Ruminants 13–31 ha/ Low to Integration of and Africa, 1984 Niger, Nigeria Ethiopia; MRA, survey TLU in medium remote sensing Milligan <25 in Mali MRH Niger and field (1984) and Niger; verification variable in methods for Nigeria animal and land-use surveys Wilson West Africa, Mali B LGA, LGH, 5000 km2 Ruminants Grazing (1986) 1980s MRA, (Niono only) management, MRH, MIA, health, forage MIH production, water von West Africa, Central Nigeria; Subhumid (Aw); Kurmin Biri: Pastoral 2500 km2; no. Cattle, Abet: wet Abet: ~25; Grazing reserves, Kaufmann 1979–1986 Abet, area of 4–12; Abet: (grazing of animals goats, season, Kurmin sown forages, et al. Ganawuri, 455,000 km2 70; reserve); studied as % sheep, ~25; dry, Biri: 5–15; manure (1986) Kurmin Biri; Ganawuri: mixed of n animals camels ~40; Ganawuri: management, study area of 85; surveys farming in the study Kurmin Biri: ~30–40 crop residue 11,225 km2 1979–1984 with cattle; zone; wet, 5; dry, management largely 2475 km2 ~10–18; sedentary; Ganawuri: some wet, NA; nomadism dry, ~20–25 Ellis and East Africa, late Northern Arid, semi-arid; Not stated LGA 9000 km2 Cattle, Low, highly Zero to very Grazing Swift 1970s– Kenya/ 200–600 mm territory camels, variable low dynamics, (1988) early1980s Turkana annual rainfall; goats, grazing LGP 60– sheep management 90 days Le Houérou Grazing lands Arid, semi-arid Variable LGA, LGH West African Cattle, ~5–20 TLUs/ Low Grazing (1989) of the Sahel Sahel sheep, km2 management, goats, pasture camels production, water development Solomon East Africa, Kajiado County, Semi-arid to LGA, LGH Cattle, goats Grazing Bekure 1980–1991 Kenya subhumid management; et al. (‘Maasailand’) group ranches (1991) Continued Table 15.1. Continued. Population Country/ density at Farming Scales of Livestock Cultivation Proposed Study Region/period sub-region/area Climate outset (n/km2) systems research Species density intensity technologies Fratkin et al. African Various (1994) pastoralist sub- systems Saharan Africa (Homewood East Africa Kenya/ Mix of Af, Am, LGA, LHG Kenya/ Ruminants ~5–40 TLUs/ Very low Water and Maasailand, Aw, BWh, Maasailand, and km2 management Rodgers Tanzania/ BSh, Cwa, Tanzania/ wildlife 2004) Serengeti Cwb, Cfib Serengeti Behnke East Africa, Variable, et al. 1970s–1980s typically (1993) <30 Coppock East Africa, Ethiopia/Borana Arid to subhumid <10 LGA, LGT, 16,000 km2; Cattle, ~20–30 Fodder banks, water (1994) 1980–1991 MRA 90,000 camels, TLUs/km2 development, people goats, grazing sheep management, social organization Williams West Africa, Niger Arid, semi-arid 10–100 LGA, MRA Mainly cattle Grazing et al. 1990s and management, (1999) 1980s social organization, manure, fertilizer, plant varieties Barbier and Niger, 1990s Niger Arid, semi-arid Variable LGA, MRA Mainly cattle ~5–20 TLUs/ Livestock Hazell km2 transhumance, (1999) seasonal labour migration, feed purchases as alternatives to transhumance Kruska et al. 2003 Global All Global range All Region Ruminants Global range Production models (2003) agroecologies and forecasts; poverty analytics; policy modelling Devereux Early 2000s Somali region, BWh, BSh Roughly Pastoral and Region, Camels, Unknown Limited, Build processing (2006) Ethiopia 10–15 agro- household ruminants maize, and marketing pastoral barley, facilities, impose wheat, international sorghum phytosanitary standards, conflict resolution, education, financial services, social safety nets, mixed urbanization, rather than sedentarization Reid et al. Various Many (2008a) regions Lesororgol East Africa, Northern Kenya, Semi-arid Pastoral and Region, Camels, Unknown Limited (2008) 2000–2005 Turkana agro- household ruminants crops pastoral Homewood Sub-Saharan Kenya, Tanzania Mix of Af, Am, Pastoral Region, Ruminants Unknown Limited Brief review of (2008) Africa Aw, BWh, household and crops modelling BSh, Cwa, wildlife literature Cwb, Cfb Hiernaux West Africa, Niger Grazing et al. 1990s management, (2009) health, forage production, water Bollig et al. Mainly East Various (2013) Africa sub- Saharan Africa Catley et al. (2013) Konczacki (2014) 542 J. McIntire, T. Robinson and C. Bosire the rangelands (Sandford, 1983a, pp. 11–19). made the introduction of higher-yielding This ‘mainstream view’ held that tropical pas- plant species unsuccessful. toralism was unproductive, for two reasons. The • There had been no commercial success in first reason was the biophysical determinants of reseeding pastures at rainfall around the animal productivity27 and the lack of adapted 550  mm isohyet, with local or introduced technology and management to raise productiv- grasses or with legumes29. ity. This technical pessimism about potential • There had been only one success30 in estab- output was attributed to the arid and variable lishing browse species at rainfall less than climates28 in the tropics, which made crop pro- the 400 mm isohyet in West Africa. duction infeasible without irrigation and which • The high cost of fencing, firebreaks and water restricted agriculture to extensive stock raising made it nearly impossible to achieve higher and oasis farming. plant yields even if browse and pasture pro- Most of the papers in Monod (1975) ex- duction could be improved under experimental pressed the ‘mainstream view’. Gourou (1966, conditions (Le Houérou, 1989, p. 150; see also p. 192), a leading post-war geographer of the trop- Montgolfier-Kouèvi and Le Houérou, 1980). ics, had earlier argued that tropical cattle product- ivity could converge to that of the temperate zones The second reason for ‘livestock pessimism’ only under strict technical conditions involving: of the ‘mainstream view’ was the persistent be- ‘…the importation, full application and adaptation lief among external observers that African pas- of all the technical progress made in the temperate toralists were inefficient managers for ‘cultural’ lands, namely selection of fodder plants, the use of reasons. The cultural reason, common around irrigation, abundant and precise application of the time of the founding of the African livestock manures, careful breeding of animals, controlled centres, was that herders kept too many animals feeding, and, above all, complete and lasting vic- relative to range capacity because they derived tory over microbial and parasitic diseases.’ utility from keeping stock rather than selling The biological reasons for pessimism were them. This interpretation impelled the view that well founded, as recognized early by the United rangelands management practices, especially Nations Educational, Scientific and Cultural Or- open-access land tenure, would lead to overgraz- ganization (UNESCO 1963, pp. 168–169), ILCA ing31 and should therefore be modified to avoid (1975), Pratt and Gwynne (1977), Le Houérou the destruction of the range. (1989) and in the Dutch–Malian study of Sahel- That overgrazing would eventually destroy ian rangelands (Penning de Vries and Djiteye, the range was a common argument in East and 1991). The economic reasons for pessimism were West Africa alike. The work of colonial scientists in given by Pratt and Gwynne (1977, pp. 100–128). Niger is representative of francophone scientific They showed that the economics of range and administrative opinion (Peyre de Fabrègues, improvement – water development, fencing, over- 1984). The same perceived risk of overgrazing seeding of the range, managed grazing and seed- was the basis of the colonists’ agricultural policy ing with new species of grasses – were in Kenya32. Pratt and Gwynne (1977, p. 38), unfavourable in drier conditions and risky in wet- working mainly in East Africa, argued that: ‘Over- ter ones. These biological and financial findings grazing may often be the direct result of human were generally confirmed by Le Houérou (1980) biological needs… It is, however, also a common for browse and in the global review by Sandford situation that the pastoralist maintains a herd far (1983a) of pastoral development. A review nearly larger than needed for his own subsistence. This is 40 years later of rigorous impact studies of CGIAR usually attributed to sheer greed, prestige, or the research in grazing systems found few productiv- concept of livestock as movable wealth.’ Dahl and ity effects of that research (Jutzi and Rich, 2016). Hjort (1976, p. 17) were unusually prescient in re- Le Houérou (1989, pp. 149–155) synthe- jecting the prestige argument about overgrazing. sized Sahelian investigations over many years A leading observer of tropical agriculture, and concluded the following: Ruthenberg (1980, p. 332) concluded, at the beginning of the modern livestock research in • Irregular rainfall, a long dry season, and Africa, that it was ‘…usually advisable to leave competition from native grasses and forbs totally nomadic systems undisturbed’ because African Livestock Systems Research, 1975–2018 543 they were unsuited for ranching for both bio- of (enforced) limited mobility in the reserves logical and managerial/cultural reasons. with erratic rainfall. Ruthenberg (1980, p. 343) was pessimistic about semi-nomadic systems, contending that Productivity of subsistence and regulation of stock numbers in semi-nomadic commercial grazing systems systems occurred through ‘disaster’ and not through rational management. Ruthenberg, The ‘mainstream view’ (as described and criti- whose books are deeply insightful nearly 40 years cized by Sandford, 1983a, pp. 11–19) was that after their final editions, was dissatisfied African rangelands, like commercial ranches in enough with semi-nomadic herding to refer to Latin America, the USA and Australia, should be its ‘malpractices’ (Ruthenberg, 1980, p. 342): open-air factories to produce meat. Following this ‘uncontrolled animal densities, over-grazing’; ‘in- view was the contention that traditional grazing, adequate fodder distribution over the year’; inef- in Asia and sub-Saharan Africa, was less product- ficient placement of camps to water, resulting in ive in live weight per hectare or per worker than long daily treks; low calf productivity owing to ranching. This contention encouraged measures competition with people for milk; and inefficient to raise stocking rates and offtake from grazing grazing practices (Ruthenberg, 1980, p. 342). systems and to shift herd composition from He concluded that neither greater efficiency nor females to males to make them more product- better environmental stability could be achieved ive in terms of meat supply (Sandford, 1983a: without human population control through pp. 123–126; Wagenaar et al., 1986, pp. 50–51) emigration. He continued to say that replace- and hence more efficient in terms of land use. ment of collective land tenure was necessary by A few early studies tested the proposition allocation to groups or individuals ‘whereby that grazing was less productive than ranching grazing rights were both allotted to the herds- in live weight per hectare (Cossins, 1985 for men and enforced’ (Ruthenberg, 1980, p. 343). Ethiopia; Sandford, 1983b, for sub-Saharan Ruthenberg’s view of the difficulty of real- Africa; Sandford, 1983a, pp. 123–126; Wilson, izing the potential of tropical pastoralism was 1986, for Mali; Behnke, 1985, for sub-Saharan part of a general pessimism about technical and Africa; Wagenaar et al. 1986, p. 47, for five sys- policy measures. Grigg (1974) had earlier con- tems in three countries; Table 15.2). These stud- cluded that ‘the conversion of pastoralism to ies tended to show that grazing productivity per ranching has proved difficult’33. Jahnke’s (1982, unit of land exceeded that of ranches. Extending p. 149) review of African livestock systems ar- the comparison, Breman and de Wit (1983) gued that extensive herding in areas under tryp- found that the protein production per hectare in anosomiasis challenge should sometimes ‘be left Mali was higher than on large ranches in Texas to the fly’ given that ranching is not economic and Australia. Ocaido et al. (2009) found much and that the recurrent costs of long-term tsetse higher annual returns per hectare for pastoral- control are unsustainable on small farms and ism than for ranching when the two were com- ranches. Coppock’s (1994, p. 273) multidiscip- pared in the same area of Uganda. linary study of southern Ethiopia found that Behnke (1985) concluded that the gap be- previous rangelands development projects had tween grazing and ranch productivity in the not produced a ‘documented increase in cattle same environment was usually small, based on offtake or a widespread and sustained improve- observations from Botswana, Kenya, Uganda ment in human welfare from veterinary cam- and Zimbabwe. Behnke further argued that the paigns or ponds, roads, and markets’. Coppock putative gap between ranch and grazing output further noted that the ‘ranch experiment’ in had been exaggerated in favour of commercial southern Ethiopia had ‘failed to transform ranches by cultural bias among external observers, traditional pastoralism’ (Coppock, 1994, p. 35). including scientists, who were too eager to find Mortimore’s (2000, pp. 104–105) review of overgrazing and low productivity among trad- northern Nigeria found that efforts to establish itional herders. Homewood (2008, pp. 63–65), ranches and grazing reserves in the 1960s had using a larger sample of studies, later restated failed because of high fixed costs per reserve, com- the critiques of the ‘mainstream view’, notably petition from cropping and the incompatibility the lack of evidence for range degradation.A se- 544 J. McIntire, T. Robinson and C. Bosire Table 15.2. Productivity comparisons of ranching and pastoralism, various years. Pastoral productivity Ranching Study Region/period Subregion/country Climate/system Species estimates (TLUs/ha) (TLUs/ha) Dyson-Hudson East Africa, Karamoja, Uganda Warm semi-arid Cattle Not specific Unknown (1966) 1966 (Bsh), tropical savannah (Aw), LGH, LGT Dyson-Hudson and Global Global Tropical, temperate, Ruminants, Not specific Not specific Dyson-Hudson highland camels (1980) ILCA (1975) 1960s and Drylands, sub- Semi-arid (Bsh), Cattle, goats, Variable; 0.05–0.2 in Verify in 1970s Saharan Africa tropical savannah sheep, camels Sahelian conditions Breman (Aw), LGA, LGH, LGT, MRA, MRH Dahl and Hjort East Africa, Western Sudan, Arid, semi-arid, Cattle, camels, Not stated Not stated (1976) 1976 Kenya LGA, LGH sheep, goats Fricke (1979) 1970s North-central Semi-arid, Mainly cattle Nigeria subhumid, LGH, MRH Pratt and Gwynne 1960s and East Africa Arid, semi-arid, Mainly cattle (1977) 1970s subhumid, LGA, LGH Jahnke (1982) Sub-Saharan Sub-Saharan Africa Very arid to humid, Ruminants 0.1–0.5 by climate, Africa, 1970s LGA to MIT ~0.35 considered max subhumid Wilson (1986) West Africa, Niono, Mali Arid to semi-arid, Ruminants 0.02–0.25 in arid and 1980s LGA, LGH, MRA, semi-arid conditions MRH Breman and 1970s and East Africa and Arid, semi-arid, Ruminants In protein kg/ha, 0.4 In protein kg/ de Wit (1983) early 1980s some temperate temperate, LGA, for nomads, 0.6–3.2 ha,0.3–0.5 countries LGH for seasonal for USA, 0.4 transhumants, 0.3 for Australia for sedentary Sandford (1983a) 1960s–1970s Developing Mainly arid to Cattle, camels, countries semi-arid goats African Livestock Systems Research, 1975–2018 545 Ellis and Swift East Africa, Northern Kenya Arid, semi-arid, Cattle, camels, Low, highly variable Greater than (1988) 1980s subhumid, LGA, goats, sheep pastoral, but LGH highly variable de Leeuw and Sub-Saharan Ethiopia, Mali, Arid, semi-arid, Ruminants Niger: 0.07–0.03; Mali Milligan (1984) Africa, 1984 Niger, Nigeria LGA, LGH, MRH (Niger Delta): 0.05–0.78 Cossins (1985) East Africa, Borana, Ethiopia, Semi-arid to Cattle ~2 kg animal protein/ ~2 (Kenya) 1980s subhumid ha/year Semi-arid Mali Arid, semi-arid Cattle ~0.6–3.2 kg ~0.5 Australia animal protein protein/ha/year Homewood and East Africa Maasailand, Kenya; Semi-arid to Ruminants and Rodgers (2004) Serengeti, subhumid wildlife Tanzania Solomon Bekure East Africa, Kajiado County, Semi-arid to Cattle, goats 0.25–1, function of Not shown et al. (1991) 1980–1991 Kenya subhumid wealth index (‘Maasailand’) Coppock (1994) East Africa, Borana, Ethiopia Semi-arid to Cattle, camels, Cattle 10–13, goats 2–4 traditional 1980–1991 subhumid goats, sheep 1–2, sheep 1–2 cattle,10–15 introduced Behnke et al. East Africa, Arid to semi-arid to Cattle, camels, Many, various Not presented (1993) 1970s–1980s subhumid goats, sheep rangeland sites Hiernaux and West Africa, Fakara, Niger Arid to semi-arid Cattle, sheep, Not presented Ayantunde (2004) 1990s goats Barbier and Hazell Niger, 1990s Niger LGA, LGH Cattle, sheep, Not presented (1999) goats Kruska et al. (2003) 2003 Global Tropical and Ruminants and Not presented temperate monogastrics Lesorogol (2008) Kenya, 2000 Turkana, Kenya LGA, MRA Cattle 0.05–0.1 and 2005 Homewood (2008) Sub-Saharan Many Ruminants and Sahel, East Africa Not presented Africa wildlife specifically Reid et al. (2008b) 1990s and Athi-Kaputiei Arid, semi-arid Ruminants and Highly variable Unknown 2000s plains, Kenya wildlife Konczacki (2014) 2014 Sub-Saharan Africa Tropical Mainly cattle Not specified Unknown 546 J. McIntire, T. Robinson and C. Bosire cond counterattack against the ‘low productiv- (1994) does not permit conclusions about ity of subsistence grazing’ argument was that long-term changes in mobility. the objective of African rangelands systems was Turner et  al. (2014) completed an un- not to produce meat for sale but to produce milk usually detailed study of livestock mobility in 32 for consumption by the herders and to provide mixed-farming villages in Mali and Niger. They security against drought and disease. This alter- found that mobility was a dominant strategy native view of the objectives of grazing systems among all groups (farmers, herders, artisans and – confirmed notably in the Borana, Maasailand, fishers) and that mobility became more domin- Kaduna and both Mali studies and in the re- ant as stock density rose. Mobility, and the type of views by Dyson-Hudson and Dyson-H udson mobility (within village, proximate or distant sea- (1969), Nicholson (1984) and Behnke (1985) sonal transhumance), were not strongly affected – is consistent with common features of grazing by species in the rainy season, although it was modes. These are: (i) low weaning weight of reduced for all species in the dry season. Appar- calves because of competition for milk with ent longer-term trends affecting mobility in this people; (ii) more females than males in the herd study were: (i) insecurity in grazing areas outside to sustain milk production with limited feed; the survey villages, whether these areas were and (iii) seasonal grazing rotations between proximate or distant, reduced mobility; (ii) higher milk and dry herds to provide milk for subsist- cultivation densities in the survey villages in- ence. creased the risk of livestock damage to crops and hence the risk of conflicts among farmers and The problem of animal mobility herders; and (iii) changes in livestock corridors related to planning of water points also ham- The ILCA systems studies (Ethiopia, Kenya, Mali, pered mobility and changed its paths. Nigeria and Niger) were not designed to estimate There is a long chain of research confirming long-term changes in the mobility of grazing ru- the importance of mobility to pastoralists and minants, although they are definitive on the role highlighting the potential costs to their liveli- of traditional mobility patterns in exploiting sea- hoods if mobility were to be restricted. This chain sonal changes in feed, water and markets. Older includes the ILCA systems studies for Mali, Kenya studies (ILCA 1975; Pratt and Gwynne, 1977; and Ethiopia (Behnke and Scoones, 1993), the Le Houérou, 1989) have the same shortcoming studies on Niger (Hiernaux and Ayantunde, as does the most recent major compendium on 2004; Turner et al., 2014), the FAO/CIRAD com- African pastoralism (Catley et al., 2013). The FAO/ pendium on the Sahel (FAO/CIRAD, 2013) and CIRAD (2013) review of Sahel grazing systems, later de Haan (2016). They confirmed the poten- covering 1970–2010, is the only long-term study tial damage to productivity from policy restric- of the issue. It found that ‘movements have tions on mobility, and the long-term adverse become longer and more dispersed, especially trends related to insecurity, cultivation density, southward’ in response to climate, markets and and the competition for water among herders cropping density (FAO/CIRAD, 2013, p. 14). and between herders and farmers. Bourn and Wint (1994) reviewed 20 aerial and ground surveys of livestock and land use be- tween 1980 and 1991 in Mali, Niger, Sudan, Chad and Nigeria. The surveys confirmed the ILCA’s programme greater importance of stock mobility in the West African subhumid climates, related to the strong- ILCA began from the general idea that not er seasonal variability of rainfall and hence of enough was known about African grazing sys- plant biomass in the ‘750–1250  mm rainfall tems, or about mixed systems in which the ac- band’ (Bourn and Wint, 1994, p. 9). They noted tual productivity of livestock appeared to be less that seasonal mobility was weaker in arid and than its potential, to permit the introduction of humid climates than in the subhumid climates, technical, managerial or policy changes. The presumably because the latter is a transition specific origin of grazing systems research was zone in West Africa north of the equator. The to test the founding hypothesis of ILCA: that single-point sampling frame in Bourn and Wint there had been ‘…a failure to integrate the African Livestock Systems Research, 1975–2018 547 biological, economic and sociological compo- The ILCA (1975) rangelands book was nents of research and development programmes’ • followed by studies of browse in Africa (Tribe et  al., 1973, p. 1). Multidisciplinary re- (Le Houérou, 1980), the rangelands of search was therefore proposed to provide the sci- Kenya (de Leeuw et  al., 1984; Solomon entific basis for programmes to expand livestock Bekure et  al., 1991), Ethiopia (Coppock, production and to improve herders’ welfare. 1994) and tropical Africa (Sandford, ILCA accordingly launched a series of live- 1983a). stock systems studies, which continued for 25– ILCA published extensive literature reviews 30 years in various parts of sub-Saharan Africa, • on trypanotolerance (Trail, 1979a,1979b), to investigate the economic, biotic, abiotic and on livestock in the subhumid zone (ILCA, organizational constraints to improving live- 1979a) and on small ruminants in Africa stock systems. ILRAD’s involvement with sys- (ILCA, 1979b; Gatenby and Trail, 1982), tems analysis was almost nil for its first decade which outlined the state of knowledge and and began only in the 1980s with its role in the suggested priorities for new research in African Trypanotolerant Livestock network these areas. (ATLN) and its leadership of epidemiology and Research on management and productivity economics work of ECF. These studies ultimately • of extensive mixed systems in semi-arid filled some of the gap between the development central Mali (Wilson et  al., 1983; Wilson, importance of livestock in sub-Saharan Africa 1986; Wagenaar et  al., 1986), with data and the research effort on livestock problems. collection from 1978 to 1984. ILCA’s studies had two generations. The Research on management and productivity first generation followed extensive communal • of mixed systems in subhumid central rangelands in Ethiopia, Kenya and Mali from the Nigeria (von Kaufmann et al., 1986), with early 1970s until the early 1990s. These studies data collection from 1978 to 1986. featured little collaboration with the crop centres. The first generation of ILRI research took place • Mixed systems in highland Ethiopia (Gry-seels and Anderson, 1983), with data col- mainly in the LGA/LGH and MRH/MRT types. lection from 1978 through 1985. The second followed smallholder mixed systems • Mixed systems in semi-arid, subhumid and in Nigeria, Ethiopia and Mali (with enhanced highland areas of sub-Saharan Africa, plus collaboration among ILCA, the International the related work of McCown et al. (1979), Institute of Tropical Agriculture (IITA) and McIntire et al. (1992), the collaboration of ICRISAT), in the Middle East (with ICARDA) and ILCA with Pingali et  al. (1987) and Win- in Central America (with CIAT). The second gen- rock International (1992). eration concentrated more on mixed systems. The productivity of pastures, crop res- The broad purposes were to describe how • idues and other feeds, with a new focus on the systems functioned, to define types of pro- browse, which had not been well studied duction units and to estimate productivity in sub-Saharan Africa before the mid- parameters and thereby to understand con- 1970s (Le Houérou, 1980; Walker, straints to productivity. A specific goal of the 1980). first studies was to collect primary field data on • Soils, water and vegetation (complemented crop and animal performance, which in the by the Dutch–Malian work of Penning de late 1970s were unavailable in most of sub- Vries and Djiteye, 1991; and Breman and Saharan Africa. A second goal was to estimate de Wit, 1983) in semi-arid Mali and by the input–output relationships and later to test compendia of ILCA (ILCA, 1975) and of new technologies. Pratt and Gwynne (1977). Paul Neate’s history of ILCA (ILCA, 1994) • The volume of Powell et al. (1995) on live-maps several milestones: stock and nutrient cycling. • ILCA’s pastoral work began with a compen- The following section sketches the main dium on mapping sub-Saharan Africa range- characteristics of the research sites and the spe- lands (ILCA, 1975), with some reference to cific experimental and survey treatments done Tunisia, India, Iran and Australia. at each site. 548 J. McIntire, T. Robinson and C. Bosire Borana, Ethiopia 325,000 cattle (21/km2) and TLUs of nearly 350,000. The latter gives an average stocking of ILCA and national partners studied an area of about 4.5 ha/TLU. 15,475 km2 in the Borana Plateau of southern A rough estimate of land use in the early Ethiopia (Coppock, 1994, p. 39)34 from 1980 to 1980s was that perhaps 4% was cropped (Cop- 1991. The site was chosen within the much pock, 1994, p. 85). Measurements of cropped larger area of a regional livestock development area per household were not reported in the project that provided background information grazing territories (madda) studied. The main on the environment, social organization and crops noted were maize, beans, teff (eragrostis technology of crop and livestock production tef) and wheat. No strict division of land use be- (Coppock 1994, pp. 38–60). Previous research tween grasslands and wooded areas was possible on the Borana areas of northern Kenya and at the time but the majority would have had southern Ethiopia had concentrated on social 10–40% woody cover (Coppock, 1994, pp. 84–85). organization and not on biological constraints to Perhaps half of the grazing territories would productivity. have been ‘unsuitable for cultivation’. Scientists identified four ecologies in the Livestock provided the largest proportion of study area: income for survey households. The average • Subhumid, with varying patterns of live- number of livestock stock units per person in a stock, forest, cropping and urban, with an- sample of 49 households from 1981 to 1983 nual rainfall as high as 750–1000 mm, and (Coppock, 1994, Table D1) was 13.8 (range some areas of bimodal rainfall (March–May 4.6–39.8) averaged across grazing territories. and September–November). There were major wealth differences across • ‘Upper semi-arid’, with about 600  mm of Borana households, reported as livestock hold- rainfall, with a mixed landscape of savan- ings per adult equivalent household member nahs and woodlands, varying in elevation. (Coppock, 1994, pp. 174–175). • ‘Lower semi-arid’, with about 400–600 mm Experimental and survey treatments in the of rainfall in a single season, varying soil Borana programme were: types and landscapes of wooded areas and • Fodder resources (introduction, conserva- grasslands. tion, feed gardens and rehabilitation of de- • ‘Arid’, with varying soil types and land- graded areas). forms, and less than 400 mm rainfall in a • Herd management (offtake, calf feeding single season. and breed improvement). Measures of LGP at five sites were between • Water surveys and experiments (Nicholson, 114 and 151 days, summed between the ‘long’ 1987a,b, 1989). and ‘short’ rains. The systems would be classi- • Novel work on the intergroup inequality of fied as LGA and MRAc. The longer growing sea- herd holdings. sons, compared with West African areas of simi- The Borana volume did not clearly present lar rainfall, were related to higher elevation and feed composition although it is evident that pasture bimodal rainfall not found in West Africa. The was the principal feed. The unusual features of the extended growing periods typically produce pas- Borana Plateau compared with West African pastor- tures with a higher CP content and with longer alism – two rainy seasons, hence less intra- annual periods of CP availability than in West Africa. variability of grazing and better forage quality The longer growing seasons and greater vari- throughout the year; a scarcity of cropping, ation in altitude of East Africa produce a much hence less crop residue production; and the con- greater diversity of plant species than in the stant battle against bush encroachment, hence West African Sahel (Le Houérou, 2009, p. 133). greater incentive to feed browse – impelled feed Household survey and aerial surveys were research into characterization of pasture and used to estimate aggregate resource use in the browse and away from work on crop residue qual- 15,475 km2 area studied. A combination of the ity, which was significant in the Kaduna study. two methods was used in the 1982–1985 period The Borana study outlined a development to arrive at estimates of 66,000 people (4.3/km2), path for the region. The first step was to raise African Livestock Systems Research, 1975–2018 549 stocking rates on undergrazed areas through has since been studied by scientists of many dis- water development. The argument was that ciplines over the past 40 years. higher well density, and lower labour costs of lift- The Maasailand studies differed from the ing water, would divert energy spent by animals others in two respects. One was that they could on trekking and watering into weight gain. use historical data on system structure and The second step proposed for Borana was to productivity that was unavailable in Mali, sub- manage stocking rates – numbers per unit of humid Nigeria and later Niger38. Another was land, species, seasonality and stock preferences that the Maasailand had been the site of dra- among forages – to prevent overgrazing, to adapt matic changes in pastoralism. Land availability stocking rates to rangeland potential and to ex- and therefore herd mobility had diminished with ploit unused feeds, notably browse36. This step the appropriation of Maasai lands by the colon- has failed as it was originally conceived, because ists at the beginning of the 20th century (as it depended on the assumption that cattle would summarized for an area north of Kajiado by Reid remain the dominant species and hence excluded et  al., 2008b). The later introduction of group species shift as an alternative to the manage- ranches by the independent Kenyan government ment of cattle stocking rates on common land. in the 1960s further limited the land and mobility The explanation for the failure to manage stock- of Maasai. ing rates was that it was uneconomic. In long The Maasailand system, classed as LGA by wet cycles with good pasture growth, there was Robinson et  al. (2011) at an altitude around no good reason to destock. When good pasture 1500  masl, was semi-nomadic with a low cycles ended, destocking would occur naturally human population density, poor market access, through sales and mortality. low stocking rates, milk as the staple and very The third step was to augment external in- little cropping. The population density was about puts, such as fertilizers and new plant species, to eight persons per km2 in 1979 (Solomon Bekure raise primary production and therefore carrying et al., 1991, p. 16). Stocking rates ranged from 1 capacity. Using external inputs to raise range TLU/ha to 0.25 TLU/ha across the study sites productivity has typically failed, as one sees lit- with significant wet- and dry-season variation. tle trace of mineral or organic fertilizer use in Cropping density was almost nil. African rangelands and/or of introduced herb- Species and breed composition of herds and aceous species. There appears to have been pro- flocks changed little in the years before and after gress in leguminous tree hay making and fodder the group ranches. Herd structure did not change banks (see Chapter 13, this volume, on planted notably over time from a two-thirds share of fe- forages including trees), but the benefit and costs males; offtake changed little as a share of herd of such methods are poorly known and cannot size. There was limited infrastructure development be attributed to research or extension. A variant and new road, water and dipping infrastructure was to use external inputs (fencing, mineral fell into disuse from lack of maintenance. The fertilizers and new forage crops) to rehabilitate use of variable inputs was uncommon except for degraded sites. acaricides and veterinary drugs. Inequality in stock ownership was high. Maasailand, Kenya The Gini coefficient of cattle and small stock ownership in three Kenya group ranches The ILCA studies of Maasailand in Kajiado (1980–1981) was about 0.50 (Solomon Bekure County of south-eastern Kenya – of water, soils, et al., 1991, p. 3) but there are no corresponding climate, animal health, herd management, vege- earlier data. The effect of group ranches on the tation, pastures, economics and labour use – inequality of stock ownership was unknown. began in 1978 and continued into the mid- A later summary of the effects of subdividing 1980s37. The outstanding study is that of group ranches found that population pressure in Solomon Bekure et  al. (1991), covering 1979– Maasailand in the final quarter of the 20th cen- 1991, followed by the later contributions of tury would have reduced livestock/person num- Homewood and Rodgers (2004), among others. bers enough to make ranch subdivisions inviable The general area of Maasailand covering much without outside income from wage labour or re- of southern Kenya and north-central Tanzania mittances (Thornton et al., 2006). 550 J. McIntire, T. Robinson and C. Bosire Experimental and quasi-experimental treat- grazing systems (Le Houérou, 1989) relied heav- ments were: (i) higher-yielding fodder resources; ily on the ILCA/Malian IER work at Niono and in (ii) policies to encourage greater offtake from the the Niger Delta of Mali and on earlier research herd; (iii) grazing and water management; and by the L’Institut d’Elevage et de Médecine Vétéri- (iv) management and health of sheep and goats. naire des pays Tropicaux (IEMVT) on Sahelian livestock and agrostology in Niger and Senegal. Niono, Mali and the Malian Delta The Mali studies differed from the others in not having formal experiments with productiv- The Mali livestock systems (LGA) research was ity treatments. They relied on statistical analysis led by ILCA and the Malian Institut d’Economie of implicit ‘treatments’ arising from surveys of Rurale (IER) around the town of Niono in cen- herd and flock effects on reproductive perform- tral Mali near the Niger River. There were two ance, nutritional status, young stock mortality studies, one of the agro-pastoral systems in and seasonal respiratory diseases. The Niono areas farther from the Niger and the second in study (Wilson, 1986, p. 108) used the statistical and around the Inner Delta of the Niger. The results to make general recommendations about principal themes of both were animal health feed, management, and prophylactic and cura- and productivity, herd management, vegetation tive health measures without benefit–cost ana- and grazing patterns, and crop agronomy. The lysis. The Inner Delta study did a more thorough areas were LGA, with generally low human analysis of calf weaning policies leading to re- population density, highly variable seasonal live- commendations that could be applied by herders stock density, monomodal rainfall, good market (Wagenaar et al., 1986, pp. 45–51) and leading access in regional towns, and a diverse mix of to other proposals for pasture management in- rain-fed and irrigated crops near the animal pro- volving existing forage species rather than the duction zones. The altitude was between 200 introduction of new forage species. and 300  masl in the Niono and Interior Delta areas. The hot dry conditions (Köppen climate Kaduna, Nigeria BSh) made trypanosomiasis rare and Theileria unknown. The Kaduna sites (von Kaufmann et al., 1986)40 The main reports were those of Wilson were chosen to contrast livestock systems under (1986) and Wagenaar et  al. (1986). Wilson different land-use patterns imposed in part by (1986) was a research project on cattle and the natural environment and in part by policy. small-ruminant productivity among agro-pas- The Kaduna area is in Köppen climate Aw (trop- toralists39 near Niono in north-central Mali with ical savannah). The livestock systems are LGT field work over the period 1976–1983. The study and MRT. Two of the three sites – Kurmin Biri, by Wagenaar et  al. (1986) from 1979 to 1983 between a grazing reserve and a national forest, covered cattle productivity among agro-pasto- Abet – were around 2500 km2. Annual rainfall ralists in the Inner Delta of the Niger River in in Kurmin Biri and Abet was 1200–1300 mm in central Mali. Both Mali studies were in areas one season. Ganawuri is a higher-cultivation- without major trypanosomiasis challenge and density site, with slightly higher rainfall, of which where internal parasites were the principal ani- 800 km2 was subject to aerial survey and 40 km2 mal health risk. The Inner Delta study area dif- was subject to ground truthing (von Kaufmann fered from the Niono one chiefly in that water et al., 1986, p. 45). Kurmin Biri and Abet were at was available year-round and that the divide be- about 600 masl and Ganawuri above 1250 masl. tween wet- and dry-season pastures was much Waters-Bayer and Taylor-Powell (1986a) less pronounced. summarized population and land use in the There were associated studies of vegetation three sites. They distinguished three groups: (i) and biomass carried out by Breman and Cissé, pure pastoralists, who do not practise cropping (1977), Hiernaux (1980, 1984) and Hiernaux and are nomadic; (ii) agro-pastoralists, whose et  al. (1983), and on primary productivity and main activity is livestock but who do have crops soil–water relationships led by the University of and are less nomadic than ‘pure pastoralists’; Wageningen (Penning de Vries and Djiteye, and (iii) mixed farmers, whose main activity is 1991). A later comprehensive review of Sahel crops but who do keep some ruminants; animal African Livestock Systems Research, 1975–2018 551 traction for power was rare in this production mainly of browse and annual grasses, and crop type, which was the largest. residues. Animal disease prevalence is low enough Cultivation intensity varied from 5–15% of that Hiernaux and Ayantunde (2004) did not all land in Kurmin Biri to 25% in Abet to as mention it at all; the principal disease problem much as 40% in Ganawuri. Wet-season live- appears to be internal parasites. The hot, dry con- stock density ranged from five heads/km2 in Kur- ditions made tsetse and therefore trypanosomia- min Biri to 25 in Abet. Dry-season stock density sis very rare. ECF is unknown in West Africa. ranged from 10–20 head/km2 in Kurmin Biri to The Fakara field studies were complemented 40 in Abet to 20–25 in Ganawuri. Kurmin Biri by station investigations of dry-season supple- had at one time been declared tsetse free, al- mentation, sheep fattening, alternative grazing though the vector was apparently still found at practices such as rotational grazing and sea- Abet during the study period. Ganawuri, at some sonal grazing, and soil fertility management in- 1000 masl, had a cooler climate and lower vec- cluding manuring and mineral fertilization of tor incidence. arable crops. The field and station studies were The Kaduna studies carried out careful experi- subsequently used in bioeconomic models of soil ments of practices to improve productivity. These fertility management and household behaviour tests focused on supplementary feeding of cattle, (Barbier and Hazell, 1999). planted forages (mainly Stylosanthes spp.) and on manure for crop production (Powell, 1986a,b; Mohamed-Saleem, 1986a,b; Bayer, 1986). What did the pastoral systems find? Fakara, Niger Sedentarization The ILRI studies in the Fakara subregion of Three tenets of the ‘mainstream view’ of pastoralism western Niger – of soils, vegetation, grazing were that it was unproductive; that it inevitably practices, soil nutrient recycling, herd manage- involved overstocking, which led to environmen- ment, vegetation, household economics and la- tal damage and to conflicts with farmers; and bour use – followed earlier work in the same that reducing nomadism was a path to higher zone by ICRISAT, ILCA and the International productivity and less pasture damage due to Fertilizer Development Center (IFDC), beginning lower grazing pressure. One result of these be- in the early 1980s. The Fakara studies (Williams liefs was that sedentarization of herders was et  al., 1999, and Hiernaux and Ayantunde, sometimes advanced as an appropriate, even ur- 2004, reporting on work from 1994 to 2005) gent, policy. In addition to raising productivity, were in Köppen climate BWh at altitudes of less sedentarization would, moreover, allow provi- than 300 masl. Annual rainfall in the Fakara is sion of housing, education and health services reported to have been 560  mm from 1905 to to herders. It followed from this view, which was 1989. There is a monsoon seasonal pattern with widely held among administrators and scien- rainfall from June to September, a transition tists, that policy should seek to sedentarize com- period in October and November, a relatively pletely, settle partially by season or otherwise cool season from December to February and a restrict the mobility of pastoral groups. very hot dry season from March until the onset A prediction from the agricultural evolu- of the rains. Surveys in the 1990s showed the tion literature is that sedentarization of nomads population density to be low, in the order of responds to two forces. The first is a push effect – 10–50 rural persons per km2 depending on the population pressure and the expansion of land proximity to water. for crops or wildlife cause herders to lose grazing The livestock system is a mix of seasonal access. This loss of grazing limits feasible herd and annual transhumance. Animal density per numbers, restricts mobility and productivity of unit of land was low, ranging from 5 to 12 TLUs/km2. the remaining animals, and eventually obliges Animal density in TLUs per person was com- pastoralists to abandon nomadism for cropping paratively high, ranging from 0.15–0.20 among in whole or in part. A related prediction was that poorer families to as high as 2.5 among richer disasters such as drought or animal disease would ones. The principal feeds were natural grazing, aggravate herd loss and therefore accelerate 552 J. McIntire, T. Robinson and C. Bosire sedentarization. The second force is a pull Enforced sedentarization policies generally effect – by sedentarizing, herders can benefit failed. Grigg (1974, p. 122) that ‘…efforts to es- from access to land, cereal production, crop res- tablish Maasai families on ‘ranches’…foundered idues, water and even political representation. on the refusal of the Maasai to cull their herds to The strength of these predictions has varied a level compatible with the grazing resources greatly by agroclimate, population density and available, and their reluctance to sell their cattle’42. the emergence of markets for milk, commercial Sandford’s (1983a) review of African and meat and cash crops. Both ‘push’ and ‘pull’ pre- other work (ex-Soviet Union, the Middle East) dictions have been confirmed in the densely found that efforts to settle nomadic herds, while populated areas of northern Nigeria by Waters- integrating their owners into cropping or mov- Bayer and Taylor-Powell (1986b), and also by ing them into ranching, usually failed because Blench (1994) who gives many examples of suc- of the cost of foregone mobility and the unwilling- cessful spontaneous sedentarization of herders ness of pastoralists to abandon their livelihoods. throughout Nigeria, and in at least three pas- The sedentarization of herders in northern toral systems of southern Ethiopia and northern Nigeria was indeed found to create conflicts Kenya (Fratkin, 2001). In more recent research with farmers, as indicated by numbers of lawsuits in semi-arid western Niger, Hiernaux et  al. for crop damages caused by animals (Waters- (2009) showed a progressive sedentarization of Bayer and Taylor-Powell, 1986b, pp. 213–214). herding groups into rain-fed farming, caused by Blench (2001, pp. iv and 61–64) found that drought, with a concurrent integration of crop policies of enforced sedentarization ‘had a very and animal activities on previously settled farms, unsatisfactory history’, citing the instances of caused by demand for land and access to fodder. Tanzania, Iran, Kenya and Somalia. Homewood Northern Nigeria, and parts of semi-arid India, (2008) cites many examples from East and West illustrate the impact of population density and Africa where explicit or implicit settlement pol- service availability on sedentarization, while icies have failed, sometimes with disastrous East Africa illustrates the impact of recurrent consequences for the rights and livelihoods of droughtsc. herders. Induced sedentarization may, moreover, A related prediction was that pastoralism, have produced land degradation by concentrat- because of herders’ knowledge of animal pro- ing people and animals in smaller areas of scarce duction, would evolve more easily into mixed resources, such as dry-season water and pas- farming than arable farming without livestock. tures, but it is not possible to quantify such The latter would involve aspects of arable farm- adverse effects with existing information43. ing, ley farming, animal traction, use of manure After enforced mobility restrictions during for cropping, and commercial livestock, such as the collective era in Mongolia, herders ‘very rap- dairying or ranching for meat. This set of predic- idly reverted to traditional mobile transhumance’ tions seems to have succeeded only in some as- when the collectives were dismantled (Suttie pects of arable farming and to have generally et al., 2005, p. 294). Arable farming was indeed failed with respect to ley farming, animal trac- taken up by pastoralists (Blench, 1994, 2001), tion, planted forages and commercial livestock with marked differences among subregions in production. The successful predictions – use of sub-Saharan Africa. The reasons for sedentariza- manure in cropping, use of crop residues to feed tion were necessity – herders had to settle after animals, use of animals for transport and culti- losing most or part of their animals, in part be- vation – were confirmed by McCown et  al. cause of long-term conversion of grazing and for- (1979), in a study of agro-pastoralists in central est to cropland and in part because of opportun- Nigeria (Powell and Taylor-Powell, 1984), in the ity to exploit interactions between crops and ILCA research on agro-pastoralists in north- livestock in the same enterprise. Sedentarization central Mali (Wilson, 1986), more generally across could not be sustainably enforced by public policy. semi-arid and subhumid Africa (McIntire et al., It was also reversible – as soon as assets would 1992), particularly Chapters 5 and 6) and Hill’s allow, settled pastoralists would revert, at least for (1982, pp. 21–23) book on two dry farming vil- part of their family, back to a mobile system. lages in Hausaland and in the international The extensive systems most carefully stud- comparisons by Baltenweck et al. (2003). ied – Borana, Maasailand, Kaduna, north-central African Livestock Systems Research, 1975–2018 553 Mali, south-western Niger, the Kenyan–Tanza- of herders to drought (or, in Mongolia, to ex- nian border – seem to have regressed in area treme cold) and related pasture loss. Second, under pressure from arable farming (Kaduna), there was the devaluation of traditional herding bush encroachment (Borana) and wildlife reser- authorities in allocating resources and in man- vations (Maasailand). One study of northern Ni- aging conflicts about pastures and water. Third, geria44 found that the principal land-use change there may have been some productivity loss was loss of woodland. Another factor, which has caused by greater exposure to epizootics as herds been more difficult to map, is the taking of land became denser in smaller areas. Last, there were for cropping, whether by governments for sugar conflicts between herding and farming groups, estates in the Awash Valley of Ethiopia or by which may again have become more frequent as small holders on what was formerly lowland herds sought access to smaller areas for water seasonal grazing in central Nigeria (von Kauf- and dry-season pasture. mann et  al., 1986). The review by Reid et  al. Organizational changes in grazing have (2005, pp. 46–48) noted that the encroachment often been proposed and have nearly as often of crop farming had reduced wildlife (wilde- failed. An early example of a failure to manage beest) populations but not those of cattle on the mobility was the attempt to introduce grazing Kenyan–Tanzanian border areas. blocks among cattle and camel herding pastoral- ists in semi-arid north-eastern Kenya (Helland, Grazing organization 1980). Helland found that the complex manage- ment required for the grazing blocks was incom- A concomitant policy to sedentarization was to patible with the traditional institutions charged modify the traditional social organization of with managing the seasonality of water and for- grazing. This traditional organization was typic- age in the area. Galaty (2013a) found that group ally labelled as ‘collective’, ‘communal’ or ‘com- ranches in Maasailand in both Kenya and Tanza- mon property’. It was believed to provide weak nia harmed the interests of the traditional herd- incentives to sell animals by lowering the private ers through a process of institutionalized land costs of grazing while simultaneously encour- grabs. Similarly, confiscation of grazing lands aging overgrazing. and associated settlement policies by the Imperial A common measure was to re-allocate pas- and derg regimes of Ethiopia from the 1950s to ture by giving selected groups exclusive access to the 1980s had adverse effects on pastoralists’ grazing. Such policies were enforced in Maasail- interests (Mulatu and Solomon Bekure, 2013). and (Solomon Bekure et  al., 1991) and in One path by which subdivision of grazing north-central Nigeria (von Kaufmann et  al., would have affected the distribution of animals 1986; Blench, 1994). In Maasailand, this was was through its effect on pasture productivity. the introduction of group ranches45. In nor- Subdivision tended not to affect productivity in thern Nigeria, it was the definition of grazing ranges of very low or very high productivity but reserves with remapping of herding family ac- tended to reduce it significantly on ranges of inter- cess to the reserves. In several francophone mediate productivity. The effect of restructuring nations of West Africa, it was the promulgation on the inherent seasonality of pastoralism, espe- of a ‘code pastoral’ that delimited (very large) cially on dry-season grazing, is not clear at the dif- grazing areas and corridors through arable ferent study sites. In Maasailand, Solomon Bekure lands but without reorganization of pastoral et al. (1991, p. 35) found that ‘the degree to which groups (FAO/CIRAD, 2013). In Mongolia, it was group ranches altered management strategies the collectivization of pasture lands during the cannot be determined with available data’. The later Communist era (1950s–1980s; Mearns, same study noted impressions of changes in water- 1996, p. 310) with reorganization of herding ing frequency, use of salt licks and acaricides, and districts into brigades under the control of the introduction of the improved Sahiwal breed, but local collective administration. could not quantify these changes or estimate their There were common problems in each at- impact on productivity46. tempt at reorganizing pastoral groups and their Ellis and Swift (1988) had earlier shown grazing rights. First, there was the cost of re- that the non-equilibrium rangelands model de- stricted mobility in terms of increased exposure pended so much on mobility that demarcation 554 J. McIntire, T. Robinson and C. Bosire of ranches would damage system productivity. literature reviews. ILCA commissioned three The effects of conversion, and related changes, literature reviews of livestock water use in Af- were carefully studied by Angassa and Oba rica. King (1983) studied water intake and me- (2008) in Borana. They found that range enclos- tabolism by different species and their effects on ures led to grazing land fragmentation; associ- productivity, Sandford (1983b) focused on water ated with new arable farming, the overall effect planning and Classen et al. (1983) characterized was a degradation of range productivity. Earlier water use in tropical Africa47. work related a ‘decline in total biomass produc- Sandford (1983b) recommended the tion and animal performance’ to ‘increases in f ollowing: human and animal populations’ and to ‘de- creases in grazing land’ (Angassa and Beyene, • Collecting new data on livestock and water 2003). The adverse effects of restricting pastoral point density, which existed then only in mobility were also observed in Inner Asia arid zones. (Sneath, 1998). • Gathering new data on watering practices A related example is from Borana. The land in semi-arid areas, which were (then) scar- reform imposed by the derg government (1974– cer than in arid areas. 1991) in Ethiopia was associated with ‘semi- • Studying actual water use and comparing it privatization’ of grazing lands, range enclosures with optimal use to test the hypothesis that and the suppression of fire as a bush manage- producers overwater. ment practice (Angassa and Oba, 2008). These • Studying site design to reduce production policies had the aggregate effect of ‘…forage costs. scarcity and greater vulnerability of stock dur- • Recovering water costs from users to pay for ing drought years’. These parallel findings from system maintenance. Maasailand and Borana are good examples of Water for pastoralism has expanded greatly policy ignoring what research had predicted since the 1970s, pushed by voluminous public about the value of traditional management in and private investments. Research has identified maintaining rangeland productivity and in redu- defects in the design and management of pas- cing the environmental costs of invasive species. toral water supply (Sandford, 1983b, pp. 63–67; Homewood and Rodgers, 2004, pp. 249–253), Water but its contribution to better designs is unproven. Water was long understood to be an essential experiments. An important point in the studies component of higher pastoral productivity, of re- by King (1983) and Sandford (1983b) is the ef- sistance to drought and of recovery after drought. fects of water availability on feed intake in arid Basic questions about water development were: zones and on the energy used in trekking to • water. Restricted water availability by season and Could potential water supply support live- because of distance incurred costs of foregone stock production growth? What are the milk and live-weight production; estimates of barriers to water development? • trekking energy and its productivity effects could Can water consumption by pastoral stock therefore be used to optimize water investments. become more efficient? • The evidence of Sandford (1983a, pp. 73–85) on What has been the impact of water devel- water-productivity effects was generally positive opment on pastoralism? • from arid zones in India, Australia and the Have there been adverse effects of water de- Middle East. velopment? Experiments on water and productivity Research on pastoral water was of four were sparse in Africa, despite the importance of types: (i) literature reviews; (ii) experiments on the water-energy–productivity relationship stations and in other controlled conditions; (iii) stressed by King (1983). A rare and important original field investigations, including the system contribution to understanding water-produc- studies; and (iv) recent surveys and modelling of tivity functions was Nicholson (1987a; 1987b, the ‘water footprint’ of livestock, which became a 1989) in Borana. He established that traditional way of estimating total water use by animals. 2-day and 3-day watering of Zebu cattle would African Livestock Systems Research, 1975–2018 555 not compromise productivity if the animals pieces on African pastoralism analyses many were adequately fed48. There are few other ex- policy/technical options but does not treat water periments that studied water-use efficiency development as one of them (de Haan, 2016); among African pastoralists. The reviews of the other (Cervigni and Morris, 2016) makes range productivity of ILCA (1975), Pratt and the rather general recommendation to develop Gwynne (1977, for East Africa), Le Houérou ‘water resources to allow better access to under- (1989, for West Africa) and de Haan (2016) exploited rangelands’ (p. 85). scarcely mention water except to apply stand- ard consumption coefficients per animal. The adverse effects of water development. There limited research on water and livestock prod- are adverse effects of water development. The uctivity in arid Africa can be partly explained by main negative effect is soil erosion caused by in- the fact that additional water is seen as unam- tensive cultivation, or by trampling stock, near biguously good. The demand for water among water points (Sandford, 1983a, pp. 76–78; pastoralists is so high that public agencies and Wilson, 2007). A 4-year station study of grazing private water producers are often willing to in- effects on soil properties in western Niger found vest beyond what is technically optimal. variable impacts on soil compaction and water infiltration as functions of grazing pressure system studies and other field investigations. (Hiernaux et  al., 1999). Pratt et  al. (1997) did The pastoral system studies (Niono, Mali Delta, not quantify the adverse effects of water devel- Maasailand, Borana and Fakara) identified water opment although they argued that such effects as a constraint, yet their contributions to an- can be managed with simple technologies. swering the water questions are surprisingly Another potential adverse effect is that limited. The Maasailand and Borana studies are mechanization of boreholes might reduce la- the only ones to answer two of the basic ques- bour demand. If labour is an important mech- tions about water adequacy for production anism of patron–client relationships among growth and the cost of new production, while animal-r ich and animal-poor herders, then the Borana study is the only one to conclude mechanization can weaken such relationships about efficiency. The Niono and Mali Interior (Coppock, 1994). Lower labour requirements Delta studies mention water only in passing and owing to mechanization would reduce the bar- have nothing quantitative on consumption, gaining power of poor pastoralists and deprive production, cost or efficiency. The subhumid them of employment, income and, potentially, Nigeria study says little about water, except to finance. The West African literature is silent on report daily livestock watering times (von Kauf- possible labour displacement effects of wells, al- mann et al., 1986, p. 431) and to make the evi- though it says a great deal about water as a dent point that seasonal water is less limiting cause of conflict between herders and farmers than it would be in a semi-arid zone such as Bora- (Wilson, 1986; Wagenaar et  al., 1986; Le na or Niono. The Maasailand study did detailed Houérou, 1989). research on pastoral water infrastructure (type, scale and location), production and consumption. water footprint and productivity. The water The principal books on African pastoralism, all footprint literature can provide information based at least in part on field investigations since about changes in feed composition and water- the mid-1970s49, say very little about technical use efficiency per animal (Hoekstra and Mekon- water issues. nen, 2012; see also Chapter 11, this volume). The pastoral studies usually made general One application of this literature is to partition qualitative recommendations to raise water pro- change in total livestock water use per unit duction, to allocate water more equitably among of output as the sum of direct consumption users and to prevent damage around wells and by animals and indirect consumption by for- ponds. Water research had little or no impact on age plants. One long-term study in arid, semi- productivity except in the sense of understand- arid and humid zones of Kenya found that ing system dynamics and explaining geographic total water consumption by cattle, sheep and patterns (e.g. Mesele et  al. 2006 on the Yabelo goats, and camels in each of the three zones was district in Borana). One of two major recent closely proportional to animal numbers (Bosire 556 J. McIntire, T. Robinson and C. Bosire et  al., 2015, p. 36, covering 1977–1990 and nomadic herders without arable crops, pasture 2001–2012). This implies that indirect water grazing and cereals residues, obtained by trans- consumption has not changed relative to the actions with farmers52, dominated. total because feed composition has not changed, Natural pastures, crop residues, browse and nor has water-use efficiency per animal. Add- planted forages were the major themes of the itional inferences are that species and breed initial ILCA systems studies53. Every pastoral composition within zones have not shifted to- study started from the understanding that aver- wards water-efficient animals. age feed availability from all sources – grazing, cut grass, shrubs, forbs, browse and crop res- Feed systems idues – limited animal productivity; and that in- terannual and interseasonal variability in feed Most African pastoral research focused on its supply was a constant threat to herd viability54 geography, society and anthropology until the and growth. volumes of ILCA (1975) and Pratt and Gwynne Every study of mobile grazing recom- (1977) and the initial ILCA literature reviews. mended improvements to feed quantity and qual- While it was long evident that feed, after disease, ities. Suggested improvements involved planting was the principal limitation to livestock produc- legumes and more productive grasses (both tion in grazing systems, older research had done African and exotic species), control of bush en- little on possible technical changes in feed, such croachment, fertilization of pastures and grazing as introducing fencing, legumes, exotic grasses, management. For example, the study by Penning fertilization or mechanization50. The technical de Vries and van Heemst (1975) argued that ‘…it aspects of older work emphasized water and vet- is expected that appropriate legumes and Rhizo- erinary problems almost exclusively, with less at- bium strains can be found to increase the soil ni- tention to other biological issues, to economics or trogen supply in Sahelian natural grasslands and to introduced technical change through research agricultural fields’. and extension51. Feed research in Africa there- fore began from a comparatively weak basis. experiments. Some experiments on feed prod- Feed systems research had four components: uctivity were done by the colonial and national (i) field characterization of available and potential programmes before the advent of international feeds and their use, with emphasis on rangelands livestock research. The general conclusion of the characterization (many of the papers in ILCA, many detailed empirical investigations of feed 1975); (ii) station characterization of the prod- improvements in pastoral zones of sub-Saharan uctivity and nutritive value of such feeds as plant- Africa is that it has been practically impossible to ed forages, crop residues, and browse; (iii) field make economic improvements in range product- experiments of these feeds; and (iv) adoption ivity under African conditions. This holds under studies of promising new materials. such treatments as planted forages, whether legumes or grasses, and with soil fertility amend- field characterization. The early pastoral ments, such as mineral fertilizers or manure. studies found that feed consisted of grass, forbs, Pratt and Gwynne (1977, pp. 110–128) arrived crop residues and tree browse, as shown in ILCA at this conclusion for East Africa. In West Africa, (1975), Fricke (1979, for Nigeria) and in the Le Houérou (1989) found the same when refer- references in Le Houérou (1989), McIntire et al. ring to more than three decades of research, as (1992) and Baltenweck et al. (2003). Purchased did Fricke’s (1979) review of field and experi- feeds and planted forages were practically non- mental work on cattle in Nigeria in the 1960s existent, even in stall-feeding systems where and the 1970s. crop residues and cut grasses dominate. Even in A careful long-term experiment was the South Asia and highland Ethiopia where the use Projet Productivité Primaire au Sahel (PPPS) of animals for power is common and hence cre- study of Mali, (Penning de Vries and Djiteye, ates a constant source of demand for feed among 1991), which found that: (i) soils could support immobile animals, the principal forage was res- higher stocking rates where water is available; idues from multidimensional cereals, such as and (ii) grazing cover management reduced soil teff, sorghum, pearl millet, maize or wheat. For degradation and bush encroachment. African Livestock Systems Research, 1975–2018 557 Studies of soil fertility management using Pratt and Gwynne (1977) concluded that re- mineral fertilizers in African grazing systems search on range reseeding using exotic grasses were rare because the null hypothesis of no showed ‘little promise’ and should be considered treatment effect was strong, even from qualitative ‘a last resort’. Ruthenberg (1980, pp. 110–125 observations. Introduced treatments – fertilizers, and 322–355) was negative on ley farming for recycling manures or residual vegetation, plant- pasture improvement for the same reason – in ed forage legumes or grasses – have not gener- the absence of dairying, expected benefits were ally succeeded. The magisterial work of Penning negative. The PPPS study in Mali found little de Vries and Djiteye (1991, p. xix, passage trans- average effect of reintroducing perennial grasses lated from the French by the editors) covering (Penning de Vries and Djiteye, 1991, p. 469) but 5 years of experimentation in semi-arid central some effect on the ‘stability’ of soils and pas- Mali noted that, ‘The immediate introduction of tures. A hypothetical model of ley farming in a new forage legume species is likely to fail for the cotton-based system of southern Mali gave in- same reasons – competition from cereals, low conclusive results and does not seem to have germination capacity, lack of soil P, and low stimulated adoption of pasture fertilization (Bos- rainfall’. The review by Thomas and Lascano ma et al., 1999). Doppler’s (1980) book on beef (1995) found benefits in soil quality, pasture cattle production in subhumid Togo found that productivity and animal productivity from mix- mineral fertilizers and pasture improvements ing legumes into acid-soil pastures in Latin were generally too unprofitable for smallholders, America; however, they did not specify the ul- although they would be suitable for ranches and timate adoption of the proposed legume innov- larger farms. ations and hence it is not possible to estimate the ILCA studied the agro-pastoral areas near development impact of their research. Macharia Niono in central Mali (Hiernaux et al., 1983; Wil- et al. (2011) showed that introduced legumes on son, 1986) over the period 1978–1984. Wilson semi-arid pastures in Kenya could improve soil et al. (1983, Chapter 6) elegantly describes vege- quality (pH, organic carbon, nitrogen and potas- tative composition, tree density and cover in the sium) on station, but the on-farm phase of the three principal land-use types near Niono. Its trial only lasted 1.5  years, did not conclusively analysis of the species composition and yield of show a soil-quality effect and apparently did not natural pastures compared treatments with ma- lead to field adoption of legume technologies55. nure, added phosphorus, grazing management Studies at one site in semi-arid central Mali and ploughing. It showed (Wilson et  al., 1983, (Penning de Vries and Djiteye, 1991) established pp. 47–50), as had the PPPS work, that the Sa- the basic abiotic (soil and water) and biotic (com- helian rangelands had a high production poten- petition from grasses) constraints to planting tial but that added mineral fertilizers at a rate of legumes in extensive grazing systems. These con- 22 kg ha-1 were ‘probably not profitable’ (Wilson clusions are replicable throughout the semi-arid et  al., 1983, p. 50, authors’ translation of the tropics of Africa and South-east Asia; exceptions French original). Farm management research in Latin America are discussed in Chapter 12 near Niono station, as reported by Toulmin and (this volume). The Mali research concluded that, Fulton (1983, pp. 107–119), found overseeding in arid situations of less than 750  mm annual of introduced forage species to be unsuccessful rainfall, the introduction of legumes into natural for both grasses and legumes in agronomic and pastures was unlikely to be economic (Penning economic terms over the years 1977–1980. de Vries and Djiteye, 1991, p. 435), in part because An example of the benefit:cost ratio of of higher temperatures that lowered legume planting a tree legume (Leucaena leucocephala), yields and in part because of adverse meat:ferti- simulated under commercial beef ranch condi- lizer price ratios (Penning de Vries and Djiteye, tions, came from the initial ILCA cattle model in 1991, p. 470). Botswana (Cartwright et al., 1978, pp. 55–58). Depending on projected tree establishment costs pasture improvement56. Pratt and Gwynne (1977, and Leucaena yields, the model projected an IRR p. 122) analysed the technical and managerial of between 2.7% and 22.4% at median aspects of sown pastures in East Africa. Despite e stablishment costs and tree yields (Cartwright not considering the costs of sown pastures, et al., 1978, p. 78). 558 J. McIntire, T. Robinson and C. Bosire Recent experience is that the use of planted means (introduction of competing crops and use or protected forages, whether as arable crops or of mineral fertilizers) are low soil fertility and as browse trees and shrubs, has grown very lack of water (Penning de Vries and Djiteye, slowly from the 1960s to the present. Such for- 1991). There is some evidence of benefits from ages are still insignificant in East African and overgrazing in terms of preservation of peren- West African pastoralism, even where formerly nials via the effect of lesser competition from highly mobile grazing systems have evolved to heavily grazed annuals and protection against include more arable farming (e.g. Beyene, 2016, invasive species, again via lessened competition. for eastern Ethiopia). impact. The impact of feed improvements in bush encroachment. Bush encroachment was grazing systems has been very weak in sub- carefully studied many years ago. The basic East Saharan Africa (Table 15.3). Over the 40 years African text on rangeland management (Pratt of research on grazing systems, novel plant ma- and Gwynne, 1977, pp. 128–138) gave an over- terials remained a small fraction of feed intake, view of ‘troublesome’ bush species and their po- although total animal production grew substan- tential control, including by fire, machines and tially. There is little evidence of planted forages grazing (e.g. using goats to graze species unpal- in sub-Saharan Africa in extensive grazing sys- atable to cattle and sheep). The Borana study tems, even on limited areas such as feed gardens (Coppock, 1994) reviewed bush encroachment for segments of mobile herds or as calf supple- and control in the context of grazing dynamics, ments. There have been successes in humid areas vegetation succession and soil. Its conclusion of Latin America (see Chapter 13, this volume), (updated in Angassa and Oba, 2008) is that but these have most often been on large ranches there is no general prediction about the effects of or on commercial dairies. From the Mediterra- bush encroachment, which might even be posi- nean mandate area of ICARDA, we know that tive in some stages of land-use evolution. Predic- long-term research on Medicago sativa was un- tion of crop encroachment reducing pasture successful in achieving agronomic or develop- area and pasture productivity was later con- ment impact on a large scale57 although there firmed by satellite image studies in Borana have been some benefits from feeding spineless (Mesele et al., 2006). cactus (Chapter 13, this volume; Jutzi and Rich, The economic analysis of bush control is 2016) in Brazil and in North Africa. not well established. Pratt and Gwynne (1997) The negligible impact of African agricul- apparently concluded that fire was the most tural research on forage productivity in grazing cost-effective technique for controlling bush en- systems has some precedent. Two studies of croachment in East Africa but did not systemat- China mention no significant impact from re- ically compare fire with other methods. Sandford search on sown pastures or planted forages. (1983a) concluded that chemical, mechanical Perkins’ (2013) study of Chinese farming over and biological methods of bush control are un- six centuries makes no reference to pastures or likely to be effective in the African tropics, as did planted forages in any subsystem – not in graz- Homewood and Rodgers (2004, p. 252) for the ing or mixed systems, in temperate or tropical rangelands of the Ngorongoro Conservation areas, or with reference to milk, meat or draught Area of Tanzania. Fricke (1979) did not really power. Even discussions of cereal crop residues discuss the financial aspects of bush manage- and multidimensional crops, such as sugar ment but noted the importance of fire for the beets and soybeans, do not mention their use sustainability of grazing systems in Nigeria. The by animals. A contemporary comparison of the central Mali study at Niono ranch examined the lessons of Chinese agriculture for Africa does effects of fire, species and cutting on vegetation, not mention livestock research – breeding, man- palatability and soil quality but did not review agement, pastures, sown forages, genetics or chemical or mechanical control (Penning de nutrition – as a factor in the more rapid agri- Vries and Djiteye, 1991, pp. 346–352). cultural growth of China compared with sub- Chief among the reasons for the failure of Saharan Africa (Li et al., 2012).Some of the fail- pasture improvement and related efforts to ure to intensify pasture production stems from control bush encroachment through biological the failure to understand that conversion of African Livestock Systems Research, 1975–2018 559 Table 15.3. Feed recommendations in systems studies, various years. Country/ Labour intensity of Study Region/period subregion/area Climate Farming system Species Feed recommendations recommendations Dahl and Hjort East Africa Western Sudan, Arid to LGA Cattle, camels, Not specific Unknown (1976) northern Kenya semi-arid sheep, goats Dyson-Hudson Mainly East Several Arid to Extremely Ruminants None Unknown and Dyson- Africa subhumid variable Hudson (1980) mobility types Jahnke (1982) Sub-Saharan Very arid to LGA, LGH, Ruminants Grazing, deferred Probably much higher Africa, 1970s humid LGT, MRA, grazing plus fertilizer MRH MRT Sandford (1983a) Sub-Saharan More on East Mainly arid to LGA, LGH, LGT Cattle, camels, Variable across Unknown Africa, Africa semi-arid goats production 1960s-1970s subsystems de Leeuw and Sub-Saharan Many Arid, semi-arid LGA, LGT, LGH Ruminants and Not specific Unknown Milligan (1984) Africa, 1984 wildlife von Kaufmann 1978–1986 Kaduna, Nigeria Semi-arid to MRH Cattle, sheep, Crop residues, pasture Mixed (highest for S. et al. (1986) subhumid goats with S. hamata hamata) Wilson (1986) West Africa, Mali Arid, semi-arid LGA, LGH, Ruminants Non-specific fodder and Unknown 1980s MRA, MRH, forage crops, MIA, MIH including browse Ellis and Swift East Africa, Northern Kenya Arid, semi-arid, LGA Cattle, camels, Not specific Unknown (1988) 1980s subhumid goats, sheep Solomon Bekure East Africa, Kajiado County, Semi-arid to LGA, LGH, Cattle Feed gardens with Mixed (highest for et al. (1991) 1980–1991 Kenya subhumid limited MRA, Panicum maximum Panicum, less for (Maasailand) MRH or Pennisetum browse) purpurem, cowpea, pigeonpea or Leucaena Homewood and East Africa Kenya/ Semi-arid to LGA, LGH, Ruminants and Natural vegetation Unknown Rodgers Maasailand, subhumid MRA wildlife (pasture, browse) (2004) Tanzania/ Serengeti Behnke et al. East Africa, Ethiopia, Kenya, Several LGA, LGH, Ruminants and Natural vegetation Unknown (1993) 1970s–1980s Tanzania MRA wildlife (pasture, browse) Continued 560 J. McIntire, T. Robinson and C. Bosire Table 15.3. Continued. Country/ Labour intensity of Study Region/period subregion/area Climate Farming system Species Feed recommendations recommendations Coppock (1994) East Africa, Ethiopia/Borana Semi-arid to LGA, LGT, MRA Cattle, camels, Hay making with grass, Highest with hay, lowest 1980–1991 subhumid goats, sheep Acacia browse, with browse pigeonpea Barbier and Niger, 1990s Niger LGA, LGH LGA, LGH, Ruminants Purchased feed of Unknown Hazell (1999) MRA, MRH unspecified origin Kruska et al. Early 2000s global Arid, semi-arid LGA, LGH, Ruminants Not specific Unknown (2003) MRH Hiernaux and West Africa, Fakara, Niger Arid to LGA, MRA Ruminants Crop residues (millet, Bran mentioned in Ayantunde 1990s semi-arid cowpea), millet passing, not included (2004) in economic analysis of traits Homewood Sub-Saharan Many LGA, LGH Ruminants and Natural vegetation Mentions division of (2008) Africa wildlife (pasture, browse) labour, not linked to feed practices Reid et al. 1990s and Athi-Kaputiei Arid, semi-arid LGA, LGH, Ruminants and Not specific Not studied (2008b) 2000s plains MRH wildlife Hiernaux and West Africa, Niger Arid, semi-arid LGA, LGH, Ruminants Crop residues Careful analysis of Ayantunde 1990s MRA, MRH labour use in crops (2004) Konczacki (2014) Literature review Global Arid, semi-arid Several Ruminants Not specific Unknown African Livestock Systems Research, 1975–2018 561 grazing land to cropland is not necessarily a net mainly young stock and having a strong inter- loss of feed. The net change is positive at lower action with animal nutrition. There was a low levels of cultivation intensity and may not be- incidence of trypanosomiasis in cattle (Coppock, come negative until quite high levels. This ap- 1994, Table E2) because herders made careful parent paradox is due to the fact that crop res- choices of grazing territories. Despite detailed idues from newly planted areas are valuable animal health status measures taken in the Mali feeds and can be stored more easily into the dry Inner Delta study on brucellosis, internal para- season(s) than pastures. Cropping would be ac- sites, contagious bovine pleuropneumonia and companied, to some degree, by destruction of in- trypanosomiasis (Wagenaar et  al., 1986), this vasive browse species during the fallow parts of work did not lead to animal health improve- the cropping cycle, thus improving the fodder ments or to associated productivity gains. The value of the fallow. In addition, such pasture same lack of productivity effect of the field substitutes as processed feeds or cut fodder are health research was seen in the Niono and sold near expanding towns (e.g. for the ground- Kaduna studies. nut basin of central Senegal (Mortimore, 2000); the Kano close-settled zone and other parts of Modelling pastoralism northern Nigeria; and the areas around Niamey, Ouagadougou, Bamako and Addis Ababa). Numerical models, based on coefficients derived Moreover, secondary feed effects occur, as noted from experiments and surveys, are useful for by Mortimore and Turner (2005), including simulating the impact of technology and policy planting trees on cropland to replace woodlands on livestock and crop production. Many models lost to cropland expansion with variable effects were developed to simulate livestock demog- on feed balances depending on climate, crop raphy and technology in the expectation that density and tree species. new methods could be adopted by herders or that such methods could be promoted by appro- animal health. Animal health was not the main priately targeted policies. Despite the effort in- point of pastoral characterization research, except vested in models, they have generally failed, in for field studies of epizootics, because so much African situations, to promote new technologies was already known and because ILRAD’s focus or policies. The reasons have been: (i) complex was on parasitology and immunology. The main data requirements; (i) failure to measure import- diseases of grazing ruminants were: (i) trypano- ant parameters in many studies; (iii) sampling somiasis, caused by protozoan parasites trans- variation in model parameters estimated from mitted by the blood-feeding tsetse flies of the surveys and experiments; and (iv) inability to genus Glossina, which tend to be more abundant use natural parameters, especially rainfall, as in the humid and subhumid zones; (ii) Theilerio- control variables for policy experiments. sis, most often known as ECF, another protozoan Significant research in rangeland mod- parasite transmitted by a tick of the genus Rhipi- elling – soils, water, pastures, live-weight re- cephalus (Boophilus), which is limited to 12 East sponse to feed intake, management, species and African countries; and (iii) rinderpest. The prin- breed effects – was made outside ILCA’s man- cipal contributions of animal health work, in date area before the mid-1970s, as presented at terms of systems research, were those on trypa- the 1975 rangelands symposium (ILCA, 1975)59. notolerance and ECF58 as shown in Chapters Much of this work was not directly applicable 2–6 and 10 (this volume). to sub-Saharan African conditions because Animal health was the domain in which model processes were derived from range situ- ILCA systems research had the weakest impact ations – climate, breed and animal disease – in precisely because the principal animal health the USA, Australia or Latin America60. Hence, work was laboratory based or was already the field modelling work began at ILCA in the object of international control campaigns. For 1970s from a narrow database and confronted Borana, Coppock (1994, p. 140) reported that difficult problems of using data and model ‘animal health was never a significant focus of structures from situations in the USA, Austra- research’. The principal afflictions of cattle were lia and Latin America that were rarely closely pasteurellosis and external parasites, affecting comparable. 562 J. McIntire, T. Robinson and C. Bosire primary productivity. Animal production in nitrogen once the ratio of phosphorus:nitrogen extensive grazing systems depended more directly reached a given level as determined by soil type. on grazing resources than it did on mixed-farming Subsequent station and field research tested systems because the latter can more econom- how two major constraints – water and soil ically exploit cut fodder and purchased con- fertility – could be managed biologically while centrates. Despite their importance, studies of producing adequate financial returns. primary productivity in grazing systems were These soil–water–plant models established rare until comparatively recently because of the the determinants of plant production in the long-term data requirements necessary for inte- grazing areas. However, it has been impossible to grated soil–water–plant–animal models and the apply their findings to technology and policy in resulting high cost of such studies61. sub-Saharan Africa given that rainfall is so vari- What did the models show about the po- able and cannot be managed in arid climate tential primary productivity gains as functions without irrigation. The inability to control the of climate, soils and technology?62 The litera- main exogenous variable – water – becomes ture review by Le Houérou and Hoste (1977) more binding in arid sites where most of the ani- found close relationships between rainfall, as a mals graze. good proxy for water availability, and pasture production in Mediterranean areas (sites ran- herd demography and stocking rate. Models ging from 70 to 900 mm of rainfall) and in the having stocking rate as a control variable were Sudano–Sahelian zone of sub-Saharan Africa typically only possible under ranching condi- (200–600 mm). The most comprehensive early tions. For this reason, several early simulations work on potential primary production around (Cartwright et  al., 1978, 1982; Konandreas the time of the founding of ILCA and ILRAD and Anderson, 1982; Konandreas et al., 1983) was that at the Niono ranch (Penning de Vries adapted a US model for ranch and rangeland and Djiteye, 1991, is a summary of many pa- conditions in arid Botswana. These efforts speci- pers). A model of primary productivity, derived fied a detailed model of cattle production in with annual rainfall levels below 600 mm in a which pasture productivity was exogenous and hot Sahelian climate, was reported by Penning in which the live-weight response to feed intake de Vries and Heemst (1975, pp. 323–328). was the main function to be estimated. They concluded that a basic plant production This model’s major result was the projected function of water availability was applicable to relationship between feed resources and a ‘per- annual grasses on sandy soils under arid and missible’ stocking rate for different classes of semi-arid conditions in Mali. Later work (van herders over a 30-year period, which included Keulen et al., 1981) on semi-arid pastures con- a sample range of very good to very bad years. cluded that the model’s predictions were not It evaluated four production alternatives – calf ‘very sensitive’ to most weather parameters, ex- weaning age under ranch conditions (Cart- cept rainfall, and could be ‘applied with reason- wright et al., 1978, pp. 46–49) and three innov- able confidence’ to sparse data situations in hot ations on traditional rangeland cattle posts: a climates. It was noted that this primary model lower weaning age, a shorter breeding period was only applicable to annual grasses and not and reserve pastures for calves (Cartwright et al., to legumes or to woody species, even in the 1978, pp. 49–55). None of the rangeland innov- same climate and soil conditions (ILCA, 1975, ations appeared profitable in the simulations pp. 336–337). and, applied jointly, the three achieved only ‘mo- The principal findings of the Niono ranch dest’ productivity gains (Cartwright et al., 1978, study were that day length limits biomass pro- p. 55). Konandreas and Anderson (1982) de- duction, and this can only be managed within fined policy experiments for supplementary feed- narrow boundaries set by other constraints, such ing, calf offtake, weaning, sales/purchases and a as water and soil nutrients. Water was identified drought response consisting of sales and supple- as the second limiting factor for biomass produc- mentation for stock remaining in the herd. Ko- tion; the growth response to water is linear until nandreas and Anderson (1982) did not report it reaches a boundary set by soil fertility. Phos- policy simulations as functions of price, weather phorus was the next limiting factor, followed by or exogenous technology. African Livestock Systems Research, 1975–2018 563 calf nutrition. Lambourne and Butterworth the additional labour demands resulting from (1983; in IDRC/ILCA, 1983) stressed the im- supplementation. portance of calf nutrition for herd growth and The work of Coppock (1994) and Solomon argued that the conflict between calves and Bekure et al. (1991) generated long-term predic- people for milk incurred an opportunity cost in tions for the viability of the Borana and Maasai- herd productivity. The initial comprehensive cat- land systems. Coppock (1994) predicted that the tle model of ILCA (Cartwright et  al., 1982, pp. Borana would be ‘increasing dependent on grain 51, 58–65) examined this argument by simulat- purchases’65, more cattle marketing and ‘less ing alternative milking strategies, with and sustainability in terms of per capita milk produc- without supplementary feeding, on traditional tion and asset accumulation’; an additional im- rangeland cattle production in Botswana. The plication of these trends would be greater culti- economic return to milking, without supple- vation of crops by the pastoralists and greater mentary feeding, was positive in the first 3 years recourse to wage labour. and fell in the following two, final, model years. A mean IRR of –14.0% was estimated for the vector control. Biting insects and ticks and no-milking strategy without supplementation, other vectors of disease can be controlled by indicating that traditional herders in Botswana spraying and trapping, by killing wildlife that are had foregone significant cash income by not reservoirs of the pathogens and by destroying milking63. vector habitats (Jahnke, 1982, pp. 142–149). Wagenaar et al. (1986, p. 48, Table 38) sub- Vector modelling to make control more effective sequently measured calf live weight by milk off- has taken many forms. Estimates of the coverage take for traditional Fulani cattle in semi-arid of spraying campaigns are quite old66; recent de- central Mali. Their major result was an elasticity tailed estimates are not available, but it is likely of calf live weight with respect to milk offtake that vector control through spraying and trap- between –0.27 and –0.38; a doubling of milk ping has been effective, although sporadic and offtake would reduce calf live weight by about on relatively small areas (Grace, 2003, p. 2, esti- one-third. The Wagenaar et  al. (1986) model mated that vector control was effective on 2% of was deterministic and did not generate a fre- the tsetse-infested area of sub-Saharan Africa). quency distribution of productivity or economic The killing of wildlife reservoirs has probably ad- benefits with respect to the milk offtake treat- versely affected the vector and hence limited the ment; hence, it could not conclude anything disease. Destruction of tsetse habitats caused by about the riskiness of offtake strategies. the expansion of human population and crop The Borana, Maasailand and both Mali cultivation has been successful but is not a man- studies all identified calf survival and weight as ageable control policy. A long-term study by Reid a major constraint to productivity; accordingly, et  al. (2000b) for 1960–2040 projected that they proposed supplementary feeding of calves many of the 23 African tsetse species would as a solution. The Mali studies reported nutritional ‘begin to disappear’ but that ‘for the foreseeable stress causing poor reproductive performance in future’ a large area in sub-Saharan Africa would both cattle and small ruminants as the chief remain ‘infested by tsetse and under threat of cause of productivity below potential (Wilson, trypanosomiasis’. 1986; Wagenaar et  al., 1986). Wagenaar et  al. (1986, p. 51) recommended calf supplementa- a new grazing model. The major change in tion but only as ‘a practice worth exploring’, al- grazing systems theory, associated with the though they concluded, given the prices of meat field observations derived from farming sys- and milk and expected live weight as a function tems research, was the shift from equilibrium of milk offtake, that the herders’ observed aver- to non-equilibrium models (Ellis and Swift, age milking period of 6–13 months was roughly 1988; Behnke et  al., 1993; Vetter, 2004). In optimal given local conditions64. Solomon Bekure the former, systems are density dependent et al. (1991, p. 145) found calf supplementation on such biotic factors as grazing intensity with purchased feed to be profitable but only if and population density; these factors deter- cow and calf mortalities were reduced ‘drastic- mined the average system performance, its ally’ and only if workers were available to meet variability and its resilience. Non-equilibrium 564 J. McIntire, T. Robinson and C. Bosire systems are density independent, and abiotic equilibrium and disequilibria reasoning into factors, such as the frequency and duration of specific situations of rangeland ecology] has drought, are determinant. Non-equilibrium fundamental policy implications for sub- rangeland systems would not revert to a steady Saharan Africa’. True as this generalization state, or to a long-term trend, as a function of may be, it has been very difficult to find ex- grazing pressure or population density67. The amples of technology or policy research to shift in the basic grazing model from ‘equilib- which economic or environmental benefits rium’ to ‘non-equilibrium’ affected the pro- may be attributed using either the equilibrium posed technical and managerial changes in the or non-equilibirum model. A ttempts to apply domains of feed, species choice, the transition non-e quilibrium grazing practices – such as ro- to settled cropping, water use and animal tational grazing from more temperate climates health. (Vetter, 2004, p. 11) – have often failed in Observations over long periods are too Africa and in the tropics more generally be- i nfrequent in semi-arid conditions to test cause rainfall is low and variable; low and vari- equilibrium and non-equilibrium models. One able rainfall makes rotational grazing riskier, comparison included a study over 27  years in because of its effects on the level, spatial and semi-arid Senegal, which found that dry periods, temporal distribution of pasture productivity, with higher rainfall variability, tended to be and enforces mobility across a much larger ‘non-equilibrium’, while wetter periods, with area than is possible for most pastoralists. lower rainfall variability, tended to be ‘equilib- rium’ and to show density dependence (e.g. over- grazing) (Miehe et  al., 2010). A subsequent paper, using data from the same sites, concluded What did the mixed-systems that vegetation dynamics are ‘driven by pre- studies find? cipitation not by grazing’ and hence are explic- able by the non-equilibrium model (Retzer, The second focus of international livestock re- 2006). Coppock’s (1994, p. 193) study of south- search was on smallholder mixed farming where ern Ethiopia argued that density dependence is crops were the dominant source of income and em- stronger in higher rainfall areas and weaker in ployment and where livestock had a secondary role. drought-prone areas, and that long-term loss of grazing reserves has tended to increase density Characteristics of mixed systems dependence, but these arguments have not been in sub-Saharan Africa rigorously tested. The policy implications of a non-equilibri- Seré and Steinfeld (1996, p. 11) defined a mixed um model would be that, because destocking system as one ‘in which more than 10% of the occurs with variations in rainfall, especially in dry matter fed to animals comes from crop droughts longer than 1 year, efforts to impose by-products or more than 10% of the total value generalized destocking independent of weather of production comes from non-livestock farming to preserve range quality fail because they im- activities. In these systems, more than 90% of pose the cost of underutilization of available the value of non-livestock farm produce comes grazing on herders. Blench (1994, 2001) from rainfed land use.’ Ruthenberg’s (1980, pp. reviewed global rangelands and argued that 1–18) classification for cropping systems applied destocking has failed and, moreover, that re- the keys of: (i) population density; (i) degree of stocking after drought or epizootics has been cultivation in the system (arable land as a ineffective because it cannot mobilize the feed share of arable plus pasture); (iii) type and dur- and water resources needed to revert to the pre- ation of fallow; (iv) share of irrigated cropping vious stocking rates. Public restocking pro- in total of irrigated plus rain-fed; and (v) grammes have sometimes failed because of in- agro-climate, typically measured by LGP and equitable distribution of new stock because of cross-tabulated against temperature, altitude corruption. and bimodal/monomodal rainfall regimes. Homewood (2008 p. 70) asserts that an A more recent classification of mixed cropping ‘expanded view [one that incorporates both systems describes them by matching details of African Livestock Systems Research, 1975–2018 565 crops and livestock enterprises against the re- Mixed farming in these three systems differed source base (Dixon et al., 2001). This classification68 from pastoralism by the presence of the following: compared with Seré and Steinfeld (1996) has the following groups: • Majority shares of cropping in total income and employment. • Irrigated farming (‘MIT’ in the model of Seré • Varying shares of livestock in income and and Steinfeld, 1996), included by Ruthenberg employment, with lower shares in food con- (1980, pp. 178–250) under ‘arable irrigation’. sumption compared with pastoralism. • Wetland rice (‘MRT’), included by Ruthen- • More unequal distribution of cattle be- berg (1980, pp. 178–250) under ‘arable tween households, although this was com- irrigation’. pensated to some degree by more equal dis- • Rain-fed farming in high-potential areas, tribution of small ruminants. such as the West African savannahs and • Less animal mobility or none, which was much of Brazil (‘MRT’). limited to seasonal transhumance or to • Mixed crop–livestock rain-fed farming in daily movements near permanent villages. highland areas, such as central Ethiopia In the Latin American systems, with private (‘MRH’). and fenced grazing lands, there would be no • Mixed crop–livestock rain-fed farming in seasonal or annual mobility. semi-arid areas, such as the West African • New livestock functions, notably nutrient Sahel north of the 500 mm annual rainfall recycling and the use of draught animals isohyet (‘MRA’). for cultivation. These functions would be • Dualistic – a mix of large commercial and much less important or even absent in Latin smallholder in the same general environ- America. ment; these were restricted to former areas • Important shares in income and employ- of European colonization in Africa but were ment of annual and semi-perennial cash common in much of Central America and crops – cotton, groundnut, rice, roots and Latin America. tubers – that did not exist in rangelands. In the Latin American and Middle Eastern Dixon et al. (2001, p. 13) defined three spe- production types, animals are the cash cific mixed smallholder systems: commodity. • Mixed crop–livestock rain-fed farming in • Important shares in income and employ- the highlands (corresponding to MRT), ment of perennial cash crops – coffee, tea, such as central Ethiopia and parts of Kenya. cocoa, rubber and oil palm – that are found Such farms were typically less than 3 ha in in mixed systems with poultry, swine and operated area, with animal traction featur- small ruminants, although much less often ing in Ethiopia although not in Kenya, and with cattle. feature mixes of annual crops, permanent • Commercial dairying with zero grazing, crops and ruminants. planted forages, purchased feeds, and culti- • Mixed crop–livestock rain-fed farming in low vation of annual and perennial crops on potential areas (corresponding to MRA), the same farm. overlapping with grazing (LGA), such as the • Commercial on-farm fattening of rumin- West African Sahel north of the 600 mm an- ants, swine or poultry, in zero-grazing or nual rainfall isohyet; farms would be much transhumant systems, using purchased larger in such areas, with low cereal and feeds and crop residues from annual and grain legume yields per hectare. perennial crops. • Rain-fed farming in high-potential areas Although it is evident that people and ani- (MRT for rain-fed and MIT for irrigated), mals are more concentrated in the mixed systems, such as the West African savannahs and in estimates of the distributions of tropical small- much of Brazil and Colombia. Farms in holders, and their livestock, are difficult because Latin America were typically much larger of a lack of accurate data linking production and than those found in the Sahel or in the input use to farm size and structure. It is even more highlands of East Africa. difficult to derive productivity e stimates for 566 J. McIntire, T. Robinson and C. Bosire individual farming systems, for example by the modern plant varieties in sub-Saharan Africa share of livestock in income or assets. from the 1960s to 2011. Technical progress in crop management, as indicated by uptake of levels and rates of change in farm sizes. Levels modern varieties in sub-Saharan Africa, was sig- and rates of change in farm sizes in the 1960s– nificant during that period. At the beginning of 1980s are impossible to estimate by agroclimate CGIAR research on mixed livestock systems in sub-Saharan Africa because of a lack of (1976–1980), the share of cropped areas in comprehensive data. However, published meta- seven major cereals – sorghum, maize, pearl mil- analyses make the trends clear for farm size, as let, rice, teff, wheat and barley – was low, ran- measured in area per worker, and in type of ging from roughly zero in teff and barley, to 6% farm. Masters et al. (2013) found that ‘Africa and in sorghum, to 37% in wheat, with an area- Asia experienced a gradual decline in total land weighted average of 3.0%. The area sown to available per rural worker’ over the period modern varieties grew at an average annual rate 1960–1970 to 1999–2000. The same decline of 1.4% (Walker and Alwang 2015, p. 389) can be seen over the period 1990–2015 (Masters from 1998 to 2011. The average for the seven et  al., 2013, p. 2). Associated with this decline cereals had grown to 31% in 2006–2010, with would have been some consolidation of subsist- values of 52% in maize, 62% in wheat and 36% ence farms into commercial farms in densely in rice; most of the area growth was in maize populated areas. An earlier paper of Hazell et al. (52% of the total growth from 1976–1980 to (2010, p. 11) found a decline in median farm 2008–2010) and in sorghum (20% of the total size in 16 countries (all regions) over periods growth from 1976–1980 to 2008–2010). ranging from 1970 to 1990 to 1998 to 2001. There were important increases in the areas Hazell (2013) compared farm size trends in planted to improved cultivars of other crops Asia and Africa from 1970 to 2011 and further such as cassava, yam, soybean, cowpea, com- projected trends to 2030 and to 2050. He noted mon beans and groundnut (Walker and Alwang, that rural population growth would fall (Hazell, 2015, p. 344). 2013, p. 1) substantially from 2011 to 2030, There are no comprehensive data on modern compared with the earlier period, and would fall variety adoption by agroclimate in sub-S aharan again from 2030 to 2050. While the expected Africa from Walker and Alwang (2015). The fall in rural population growth reduces pressure national averages (Walker and Alwang, 2015, on farm sizes, it is associated with other problems p. 346) show rapid growth of area under mod- of farm consolidation, the transition from sub- ern varieties in countries and crops – maize and sistence to commercial farming and the poten- sorghum in Nigeria; maize, sorghum and teff in tial for extreme inequality among smallholders. Ethiopia; and pearl millet in Mali, Senegal and A global summary (Lowder et  al., 2016) Niger – where mixed crop–livestock are the prin- found the usual problem with data coverage in cipal farm types in humid (Köppen climate A) tropical countries. It confirmed other findings and semi-arid (Köppen climate B). These pat- about the fall in mean and median holdings in terns imply that important benefits accrued in sub-Saharan Africa, while stressing the fact that mixed livestock systems from additional crop land abundance in sub-Saharan Africa was residue output produced by modern varieties, limited to a few countries. The great majority of although we cannot quantify these benefits or sub-Saharan African smallholdings would be less attribute them to (crop) research fields other than 5 ha (Lowder et al., 2016, Fig. 5) and such than plant breeding and pathology69. holdings would be 50–70% of agricultural area. Data on the yields of modern cultivars in sub-Saharan Africa are less comprehensive than improved plant cultivars. Improved cultivars data on area planted. Walker and Alwang of cereals have become nearly universal on (2015, p. 354) estimated that a 1% increase in farms of all types in India, Indonesia, China and the share of area under improved cultivars was Brazil. Sub-Saharan Africa has lagged East and associated with an increase in TFP of about 0.47 South Asia in adopting modern cultivars but has in a sample of 30 countries covering most of recently begun to converge. Walker and Alwang sub-Saharan Africa. Older data reported by (2015, pp. 265–293) studied diffusion of Walker and Alwang (2015, p. 355) gave average African Livestock Systems Research, 1975–2018 567 yield increases of 41% for nine crops and 38% between plant and animal biomass; (ii) positive for 21 crops at varying periods from 1970 to relationships between livestock distribution and 2000. land-use intensity, rural settlement density and mean annual rainfall (Bourn and Wint 1994, fertilizer use. Fertilizer use in the initial ILCA p. 6); (iii) weak seasonal effects on animal bio- studies of mixed farming – Debre Zeit and Debre mass in arid and humid climates, implying less Berhan in highland Ethiopia, Kaduna in subhu- seasonal animal mobility in both; (iv) stronger mid central Nigeria, Ibadan in southwest Nigeria, seasonal effects on animal biomass in the ‘750– and Niono and the Inner Delta of the Niger in 1250 mm rainfall band’ (Bourn and Wint 1994, semi-arid Mali – probably did not exceed 10 kg/ p. 9), implying more animal mobility; (v) a de- ha at any site before 1990. Input use grew as clining threat from tsetse and trypanosomiasis average cultivated land per person fell. While because of ‘agricultural expansion, deforestation fertilizer-use data are unreliable before 2000, it and the removal of wildlife’ (Bourn and Wint was clearly very low across sub-Saharan Africa 1994, p. 15); and (vi) in Nigeria, probable under- before this century. Current estimates are less counts of 25% of domesticated livestock popula- than 15  kg/ha in West Africa and less than tions compared with official figures, indicating 20 kg/ha in East and southern Africa. that the hypothesized conflict between crop and The delay in developing and extending pro- livestock production had been overstated.The ductive seed/fertilizer packages for crop farming principal research goals on mixed systems with in sub-Saharan Africa, compared with South comparatively weak integration between crop Asia and East Asia, had a negative indirect effect and animal production were as follows: on livestock production. Seed and fertilizer pack- To define constraints to higher productivity ages were not always highly profitable under ex- • in the power, soil nutrient, feed production perimental conditions, especially in more arid and animal management components of sites, indirectly reducing potential feed produc- mixed farms. tion for ruminants. New seeds and fertilizer To introduce or improve animal draught were less profitable under farmers’ conditions in • power with a research focus on higher sub-Saharan Africa; strong evidence for this fact work output through health and feed is the 20–40-year lag between the rapid growth interventions. of high-yielding variety/fertilizer packages in South Asia and the slower growth of these pack- • To improve nutrient cycling by using ma-nure on crops. ages in sub-Saharan Africa. • To feed crop residues and higher-quality planted forages to ruminants. Crop–livestock interactions in mixed systems • To introduce dairying and small-ruminant Most of the work on mixed systems was done in fattening into mixed farms. the semi-arid tropics of central India, the Sahel • Ultimately to raise livestock and crop prod- of West Africa, the subhumid savannahs of uctivity jointly by exploiting positive inter- sub-Saharan Africa and the highlands of East actions between the two components. Africa, mainly in Ethiopia and Kenya. Examples Some examples of feasible technical changes of such studies in the IARCs were chiefly, but not in mixed enterprises included the following: only, from ILRI, ICRISAT and the Centro Inter- nacional de Mejoramiento de Maíz y Trigo (CIM- • Applying seed/fertilizer packages. MYT) as shown in Table 15.4. • Shifting the livestock enterprise mix to Grazing and mixed systems present many dairying and intensive stock raising (poultry, complex interactions. Bourn and Wint (1994, pigs, small ruminants) components. pp. 4–5) reviewed 20 aerial and ground surveys • Improving feed production by planting forages of livestock and land use carried out between or by treating crop residues in densely popu- 1980 and 1991 in Mali, Niger, Sudan, Chad and lated areas (semi-arid tropics of India, Ethi- Nigeria, covering a broad range of grazing and opia, northern Nigeria, and cotton-producing mixed smallholder systems. The surveys found areas of West Africa), whether for animal trac- many common elements: (i) positive relationships tion, dairying or on-farm fattening. Table 15.4. Mixed crop–livestock systems studies, various years. Country/ Farming Scale of Study Region/period subregion Climate systems research Animals Main crops Biotic factors Abiotic factors Technologies Von Kaufmann West Africa, Nigeria, Arid to MRH Household, Cattle Pearl millet, Soil, water, Rainfall, Subsistence et al. 1979–1986 Kaduna semi-arid territory sorghum, health, feed temperature dairy (1986) maize, cassava Wilson West Africa, Mali Arid to MRA, MIA Household, Cattle, Pearl millet, Soil, water, (1986) 1978–1986 semi-arid territory camels, cowpea health, feed sheep, goats Powell Sub-Saharan West Africa Arid to MRA, MIA Household, Ruminants Pearl millet, Soil, water, Rainfall, Animal traction (1986a,b) Africa, semi-arid territory sorghum, microbes temperature 1980s–1980s maize, cassava MRH (Primary Water, fire (?) Mixed farming studies) without animal traction (Primary Manuring crops studies) Nutrient cycle Altitude Animal fattening (Primary Grazing reserves studies) Pingali et al. Sub-Saharan Many Semi-arid MRA, MIA Sub- Ruminants Maize, cotton, pearl Grazing, feed Animal traction (1987) Africa Saharan millet, sorghum, Africa paddy, cassava, cowpea Subhumid MRH Tractor mechanization Highlands MRT Mineral fertilizer Humid MRH Gryseels East Africa, Ethiopia Highlands MRH, MRT Territory Draft oxen, Wheat, barley, teff, Grazing, feed Rainfall, Dairy (1988); 1970s early cattle, faba temperature, Gryseels 1980s sheep frost risk and Anderson (1983); Getachew Asamenew et al. (1993) Norman et al. Southern Africa Botswana Arid to MRA Household cattle Sorghum, Heat, aridity (1988) and West semi-arid cowpea, Africa, watermelon, 1970s–1980s maize, groundnut Debre Cool MRT Household Dairy, Animal traction Berhan highlands work oxen, sheep Nigeria MRH Millet, sorghum Debre Zeit Subhumid to MRT Teff, wheat, faba Land cool management highlands Walker and South Asia, India, Semi-arid to MIA, MRA Household, Work oxen Paddy, pearl Soil, feed Water Mechanization, Ryan 1975–1990 semi-arid subhumid village millet, sorghum, high-yielding (1990) tropics pigeonpea, varieties, chickpea fertilizer water harvesting, tractorization Bartholomew West Africa, North-west Semi-arid MRA, LGA Households Cattle, Mixtures of millet Raghuva spp. Rainfall, Animal traction, (1988) 1984–1989 Mali work and cowpea temperature manurefor oxen crops McIntire et al. Sub-Saharan 22 Semi-arid MRA, MIA Maize, cotton, Grazing, feed Heat, aridity Animal traction (1992) Africa countries pearl millet, in sorghum, sub- paddy, cassava, Saharfan cowpea, teff, Africa wheat Subhumid MRH Sorghum and Stiga Short growing Cattle fattening cowpea hermonthica season on farm, animal fattening Highlands MRT Millet, sorghum Crop residue use Humid MRH Manure use Ehui and Nigeria, 1980s Nigeria Subhumid to MRT Households Cassava, maize, Fertilizer, hand Spencer humid rice, cowpea, cultivation, (1993) yam, melon, tractors plantains Continued Table 15.4. Continued Country/ Farming Scale of Study Region/period subregion Climate systems research Animals Main crops Biotic factors Abiotic factors Technologies Sanders 1970s–1990s Sub- Semi-arid MIA, MRA Households Pearl millet, Soil, fertilizer, Water Irrigation et al. Saharan sorghum, crop (1996) Africa, maize, cowpea cultivars semi-arid tropics Bosma et al. West Africa, Southern Semi-arid to MRA, MRH Households, Cattle, Maize, cotton, Fertilizer, hand (1999) 1977–1987 Mali subhumid territory work pearl millet, cultivation, oxen sorghum,paddy animal traction Hill (1982); Sub-Saharan Northern Semi-arid MRA, MRH Hand cultivation, Mortimore Africa Nigeria fadama (2000) irrigation Dixon et al. Global tropics Many Arid/ MRA, MRH, (2001) semi-arid MRT Humid MIA, MIH, MIT Highland LGA, LGH, tropics LGT Hiernaux and West Africa, Niger Arid to MRA Household, Cattle, Pearl millet, Soil, grazing, Ayantunde 1990s and semi-arid territory camels, cowpea feed (2004) 2000s sheep, goats Baltenweck Sub-Saharan Colombia, Semi-arid MRA, MIA (Primary studies) et al. Africa, Latin India, (2003) America Kenya, Sri Lanka, Nigeria, Niger South Asia Subhumid MRH, MIG Highlands MRT, MIT Humid MRH African Livestock Systems Research, 1975–2018 571 • Diversifying crop production and process- A third prediction was that smallholdings ing by adding annual cash crops, such as would gradually consolidate into larger com- oilseeds, cotton and grain legumes, which mercial enterprises because such units could ex- would have a secondary effect of providing ploit economies of scale in technologies and more feed. management. This prediction was a stimulus to • Cycling nutrients70 in mixed grain–l ivestock mechanization research, which varied among farms by feeding crop residues to stock and regions, being stronger in Asia and Latin Amer- by restoring manure to fields. ica than in Africa. This evolutionary model influenced the Predictions about the evolution of mixed three principal themes of research on mixed systems strongly influenced the initial inter- systems: (i) mechanization; (ii) soil fertility man- national research in the 1970s. New theories of agement; and (iii) feeding systems. A fourth farming systems (e.g. Boserup, 1965; Ruthen- theme – improved animal health – was viewed as berg, 1980; Binswanger and Rosenzweig, 1986) necessary to achieving results in the other com- provided a deeper understanding of how trop- ponents. ical agriculture evolved under the influence of population density, market access, information mechanization. Animal traction had been prac- costs and incentives. tically universal among smallholders in the The first prediction was that nomads would semi-arid tropics of India and in highland Ethi- settle to become crop farmers. This prediction opia for centuries. It had begun to grow rapidly depended on the view that sedentarization was across sub-Saharan Africa in the 1970s. Despite desirable because it was more productive and the infrequency of use of animal traction in would allow better provision of social and infra- most of tropical Africa, it was believed that structure services to the former nomads. Some land-abundant areas within sub-Saharan Africa mixed-systems research therefore sought to had the potential for mechanization with ani- identify the policies and technologies that would mals. Mechanization would allow an expansion reduce the transition costs of the inevitable of cultivated areas and a concomitant increase sedentarization. in crop yields. The efforts to realize these poten- Related to the settlement prediction was a tial gains made farm mechanization with ani- later view from the characterization literature. mals an important part of research at ILCA, and Thornton et al. (2002, maps 16a and 16d) con- to a much lesser extent of ICRISAT, from the tended that there would be major long-term 1970s and to the mid-1990s71. changes in growing periods and in locations of Mechanization research started from three dense livestock and human populations related assumptions: to climate. The combined effects of these changes would be to shift the West African rangeland 1. Animal traction was the core component of units into mixed systems and to eliminate mixed mixed farming and could create productive highland systems in East and southern Africa. interactions among other components that Jones and Thornton (2009) noted that climate would not occur if those components were sep- change might push marginal, low-productivity arate; Jahnke’s (1982, pp. 134–140) summary farmers in West Africa into livestock activities, of livestock development problems in tropical thereby increasing pressure on rangelands. Africa is an example of this view. A second prediction was that closer integra- 2. Animal draught power could increase tion of crops and livestock would be more product- the cultivated area per worker, could stimu- ive than separate enterprises. Integration would late diversification into more profitable crops allow more efficient use of resources – animal and, by improving the quality and timeliness power, manure and crop residues – that were of farm tasks, could raise yields per unit of underused when crops and livestock were man- land. aged separately. Many development projects, and 3. Better animal nutrition could allow more their associated research components, acted on power from livestock and hence extend the hy- this prediction and attempted to accelerate this pothesized benefits of area, cropping pattern integration (McIntire et al., 1992, pp. 4–5). and yield. 572 J. McIntire, T. Robinson and C. Bosire The second assumption – that animal traction assumption was tested in many experiments would produce substantial area, yield and crop- examining animal breed and type (dairy cows or ping-pattern benefits – led to many studies of oxen) and feed type (quantity and quality of add- constraints to animal power. These studies often itions to basal diet of crop residues) at sites in failed to find adoption of animal traction, even Mali, Ethiopia, Niger and semi-arid India (Renard, where livestock disease was manageable and 1997). The ILCA research in Mali summarized where adequate feed was available. Pingali et al. their findings as: ‘…dry-season supplementation (1987) explained the mixed adoption of animal and weight gain would not improve work output traction across sub-Saharan Africa in terms of and would be unlikely to increase the amount of labour use and fallow type; they concluded that land cropped or crop production’ (ILCA, 1994, the driving force for mechanization with ani- p. 134). The lead scientist for the Malian trials mals was not output benefits but labour savings under experimental conditions concluded that in areas of annual cultivation where the transi- work output during the brief 2-week ploughing tion from forest or bush fallow to grass fallow season ‘…seem[s] to be unrelated to an animal’s had taken place and where heavy soils created body energy reserves at the start of work’ (Bar- additional demand for power over what hand tholomew 1988: p. 59). An experiment using hoes could provide. dairy cows for traction in Ethiopia found that Pingali et al. (1987) complemented a litera- supplementation allowed additional work while ture review with original farm interviews about (almost) maintaining the milk output and repro- the gains from animal traction. They found that ductive performance of the cows (ILCA, 1994, area benefits were positive (Pingali et  al., 1987, pp. 134–136). Further work (ILRI, 1998, pp. pp. 99–101) and averaged about 25% per person; 178–183) in Ethiopia compared working that yield gains (animal-cultivated fields over cross-bred dairy cows with non-working dairy hand-cultivated fields for the same crop) were cows and traction with local oxen alone, plus im- often zero because the quality of tillage with ani- proved management of feed and animal hous- mal traction was not better than that with hand ing. The working cross-bred treatment did not hoes; and that impacts on cropping-pattern diver- produce higher incomes compared with the sity were important only where cotton and non-working cross-bred treatment or the local groundnut had been introduced. oxen treatment. An ILRI experiment on a semi- Animal and machine power was introduced arid experiment station with sandy soils in Niger into farming systems as a function of population measured energy expenditure of working bulls density, given the need to cultivate the same plots and oxen (Fall et al., 1997), feed intake and the repeatedly to suppress weeds, and in response to effects of body condition on work; two import- better market access from proximity to cities, ant conclusions were that roughage intake did higher population density areas and the intro- not increase during work and that weight losses duction of cash row crops, notably cotton and during work did not cause reductions in power groundnut. The presence of trypanosomiasis, output (Fall et  al., 1997, Chapter 6)72. Station long held to block draught power by restricting and field research on animal power in cattle production, was shown to be a secondary sub-Saharan Africa did not generally find that constraint in the humid and subhumid areas nutrition was a serious constraint to adoption or where other climate and population factors to optimal use of animal power. weakened the demand for animal power. The core conclusion – the introduction of animal animal power and mixed farming in sub-saharan traction into farming systems was a long-term africa. The view that animal traction had a response to growing population and cultivation leading role in developing mixed systems in density, rising wages and the introduction of sub-Saharan Africa was examined by McIntire cash crops (Pingali et  al., 1987; McIntire et  al., et al. (1992, pp. 47–72), following the study by 1992) – greatly weakened the research emphasis Pingali et al. (1987). McIntire et al. (1992) found on animal draught power in mixed rain-fed sys- that components of mixed farming – feeding tems where animal power was absent or nascent. crop residues to stabled or mobile animals, ma- The second assumption was that poor ani- nuring crops, and investing in dairying and on- mal nutrition constrained draught power. This farm fattening – existed in many sites without African Livestock Systems Research, 1975–2018 573 animal traction. They concluded that using ani- public research on mechanization either added mals for power was not a necessary condition for too little to farmers’ knowledge of crop manage- mixed farming. An example was found in a field ment – the case of most animal traction research – study of mixed smallholders in south-eastern or could be done better and faster through adap- Burkina Faso, where Delgado (1980) found that tations in the private sector – the case of most sample farmers practised some elements of mixed tractor research. Herbicides were proposed as an farming (manuring, crop residue grazing, fatten- alternative to animal or tractor mechanization ing animals on farm) but not animal traction. (Le Moigne, 1979, pp. 219–220), but herbicides The mechanization research at ILRI with have only recently become competitive in coun- the greatest development impact was that lead- tries where rural wages have risen rapidly; if ing to the broad-bed and furrow maker (BBM), herbicides have replaced hand weeding in sub- as recounted by Rutherford (2001, 2008). Ru- Saharan Africa, it would be a recent development. therford’s 2008 analysis found a rate of return In 40  years of national and international to ILCA/ILRI research and development of the work on animal traction in West Africa, there have BBM to be in the order of 0.1%. This was much been no significant technical changes induced by higher than the negative return found in the research findings that are clearly distinguishable 2001 study, the difference being attributed to from the changes induced by rising population the greater availability of credit, farmers’ adap- density, cheaper market access and external econ- tation of the BBM tool to their circumstances, omies arising from lower implement production and the resulting area and yield effects from costs. The arguments of Pingali et al. (1987) and greater experience with the tool. McIntire et al. (1992) derived from the evolution- ary models of Boserup (1965) and Ruthenberg animal power and mixed farming in the semi-arid (1980), and the mixed findings of station and field tropics of india. ICRISAT began work on studies of animal power, effectively ended the era mixed irrigated and rain-fed cropping systems in of projects seeking to introduce animal power as a the semi-arid tropics of central India (Jodha necessary component of mixed farming. et  al., 1977). These systems were in a hot and usually dry climate, with a single monsoon sea- soil fertility management. Low soil fertility son, and consisted of small family farms grow- was long understood to be a major cause of the ing ICRISAT mandate crops (sorghum, pearl lagging productivity of tropical agriculture. Sig- millet, groundnut, chickpea and pigeon pea) nificant work was started in the 1960s and and others (irrigated paddy, cotton, castor and 1970s, in Africa, South Asia and Latin America, vegetables), usually with bullock power. The on soil fertility management with different em- chief livestock research component was on the phases for the systems found in each region. Soil role of animal power for management of Verti- fertility work had substantial scientific impact sols, cultivation, transport and processing with on all continents, but its development impact some work on crop residues. Other national and varied greatly by region, farm type and source of international research in South Asia concen- water for cropping. trated on food crops, on technical changes in ir- The Africa soil fertility focus was on nutri- rigation, and on mechanization compatible with ent cycling in mixed systems because of the smallholdings; livestock research concentrated prediction that animal manures were ‘slack on animal traction and smallholder dairying resources’ that could be used more efficiently (e.g. Singh, 1990, pp. 204–232, on rural pov- where crops and animals were managed on the erty in South Asia; Walker and Ryan (1990, on same farm73. The integration of crop and animal the semi-arid tropics of India). production would occur in part through nutri- The centrepiece of ICRISAT’s limited live- ent cycling through linked mechanisms: the ‘ex- stock work in India were the village-level studies change’ of plant residues from crops to animals, (Walker and Ryan, 1990). These studies allowed and the ‘exchange’ of manure and urine from a better understanding of the roles of livestock animals to crops. in the semi-arid tropics of South Asia and led Nutrient cycling was studied by Powell et al. to a reorientation of mixed-system smallholder (1996). The orientation of ILCA/ILRI work on research. This reorientation conceded that nutrient cycling was on smallholder farms, 574 J. McIntire, T. Robinson and C. Bosire where the cycle was soil to food crop to animal to benefits surpass those of sown forages (Sum- soil and crop vegetation74. berg, 2002; McIntire and Debrah, 1987). Where ley farming did emerge, it did so al- Feed systems most exclusively in ‘unregulated’ form (Ruthen- berg, 1980). Where it does exist, it evolved as part All characterization research found feed scar- of smallholder dairying in temperate highland city, in quality and in dry-season quantity, to areas where cross-bred European–African cattle constrain the livestock component whether ani- could be raised, rather than from progressive mals were used for milk, power or meat produc- adoption by sedentarized pastoralists. One limited tion. Three interventions, with increasing levels success was achieved around Kaduna, in subhu- of management costs, were proposed to unbind mid central Nigeria, where Stylosanthes hamata this constraint: (i) crop residue management; was introduced as ‘fodder banks’ in mixed sys- (ii) sown forages, including alley farming; and tems (von Kaufmann et  al., 1986). Despite the (iii) ley farming. technical promise of fodder banks, their long-term area and yield effects have not been precisely esti- crop residue management. Crop residue man- mated in Nigeria or elsewhere in West Africa. agement – harvesting, storing, chopping, making Other obstacles to ley farming were seed costs and hay, field grazing or the many other possibilities – farm size (Christiansen et al., 2000, p. 191, and was found to be quite common but has been dif- the papers cited therein for the Mediterranean; ficult to improve through research. ILCA’s com- Ruthenberg, 1980, for East Africa; Tiffen et  al., pilation of research on ‘Plant Breeding and the 1994, pp. 164–166, for Kenya; Powell, 1986a for Nutritive Value of Crop Residues’ (Reed et  al., central Nigeria). 1988) cited no successful examples of improve- Nordblom et  al. (1994) reviewed field and ment of crop residues in the tropics (nor did the station studies of a rotation of wheat and M. sa- book by Renard, 1997), through plant selection tiva with sheep grazing in north-west Syria. or breeding, or by chemical or urea treatment They found historical adoption of M. sativa as a beyond chopping or other practical methods pasture crop to be limited in West Asia and that farmers can already use (see Chapter 14, North Africa. Their modelling work showed M. this volume). The Walker and Ryan (1990) book sativa to be ‘less profitable than traditional rota- covering a decade of on-farm research in the tions’ and its adoption to be sensitive to farm semi-arid tropics of India noted that the unit size, milk prices and soil nutrient carryover ef- value of fodder in the study villages was some- fects to the following wheat crop. times as high as that of grain, but found few ex- amples of crop residue improvement or of plant- fodder trees. Fodder trees started with the pur- ing of specialized fodder crops to replace crop pose of adding nitrogen to leached or otherwise residues in ruminant diets, even where the fod- infertile soils, thereby improving the soil nutri- der value of crop residues was high for dairy pro- ent stock while contributing nitrogen to cereal duction or for animal power. and tuber intercrops. The field model of this idea is ‘alley farming’ – typically planting rows of ni- sown forages and ley farming. Sown forages trogen-fixing trees between rows of food crops, and ley farming have usually failed in the Afri- such as maize or cassava. Several synthetic can tropics. Chapter 13 (this volume) reviews works established the basic science of the alley attempts to introduce sown forages into mixed farming model (Kang et  al., 1990; Sanchez, systems in Africa, with reference to the success 1995; Giller, 2001), followed by many subse- of pasture grasses in Latin America, while Chap- quent applications and extensions (Sumberg ter 14 (this volume) covers multidimensional et al. 1987; Jabbar et al., 1996). Promising work crops. Nordblom et al. (1992, 1994), in analys- was done on economic and environment bene- ing a Syrian site, identified land cost as the most fits of silvopastoral systems in Latin America direct reason for the failures of sown forages in (Ibrahim et al., 2010, p. 189–196). the tropics. The available alternatives – weeds, One extension of the alley farming model was field boundaries, crop residues and browse – are to use nitrogenous browse as a supplementary cheap and of high enough quality that their feed for small ruminants, given the lack of CP African Livestock Systems Research, 1975–2018 575 available in grasses or crop residues in the humid has been rapid in this century but is a function tropics. This extension was successful in raising of income growth, urbanization and falling soil nitrogen stocks, in lifting intercrop yields intermediation costs, and does not appear to be and in providing higher CP content to livestock related to technical packages developed by (Kang et  al., 1990, pp. 340–345), while being r esearch. profitable in an economic sense (Kang et  al., 1990). Experiments with livestock were done at two ILCA sites in Nigeria, on alley farming using Conclusions nitrogenous trees in the humid zone near Iba- dan and using leguminous forages (without al- What were the impacts of LSR, done mainly by leys) and manure recycling at the subhumid site ILRI and its predecessors, on the scientific and near Kaduna. The latter had greater impact po- development problems of grazing and mixed tential because it took place in an agroclimate livestock systems in Africa? How did new know- with lower trypanosomiasis pressure and less ledge improve productivity or equity in the prin- heat and humidity, allowing higher ruminant cipal livestock systems? density and productivity. Jabbar et  al. (1994) showed that continu- Scientific impact ous alley farming would be more profitable than alley farming with short fallows or farming LSR has had a strong scientific impact in pas- with fallow but without the alleys nitrogen- toral and mixed systems. This scientific impact fixing trees. Reynolds and Jabbar (1994) was notable in: (i) dismantling the ‘mainstream showed that the major benefit of supplement- view’ of pastoralism and replacing it with a new ing free-roaming small ruminants in West Africa model of pastoralism that is biologically and eco- with the foliage of leguminous trees (Leucaena nomically more credible than the mainstream and Gliricidia spp.) was an increase in survival, view, which has been antiquated for 50  years; and that the forage was best directed at late- (ii) classifying and mapping systems; (iii) estimating pregnant and lactating females. In East Africa, productivity parameters to develop bioeco- cross-bred dairy cows showed a significant re- nomic models, which have been used particu- sponse in milk production to supplementation larly in modelling climate change effects; and with Leucaena. (iv) applying the network research model to the The International Centre for Research on study of trypanotolerance at disparate field sites. Agroforestry (ICRAF) estimated that ‘fodder shrubs have been widely adopted in East Africa, a new view of pastoralism. The core of the by an estimated 205,000 smallholder dairy ‘mainstream view’ was that African herders farmers by 2005’ (Place et al. 2009). There is no kept too many animals for cultural reasons and reliable estimate of direct tree yields or of their that such overstocking caused overgrazing. The indirect effects on crop or livestock yields. systems studies and related work destroyed this view. A first attack on the ‘mainstream view’ animal fattening on farm. On-farm animal fat- was the demonstration that animal productivity tening was studied as a familiar technology with per hectare in the grazing systems was not uni- potential for improvement by using slack re- formly worse, and was sometimes better, than sources, especially crop residues and other cut that of ranching systems. Such research further fodder, with local animal breeds. Research demonstrated that herd structures did not themes included: (i) the productivity of supple- consist excessively of older males kept for senti- mentary feeds to crop residues and cut grasses; mental reasons or held against risks beyond (ii) the introduction of new feeds, notably tree their optimal sale age75. The age/sex compos- and herbaceous legumes; and (iii) the timing of itions of herds did not consist of ‘excess’ males, fattening and sales. Examples of performance as the ‘mainstream view’ had contended, given trials are given by Bartholomew (1988) and the milk production objectives of herders and those cited in McIntire et  al. (1992, pp. 135– the low marginal costs of feed and labour 164). The emergence of many small peri-urban needed to maintain male stock to a roughly fattening units, for ruminants and for poultry, optimal sale age. 576 J. McIntire, T. Robinson and C. Bosire A second part of the ‘mainstream view’ was rigorous enough to permit recommenda- that pastoralists’ management practices deter- tions of alternative management practices mined stocking rates. The ‘mainstream view’ to stop overgrazing; held that maintenance of stocking rates above • estimates of overgrazing were biased up- biologically sustainable values led to overgraz- wards because measures of animal num- ing and eventually to destruction of the range. bers and of pasture quantity and quality The rise of ‘non-equilibrium’ models of pasture were sparse or inaccurate; dynamics (led by Ellis and Swift, 1988) fortified • the ‘mainstream view’ of the dynamics of the new view by stating how pasture dynamics overgrazing – a movement from undis- depend on abiotic factors and not primarily on turbed vegetation subject to grazing to dis- the decisions of pastoralists themselves. turbed vegetation of lowered grazing cap- A third tenet of the ‘mainstream view’ was acity – was contradicted by observations of that crops and livestock would inevitably face a pastures whose productivity had been im- general land conflict as population density and proved by heavy grazing; and cultivation density rose. This hypothetical con- • there was a logical inconsistency in positing flict would reduce grazing areas and, with fixed overgrazing where stocking rates were high- pasture yields for given rainfall, would reduce est while simultaneously assuming that livestock production. The attack on this tenet such high stocking rates were infeasible on came from the practical fact that cereal and leg- pastures degraded from overgrazing. ume crops in the semi-arid and subhumid zones, which house the majority of African ruminants, The critique of the ‘mainstream view’ led to produce crop residues that are valuable as feed. a new development path for pastoralism. The The growth of crop production, even in areas new path allowed herd expansion while promot- marginal for arable farming and even in the ab- ing public investments around grazing areas sence of feeding grain directly to animals, al- and regulating conflicts over land and water. lowed more livestock production, not less, by the This new development path proceeded by: indirect channel of additional crop residues. • improving animal health by controlling dis- It should be acknowledged here that part of ease, mainly rinderpest and trypanosomiasis; the scientific impact achieved through the new • defending grazing lands by laws and regu- view of pastoralism derives from the older work lations; new laws and regulations reduced of anthropologists, notably the work of Dupire conflicts between herding and farming (1972), Monod (1975), Dyson-Hudson and Dy- groups and in so doing limited the costs of son-Hudson (1969), Oxby (1975) and Stenning violence; (1994), among others. That work, in East and in • introducing water interventions to raise West Africa, laid an empirical foundation for the productivity per animal by reducing trek- new view of pastoralism which recognized the king times; rationality of extensive grazing (as clearly stated • introducing planted forages, as arable crops by Barbara Grandin in 1987 in an early ILCA or as browse trees and shrubs, and supple- paper on Maasailand). mentary feeds targeted to a subset of the herd Sandford (1983a, pp. 11–18) best sum- or flock, such as sedentary milk herds or marized the refutation of the ‘mainstream small ruminants for fattening; and view’. He argued that the ‘mainstream view’76 of • accommodating mixed land use of crops, grazing systems was wrong, anticipated the non- livestock and wildlife. equilibrium critique and asserted that policies to reduce overgrazing were misguided because The studies of Pratt and Gwynne (1977), Ru- the evidence for overgrazing was weak. Sand- thenberg (1980), Sandford (1983a), Le Houérou ford (1983a) contended that: (1989), Solomon Bekure et  al. (1991), Coppock (1994), Blench (2001) and Lesorogol (2008) • definitions of overgrazing – changes in projected variants of that generic path. A detailed vegetation from an undisturbed state (‘cli- development pathway for pastoralism was drawn max vegetation’) to a disturbed one, or loss for the Borana system of southern Ethiopia (Cop- of productive capacity over time – were not pock, 1994, pp. 272–295; on water, pp. 202–209). African Livestock Systems Research, 1975–2018 577 mapping systems. The outstanding scientific Development of better survey and analytic achievement of the various LSR studies, carried • tools for estimating system scale and poten- out by ILRI and many partners from 1975 to the tial across continents. present, has been to develop a new view of graz- Creating methods to compare and reconcile ing systems that can generate better policies for • national statistical estimates and remote rangeland management. This view started from sensing estimates. the observation that direct efforts to improve • Reducing errors in estimates of woodland, rangelands – pasture and breed improvement wildlife, crop and grazing areas, permitting and grazing restrictions – had failed. It continued closer analysis of actual and potential re- with the contention that rangelands could be- source conflicts. come more productive if herders benefitted from • Better projections of animal numbers public investments in human and animal health, across climates and countries, which even- water, transport and markets, communications tually produced better estimates of global and social protection. The findings of rangeland environmental effects77 and, very recently, scientists about the potential of arid areas did refinement of Tier 2 estimates of green- not differ greatly from those of earlier scientists house gas emissions from livestock. in terms of raising biological potential, but they Planning investments in irrigation, graz- did differ in terms of defining an appropriate • ing, protected areas, disease management, sequence of policy and public investment to vector behaviour and control, predator achieve that potential. control and wildlife interactions. Landmark papers – Coppock (1994) for the Understanding seasonal effects on plant Borana system in southern Ethiopia; Solomon • and animal biomass, wildlife and domestic Bekure et  al. (1991), for Maasailand in Kenya; stock biomass, and animal mobility. Wilson (1986) in north-central Mali; Wagenaar et al. (1986) in the Niger Delta of Mali; von Kau- fmann et al. (1986) in subhumid central Nigeria; development constraints in african livestock and Hiernaux et al. (2009) in semi-arid western systems. An important scientific impact, from Niger – not only contributed to knowledge about work not always led by ILRI but often done with these areas but also established methods of its support and advice, was in the transversal studying them and induced changes in related systems studies. These include the publications research and development programmes. of Sandford (1983a) 78 and others on pastoral- The major system papers over the past ism; Pingali et al. (1987) on animal traction in 20 years – Seré and Steinfeld (1996); Thornton the farming systems of sub-Saharan Africa, et al. (2002); Otte and Chilonda (2002); Kruska McIntire et al. (1992) on crop–livestock integra- et al. (2003); Robinson et al. (2011, 2014), plus tion in sub-Saharan Africa, Thornton et al. (2002) the many papers on climate change as dis- on livestock and poverty mapping, Baltenweck cussed in Chapter 16 (this volume) – modern- et  al. (2003) for crop–livestock interactions on ized the methods for defining production sys- three continents, Thornton and Herrero (2001) tems and reduced the estimation errors for on crop–livestock simulation models, McDer- areas, animal numbers and feed balances. This mott et al. (2010) on sustainable mixed systems research has clearly had an impact on scientific and Herrero et al. (2012). Moreover, they have understanding (Class I) even if it is impossible added to the scientific and development consen- to estimate the development benefits of this sus about what to avoid in grazing systems – type of research. Work in this century on live- introduction of exotic animal breeds, hurried stock and the global environment has had a privatization of communal grazing tenure, re- similarly large Class I effect (see Chapter 16, striction of mobility and oversowing of pasture this volume). in conditions unfavourable to the viability of The specific scientific achievements were as introduced plants. follows: productivity parameters. The representative- • Estimating the economic and environmental ness and accuracy of estimated productivity weight of livestock systems. parameters – such as fertility, mortality, morbidity, 578 J. McIntire, T. Robinson and C. Bosire milk production and animal growth – were poor A recent book (de Haan, 2016, pp. 79–122) when the older systems studies were launched. presents an integrated bioeconomic model that The goal of estimating factor and input produc- uses and extends some of the research discussed tivities for the main African livestock systems in this chapter. This book made the first effort to was completed in the first half of the modern compare the benefits and costs of technical era, as shown in ILCA (1979a) for the subhumid interventions to preserve livelihoods in East and zone of West Africa, Wilson (1986) for Mali, West African drylands. While it is premature to Wagenaar et al. (1986) for the interior Delta estimate the economic effects of the model, of Mali, von Kaufmann et al. (1986) for central which depend on application of the proposed Nigeria, Coppock (1994) for Borana, and Solomon interventions by governments (health, market Bekure et al. (1991) for Maasailand79. Following integration and promoting recovery from drought the work begun in the 1960s and 1970s, the through small ruminants), this book is a landmark scientific understanding of grazing systems prod- contribution to more scientific policy making, uctivity has deepened such that policies and pro- which uses many of the historical efforts of ILRI jects can be prepared with comparatively cheap and its predecessors. additional background work. One important example is the contemporary use (e.g. de Haan, the network research model80. A further 2016) of parameters derived from the ILCA sys- methodological contribution was the study of tems studies in Mali (Wilson et al., 1983; Wilson, trypanotolerance and trypanotolerant animals 1986). The policy recommendations of de Haan through a novel international network of scien- (2016) depend in part on data and analysis from tists (Trail et al., 1979a,b; Murray et al., 1990; earlier field studies. ILCA/ILRAD, 1988; Rowlands and Teale, 1994; A related use of the data – to apply the esti- Itty, 1992). The purpose of ATLN (ILCA/ILRAD, mated input–output parameters in systemic models 1988, p. 32) was to provide ‘a better understand- that can simulate technical, policy and manage- ing of genetic resistance, acquired resistance, ment changes – has scientific validity but has environmental factors which affect susceptibil- had little or no productivity impact. Reasonably ity and the efficacy of present control measures, detailed and accurate policy analyses have used and second by ensuring optimal application of the input–output parameters (Thornton et  al., both existing knowledge and recent research 2006; Nelson et al., 2009; Thornton and Herre- findings’ (see also Chapter 2, this volume) ro, 2010; de Haan, 2016), but resulting policy ATLN allowed a rapid growth of knowledge recommendations to date have not had measur- about trypanotolerance across 18 countries, able and attributable welfare impacts. mainly in village situations, in the humid and subhumid climates of sub-Saharan Africa. The bioeconomic models. A second methodological work included status reports beginning in 1979, contribution was the application of bioeco- the establishment of field sites in the early 1980s nomic models to grazing and mixed systems, and ultimately the production of a series of land- using station, field and remote sensing data. mark papers. Published work after more than a Examples are Penning de Vries and Heemst decade of the ATLN had established the genetic (1975), Cartwright et al. (1978, 1982), Ko- basis of trypanotolerance, clarified the relationship nandreas and Anderson (1982), Konandreas between the animal’s ability to control parasit- et al. (1983), Itty (1992), Itty et al. (1995a,b,c) aemia and to control anaemia, developed novel and Solomon Bekure et al. (1991). The influence diagnostic tools and laid the basis for selecting of this work on poverty analytics is discussed for trypanotolerance in young stock (Murray by Thornton and Herrero (2001); Thornton et al., 1990, p. 381). et  al. (2002, 2006) and Rich and Perry (2011). This work has achieved much in estab- Development impact lishing empirical models and in applying them to targeting and analysing effective treat- the direct development impact of lsr in pastoral ments. Many applied examples to the field of areas. The direct development impact of LSR livestock and climate change are discussed in in pastoral areas has generally been weak. It is Chapter 16 (this volume). generally impossible to measure the ex post African Livestock Systems Research, 1975–2018 579 economic impact of grazing system studies by Niger, Peyre de Fabrѐgues (1984) advocated a estimating a function relating output or product- pastoral code to define the rights and responsi- ivity to research or by using indirect methods, bilities of herders and farmers, thereby allowing such as analysis of citations. Attempting to map the development and preservation of pastures. the use and effects of new technologies, as The later study of FAO/CIRAD (2013) delin- proposed by LSR, is also impossible. For these eated years of progress in establishing and pro- reasons, we conclude that LSR in sub-Saharan tecting the legal and administrative rights of Africa has failed to contribute significantly to West African pastoralists. The FAO/CIRAD (2013) technical change. study showed advances in national laws and re- The tropical examples of how knowledge of gulations in Mauritania, Senegal, Mali, Burkina pastoralism, as derived from characterization re- Faso, Niger and Chad, and in international trea- search or from integrated work on station and ties and local administrative practices, all of which on farm, have changed technologies, policy and have served to protect pastoralist economies against productivity are from ranches in Latin America. encroachment by arable farming, commercial Despite more than three generations of research in ranching, urbanization and land grabs by out- pastoral systems of sub-Saharan Africa, no prod- siders. While pastoralists’ rights are still precar- uctivity effects have been observed correspond- ious in much of Africa, research by national and ing to those achieved in soil fertility, pasture international programmes has contributed to management or beef breeds in Latin America. the understanding of pastoralism and to the The chief development impact of grazing defence of pastoralists’ rights and welfare. systems studies since the 1960s has been to Another successful example is the long d efend the economies and rights of pastoral tradition of research on the economic aspects of peoples. This defence, which is an indirect effect controlling tsetse and trypanosomiasis, to which of research and extension, has taken the form of ILCA/ILRAD/ILRI work has contributed in part grazing rights legislation and consultation with over many years (see Chapter 3, this volume). pastoral peoples on their economic, demographic This began with Hans Jahnke’s path-breaking and political interests. Such legislation may have work (1974, 1976, 1982), continued through the contributed to higher productivity in some African Trypanotolerance Livestock Network grazing situations, but this impact cannot be (ILCA/ILRAD, 1988; Itty and Swallow, 1994), measured without detailed biological, economic the summary of Swallow (2000), and the de- and cultural studies (e.g. Lesorogol, 2008, on tailed investigations by Shaw (2004) and Shaw northern Kenya, which is an unusual study be- et al. (2015), for example. This work has allowed cause it used two survey rounds and stratified by the definition of control models and has provided ‘privatized’ and ‘communal’ tenure). valuable advice to extension services on the ap- Part of the defence of pastoralists’ rights has plication of those models. been to enforce pastoralists’ rights or to stop bad Other possible exceptions are usually not ap- policies (de Haan, 2016, pp. 59–61 and 76–77). plicable to livestock whether in grazing or in mixed An example of this defence is the research find- systems. These include: (i) research in which ing that old legislation, such as the 1965 Graz- crops, not livestock, were the principal treatment; ing Reserve Law of Nigeria (Waters-Bayer and (ii) the finding of positive returns to new livestock Taylor-Powell, 1986a) had set aside land for herd- technologies using ex ante rather than ex post tech- ers use but that the areas actually allocated were niques (Bryant and Snow, 2008; Kristjanson et al., much smaller than projected. A recent example 1999a, for a trypanosomiasis vaccine; Kristjan- of the new rights of pastoralists has been the son et al., 1999b, for genetic enhancement of cer- genesis of new laws and regulations governing eal crop residues; Thornton et  al., 2003, for pastoralism in West Africa. IEMVT, relying on its dual-purpose crops); (iii) efforts to adapt research long tradition of work in the Inner Delta of the in temperate countries, where there is a long his- Niger in Mali, proposed a ‘Code Pastoral’ (Gallais tory of research, data and controlled conditions and Boudet, 1980) specifying political, legal and for estimation and attribution of treatment effects; institutional reforms designed to regulate land and/or (iv) research on tropical ranches under use and to protect the rights of all users – arable more controlled conditions and in more favour- farmers, herders and fishers – in the Delta. In able climatic conditions at larger scale. 580 J. McIntire, T. Robinson and C. Bosire the weak impact of lsr on productivity in grazing instances, specialization in dairying. The logic areas. This weak or nil impact is often attrib- behind specific technical interventions – adapting utable to poor experimental design. For example, animals for farm mechanization, using crop the treatments studied had no effect, because the residues as feed, recycling soil nutrients through treatment effect was not measured correctly or application of manure or crop residues and because the original research was not designed introducing high-yielding dairy cattle – was that with economic impact in mind. In some instances, on-farm resources were underused and could be treatments spread through extension and hence made more efficient if managed in an integrated the research effect could not be separated from crop and livestock enterprise. the extension effect. This conclusion about the weak impact of technology innovations applies Farm mechanization more forcefully to grazing systems than to mixed systems, where crop research has had a signifi- One core assumption of LSR was that a lack cant indirect effect on animal production via of adapted tools and quality feed for draught the pathway of higher grain and crop residue animals blocked crop–livestock integration output. One notable failure of grazing systems (Le Moigne, 1979). This assumption was inves- research was testing treatments under unrepre- tigated by ICRISAT in India and in West Africa sentative conditions; the prominent example of and by ILCA in Ethiopia and West Africa along this failure was grazing reserves or group ranch- two lines – developing new farm tools and im- es that were too small or that lacked adequate proving the condition of work animals. Neither water or land for stock movements. the tool line nor the work conditions line pro- An important example of LSR’s effect on duced widely adopt innovations by farmers animal health is the development of the infec- beyond the intensification response to the increas- tion-and-treatment method vaccine against ECF, ing value of draught power at higher cropping as discussed in Chapter 6 (this volume)81. The intensities. Station work on tools did not pro- targeting of this method to specific types of live- duce successful innovations and was ultimately stock production in East and southern Africa abandoned in the 1980s by ICRISAT and in the has resulted from the work of Brian Perry, Ad- 1990s by ILCA/ILRI. Station research on feed- rian Mukhebi and colleagues on the epidemi- ing draught animals was more successful in that ology of Theileria. a positive effect was observed on work capacity, LSR has generally failed to contribute to but this result did not translate into adoption be- successful project preparation and management cause alternatives to crop residues for supple- in grazing and mixed systems (Wanyoike and mental feeding of work animals were too costly. Baker, 2013). The contribution of research to The books by Pingali et  al. (1987) and the development impact of pastoral projects McIntire et  al. (1992) prompted a rethinking funded and/or managed by the International about agricultural mechanization in Africa, in- Fund for Agricultural Development (IFAD) or cluding views on animal nutrition as a factor in FAO, as measured in a sample of 194 projects stimulating demand for work animals. The au- from Latin America, East Africa, the Middle East thors of these publications found many sites and North Africa, was not significant. where animals were commonly used for tillage Components of higher plant productivity and cultivation, even with feed being scarce in under ranching conditions – fencing, rotational the same sites. They also found many sites with grazing, notably pasture improvements using adequate feed, in terms of quality and seasonal legumes or exotic grasses, mineral fertilizers and availability, with little or no draught animal use of exotic animal breeds – and attempts in power in use. They concluded that livestock nu- projects to introduce them usually failed. trition was not a significant constraint to the adoption of animal traction in sub-Saharan Mixed systems Africa and that research on the topic should accordingly be limited. Proposed development pathways in mixed crop– The sustained growth in farm mechaniza- livestock systems had the common objective of tion in sub-Saharan Africa over the past 30 years integrating crops and livestock with, in some is strong evidence that mechanization, with African Livestock Systems Research, 1975–2018 581 t ractors or animals, does not face insurmountable A major indirect effect on livestock prod- biological or cultural obstacles and has not uctivity occurred via the adoption of high- blocked growth. The present consensus is that er-yielding crop cultivars. Although the Walker animal traction is a viable path for higher farm and Alwang (2015) record of area and yield productivity, but that research has done, and can increases of improved cultivars did not discuss do, little to widen that path. Where conditions for effects on fodder or by-product yields in any of the use of animal draught power do not exist – the species studied, Indian experience with because of low cropping intensity, a domination of major cereals (rice, wheat, maize, sorghum and bush fallow, animal disease and/or poor market pearl millet) has shown such effects to be im- access – then research on health and feed can do portant (Blümmel et  al., 2013, 2014; see also little to expand animal draught power. The same Chapter 14, this  volume). A related effect, lack of demand would, of course, occur if there is a although not well measured or related to re- weak yield, area or cropping-pattern effect attrib- search or extension investments), was the ex- utable to animal power on farm, but research to pansion of irrigation and the associated growth strengthen these effects has not succeeded where of double or triple cropping, allowing higher the systems conditions are unfavourable. Even as fodder output per season and longer periods of seed and fertilizer packages became more widely fodder abundance. used, beginning around 1990 (Walker and Alwang, 2015), they could be used without the legacy of inadequate crop residue animal traction and hence did not require or research. The older work found that crop res- induce greater crop–livestock integration. idues were important shares of animal diets in An exception to generalizations about the livestock systems studied by ILRI, ICARDA agricultural mechanization is the simultaneous and ICRISAT, and in South Asia (Reed et  al., introduction of animal power with a cash row 1988; Kelley et  al., 1993; Renard, 1997)82. crop, such as cotton in the subhumid zone of Recent studies in sub-Saharan Africa (e.g. Hier- West Africa, groundnut in the sandy soils of naux and Ayantunde, 2004) have confirmed Senegal, Mali, Niger and Nigeria, or maize in the this pattern. Crop residues have always been a subhumid savannah. In such areas, simple ex- significant share of ruminant feed intake on tension programmes, combined with profitable small mixed farms; there is significant growth crop production packages, have promoted mech- potential from better use of crop residues, and it anization without significant new research. is likely that the increase in crop residues associ- ated with higher grain production has stimulated crop residues. There is scant evidence of any livestock production. research impact on animal nutrition, livestock IARC research has done too little on crop productivity or on soil quality despite the effort residues as feed, despite their importance invested in feed in mixed (and pastoral) systems. for smallholder livestock throughout sub- Proposed improvements in feed resources have Saharan Africa, West Asia and North Africa, tended to fail to be completely adopted and have and South Asia. A recent and major example only been partially adopted (see Chapter 13, this of this shortcoming is the Africa book by volume, for an explanation of adoption failures Walker and Alwang (2015), which does not of planted forage grasses and legumes in the report crop residue yields83 or relate them to tropics). Ley farming, in the Mediterranean or in grain yields. While the allocation of crop res- the highlands of East Africa, generally failed be- idues between crops and animals was a major cause farm size and environmental conditions research theme at ILCA, in the semi-arid trop- were unfavourable to this technology for any- ics and the Ethiopian highlands, and to a thing other than specialized dairy production. lesser extent at ICARDA in Mediterranean The failure of such proposed improvements in c limates, the impact of this work on livestock communal systems contrasts with the success in systems productivity has been weak. The fail- ranching, where private land tenure and access ure to devote adequate research to crop residue to markets, finance and veterinary care made quality, and its relation to the harvest index, is investment in primary productivity more the single most important gap in African live- remunerative. stock research. 582 J. McIntire, T. Robinson and C. Bosire nutrient cycling. The extensive soil fertility nitrogenous trees in the humid tropics. The straw research for mixed farms in sub-Saharan Africa components of multidimensional crops have been has had an indirect and weak development im- highly successful as by-products of research on pact. Nutrient cycling in mixed systems has been cereals, but breeding programmes targeted at studied by Powell et al. (1995) for sub-Saharan changing plant architecture to lower the harvest Africa, and Boddey et  al. (1996) for Brazil and index have not done well. An indirect effect of Colombia. This is not because mineral fertilizer the failure to raise the quantity and quality of use has not grown; there has been significant feed production was the limited profitability growth of fertilizers and other modern inputs in of animal fattening because of the high costs of sub-Saharan agriculture since the mid-1970s. feed. If the profitability of fattening is changing First, most of the stimulus for mineral fertilizer (e.g. de Haan, 2016), it is more the result of higher use has come from economic growth and market incomes shifting demand than it is to research development in many African countries, which reducing production costs. had both demand and supply effects on fertilizers applications (Townsend, 1999; Morris et  al., the iarc mandate. The mandate of the inter- 2009). A second stimulus has been intensifica- national centres in sub-Saharan Africa and tion of farming systems under the pressure of South Asia – to concentrate on small undercap- population growth and greater market access. italized farms that used few purchased inputs An indirect effect of research has been on the re- and not on large specialized ranches, tree crop cent growth in areas sown with new plant var- estates or well-capitalized arable farms84 – explains ieties, especially of maize, which has stimulated some of the failure to observe a higher develop- fertilizer use. ment impact of systems research. This is not a criticism of the mandate – it is a statement of What blocked the translation of scientific one of its inevitable consequences – nor does impact into development impact? this mean that the international centres should have concentrated on larger, more capitalized The contradiction of strong scientific impact farms; it means that the choice of the small and weak development impact was seen in both mixed farmer as the principal IARC client neces- pastoral and mixed systems. This contradiction sarily reduced returns to research because the had several causes: (i) the difficulty of raising mandate area is more difficult. The long-term primary productivity in arid and semi-arid re- success of private animal breeding research in gions; (ii) the small farm mandate of the inter- the USA, for example, is a complement to public national centres, especially in sub-Saharan Af- research in the USA that has no analogue in rica; (iii) the persistent misapplication of models sub-Saharan Africa. from other agricultural systems; (iv) policy bias against agriculture; and (v) lack of background factor proportions and the misapplication of data at the outset. external models. One obstacle to applying research results in sub-Saharan Africa was the low primary productivity in arid and semi-arid misapplication of principles from ranching sys- areas. The translation of knowledge into prod- tems in the USA, Australia, New Zealand and uctivity has largely failed in the dry areas. One Latin America. These principles – commercial strong indicator of this failure is that there have orientation, private land tenure, pasture improve- been few sustained investments in plant produc- ments and limited animal mobility – were generally tion (e.g. de Haan, 2016). Hypotheses about the not applicable to African pastoralism. A con- share of primary productivity in the yield gap firmation of this was the finding, long ago, that between actual and potential yield, for both African rangeland systems were already effi- grazing and mixed smallholder systems, proved cient in terms of live-weight production per fruitful in terms of scientific impact but less so in hectare, as shown by Cossins (1985) comparing terms of economic impact. New planted forages East Africa and Australia, and Breman and de have not been generally successful in the African Wit (1983) for the USA, Australia and Mali. tropics, with the chief exceptions of cut and Other examples of technology transfer failure carry forages in highland dairying systems and include planted forages and fencing. Despite African Livestock Systems Research, 1975–2018 583 their efficiency at low cash investment and high policies would have discouraged the production labour input, the African pastoral economies of tradable beef and encouraged production of have not grown as fast as other sectors in the non-tradable milk). same economies; this is not because of a failure Lack of complementary policies and exten- to apply research results but rather from the sion effort reduced the development impact of inability to generate research results that raise technologies. An example is trypanosomiasis productivity in arid and semi-arid climates. control. In the absence of a vaccine against this A contributor to weak research impact has disease, an integrated campaign was needed been the persistence of errors in defining prob- to apply the knowledge derived from systems lems. Beginning with the fundamental error of research – drug use, management of resistance, ILCA – that technologies existed in 1975 that vector control and management of trypanotol- could raise productivity of African livestock – erant stock – and practically all elements of such the pattern continued for decades and was seen an integrated campaign have been difficult to as recently as the recommendation of Little and sustain because of fiscal choices made by African Dube (2011) for species and breed diversifica- countries. tion in pastoralism. This recommendation has failed because breed selection and upgrading for lack of background data at the outset. Sys- cattle in the African tropics has long been blocked tem information – soils, rainfall, LGP, land-use by heat, thirst, disease and insects; none of these mapping and productive capacity – only became adverse factors has been removed by research generally available 10–20 years after independ- (with again the notable exception of the eradication ence in sub-Saharan Africa85. It took many of rinderpest in 2011). Beyond breed diversifica- years for such information to spread and to be us- tion, cross-breeding in cattle has been unsuc- able to policy makers and producers; lack of infor- cessful in grazing and mixed systems. The only mation about resources would have limited up- major success is the use of higher-yielding dairy take of technologies adapted to specific resource animals in the cooler highlands of East Africa. types. The reason, as shown in experiments and in the The lack of background data at the outset opinions of the herders, was that exotic races of the systems studies was aggravated by some were more susceptible to heat, thirst and disease. features of older research. Many older studies Contrasting factor proportions – historically were too qualitative, had data quality problems, low labour:land ratios in some parts of sub- notably small sample sizes, and demonstrated Saharan Africa and much higher ratios in failure to conduct or use station research jointly others – produced contrasting obstacles to re- with field studies. In sharp contrast to the Afri- search. It may be argued that rural labour scar- can situation 50 years ago, environmental data city has disappeared in Africa with the doubling in 2020 is much more widely and cheaply of rural population density over the past two available, and lack of such data can no longer be generations, but this is not the way to look at adduced as a reason for costly research delays. labour scarcity. Labour scarcity is a function of This chapter has found few examples of ex relative wages. If urban productivity, and the post economic analysis of grazing systems research purchasing power of urban labour, are higher and development in sub-Saharan Africa. Such than the corresponding rural values, then la- analysis rarely shows value in LSR because the bour will be scarce in rural areas even if rural data were not collected or because returns were population density has grown. If rural tech- negative where the treatment effect was nil or nologies lag behind urban ones, then so will even negative86. Where there were data on inputs rural productivity, producing the appearance of and outputs needed to estimate productivity, it is rural labour scarcity. difficult or impossible to establish a functional link between research effort and productivity at policy bias against agriculture. The long his- the level of the decision-making unit of the farm tory of bias against market agriculture in Africa or the policy unit of the sector. Lastly, in most is well documented. Macro-policies were un- grazing systems, detailed studies are not needed favourable to technologies that were profitable to show that research has failed to raise product- under experimental conditions (e.g. exchange- rate ivity (whether by increasing output or by lowering 584 J. McIntire, T. Robinson and C. Bosire input costs) and hence the necessary work to Another prominent example of a lack of make productivity research estimates has never quantitative work is Sandford’s (1983a) land- been done. mark book on pastoralism in the third world, which has no quantitative analysis of technolo- data problems in livestock studies. Many live- gies, such as grazing management, pasture improve- stock studies had data problems that vitiated ment, animal breeding, calf supplementation, their use in technology or policy analysis. Many water investment or public infrastructure. Early studies were done over periods that were too quantitative work, such as the cattle model of short to capture interseasonal and interannual Konandreas and Anderson (1982) or the Thorn- variability because longer studies were thought ton (1987) model for Colombia, and others to be too costly. Given the interannual variability tended to have short lives and apparently were in rainfall and feed supply, observations of even not applied to generate or to induce gains in as long as 5  years produced noisy estimates of productivity. productivity and its determinants. Even where the biophysical data were of adequate quality to discontinuation of long-term bioeconomic stud- allow economic analysis (e.g. the Nigeria study ies. Work after the merger of ILCA and ILRAD by von Kaufmann et al., 1986, or the Mali stud- has largely dropped integrated long-term stud- ies of Wilson, 1986, Wagenaar et al., 1986, and ies, with the exception of the work in Fakara and Penning de Vries and Djiteye, 1991), their sam- the Kenyan dairying studies of the mid-1990s. ples across units and time were usually too small Despite the changes in population density, crop- to make reliable parameter estimates for periods ping intensity, market access, animal numbers, longer than 2 years. Even where data were of ad- herd composition and infrastructure (transport, equate coverage, problems of selection bias, def- communications, water and energy) and urban- inition of the control or relevance of the treat- ization since the 1980s, the only recent inte- ment made data analysis inconclusive at best. grated work in African grazing systems is that of Homewood and Rodgers (2004) and Homewood lsr was sparse and too often qualitative. There (2008) for East Africa, Lesorogol (2008) for nor- are few examples of rigorous productivity estimates thern Kenya, FAO/CIRAD (2013) for the Sahel in LSR, even of single livestock technologies. and some chapters in Catley et  al. (2013). The One example is A History of Farming Systems decline in long-term field research has stopped Research (Collinson, 2000), which has 12 chap- the generation of adequate data on research and ters and more than 50 authors on dozens of productivity and has therefore prevented com- topics under the general theme of farming sys- parisons over time. Replacements with ex ante tems research in the tropics, covering work from projections over very long periods will continue 1960 to the late 1990s. The book has only three to give uncertain results and therefore will con- empirical studies of animal production: goat tinue to be unreliable policy guides. mange in Kenya (pp. 130–137), and alpaca and dual-purpose cattle in Peru (pp. 341–354). It fails to mention ILRI and its predecessors, while devoting only a few pages (pp. 110–111) to on- The Future farm research with livestock. The ICRISAT compendium on ‘Socioeco- The long-term data requirements of livestock nomic Constraints to Development of Semi-arid systems studies are made more urgent by secular Tropical Agriculture’ (Ryan and Thompson, 1980) changes in crop and animal productivity im- published some 5 years after ICRISAT’s founding posed by climate shifts. Therefore, for mixed and has more than 40 papers, which focus almost en- pastoral systems alike, it will be necessary to tirely on crops. There is one section of the paper conduct new field data collection and analysis on animal traction for smallholdings, one chap- while respecting the hypothesis testing require- ter referring to milk production, scant mention ments of climate change models, especially with of forage crops and one paper (Le Moigne, 1979) respect to parasitology and resistance to abiotic on specific problems of on-farm research with stresses, notably heat and water scarcity. The livestock or with animal-powered implements. long-term effort to sustain and use the ICRISAT African Livestock Systems Research, 1975–2018 585 village-level studies in India should be a model data collection needs to be linked to verifying for LSR in sub-Saharan Africa. The fact that we are predictions of climate change and disease dis- unable to make reliable estimates of the economic tribution models as a way of contributing to return to agricultural research, and especially to policy debates and of estimating the costs of livestock research, in sub-Saharan Africa means policy measures. that a new dedicated effort is required over many years; this effort will be expensive. Mixed systems Pastoral systems Efforts should be made to strengthen or restart long-term data collection and analysis in Governance research should be extended in densely cultivated areas, such as northern pastoral areas with the explicit political goals of Nigeria, the East Africa highlands and the river defending the land rights of pastoralists and basins in the Sahel. preserving minimum areas for these groups. Long-term data collection and analysis should be restarted for pastoral systems in nor- Acknowledgements thern Kenya, southern Ethiopia and in parts of Mali, Niger and Sudan. Special efforts must be The authors thank Jock Anderson, Cees de made where the conflict between wildlife and Haan, Peter Hazell, Mark Powell, Jim Sumberg domestic livestock is acute, in both economic and Trevor Wilson for valuable criticisms. They and biological terms; this is pointedly true thank Catherine Pfeiffer for research assistance along the entire length of the Rift Valley. Such and Susan MacMillan for editorial support. Notes 1 ILRI refers to its predecessors, the International Livestock Centre for Africa (ILCA) and the International Laboratory for Research on Animal Diseases (ILRAD), unless otherwise noted. The English acronym ILCA is used in place of the French acronym CIPEA (Centre International pour l’Elevage en Afrique). 2 One member of ILCA senior management in the 1980s, on his first visit to a Nigerian research station, was informed that cross-bred dairy cattle often suffered from heat stress in the tropics; he suggested that they be put in air-conditioned stalls ‘as the Saudis do.’ 3 The expressions ‘grazing systems,’ ‘pastoral systems’ and ‘pastoralism’ are used interchangeably to refer to farming systems in which: (i) ruminants are the main stock and are raised largely for subsistence; (ii) open rangelands, including browse from trees and shrubs, are the principal source of feed; and (iii) herds are mobile, across places, seasons and years. The use of ‘pastoral’ here is distinguished from the term ‘pastoral farming’ used in parts of the USA, Canada, Argentina, Brazil, Australia and New Zealand, which is a form of private commercial livestock production in which the animals are not mobile and often feed on sown enclosed pastures. 4 Discussion of Livestock’s Long Shadow (Steinfeld et al., 2006) is in the chapter on climate and tropical livestock production (Chapter 16, this volume). 5 Chapter 16 (this volume) discusses the relationship between climate change and tropical livestock production. 6 We use ‘characterization research’ and ‘farming systems research’ interchangeably. Most characterization work fits into the Zilberman and Heiman (2004) classification of policy studies as Class 1 (provision of economic information) or Class II (devising innovations). 7 Risk management and system resilience goals were infrequently stated in earlier livestock research at IARC, although there are exceptions (Binswanger, 1980, for mixed farming in the semi-arid tropics of Central India; Anderson and Dillon, 1992, for the global drylands). 8 We refer to research under the controlled conditions of research stations and ranches as ‘experiments’ and to research under on-farm conditions as ‘surveys’ or as ‘farming systems research.’ 9 The creation of farm typologies was a staple of farming systems research in the francophone countries and such typologies were, in theory, used to target extension messages. 586 J. McIntire, T. Robinson and C. Bosire 10 LGP is the ‘period in days during a year when precipitation exceeds half the potential evapotranspiration’ (FAO, 2014). 11 There are some Köppen cold arid climates (Bwk) in southern Africa but there is very little IARC livestock research on them. 12 Mongolia and Central Asia are not tropical but their livestock systems are often characterized by aridity and nomadism, so they are appropriate comparisons with tropical systems in these respects. 13 There are two higher-elevation livestock types in West Africa: the Mandara mountains bordering Nigeria and Cameroon (Requier-Desjardins, 2011), and the Fouta Djallon of Guinea. 14 Global estimates of livestock systems scale were unreliable until the path-breaking work of Seré and Steinfeld (1996), which became the base of subsequent classifications. While Seré and Steinfeld included non-ruminants in their classification, we concentrate only on ruminants. 15 Seré and Steinfeld (1996) presented two systems – landless livestock monogastric production system; and landless livestock ruminant production system – which are not discussed here. 16 Estimates of farm incomes are unavailable at earlier dates. 17 An example of spatial heterogeneity from highland Kenya is given by de Steeg et al. (2009). An example of managerial heterogeneity is given by Norman et al. (1979, p. 57) who found 230 distinct crop mixtures grown in the semi-arid tropics of northern Nigeria. 18 ‘Mixed farming’ and ‘agro-pastoral’ are used interchangeably throughout this book. 19 A tropical livestock unit or TLU is ‘equivalent to one bovine of 250 kg live weight.’ Typical conversion factors in sub-Saharan Africa are: cattle = 0.7 TLUs, sheep and goats = 0.10 TLUs, pigs = 0.20 TLUs and chicken = 0.01 TLUs. 20 East and southern Africa includes Angola, Burundi, Ethiopia, Kenya, Madagascar, Malawi, Mozambique, Rwanda, South Africa, Tanzania, Uganda, Zambia and Zimbabwe. West and Central Africa here includes Burkina Faso, Cameroon, Chad, Côte d’Ivoire, Ghana, Mali, Mauritania, Niger, Nigeria and Senegal; the other West and Central Africa nations of Guinea, Guinea-Bissau, Liberia, Sierra Leone and Togo have done little work with ILRI and its predecessors. 21 The countries surveyed were Bangladesh, Ecuador, Ghana, Guatemala, Madagascar, Malawi, Nicaragua, Nigeria, Nepal, Pakistan, Panama and Vietnam. 22 Catley et al. (2013), including studies of the Republic of Sudan, Tanzania, Ethiopia and Kenya, was inconclusive on changes in inequality. Devereux (2006, p. 75) reported a Gini coefficient of 0.74 for income among pastoralists in the Somali Region of Ethiopia. 23 Alene’s calculation depends on the weighting structure of the productivity data. An alternative weighting structure gave only a 0.1% annual productivity increase (Alene, 2010, p. 229). 24 The study by Wilson (1986, p. 14) over 6 years in semi-arid north central Mali argued that the ‘degree of de- pendence on the animal raised’ was the primary criterion, not mobility, which he contended was ‘contingent’ on the degree of dependence on animal products in food consumption and in income. In practice, high depend- ence on animal products in consumption and income is always associated with mobile grazing in Africa. 25 Global environmental effects are estimated in Chapter 16 (this volume) on ‘Ruminant Livestock and Climate Change in the Tropics.’ 26 Tribe et al. (1973), whose authors included at least one future ILCA Board Chair, dismissed African pas- toralism as lacking the capacity for rational management (see pp. 14, 23, 24 and 53). The report alleged that such incapacity caused overgrazing (p. 11) and called for large-scale public intervention to manage livestock movement, marketing, soil, water and vegetation (pp. 23–24). 27 The first Board chair of ILCA declared the standard view of African rangelands in 1975 (Hodgson, 1975, p. 19): ‘The productivity of the animal population [in Africa] is low and inefficient. A significant cause is the low and deteriorating productivity of the rangelands.’ 28 Leaving aside Arctic and high-altitude grazing. 29 ‘…of the 80 herbaceous tropical arid zone species of forages which have been tried at Niono (550 mm) in Mali in 1977–1980, not one single species became established and amenable to produce a grazing impact’ (Le Houérou,1989, p. 149). 30 Le Houérou (1989, p. 149) mentioned the successful implantation of Acacia tortilis and Acacia senegal near M’Bidi, Senegal. 31 Hodgson’s address to the 1975 Sub-Saharan Africa Rangelands Seminar, which was the first major international scientific meeting of the new ILCA, explicitly declared the ‘cultural’ reasons for the supposed overgrazing and supposed low productivity of rangelands. 32 As stated in the Swynnerton Plan for Kenyan agriculture, as summarized by de Wilde (1967, p. 4, pp. 174–187) for Kenya, Smith (1976, pp. 110–151) and Thurston (1987). The biased and ultimately African Livestock Systems Research, 1975–2018 587 unproductive notion of pastoralist mismanagement was insightfully criticized by ILCA/ILRI scientist Ralph von Kaufmann (von Kaufmann,1976). 33 Grigg continued to say, correctly, that it would be unwise to enforce sedentarization of pastoralists. 34 Many distinguished scientists were involved in the Borana work including Jean-Claude Bille, Assefa Eshete, Michel Corra, C.S. Kamara, Mark Nicholson, Jess Reed, Solomon Desta and Andrea Woodward. 35 Recent work (de Haan, 2016) introduces subclasses of LGA and livestock grazing (LG) semi-arid and LG sub-humid, which would apply to parts of the Borana study area. 36 Toutain and Boudet (1980, pp. 427–432) mentioned historical restrictions on cutting browse from Acacia albida, imposed either by the traditional authorities in eastern Niger, by the French colonists, or by the government of independent Mali in the 1960s, following the colonial model. 37 ILCA abandoned a study of Afar pastoralists in eastern Ethiopia for security reasons in the 1970s. 38 One example of the information available to the Maasailand researchers is von Kaufmann (1976) covering pastoral problems and proposed solutions in Kenyan rangelands up to the mid-1970s. 39 ‘Agro-pastoralists’ in West Africa usually means Fulani farmers who practised a mix of settled rain-fed farming and seasonally nomadic grazing. In some instances (Blench, 1999, for northern Nigeria; Waters-Bay- er and Taylor-Powell, 1986b) such groups had been settled for many years. In others (Benoit, 1979, for western Burkina Faso), herders had recently settled or were in the process of settling as recently as the 1970s. 40 This book is an outstanding work that is regrettably barely cited in Google Scholar and not cited at all in Scopus. It made a major contribution to the scientific literature, notably through the work of Wolfgang Bayer, David Bourn, M.A. Ibrahim, Salisu Ingawa, J.A. Maina, E O. Otchere, Mark Powell, Mohammed Saleem, Ellen Taylor-Powell, Ralph von Kaufmann, Ann Waters-Bayer and William Wint. 41 Ellis and Swift (1988) stressed that herders could recover well from droughts of one year, but that longer dry periods would cause greater losses of animals, work and income, forcing herders into other jobs. 42 Grigg's observation is correct in that the ranches did fail, but his explanation of the failure is wrong. 43 The studies by Fratkin et al. (1994) and Galaty (1994, p. 189) did not quantify the losses in product- ivity and equity from enforced sedentarization, but it is clear from these works that such losses were substantial. 44 Mortimore (2000, p. 4), in three states in northern Nigeria, found a decline in agricultural land (cropped) of 8.4% from 1976–1978 to 1993–1995, of which –12.4% was woodlands, +1.3% was grassland and +3.0% was degraded. 45 This was decades after the British had taken much of the Maasai grazing land, which had narrowed the resource base of the group ranches before they were ever constituted. 46 See Rutten (1992) on the individualization of ownership in Kajiado district of southern Kenya from 1890 to 1990. 47 The technical reviews by King (1983) and Sandford (1983a) concerned cattle for meat and dairy. Sandford (1983a, p. 49) noted a lack of information about water use by draught animals in highland Ethiopia. 48 Homewood and Rodgers (2004, p. 252) reported from Tanzania that energy loss incurred by trekking for water was the ‘single biggest constraint to milk production.’ 49 The books of Smith (1992), Fratkin et al. (1994); Scoones and Wolmer (2002), Behnke and Scoones (1993), Homewood (2008), Bollig et al. (2013) and Catley et al. (2013) do not mention water experiments, on station or in the field. The range and herd modelling in Behnke and Scoones (1993) did not use water as an objective or a constraint, nor did that of Konandreas and Anderson (1982). Mengistu et al. (2007) observed successful adaptive physiological mechanisms in Ethiopian Somali cattle when subjected to intermittent watering. 50 This statement does not apply to Latin America, where the market potential of beef had stimulated much earlier research on soils, pastures, livestock disease and animal breeds for commercial ranching. 51 Examples are Monod (1975) and papers cited in Sandford (1983a). 52 Chapter 13 discusses the research impact of quality improvements in crop residues for mixed systems, mainly in semi-arid India and in semi-arid and subhumid West Africa. 53 The book by Pratt and Gwynne (1977), while limited geographically to East Africa, covered more than 30 years of work and made careful reference to rangeland studies in the USA and Australia. 54 Feed quality was a major theme in the early CIAT Beef Program (CIAT, 1973). 55 The papers cited by Macharia et al. (2011) confirm their findings: station results are positive, field trials are mixed, adoption data are rare and benefit–cost analysis is absent. The study by Nicholson and Mengis- tu (2016) confirmed the lack of adoption of forage legumes in grazing systems and on mixed farms in Kenya and Ethiopia using surveys in 2014 and 2015. 588 J. McIntire, T. Robinson and C. Bosire 56 Chapter 11 (this volume) covers the African range ecology work of ILCA, ILRI and partners. Successes with planted forage grasses in Latin America were mainly on large commercial farms in Brazil and on some smallholder areas in central America, as discussed in Chapter 12 (this volume). 57 Medicago sativa (known as alfalfa in the USA and lucerne in much of Europe) is the most common tem- perate forage legume but is rarely planted in the semi-arid and subhumid tropics. 58 Detailed papers on these diseases are in the respective chapters of Part I. Screw worm was accidentally introduced into Libya from the new world in the late 1980s and was eradicated with sterile insect techniques and quarantine measures by the mid-1990s.The eradication of screw worm in Libya and the global eradi- cation in 2011 of rinderpest, a viral disease of cattle and some ungulates, are special cases and are not discussed here. 59 Konandreas and Anderson (1982, pp. 3–5) reviewed earlier models, including that of CIAT for ranching in Latin America and that of Texas A&M University (Cartwright et al., 1982). 60 Norton (1975, pp. 313–322) wrote that it would be ‘unwise to apply our understanding of American deserts directly to the Sahel’ because of differences in stocking rates, climate and flora. 61 The bibliography in Penning de Vries and Djiteye (1991) cites many individual studies of soils, water, plants and animals, but few analysed these elements jointly over time, at the same site, or applied simula- tion models to validate and project the field results; a significant exception is Hiernaux and Ayantunde (2004). 62 The IFPRI IMPACT model was not detailed enough in 2009 to project outcomes for individual livestock systems of any type (Nelson et al., 2009). 63 It is not possible to calculate an IRR for the ‘milking with supplementation’ strategy from the original paper, but it appears from Cartwright et al. (1978, p. 60, Table 5.4) that it would be positive. 64 Water was not a major cost of cattle production in the Inner Delta of the Niger River in Mali and hence the studies there had at least one large difference from other work in semi-arid West Africa. Wagenaar et al. (1986, p. 5) found that, ‘Throughout the year, the herds are watered at least twice daily.’ 65 Coppock (1994, p. 196) referred to ‘grain imports’ but his context implied not ‘imports’ in the sense of pur- chases from foreign suppliers but ‘imports into the grazing system’ in the sense of purchases from domes- tic and/or foreign suppliers. 66 Jahnke (1982, p. 143) estimated that roughly 275,000 km2 had been cleared of tsetse in Nigeria, Zim- babwe, Tanzania and Uganda between 1947 and 1978 in a total endemic area of 10.3 million  km2 in sub-Saharan Africa (see Box I.1 in the Introduction in this volume). 67 Vetter (2004) summarized the origins of the equilibrium model, the critique by advocates of the non- equilibrium model and the situations in which one or both might apply. 68 Dixon et al. (2001) defined eight groups, of which six are relevant here; the others are coastal artisanal fishing and urban agriculture. 69 Walker and Alwang (2015, p. 210) found that fewer than 10% of maize scientists in East and southern Africa were from ‘social science’ or ‘farming systems.’ They also found that more than 60% of scientists in 20 crops were in plant breeding, plant pathology, molecular biology and tissue culture (p. 378). 70 The term ‘conservation farming’ has been introduced more recently as a model of nutrient cycling (e.g. Dixon et al., 2001, pp. 51–52). 71 Farm mechanization with power tillers and tractors was part of the research portfolios of the International Rice Research Institute (IRRI; Chancellor, 1998), and of ICRISAT at times. The goal of this research was to accelerate field tasks by replacing animals to achieve higher cropping intensity. IRRI and other centres also studied mechanization of water supply, threshing and milling, but this work typically involves engines rather than animal power. 72 Some loss of potential impact from the Niger study was due to using a sample size of only seven animals in its trial of energy expenditures, and to sampling animals who weighed roughly one-third more than those animals typically found on farms in western Niger (Fall et al., 1997). 73 There was little soil fertility work ‘on farm’ in the pastoral areas. 74 CIAT’s orientation was on improving soil fertility by adding mineral fertilizers and raising the pH of acid soils without returning manure to the soil in a systematic way. The CIAT work had a major development impact on the acid soils of the Latin American savannahs, estimated to be more than 250 million ha in Brazil, Colombia and Venezuela alone (Lynam and Byerlee, 2017, p. 83) and is one of the principal successes of natural resource management work from the international agricultural research centres. 75 The cattle age/sex pyramid for the Malian Delta in Wagenaar et al. (1986, p. 10) showed the ratio of females to males increased sharply after the age of 3–4 years. The cattle age/sex data for Mali (Wilson,1986, p. 36) found higher female:male ratios in herds not used for animal draught; in herds used for draught power, the African Livestock Systems Research, 1975–2018 589 female:male ratios were roughly equivalent in one sedentary herd for milk and draught and much less than 1 in another sedentary herd. The female:male ratio in the Maasailand study (Solomon Bekure et al., 1991, p. 83) ranged from 1.86 (rich households) to 2.23–2.30 (poor households) with a mean of 1.97. The ratio of females:males was greater than 3 before the 1983 drought in southern Ethiopia and about 4 in 1985 after the drought (Coppock, 1994, p. 168) 76 The foreword to Pratt and Gwynne (1977, p. vii) illustrated the ‘mainstream view’: ‘…in the last few dec- ades, stock numbers have increased so much that extensive areas have been severely overgrazed and now have an extremely low annual productivity, far below that which the land is capable of producing under good management.’ 77 Examples are Thornton et al. (2006), Thornton et al. (2009) and Thornton and Herrero (2010). 78 The books of Katherine Homewood (notably, Homewood and Rodgers, 2004; Homewood, 2008), on East African pastoralism, although not ILRI work, are in the same tradition. 79 An exception is parts of East and southern Africa, where ILCA and ILRAD did little apart from early cattle modelling on Botswana (Konandreas and Anderson, 1982) and the epidemiological work of Brian Perry and colleagues in the second half of the 1980s. 80 Network models have also been used in animal traction, forages, animal genetics and economics. 81 The exception to this generalization about the lack of a productivity effect from livestock research in sub-Saharan Africa is the eradication of rinderpest (Roeder and Rich, 2009). Even for rinderpest, the mod- ern research effect was limited to a policy for delivery systems. The vaccine was developed long before eradication was declared in 2011 and the role of systems research in the campaign was very small. 82 There is a long history of crop research in the USA, Canada and Australia on the value of crop residues as a soil amendment. A review by Wilhelm et al. (2004) on removal of stover in the Maize Belt in the USA found studies covering nearly 70 years. 83 Chapter 13 (this volume) reports rough estimates of the contribution of improved cultivars of dryland cereals (maize, pearl millet and sorghum) to animal feed availability for India and some countries in sub-Saharan Africa. Walker and Alwang (2015, pp. 32, 228 and 232) made occasional reference to straw as feed but did not report data on crop residues as feed, changes in harvest index in modern varieties or changes in feed quality of crop residues. 84 There was a focus on large farms at CIAT in research on rice cultivars, beef ranching, forage grasses and acid soil management. 85 This had been available for decades in the livestock systems in South Asia and parts of Latin America. The ILCA (1975) proceedings of ‘Inventaire et cartographie des Pâturages Tropicaux Africains: Actes du Colloque Bamako’ was a landmark in compiling and publishing this information. 86 The chapter by Evenson (2001) in the Handbook of Agricultural Economics cited one estimate on pasture research among more than 100 estimates of returns to research, and fewer than five papers on livestock in a global total of more than 100 papers on crops and livestock. 84This had been available for decades in the livestock systems in South Asia and parts of Latin America. The ILCA (1975) proceedings of ‘Inventaire et cartographie des Pâturages Tropicaux Africains: Actes du Colloque Bamako’ was a landmark in compiling and publishing this information. References Alene, A.D. (2010) Productivity growth and the effects of R&D in African agriculture. Agricultural Economics 41, 223–238. Alston, J.M., Chan-Kang, C., Marra, M.C., Pardey, P.G. and Wyatt, T. (2000) A meta-analysis of rates of return to agricultural R&D: ex pede herculem? Research Report No. 113. IFPRI, Washington. DC. Anderson, J.R. and Dillon, J.L. (1992) Risk analysis in dryland farming systems. FAO, Rome. Angassa, A. and Beyene, F. (2003) Current range condition in southern Ethiopia in relation to traditional management strategies: the perceptions of Borana pastoralists. Tropical Grasslands 37, 53–59. Angassa, A. and Oba, G. (2008) Herder perceptions on impacts of range enclosures, crop farming, fire ban and bush encroachment on the rangelands of Borana, southern Ethiopia. Human Ecology 36, 201–215. Baltenweck, I., Staal, S.J., Ibrahim, M.N.M., Herrero, M., Holmann, F., et al. (2003) Crop-livestock intensifi- cation and interactions across three continents: main report. System-wide Livestock Program (SLP) project on transregional analysis of crop-livestock systems. ILRI, Nairobi. 590 J. McIntire, T. Robinson and C. Bosire Barbier, B. and Hazell, P. (1999) Implications of declining access to transhumant areas and sustainability of agro-pastoral systems in the semi-arid areas of Niger: a bioeconomic modeling approach. In: McCarthy, N., Swallow, B., Kirk, M. and Hazell, P. (eds) Property Rights, Risk, and Livestock Development in Africa. IFPRI, Washington. DC. Bartholomé, E. and Belward, A.S. (2005) GLC 2000: a new approach to global land cover mapping from Earth observation data. International Journal of Remote Sensing 26, 1959–1977. Bartholomew, P. (1988) Livestock production in central Mali: studies on meat and milk production, animal traction and forage production. ILCA. Bamako, Mali. Bayer, W. (1986) Utilization of fodder banks. In: von Kaufmann, R., Chater, S. and Blench, R. (eds) Livestock Systems Research in Nigeria’s Subhumid Zone, 29 October–2 November 1984, Kadu- na, Nigeria. ILCA, Addis Ababa. Behnke, R.H. (1985) Measuring the benefits of subsistence versus commercial livestock production in Africa. Agricultural Systems 16, 109–135. Behnke, R.H. and Scoones, I. (1993) Rethinking range ecology: implications for rangeland management in Africa. Environment Working Paper No. 53. World Bank, Washington, DC. Behnke, R.H, Scoones, I. and Kerven, C. (eds) (1993) Range Ecology at Disequilibrium. ODI, London. Benoit, M. (1979) Le Chemin des Peul du Boobola: Contribution é l’Ecologie du Pastoralisme en Afrique des Savanes. Orstom, Paris. Beyene, F. (2016) Land use change and determinants of land management: experience of pastoral and agro-pastoral herders in eastern Ethiopia. Journal of Arid Environments 125, 56–63. Binswanger, H.P. (1980) Attitudes towards risk: experimental measurement in rural India. American Journal of Agricultural Economics 62, 395–407. Binswanger, H.P. and Rosenzweig, M.R. (1986) Behavioral and material determinants of production rela- tions in agriculture. Journal of Development Studies 22, 503–539. Blench, R. (1994) The expansion and adaptation of fulbe pastoralism to subhumid and humid conditions in Nigeria. Cahiers d’Etudes Africaines 135, 197–212. Blench, R. (1999) Why are there so many pastoral groups in eastern Africa. In: Pastoralists Under Pressure: Fulbe societies confronting change in West Africa, Brill, Leiden, Netherlands, pp. 29–50. Blench, R. (2001) ‘You can’t go home again’: pastoralism in the new millennium. Overseas Development Institute, London. Blümmel, M., Grings, E.E. and Erenstein, O. (2013) Potential for dual-purpose maize varieties to meet changing maize demands: Synthesis. Field Crops Research 153, 107–112. Blümmel, M., Zaidi, P.H., Vinayan, M.T., Babu, R. and Yadav, O. P. (2014) The fodder value of maize stover vis-à-vis other cereals residues and opportunities for improvement through crop breeding and value addition. In: Prasanna, B.M., Vivek, B.S., Sadananda, A.R., Jeffers, D., Zaidi, P.H., et al. (eds) Maize for Food, Feed, Nutrition and Environmental Security, 30 October–1 November, Bangkok, Thailand. Boddey, R.M., Rao, I.M. and Thomas, R.J. (1996) Nutrient cycling and environmental impact of Brachiaria pastures. In: Miles, J.W., Maass, B.L., do Valle, C.B. and Kumble, V. (eds) Brachiaria: Biology, Agron- omy, and Improvement. CIAT; Cali, Columbia, pp. 72–86. Bollig, M., Schnegg, M. and Wotzka, H.-P. (eds) (2013) Pastoralism in Africa: Past, Present and Future. Berghahn Books, New York/Oxford. Bontemps, S., Defourny, P., Bogaert, E.V., Arino, O., Kalogirou, V. and Perez, J.R., (2011) GLOBCOVER 2009 – Products description and validation report. Available at: https://epic.awi.de/id/eprint/31014/16/ GLOBCOVER2009_Validation_Report_2-2.pdf (accessed 22 February 2020) Boserup, E. (1965) The Conditions of Agricultural Growth: The Economics of Agrarian Change Under Population Pressure. Aldine, Chicago, Illinois. Bosire, C.K., Ogutu, J.O., Said, M.Y., Krol, M.S., de Leeuw, J. and Hoekstra, A.Y. (2015) Trends and spatial variation in water and land footprints of meat and milk production systems in Kenya. Agriculture, Ecosystems & Environment 205, 36–47. Bosma, R., Bos, M., Kante, S., Kebe, D. and Quak, W. (1999) The promising impact of ley introduction and herd expansion on soil organic matter content in southern Mali. Agricultural Systems 62, 1–15. Bourn, D. and Wint, W. (1994) Livestock, land use and agricultural intensification in sub-Saharan Africa. Pastoral Development Network Discussion Paper. Overseas Development Institute, London. Breman, H. and Cissé, A.M. (1977). Dynamics of Sahelian pastures in relation to drought and grazing. Oecologia 28(4), pp. 301–315. Breman, H., and de Wit, C.T. (1983) Rangeland productivity and exploitation in the Sahel. Science 221, 1341–1347. African Livestock Systems Research, 1975–2018 591 Bryant, J.R. and Snow, V.O. (2008) Modelling pastoral farm agro-ecosystems: a review. New Zealand Jour- nal of Agricultural Research 51, 349–363. Cartwright, T.C., Anderson, F.M., Buck, N.G., Nelsen, T.C., Trail, J.C.M., et al. (1978) Mathematical modelling of livestock production systems: application of the Texas A&M University beef cattle production model to Botswana. ILCA Systems Study No. 1. ILCA, Addis Ababa. Cartwright, T.C., Anderson, F.M., Buck, N.G., Nelsen, T.C., Trail, J.C.M., et al. (1982) Systems modelling in cattle production – an application in Botswana. World Animal Review 41, 40–45. Catley, A., Lind, J. and Scoones, I. (eds) (2013) Pastoralism and Development in Africa: Dynamic Change at the Margins. Routledge, London. Cervigni, R. and Morris, M. (eds) (2016) Confronting drought in Africa’s drylands: opportunities for enhan- cing resilience. World Bank, Washington, DC. Chancellor, W. (1998) Agricultural mechanization: a history of research at IRRI and changes in Asia. In: Bell, M.A., Dawe, D. and Douthwaite, M.B. (eds) Increasing the Impact of Engineering in Agricultural and Rural Development. ILRI, Manila, Philippines, pp. 31–38. Christiansen, S., Bounejmate, M., Sawmy-Edo, H., Mawlawi, B., Shomo, F., et al. (2000) Tah village project in Syria: another unsuccessful attempt to introduce ley-farming in the Mediterranean basin. Experimental Agriculture 36, 181–193. CIAT (1973) Report of the external review team: beef production systems program. CIAT, Cali, Colombia. Classen, G.A., Edwards, K.A. and Schroten, E.H.J. (1983) The water resource in tropical Africa and its exploitation. Research Report No. 6. ILCA, Addis Ababa. Collinson, M. (ed.) (2000) A History of Farming Systems Research. FAO, Rome, and CAB International, Wallingford, UK. Coppock, D.L. (1994) The Borana plateau of southern Ethiopia: synthesis of pastoral research, development, and change, 1980–91. Systems Study No. 5. ILCA, Addis Ababa. Cossins, N. (1985) The productivity and potential of pastoral systems. ILCA, Addis Ababa. Dahl, G. and Hjort, A. (1976) Having herds: pastoral herd growth and household economy. Department of Social Anthropology, University of Stockholm. de Haan, C. (ed.) (2016) Prospects for Livestock-based Livelihoods in Africa’s Drylands. World Bank, Washington, D.C. de Leeuw, P.N. and Milligan, K. (1984) A review of integrated surveys for resource inventory and monitoring of pastoral production systems in sub-Saharan Africa. ILCA Bulletin No 20. ILCA, Addis Ababa. de Leeuw, P.N., Solomon Bekure and Grandin, B.E. (1984) Aspects of livestock productivity in Maasai group ranches in Kenya. ILCA, Nairobi. de Steeg, J.A., Verburg, P.H., Baltenweck, I. and Staal, S.J. (2009) Characterization of the spatial distribu- tion of farming systems in the Kenyan Highlands. Applied Geography 30, 239–253. de Wilde, J. (1967) Experiences with Agricultural Development in Tropical Africa. Johns Hopkins Univer- sity Press, Baltimore, Maryland. Delgado, C.L. (1980) L’élevage par rapport à l’agriculture au sud-est de la Haute-Volta: analyse de l’alloca- tion des ressources au niveau de l’exploitation. Monograph 1. Center for Research on Economic Development, University of Michigan, Ann Arbor, Michigan. Devereux, S. (2006) Vulnerable Livelihoods in Somali Region, Ethiopia. Institute of Development Studies, Brighton, UK. Dixon, J.A., Gibbon, D.P. and Gulliver, A. (2001) Farming systems and poverty: improving farmers’ livelihoods in a changing world. World Bank, Washington, D.C. Doppler, W. (1980) The Economics of Pasture Improvement and Beef Production in Semi-humid West Africa. Deutsche Gesellschaft für Technische Zusammenarbeit, Eschborn, Germany. Dupire, M. (1972) Les Facteurs Humains de l’Économie Pastorale, Centre Nigérien de Recherches en Sciences Humaines, Niamey. Dyson-Hudson, N. (1966) Karimojong Politics. Clarendon Press, London. Dyson-Hudson, R. and Dyson-Hudson, N. (1969) Subsistence herding in Uganda. Scientific America 220, 76–89. Dyson-Hudson, R. and Dyson-Hudson, N. (1980) Nomadic pastoralism. Annual Review of Anthropology 9, 15–61. Ehui, S.K. and Spencer, D.S. (1993) Measuring the sustainability and economic viability of tropical farming systems: a model from sub-Saharan Africa. Agricultural Economics 9, 279–296. Ellis, J.E. and Swift, D.M.J. (1988) Stability of African pastoral ecosystems: alternate paradigms and implications for development. Journal of Range Management Archives 41, 450–459. 592 J. McIntire, T. Robinson and C. Bosire Evenson, R.E. (2001) Economic impacts of agricultural research and extension. In: Gardner, B.L. and Rausser, G.C. (eds) Handbook of Agricultural Economics, Vol. 1. Elsevier Science, Amsterdam, pp. 574–616. Evenson, R.E. and Rosegrant, M. (2003) The economic consequences of crop genetic improvement programmes. In: Evenson, R.E. and Gollin, D. (eds) Crop Variety Improvement and Its Effect on Productivity. CAB International, Wallingford, UK, pp. 473–498. Fall, A., Pearson, R.A., Lawrence, P.R. and Fernández-Rivera, S. (1997) Feeding and working strategies for oxen used for draft purposes in semi-arid West Africa. ILRI, Nairobi. FAO (1975) Agro-ecological Zones Project. FAO, Rome. FAO (2014) Climpag: Climate Impact on Agriculture. FAO, Rome. Available at: http://www.fao.org/nr/ climpag/ (accessed 22 February 2020). FAO/CIRAD (2013) Information system on pastoralism in the Sahel: atlas of trends in pastoral systems in the Sahel, 1970–2012. FAO, Rome. Fratkin, E., Galvin, K.A. and Roth, E.A. (1994) African Pastoralist Systems: An Integrated Approach. Lynne Reinner Publishers, Boulder, CO. Fratkin, E. (2001) East African pastoralism in transition: Maasai, Boran, and Rendille cases. African Stud- ies Review 44, 1–25. Fricke, W. (1979) Cattle Husbandry in Nigeria: A Study of its Ecological Conditions and Social-geograph- ical Differentiations. Geographischen Arbeiten, Heidelberg, Germany. Fritz, S., See, L., McCallum, I., You, L., Bun, A., Moltchanova, E., et al. (2015) Mapping global cropland and field size. Global Change Biology 21, 1980–1992. Galaty, J.G. (1994) Rangeland tenure and pastoralism in Africa. In: Fratkin, E., Galvin, K.A. and Roth, E.A. African Pastoralist Systems: An Integrated Approach. Lynne Reinner Publishers, Boulder, CO, pp. 185–204. Galaty, J.G. (2013b) The collapsing platform for pastoralism: land sales and land loss in Kajiado County, Kenya. Nomadic Peoples 17, 20–39. Gallais, J. and Boudet, G. (1980) Projet de code pastoral concernant plus spécialement la région la région du delta central du Niger au Mali. GERDAT-IEMVT, Maisons-Alfort, France. Gatenby, R.M. and Trail, J.C.M. (1982) Small Ruminant Breed Productivity in Africa. Proceedings of a seminar held at ILCA, Addis Ababa, Ethiopia, October 1982. ILCA, Addis Ababa. Gerber, P., Steinfeld, H., Henderson, B., Mottet, A., Opio, C., et al. (2013) Tackling climate change through livestock: a global assessment of emissions and mitigation opportunities. FAO, Rome. Getachew Asamenew, Hailu Beyene, Negatu, W. and Ayele, G. (1993) A survey of the farming systems of Vertisol areas of the Ethiopian highlands. ILCA, Addis Ababa. Giller, K. (ed.) (2001) Nitrogen Fixation in Tropical Cropping Systems. CAB International, Wallingford, UK. Gourou, P. (1966) The Tropical World: Its Social and Economic Conditions and its Future Status, 4th edn. Longmans Green, London. Grace, D. (2003) Rational drug use to better manage trypanosomosis and trypanocide resistance. ILRI Working Paper No 4. ILRI, Nairobi. Grandin, B.E. (1987) Pastoral culture and range management: recent lessons from Maasailand. ILCA Bulletin No. 28. ILCA, Addis Ababa. Grandin, B.E. (1988) Wealth and pastoral dairy production: a case study from Maasailand. Human Ecology 16, 1–21. Grigg, D.B. (1974) The Agricultural Systems of the World: An Evolutionary Approach. Cambridge Univer- sity Press, London. Gryseels, G. (1988) Role of livestock on mixed smallholder farms in the Ethiopian highlands. PhD thesis, State Agricultural University, Wageningen, The Netherlands. Gryseels, G. and Anderson, F.M. (1983) Research on farm and livestock productivity in the central Ethiopian highlands: Initial Results, 1977–1980. ILCA Research Report No. 4. ILCA, Addis Ababa. Hazell, P.B.R. (2008) An assessment of the impact of agricultural research in South Asia since the Green Revolution. Science Council Secretariat, Rome. Hazell, P.B.R. (2013) Comparative study of trends in urbanization and changes in farm size in Africa and Asia: implications for agricultural research. ISPC/CGIAR. Hazell, P.B.R., Poulton, C., Wiggins, S. and Dorward, A. (2010) The future of small farms for poverty re- duction and growth. 2020 Discussion Paper 42. IFPRI, Washington, DC Helland, J. (1980) Some aspects and implications of the development of grazing blocks in northeastern province, Kenya. ILCA Kenya Working Document No. 18. ILCA, Nairobi. African Livestock Systems Research, 1975–2018 593 Herrero, M., Thornton, P. Notenbaert, A., Msangi, S., Wood, K.R., et al. (2012) Drivers of change in crop– livestock systems and their potential impacts on agro-ecosystems services and human wellbeing to 2030: a study commissioned by the CGIAR Systemwide Livestock Programme. ILRI, Nairobi. Herrmann, S.M., Anyamba, A. and Tucker, C.J. (2005) Recent trends in vegetation dynamics in the African Sahel and their relationship to climate. Global Environmental Change 15, 394–404. Hiernaux, P. (1980) La carte des ressources fourragères des parcours du delta intérieur du Niger. Notice. CIPEA, Bamako. Hiernaux, P. (1984) Distribution des pluies et production herbacée au Sahel: une méthode empirique pour caractériser la distribution des précipitations journalières et ses effets sur la production herbacée. Premiers résultats acquis dans le Sahel malien. Document de Programme AZ 98. CIPEA, Bamako. Hiernaux, P., Bielders, C.L., Valentin, C., Bationo, A. and Fernandez-Rivera, S. (1999) Effects of livestock grazing on physical and chemical properties of sandy soil in sahelian rangelands. Journal of Arid Environments 41, 231–245. Hiernaux, P. and Ayantunde, A. (2004) The Fakara: a semi-arid agro-ecosystem under stress. Report of research activities of International Livestock Research Institute (ILRI) in Fakara, South-western Niger between 2002 and 2004. ILRI, Nairobi. Hiernaux, P., Diarra, L. and Cisse, M.I. (1983) Ressources végétales et leur productivité. In: Wilson, R.T., de Leeuw, P.N. and de Haan, C. (eds) Recherches sur les Systèmes des Zones Arides du Mali: Résultats Préliminaires. Rapport de Recherche 5. CIPEA, Addis Ababa, pp. 28–46. Hiernaux, P., Ayantunde, A., Kalilou, A., Mougin, E., Gérard, B., et al. (2009) Trends in productivity of crops, fallow and rangelands in Southwest Niger: Impact of land use, management and variable rain- fall. Journal of Hydrology 375, 65–77. Hill, P. (1982) Dry Grain Farming Families: Hausaland (Nigeria) and Karnataka (India) Compared. Cambridge University Press, Cambridge. Hodgson, R.E. (1975) The sub-Saharan Africa rangeland resource. In: Evaluation and Mapping of Tropical African Rangelands. Proceedings of the seminar, 3–8 March, Bamako, Mali. ILRI, Addis Ababa, pp. 19–20. Hoekstra, A.Y. and Mekonnen, M.M. (2012) The water footprint of humanity. Proceedings of the National Academy of Sciences USA 109, 3232–3237. Homewood, K. (2008) Ecology of African Pastoralist Societies. James Currey, Oxford, UK. Homewood, K. and Rodgers, W.A. (2004) Maasailand Ecology: Pastoral Development and Wildlife Ecol- ogy in Ngorongoro, Tanzania. 2nd ed. Cambridge University Press, Cambridge. Hulme, M., Doherty, R., Ngara, T., New, M. and Lister, D. (2001) African climate change: 1900–2100. Climate Research 17, 145–168. Ibrahim, M., Guerra L., Casasola, F. and Neely, C. (2010) Importance of silvopastoral systems for mitiga- tion of climate change and harnessing of environmental benefits policy and economics. Proceedings of the workshop on the role of grassland carbon sequestration in the mitigation of climate change. In: Abberton, M., Conant, R. and Batello, C. (eds). Grassland Carbon Sequestration: Management, Pol- icy and Economics. FAO, Rome. IDRC/ILCA (1983) Pastoral systems research in sub-Saharan Africa. Proceedings of a Workshop held at ILCA, Addis Ababa, Ethiopia. ILCA, Addis Ababa. ILCA (1975) Evaluation and Mapping of Tropical African Rangelands. Proceedings of the seminar, 3–8 March, Bamako, Mali. ILCA, Addis Ababa, Ethiopia. ILCA (1979a) Livestock production in the subhumid zone of West Africa: A Regional Review. ILCA Sys- tems Study 2. Addis Ababa, Ethiopia. ILCA (1979b) Small ruminant production in the humid tropics. ILCA Systems Study No. 3. Addis Ababa, Ethiopia. ILCA (1994) Improving Livestock Production in Africa: Evolution of ILCA’s Programme, 1974–94. ILCA, Addis Ababa. ILCA/ILRAD (1988) Livestock Production in Tsetse Affected Areas of Africa. Proceedings of a meeting held in Nairobi, Kenya, 23–27 November 1987. ILRI, Addis Ababa, and ILRAD, Nairobi. ILRI (1998) Annual Project Progress Report 1997. ILRI, Addis Ababa. Itty, P. (1992) Economics of Village Cattle Production in Tsetse Affected Areas of Africa: A Study of Tryp- anosomiasis Control Using Trypanotolerant Cattle and Chemotherapy in Ethiopia, Kenya, Côte d’Ivo- ire, The Gambia, Zaire and Togo. Hartung-Gorre Verlag, Konstanz. Itty, P. and Swallow, B.M. (1994) The economics of trypanotolerant cattle production in regions of origin and areas of introduction. In: Rowlands G.J. and Teale A.J. (eds) Towards Increased Use of Trypanotolerance: 594 J. McIntire, T. Robinson and C. Bosire Current Research and Future Directions. Proceedings of a Workshop organized jointly by the International Laboratory for Research on Animal Diseases and the International Livestock Centre for Africa held at IL- RAD, Nairobi, Kenya, 26–29 April 1993. ILRAD, Nairobi, and ILCA, Addis Ababa, pp. 115–121. Itty, P., Rowlands, G.J., Minengu, M., Ngamuna, F., van Winckel, F. and d’Ieteren, G.D.M. (1995a) The economics of recently introduced village cattle production in a tsetse affected area (I): trypanotolerant N’Dama cattle in Zaire. Agricultural Systems 47, 347–366. Itty, P., Rowlands, G.J., Morkramer, G., Defly, A. and d’Ieteren, G.D.M. (1995b) The economics of recently introduced village cattle production in a tsetse affected area (II): trypanotolerant cattle in Southern Togo. Agricultural Systems 47, 473–491. Itty, P., Swallow, B.M., Rowlands, G.J., Woudyalew, M. and d’Ieteren, G.D.M. (1995c) The econom- ics of village cattle production in a tsetse-infested area of southwest Ethiopia. Preventive Veterinary Medicine 22, 183–196. Jabbar, M., Larbi, A. and Reynolds, L. (1994) Profitability of alley farming with and without fallow in south- west Nigeria. Experimental Agriculture 30, 319–327. Jabbar, M.A., Larbi, A. and Reynolds, L. (1996) Alley farming for improving small ruminant productivity in West Africa. ILRI’s experiences. Socioeconomic and Policy Research Working Paper No. 20. Live- stock Policy Analysis Programme, ILRI, Addis Ababa. Jahnke, H.E. (1974) The Economics of Controlling Tsetse Flies and Cattle Trypanosomiasis in Africa: Exam- ined for the Case of Uganda. No. 48. Weltforum Verlag, Munich, Germany. Jahnke, H.E. (1976) Tsetse Flies and Livestock Development in East Africa. A Study in ENVIRONMENTAL economics. Weltforum Verlag, Munich, Germany. Jahnke, H.E. (1982) Livestock Production Systems and Livestock Development in Tropical Africa. Wis- senschaftsverlag Vauk Kiel, Kiel, Germany. Jodha, N.S., Asokan, M., and Ryan, J.G. (1977) Village study methodology and resource endowments of the selected villages in ICRISAT’S village level studies. Occasional Paper No. 16. Economics Programme. ICRISAT. Hyderabad, India. Jones, P.G and Thornton, P.K. (2009) Croppers to livestock keepers: livelihood transitions to 2050 in Africa due to climate change. Environmental Science & Policy 12, 427–437. Jutzi, S.C. and Rich, K.M. (2016) An evaluation of CGIAR Centers’ impact assessment work on live- stock-related research (1990–2014). Standing Panel on Impact Assessment (SPIA), CGIAR Inde- pendent Science and Partnership Council (ISPC). Rome. Kang, B.T., Reynolds, L. and Atta-Krah, A.A. (1990) Alley farming. Advances in Agronomy 43, 315–359. Kelley, T.G., Parthasarathy Rao, P. and Walker, T.S. (1993) The relative value of cereal straw fodder in India: implications for cereal breeding programs at ICRISAT. In: Dvorak, K.A. (ed.) Social Science Research for Agricultural Technology Development. CAB International, Wallingford, UK, pp. 88–105. King, J.M. (1983) Livestock water needs in pastoral Africa in relation to climate and forage. ILCA Research Report No. 7. ILCA, Addis Ababa. King, J.M., Sayers, A.R., Peacock, C.P. and Kontrohr, E. (1984) Maasai herd and flock structures in relation to livestock wealth, climate and development. Agricultural Systems 13, 21–56. Konandreas, P.A. and Anderson, F.M. (1982) Cattle herd dynamics: an integer and stochastic model for evaluating production alternatives. ILCA, Addis Ababa. Konandreas, P.A., Anderson, F.M., and Trail, J.C.M. (1983) Economic trade-offs between milk and meat production under various supplementation levels in Botswana. ILCA, Addis Ababa. Konczacki, Z.A. (2014) The Economics of Pastoralism: A Case Study of Sub-Saharan Africa. Routledge. Kristjanson, P., Rowlands, J., Swallow, B., Kruska, R., de Leeuw, P. and Nagda, S. (1999a) Using the economic surplus model to measure potential returns to international livestock research. The case of trypanosomosis vaccine research. ILRI Impact Assessment Series 4. ILRI, Nairobi. Kristjanson, P.M., Swallow, B.M., Rowlands, G.J., Kruska, R.L. and de Leeuw, P.N. (1999b) Measuring the costs of African animal trypanosomiasis, the potential benefits of control and returns to research. Agricultural Systems 59, 79–98. Kruska, R.L., Reid, R.S., Thornton, P.K., Henninger, N. and Kristjanson, P. (2003) Mapping livestock-oriented agricultural production systems for the developing world. Agricultural Systems 77, 39–63. Lambourne, L.J. and Butterworth, M.H. (1983) Biotechnical options. In: Pastoral Systems Research in Sub-Saharan Africa. Proceedings of the IDRC/ILCA Workshop, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 383–401. Le Houérou, H.N. (ed.) (1980) Browse in Africa: The Current State of Knowledge. ILCA, Addis Ababa. Le Houérou, H.N. (1989) The Grazing Land Ecosystems of the African Sahel. Ecological Book Series, Vol. 75. Springer, Berlin. African Livestock Systems Research, 1975–2018 595 Le Houérou, H.N. (2009) Bioclimatology and Biogeography of Africa. Springer, Berlin. Le Houérou, H.N., and Hoste, C.H. (1977) Rangeland production and annual rainfall relations in the Medi- terranean Basin and in the African Sahelo-Sudanian Zone. Journal of Range Management 30, 181–188. Le Moigne, M. (1979). Animal-draft cultivation in Francophone Africa. In: Ryan, J.G. and Thompson, H. (eds) Socioeconomic Constraints to Development of Semi-Arid Tropical Agriculture. ICRISAT, Hyderabad, India. pp. 213–220. Lesorogol, C.K. (2008) Contesting the Commons: Privatising Pastoral Lands in Kenya. University of Mich- igan Press, Ann Arbor, Michigan. Li, X., Qi, G., Tang, L., Zhao, L., Jin, L., Guo, Z. and Wu, J. (2012) Agricultural Development in China and Africa: A Comparative Analysis. Routledge, London/New York. Little, P.D. and Dube, S. (2011) Centre Commissioned External Review (CCER) of Pastoral Systems. ILRI, Nairobi. Lowder, S.K., Skoet, J. and Raney, T. (2016) The number, size, and distribution of farms, smallholder farms, and family farms worldwide. World Development 87, 16–29. Lynam, J. and Byerlee, D. (2017) Forever Pioneers: CIAT: 50 Years Contributing to a sustainable food future… and counting, CIAT, Cali, Colombia. Macharia, P.N., Gachene, C.K.K., Mureithi, J.G., Kinyamario, J.I., Ekaya, W.N. and Thuranira, E.G. (2011) The effect of introduced forage legumes on improvement of natural pastures of semiarid rangelands of Kajiado District, Kenya. Tropical and Subtropical Agrosystems 14, 221–227. Masters, W.A, Djurfeldt, A.A., de Haan, C., Hazell, P. and Jayne, T. et al. (2013) Urbanization and farm size in Asia and Africa: implications for food security and agricultural research. Global Food Security 2, 156–165. Matlon, P., Cantrell, R., King, D. and Benoit-Cattin, M. (1984) Coming Full Circle: Farmers’ Participation in the Development of Technology. IDRC, Ottawa, Canada. Mayaux, P., Bartholomé, E., Fritz, S. and Belward, A. (2004) A new land-cover map of Africa for the year 2000. Journal of Biogeography 31, 861–877. McCown, R.L., Haaland, G. and de Haan, C. (1979) In: Hall A.E., Cannell G.H., Lawton H.W. (eds) Agri- culture in Semi-Arid Environments. Ecological Studies (Analysis and Synthesis), Vol. 34. Springer, Berlin/Heidelberg. McDermott J.J., Staal S.J., Freeman H.A., Herrero M. and van de Steeg J.A. (2010) Sustaining intensifica- tion of smallholder livestock systems in the tropics. Livestock Science 130, 95–109. McIntire, J.M. and Debrah, S.K. (1987) Forage research in smallholder and pastoral production systems. In: Little, D.A. and Said, A.N. (eds) Utilization of Agricultural By-products as Livestock Feeds in Africa. ILCA, Addis Ababa. McIntire, J., Bourzat, D. and Pingali, P. (1992) Crop–Livestock Interaction in Sub-Saharan Africa. World Bank, Washington, D.C. Mearns, R. (1996) Community, collective action and common grazing: the case of post-socialist Mongolia. Journal of Development Studies 32, 297–339. Mengistu, U., Dahlborn, K. and Olsson, K. (2007) Mechanisms of water economy in lacating Ethiopian So- mali goats during repeated cycles of intermittent watering. Animal 1, 1009–1017. Mesele, S., Gebrekidan, H., Gizachew, L. and Coppock, D.L. (2006) Changes in Land Cover and Soil Con- ditions for the Yabelo District of the Borana Plateau, 1973–2003. Research Brief No. 6. Utah State University, Logan, Utah. Miehe, S., Kluge, J., von Wehrden, H. and Retzer, V. (2010) Long-term degradation of Sahelian rangeland detected by 27 years of field study in Senegal. Journal of Applied Ecology 47, 692–700. Mohamed-Saleem, M.A. (1986a) Integration of forage legumes into the cropping systems of Nigeria’s subhumid zone. In: von Kaufmann, R., Chater, S. and Blench, R. (eds) Livestock Systems Research in Nigeria’s Subhumid Zone. ILCA/NAPRI Symposium. ILCA, Addis Ababa, Paper 15. Mohamed-Saleem, M.A. (1986b) The establishment and management of fodder banks. In: von Kaufmann, R., Chater, S. and Blench, R. (eds) Livestock Systems Research in Nigeria’s Subhumid Zone. ILCA/ NAPRI Symposium. ILCA, Addis Ababa, Paper 16. Monod, T. (ed.) (1975) Les Societies Pastorales en Afrique Tropicale. Oxford University Press, London. Montgolfier-Kouèvi, C. and Le Houérou, H.N. (1980) Study on the economic viability of browse plantations in Africa. In: Le Houérou, H.N. (ed.) Browse in Africa: The Current State of Knowledge. ILCA, Addis Ababa, pp. 449–464. Morisette, J.T., Privette, J.L. and Justice, C.O. (2002) A framework for the validation of MODIS land prod- ucts. Remote Sensing of Environment 83, 77–96. 596 J. McIntire, T. Robinson and C. Bosire Morris, M., Binswanger-Mkhize, H.P. and Byerlee, D. (2009) Awakening Africa’s sleeping giant: pro- spects for commercial agriculture in the Guinea Savannah Zone and beyond. World Bank, London. Mortimore, M. (2000) Hard questions for pastoral development: a northern Nigerian perspective. In: Tielke, E., Schlecht, E. and Hiernaux, P. (eds) Elevage et Gestion de Parcours au Sahel, Implications Pour le développement. Verlag Grauer, Stuttgart, Germany, pp. 101–114. Mortimore, M. and Turner, B. (2005) Does the Sahelian smallholder’s management of woodland, farm trees, rangeland support the hypothesis of human-induced desertification? Journal of Arid Environ- ments 63, 567–595. Mulatu, A. and Bekure, S. (2013) The need to strengthen land laws in Ethiopia to protect pastoral rights. In: Catley, A., Lind, J. and Scoones, I. (eds) Pastoralism and Development in Africa: Dynamic Change at the Margins. Routledge, London, pp. 186–194. Murray, M., Trail, J.C. and d’Ieteren, G.D. (1990) Trypanotolerance in cattle and prospects for the control of trypanosomiasis by selective breeding. Revue Scientifique et Technique 9, 369–386. Nelson, G.C., Rosegrant, M.W., Koo, J., Robertson, R., Sulser, T.Z., et al. (2009) Food Security, Farming and Climate Change to 2050: Scenarios, Results and Policy Options. IFPRI, Washington, D.C. Nicholson, M.J. (1984) Pastoralism and milk production.ILCA Bulletin No. 20. ILCA, Addis Ababa. Nicholson, M.J. (1987a) Effects of night enclosure and extensive walking on the productivity of Zebu cattle. Journal of Agricultural Science 109, 445–452. Nicholson, M.J. (1987b) The effect of drinking frequency on some aspects of the productivity of Zebu cat- tle. Journal of Agricultural Science 108, 119–128. Nicholson, M.J. (1989) Depression of dry-matter and water intake in Boran cattle owing to physiological, volumetric and temporal limitations. Animal Production 49, 29–34. Nicholson, M.J. and Mengistu, S. (2016) ILRI Impact Book Background Paper. Ethiopia Fodder Adoption Study. ILRI, Nairobi. Nordblom, T.L., Christiansen, S., Nersoyan, N. and Bahhady, F. (1992) A whole-farm model for economic analysis of medic pasture and other dryland crops in two-year rotations with wheat in northwest Syria. ICARDA, Aleppo, Syria. Nordblom, T.L., Pannell, D.J., Christiansen, S., Nersoyan, N. and Bahhady, F. (1994) From weed to wealth? Prospects for medic pastures in the Mediterranean farming system of north-west Syria. Agri- cultural Economics 11, 29–42. Norman, D. and Collinson, M. (1985) Farming systems research in theory and practice. In: Remenyi, J.V. (ed.) Agricultural Systems Research for Developing Countries. ACAIR Proceedings No. 11, Hawkes- bury, Australia, pp. 16–30. Norman, D., Pryor, D. and Gibbs, C.J.N. (1979) Technical change and the small farmer in Hausaland, northern Nigeria. African Rural Economy Paper No. 21. African Rural Economy Program, Department of Agri- cultural Economics, Michigan State University, East Lansing, Michigan. Norman, D.W., Sigwele, H. and Baker, D. (1988) Reflections on two decades of research on sorghum-based farming systems in Northern Nigeria and Botswana. In: Rukuni, M. and Bernsten, R.H. (eds.) South- ern Africa: Food Security Policy Options. Proceedings of the Third Annual Conference on Food Se- curity Research in Southern Africa, 1–5 November 1987. University of Zimbabwe/Michigan State University Food Security Research Project, Harare, pp. 235–257. Norton, B.E. (1975) Modelling primary production. In: Evaluation and Mapping of Tropical African Rangelands. Proceedings of the seminar, 3–8 March, Bamako, Mali. ILCA, Addis Ababa, pp. 313–322. Ocaido, M., Muwazi, R. and Opuda-Asibo, J. (2009) Financial analysis of livestock production systems around Lake Mburo National Park, in southwestern Uganda. Livestock Research for Rural Develop- ment 21, 70. Otte, M.J. and Chilonda, P. (2002) Cattle and small ruminant production systems in sub-Saharan Africa: a systemic review. FAO, Rome. Oxby, C. (1975) Pastoral Nomads and Development. International African Institute, London. Penning de Vries, F.W.T. and Heemst, H.D. (1975) Potential primary production of unirrigated land in the Sahel. In: Evaluation and Mapping of Tropical African Rangelands. Proceedings of the Seminar: Bamako, Mali, 1975. ILCA. Bamako, Mali, pp. 323–328. Penning de Vries, F.W.T. and Djiteye, M.A. (1991) La Productivité des Pâturages Sahéliens: Une Étude des Sols, des Végétations et de l’Exploitation de cette Ressource Naturelle, 2nd edn. Centre for Agricultural Publishing and Documentation, Waaginengen, Netherlands. Perkins, D.H. (2013) Agricultural Development in China 1368–1968. Aldine, New Brunswick/London. African Livestock Systems Research, 1975–2018 597 Perry, B.D., Randolph, T.F., McDermott, J.J., Sones, K.R. and Thornton, P.K. (2002) Investing in animal health research to alleviate poverty. ILRI, Nairobi. Peyre de Fabrѐgues, B. (1984) Quel avenir pour l’élevage au Sahel? Revue d’Elevage et de Medecine Vétérinaire des Pays tropicaux 37, 500–508. Pingali, P., Bigot, Y. and Binswanger, H.P. (1987) Agricultural Mechanization and the Evolution of Farm- ing Systems in Sub-Saharan Africa. Johns Hopkins University Press, Baltimore/London. Place, F., Roothaert, R., Maina, L., Franzel, S., Sinja, J. and Wanjiku, J. (2009) The impact of fodder trees on milk production and income among smallholder dairy farmers in East Africa and the role of research. ICRAF Occasional Paper No. 12. World Agroforestry Centre, Nairobi. Powell, J.M. (1986a) Cropping systems in the subhumid zone of Nigeria. In: von Kaufmann, R., Chater, S. and Blench, R. (eds) Livestock Systems Research in Nigeria’s Subhumid Zone, 29 October–2 No- vember 1984, Kaduna, Nigeria. ILCA, Addis Ababa, Paper 13. Powell, J.M. (1986b) Crop–livestock interactions in the subhumid zone of Nigeria. In: von Kaufmann, R., Chater, S. and Blench, R. (eds) Livestock Systems Research in Nigeria’s Subhumid Zone, 29 October–2 Novem- ber 1984, Kaduna, Nigeria. ILCA, Addis Ababa, Paper 14. Powell, J.M. and Taylor-Powell, E. (1984) Cropping by Fulani agro-pastoralists in central Nigeria. ILCA Bulletin 19, 21–27. Powell, J.M., Fernández-Rivera, S., Williams, T.O. and Renard, C. (eds) (1995) Livestock and Sustainable Nutrient Cycling in Mixed Farming Systems of sub-Saharan Africa, Vol. II: Technical Papers. Proceed- ings of an International Conference 22–26 November 1993, Addis Ababa, Ethiopia. ILCA, Addis Ababa. Powell, J.M., Fernandez-Rivera, S., Hiernaux, P. and Turner, M.D. (1996) Nutrient cycling in integrated rangeland/cropland systems of the Sahel. Agricultural Systems 52, 143–170. Pratt, D.J. and Gwynne, M.D. (1977) Rangeland Ecology and Management in East Africa. Robert E. Krieger, Huntington, New York. Pratt, D.J., Le Gall, F. and de Haan, C. (1997) Investing in pastoralism: sustainable natural resource use in arid Africa and the Middle East. World Bank Technical Paper No. 35. World Bank, Washington, D.C. Reed, J.D., Capper, B. and Neate, P.J.H. (1988) Plant Breeding and the Nutritive Value of Crop Residues. Proceedings of a Workshop, 7–10 December 1987, ILCA, Addis Ababa. ILCA, Addis Ababa. Reid, R.S., Kruska, R.L., Muthui, N., Taye, A., Wotton, S., et al. (2000a) Land-use and land-cover dynam- ics in response to changes in climatic, biological and socio-political forces: the case of southwestern Ethiopia. Landscape Ecology 15, 339–355. Reid, R.S., Kruska, R.L., Deichmann, U., Thornton, P.K. and Leak, S.G.A. (2000b) Human population growth and the extinction of the tsetse fly. Agriculture, Ecosystems & Environment 77, 227–236. Reid, R.S., Serneels, S., Nyabenge, M. and Hanson, J. (2005) The changing face of pastoral systems in grass-dominated ecosystems of eastern Africa. In: Suttie, J.M., Reynolds, J.G. and Batello, C (eds) Grasslands of the World. FAO, Rome, pp. 19–76. Reid, R., Galvin, K. and Kruska, R. (2008a) Global significance of extensive grazing lands and pastoral societies: an introduction. In: Galvin, K.A., Reid, R.S. Behnke, R.H. Jr and Thompson Hobbs, N. (eds) Fragmentation in Semi-arid and Arid Landscapes. Springer, New York, pp. 1–25. Reid, R.S., Gichohi, H., Said, M.Y., Nkedianye, D., Ogutu, J.O., et al. (2008b) Fragmentation of a peri- urban savanna, Athi-Kaputiei plains, Kenya. In: Galvin, K.A., Reid, R., Behnke, R.H Jr and Hobbs, N.T. (eds) Fragmentation in Semi-arid and Arid Landscapes. Springer, New York, pp. 195–224. Renard, C. (ed.) (1997) Crop Residues in Sustainable Mixed Crop/livestock Farming Systems. CAB Inter- national, Wallingford, UK. Requier-Desjardins, M. (2011) Elevages et transhumances à l’extrême-nord du Cameroun, enquêtes en milieu pastoral et essai de modélisation contractuelle. PhD thesis, Versailles Saint-Quentin-en-Yve- lines University, Versailles, France. Retzer, V. (2006) Impacts of grazing and rainfall variability on the dynamics of a Sahelian rangeland revisited. Journal of Arid Environments 67, 157–164. Reynolds, L. and Jabbar, M.A. (1994) The role of alley farming in African livestock production. Outlook on Agriculture 23, 105–113. Rich, K.M., and Perry, B.D. (2011) The economic and poverty impacts of animal diseases in developing countries: new roles, new demands for economics and epidemiology. Preventive Veterinary Medicine 101, 133–147. Rivas Rios, L. (1974) Some aspects of the cattle industry on the North Coast Plains of Colombia. Technical Bulletin No. 3. CIAT, Cali, Colombia. Robinson, T., Thornton, P., Franceschini, G., Kruska, R., Chiozza, F., et al. (2011) Global Livestock Production Systems. FAO, Rome and ILRI, Nairobi. 598 J. McIntire, T. Robinson and C. Bosire Robinson, T.P., Wint, G.R.W., Conchedda, G., van Boeckel, T.P., Ercoli, V., et al. (2014) Mapping the global distribution of livestock. PLoS One 9, e96084. Roeder, P. and Rich, K.M. (2009) Conquering the cattle plague: the global effort to eradicate rinderpest. In: Spielman, D.J. and Pandya-Lorch, R. (eds) Millions Fed: Proven Successes in Agricultural Develop- ment. IFPRI, Washington, D.C., pp. 109–116. Rowlands, G.J. and Teale, A.J. (eds) (1994) Towards Increased Use of Trypanotolerance: Current Re- search and Future Directions: Proceedings of a Workshop. ILRI, Addis Ababa. Ruthenberg, H. (1980) Farming Systems in the Tropics, 3rd edn. Clarendon Press, Oxford. Rutherford, A.S. (2001) Broad bed maker technology package innovations in Ethiopian farming sys- tems. An ex post impact assessment. ILRI Impact Assessment Series 7. ILRI, Nairobi. Rutherford, A.S. (2008) Broad bed maker technology package innovations in Ethiopian farming sys- tems: An ex post impact assessment. ILRI Research Report 20. ILRI, Nairobi. Rutten, M.M.E.M. (1992) Selling Wealth to Buy Poverty: The Process of the Individualization of Landowner- ship Among the Maasai Pastoralists of Kajiado District, Kenya, 1890–1990. Breitenbach, Saarbrücken/ Fort Lauderdale. Ryan, J.G. and Thompson, H. (eds) (1980). Socioeconomic Constraints to Development of Semi-Arid Tropical Agriculture. ICRISAT, Hyderabad, India. Sanchez, P.A. (1995) Science in agroforestry. Agroforestry Systems 30, 5–55. Sanders, J.H., Shapiro, B.I. and Ramaswamy, S. (1996) The Economics of Agricultural Technology in Sem- iarid Sub-Saharan Africa. Johns Hopkins University Press, Baltimore, M.D. Sandford, S. (1983a) Management of Pastoral Development in the Third World. Wiley, Chichester, UK. Sandford, S. (1983b) Organization and management of water supplies in tropical Africa. ILCA Research Report No. 8. ILCA, Addis Ababa. Scoones, I. and Wolmer, W. (2002) Pathways of Change in Africa: Crops, Livestock and Livelihoods in Mali, Ethiopia and Zimbabwe. James Currey, Oxford. See, L., Schepaschenko, D., Lesiv, M., McCallum, I., Fritz, S., et al. (2015) Building a hybrid land cover map with crowdsourcing and geographically weighted regression. ISPRS Journal of Photogrammetry and Remote Sensing 103, 48–56. Seré, C., and Steinfeld, H. (1996) World Livestock Production Systems: Current Status, Issues and Trends. FAO, Rome. Shaw, A.P. (2004) The economics of African trypanosomiasis. In: Maudlin, I., Holmes, P.H. and Miles, M.A. (eds) The Trypanosomiases. CAB International, Wallingford, UK, pp. 369–402. Shaw, A.P.M., Wint, G.R.W., Cecchi, G., Torr, S.J., Mattioli, R.C. and Robinson, T.P. (2015) Mapping the benefit–cost ratios of interventions against bovine trypanosomosis in Eastern Africa. Preventive Veter- inary Medicine 122, 406–416. Singh, I. (1990) The Great Ascent: The Rural Poor in South Asia. Johns Hopkins University Press, Balti- more, Maryland. Smith, A.B. (1992) Pastoralism in Africa: Origins and Development Ecology. Ohio University Press, Athens, Ohio. Smith, L.D. (1976) An overview of agricultural development policy. In: Heyer, J., Maitha, J.K. and Senga, W.M. (eds) Agricultural Development in Kenya: An Economic Assessment. Oxford University Press. Nairobi, pp. 111–151. Sneath, D. (1998) State policy and pasture degradation in Inner Asia. Science 281, 1147–1148. Solomon Bekure, S., de Leeuw, P.N., Grandin, B.E., and Neate, P.J.H. (1991) Maasai Herding: An Ana- lysis of the Livestock Production System of Maasai Pastoralists in Eastern Kajiado District, Kenya. ILCA Systems Study 4. ILCA, Addis Ababa. Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M. and de Haan, C. (2006) Livestock’s Long Shadow: Environmental Issues and Options, FAO, Rome. Stenning, D.J. (1994) Savannah Nomads: A Study of the Wodaabe Pastoral Fulani of Western Bornu Pro- vince Northern Region, Nigeria. LIT Verlag, Münster, Germany. Sumberg, J. (2002) The logic of fodder legumes in Africa. Food Policy 27, 285–300. Sumberg, J., McIntire, J., Okali, C. and Atta-Krah, A.N. (1987) Economic analysis of alley farming with small ruminants. ILCA Bulletin 28, 2–6. Suttie, J.M., Reynolds, J. G., and Batello, C (eds). (2005). Grasslands of the World. FAO, Rome. Swallow, B.M. (2000) Impacts of trypanosomiasis on African agriculture. PAAT Technical Scientific Series. FAO, Rome. African Livestock Systems Research, 1975–2018 599 Thirtle, C., Townsend, R.F., Amadi, J., Lusigi, A. and van Zyl, J. (1998) The rate of return on expenditures of the South African Agricultural Research Council (ARC). Agricultural Economics Research, Policy and Practice in Southern Africa 37, 612–622. Thomas, R.J. and Lascano, C.E. (1995) The benefits of forage legumes for livestock production and nutrient cycling in pasture and agro-pastoral systems of acid-soil savannahs in Latin America. In: Powell, J.M., Fernández-Rivera, S., Williams, T.O. and Renard, C. (eds) Cycling in Mixed Farming Systems of sub-Saharan Africa, Vol. II: Technical Papers. Proceedings of an International Conference 22–26 No- vember 1993, Addis Ababa, Ethiopia. ILCA, Addis Ababa, pp. 272–286. Thornton, P.K. (1987) Experimentation with a beef production simulation model for the savannas of Colom- bia. Cattle Production Systems Simulation Project Document No. 4. CIAT, Cali, Columbia. Thornton, P.K. and Herrero, M. (2001) Integrated crop–livestock simulation models for scenario analysis and impact assessment. Agricultural Systems 70, 581–602. Thornton, P.K. and Herrero, M. (2010) The potential for reduced methane and carbon dioxide emissions from livestock and pasture management in the tropics. Proceedings of the National Academy of Sci- ences USA 107, 19667–19672. Thornton, P.K., Kruska, R.L., Henninger, N., Kristjanson, P.M., Reid, R.S., et al. (2002) Mapping Poverty and Livestock in the Developing World. ILRI, Nairobi. Thornton, P.K., Kristjanson, P.M. and Thorne, P.J. (2003) Measuring the potential impacts of improved food- feed crops: methods for ex ante assessment. Field Crops Research 84, 199–212. Thornton, P.K., BurnSilver, S.B., Boone, R.B. and Galvin, K.A. (2006) Modelling the impacts of group ranch subdivision on agro-pastoral households in Kajiado, Kenya. Agricultural Systems 87, 331–356. Thornton, P.K., van de Steeg J., Notenbaert A. and Herrero, M. (2009) The impacts of climate change on livestock and livestock systems in developing countries: a review of what we know and what we need to know. Agricultural Systems 101, 113–127. Thurston, A. (1987) Smallholder Agriculture in Colonial Kenya: The Official Mind and the Swynnerton Plan. Cambridge African Monographs 8. African Studies Centre, Cambridge. Tiffen, M., Mortimore, M. and Gichuki, F. (1994) More People, Less Erosion: Environmental Recovery in Kenya. John Wiley & Sons, Chichester, UK. Toulmin, C. and Fulton, D. (1983) Les sous-systèmes du mil. In: Wilson, R.T., de Leeuw, P.N. and de Haan, C. (eds) Recherches sur les Systèmes Arides: Résultats Préliminaires. Rapport de Recherche No. 5. ILCA. Addis Ababa. Toutain, G.G. and Boudet, B. (1980) Integration des fourrages ligneux dans les systèmes pastoraux et agropastoraux en Afrique. In: Le Houérou, H.N. and Montgolfier, A. (eds) Browse in Africa. ILCA, Addis Ababa, p. 415–420. Townsend, R.F. (1999) Agriculture incentives in sub-Saharan Africa. World Bank Technical Paper No. 444. World Bank, Washington, D.C. Trail, J.C.M., Hoste, C.H., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979a) Trypanotolerant Livestock in West and Central Africa, Vol. 1: General Study. FAO Animal Production and Health Paper 20/1. FAO, Rome. Trail, J.C.M., Hoste, C.H., Wissocq, Y.J., Lhoste, P. and Mason, I.L. (1979b) Trypanotolerant Livestock in West and Central Africa: Vol. 2: Country Studies. FAO Animal Production and Health Paper 20/2. FAO, Rome. Tribe, D., Nestel, B., Pratt, D.J. and Thome, M. (1973) Animal production and research in tropical Africa. Report of the Task Force commissioned by the African Livestock Sub-Committee of the Consultative Group on International Agricultural Research. ILCA Accession number 00129. CGIAR, Washington, D.C. Turner, M.D., McPeak, J.G. and Ayantunde, A. (2014) The role of livestock mobility in the livelihood strat- egies of rural peoples in semi-arid West Africa. Human Ecology 42, 231–247. UNESCO (1963) Nomades et nomadisme au Sahara. UNESCO, Paris. van Keulen, H., Seligman, N.G and Benjamin, R.W. (1981) Simulation of water use and herbage growth in arid regions: a re-evaluation and further development of the model ‘arid crop.’ Agricultural Systems 6, 159–193. van Wijk, M.T., Tittonell, P., Rufino, M.C., Herrero, M., Pacini, C., et al. (2009) Identifying key entry-points for strategic management of smallholder farming systems in sub-Saharan Africa using the dynamic farm- scale simulation model NUANCES-FARMSIM. Agricultural Systems 102, 89–101. Vetter, S. (ed.) (2004) Rangelands at equilibrium and non-equilibrium: recent developments in the debate around rangeland ecology and management. Papers from the international workshop on ‘Rangelands 600 J. McIntire, T. Robinson and C. Bosire at equilibrium and non-equilibrium’ held at the Viith International Rangelands Congress in Durban, 26–27 July 2003. von Kaufmann, R. (1976) The development of the range land areas. In: Heyer, J., Maitha, J.K. and Senga, W. M (eds) Agricultural Development in Kenya: An Economic Assessment. Oxford University Press, Nairobi, pp. 255–288. von Kaufmann, R., Chater, S. and Blench, R. (eds) (1986) Livestock Systems Research in Nigeria’s Sub- humid Zone. ILCA/NAPRI Symposium. ILCA, Addis Ababa. Wagenaar, K.T., Diallo, A.K. and Sayers, A.R. (1986) Productivity of transhumant Fulani cattle in the inner Niger delta of Mali. ILCA Research Report No. 13. ILCA, Addis Ababa. Walker, B.H. (1980) A review of browse and its role in livestock production in southern Africa. In: Le Houérou, H.N. (ed.) Browse in Africa: The Current State of Knowledge. ILCA, Addis Ababa, pp. 7–24. Walker, T.S. and Alwang, J. (eds) (2015) Crop Improvement, Adoption and Impact of Improved Varieties in Food Crops in sub-Saharan Africa. CAB International, Wallingford, UK. Walker, T.S. and Ryan, J.G. (1990) Village and Household Economics in India’s Semi-arid Tropics. Johns Hopkins University Press, Baltimore, Maryland. Wanyoike, F. and Baker, D. (2013) Pro-poor development performance of livestock projects: analysis and lessons from projects’ documentation. Development in Practice 23, 889–907. Waters-Bayer, A. and Taylor-Powell, E. (1986a) Population and land use in the subhumid zone of Nigeria. In: von Kaufmann, R., Chater, S. and Blench, R. (eds) Livestock Systems Research in Nigeria’s Sub- humid Zone, 29 October–2 November 1984, Kaduna, Nigeria. ILCA, Addis Ababa, pp. 37–58. Waters-Bayer, A., and Taylor-Powell, E. (1986b) Settlement and land use by Fulani pastoralists in case study areas. In: von Kaufmann, R., Chater, S. and Blench, R. (eds) Livestock Systems Research in Nigeria’s Subhumid Zone, 29 October–2 November 1984, Kaduna, Nigeria. ILCA, Addis Ababa, pp. 213–226. Wilhelm, W.W., Johnson, J.M.F., Hatfield, J.L., Voorhees, W.B. and Linden, D.R. (2004) Crop and soil prod- uctivity response to corn residue removal: a literature review. Agronomy Journal 96, 1–17. Williams, T.O., Fernandez-Rivera, S. and Hiernaux, P. (1999) Crop–livestock systems in sub-Saharan Africa: determinants and intensification pathways. In: McCarthy, N., Swallow, B., Kirk, M. and Hazell, P. (eds) Property Rights, Risk and Livestock Development in sub-Saharan Africa. IFPRI, Washington, D.C., and ILRI, Nairobi. Wilson, R.T. (1986) Livestock production in central Mali: long-term studies on cattle and small ruminants in the agro-pastoral system. ILCA Research report No. 14. ILCA, Addis Ababa. Wilson, R.T. (2007) Perceptions, practices, principles and policies in provision of livestock water in Africa. Agricultural Water Management 90, 1–12. Wilson, R.T. and Maki, M.O. (1989) Goat and sheep population changes on a Masai group ranch in South-Western Kenya, 1978–1986. Agricultural Systems 29, 325–337. Wilson, R.T., de Leeuw P.N. and de Haan, C. (1983) Recherches sur les Systèmes Arides: Résultats Prélim- inaires. Rapport de Recherche No. 5. ILCA. Addis Ababa. Winrock International (1992) Assessment of Animal Agriculture in Sub-Saharan Africa, Morrillton, Arkan- sas. Winrock International Institute for Agricultural Development, Morrilton, Arkansas. Zilberman, D. and Heiman, A. (2004) The value of economics research. In: Pardey, P.G. and Smith, V.H. (eds) What’s Economics Worth? Valuing Policy Research. IFPRI, Washington, D.C., pp. 275–299. 16 Ruminant Livestock and Climate Change in the Tropics Polly Ericksen1, Philip K. Thornton2 and Gerald C. Nelson3 1International Livestock Research Institute, Nairobi, Kenya; 2CGIAR Research Programme on Climate Change, Agriculture and Food Security (CCAFS) and International Livestock Research Institute, Nairobi, Kenya, University of Edinburgh, UK; 3University of Illinois, Urbana-Champaign, Illinois, USA Contents Executive Summary 602 The problem 602 ILRI research 602 Research spending and bibliometrics 603 Scientific impact 603 Models 603 Data sets 603 Scientific results 603 Development impact 603 Capacity development and partnerships 604 The future 604 Livestock productivity 604 Mitigation and supply- and demand-side efforts 604 Introduction 604 Livestock Production Systems and Resource Use 605 Land 606 Water quantity 607 Water quality 607 Air 607 Ecosystem services 607 Climate Change Impacts on Ruminant Livestock 608 CGIAR research on climate change impacts in livestock systems 610 Weather data in climate analyses 610 Impacts on livestock systems 611 Current knowledge gaps on impacts 613 Adaptation of Livestock Systems to Climate Change 617 Costs of adaptation 617 CGIAR research on climate change adaptation in livestock systems 617 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 601 602 P. Ericksen, P. Thornton and G. Nelson Knowledge gaps on adaptation 618 Adaptation in mixed crop–livestock systems 619 Adaptation in pastoral systems 619 Mitigation of Greenhouse Gas Emissions from Livestock 620 Estimates of emissions from livestock 620 Mitigation via supply- and demand-side options 620 Supply-side options 621 Mitigation research in livestock systems 623 Demand-side options 628 The Future 630 References 631 Executive Summary therefore GHG emissions but mechanisms to gen- erate widespread change are not clear. The problem Adaptation to climate change will become more challenging with growing GHG concentra- The temperature and humidity changes associ- tions. At some point in this century, and in some ated with climate change will have direct and regions, temperature and humidity increases will for the most part adverse effects on tropical ani- make production biologically impossible. Well mal productivity. Related changes in pasture before that point, the adaptation costs are likely and feed productivity will have further indirect to outweigh economic benefits of producing live- adverse effects on productivity. Collectively, stock in many current producing regions. these effects will become increasingly negative as climate change progresses, reducing in- comes to both specialized and mixed livestock ILRI research producers and possibly reducing the incomes of consumers. The work of the International Livestock Research At the same time, livestock production ac- Institute (ILRI) on livestock and climate change tivities contribute to climate change. In the early began with studies of livestock water use 2000s, livestock production accounted for 18% (King, 1983; Sandford, 1983), agroecology, and of global greenhouse gas (GHG) emissions, with drought (Henricksen and Durkin, 1986) before enteric emissions about 25% of the total, emis- evolving into crop growing period models. sions from manure a further 24% and conver- Deeper global research efforts were stimulated sion of forests to pasture another 34%. Livestock by the publication of Livestock’s Long Shadow: En- also have negative environmental effects on vironmental Issues and Options by Steinfeld et al. water availability and quality, biodiversity and (2006), which was the first book to comprehen- other ecosystem services. sively address the environmental costs of live- A recent study showed per capita consump- stock. Steinfeld et  al. (2006) found that, while tion of animal products is likely to increase by livestock contributed significant shares of na- about 50% for low-income countries and about tional income, employment and protein supply, 10% for higher-income countries between 2010 it also had adverse effects on water and air qual- and 2050. This demand growth implies rising ity, contributed to deforestation and to loss of livestock GHG emissions unless cost-effective other ecosystem services, and generated 18% of mitigation options can be found. Options to re- anthropogenic GHG emissions. duce emissions include both supply and de- A second major impetus for ILRI research mand-side changes. On the supply side, technical was the Fifth Assessment Report (AR5) of the mitigation options can reduce emissions per unit Intergovernmental Panel on Climate Change of output substantially, but their economic feasi- (IPCC), published in 2014 (IPCC, 2014). AR5 bility varies by location and is generally under- was the first IPCC assessment to evaluate studied. On the demand side, changes in dietary climate change and livestock interactions. ILRI patterns can reduce meat consumption and researchers played important roles in this Ruminant Livestock and Climate Change in the Tropics 603 evaluation and extended their contributions in literature. Scientists in these activities have pub- subsequent work. lished extensively in prestigious scientific jour- This chapter provides an overview of both nals and their papers are widely cited. the scientific and the development impacts of ILRI research on climate change. Scientific Models i mpact is measured by advances in research methods and in research output, such as publi- MarkSim is a stochastic weather generator de- cations that advance our understanding of cli- veloped at Centro Internacional de Agricultura mate change and options to manage it. Tropical (Jones and Thornton, 2000) in the Development impact is about making a direct 1990s with ILRI input. It is used to downscale and positive contribution to welfare, directly or climate outputs from global climate models tem- indirectly. In the case of climate change, devel- porally to daily weather data and spatially from opment impact activities would include adop- large grid sizes of 2° latitude/longitude or more tion of adaptation and mitigation methods. down to a few kilometres. Scientific impact is relatively straightforward to The RUMINANT model, initially developed measure with bibliometric approaches. Measur- in the mid-1990s, was used to predict feed in- ing development impact is much more challen- take, nutrient supply and methane (CH 4) emis- ging because it often develops through long sions. These numbers are then aggregated to chains of causality; for example, an adviser to a systems, countries, regions and continents policy maker in a country reads a key academic using animal population projections, allowing reference that draws on a data set on coun- refinement of GHG emissions estimates related try-specific livestock systems that was generated to animal production. using a simulation model. Hence, assessments of development impact tend to be somewhat an- Data sets ecdotal in the absence of studies of adoption of new methods generated by research. Data sets produced by MarkSim and RUMINANT have been used by a wide range of researchers as well as by the model developers themselves. Research spending and bibliometrics Scientific results It was not possible to separate ILRI spending on Scientific results include the following: climate change from other spending in detail, but • Initial and periodic revisions of estimates of we do know that the climate research is the prod- area, production, livestock numbers and uct of a small number of scientists and hence the feed sources by systems in the tropics. budget share would be small in relation to the • Impacts of climate change on livestock prod- total. The productivity of climate change research, uctivity and production from changing tem- as shown in the Altmetric (www.altmetric.com/; perature and humidity, growing period accessed 7 March 2020) and bibliometric ana- shifts, and pest and disease distributions. lyses, is quite high and suggests that the field is • Identification of adaptation options. seriously underfunded when considering the im- • Estimates of GHG per animal and per system. portance of the problem, the scale of other inter- • Estimates of mitigation possibilities (e.g. national efforts and the productivity to date of percentage changes below the trend of ILRI research in this area. different climate scenarios). • Feed quality and its GHG impact in the tropics. Scientific impact Development impact The key scientific impacts of ILRI research on climate change arose from the development of The two main development impacts have been: (i) two models – MarkSim and RUMINANT – and appropriate animal selection, breeding and man- the use of these models to generate a range of agement techniques to reduce GHG per unit of data sets that are now widely used in the scientific output in the OECD countries; and (ii) modelling 604 P. Ericksen, P. Thornton and G. Nelson of supply and demand management options, Livestock productivity even if not yet broadly applied, which may have policy effects on GHG mitigation and ultimately The following questions need to be addressed: on global warming. • What are the productivity impacts in trop- ical livestock given temperature and hu- midity levels in producing regions under a Capacity development and partnerships range of climate scenarios? • How will climate change effects on livestock The principal capacity development impact of pests and diseases spill over into effects on the modelling work originating at ILRI has been livestock productivity? a wide range of partners who now use the • What types of livestock systems are most models, or the data sets generated from them. resilient to changes in both mean changes Some of these partners have been involved in and variability of temperature and humid- model development, validation and application; ity? One system in particular, confined ani- others have been trained to use the models for mal feeding operations (CAFOs), is likely to their own research needs. The creation of a web- grow rapidly in the tropics. How vulnerable site that generates MarkSim results for any arbi- are CAFOs to climate change? trary location extends the reach of the models • How cost-effective are existing adaptation dramatically. options? The second has been the development of • What are the biological limits to adapta- data and models under tropical conditions and tion? Are they likely to be reached in im- their applications. An example is the Mazingira portant producing areas? Centre at ILRI, which develops the capacity of • What are the potential effects of climate national and regional scientists to study inter- change on the use of livestock as a risk- actions between livestock and climate. management asset? • Will climate change alleviate or exacerbate livestock’s negative effects on water quality and quantity and on ecosystem services? The future Mitigation and supply- and The future for ruminant livestock is more cer- demand-side efforts tain on the demand side because of expected ris- ing incomes in developing countries and the Outstanding questions in this area include the high income elasticity of demand for animal following: products. It is projected that demand for all ani- • Are there technical and cost-effective op- mal products will grow globally, although there tions for reducing GHG emissions from will be composition effects as demand shifts existing livestock systems? among animal types and as competition from • What kind of changes to existing systems plant sources of protein grows. would achieve cost-effective mitigation? The future on the supply side is uncertain in • What policy activities could contribute to part because of the interactions between climate adoption of mitigation technologies? change and animal agriculture. Information on • What demand-side actions would be needed mitigation of, and adaptation to, climate change to have a substantial reduction in emissions is inadequate in the tropics compared with what and in what regions of the world? is known about the temperate zone. The re- search agenda stated here will require more de- tailed information on existing systems, on the Introduction potential for technical changes that contribute to adaptation and mitigation, on modelling such This chapter explores our understanding of the changes as new GCM outputs become available, evolving interactions between climate change and on policy changes that have the potential for and ruminant livestock in the tropics. It analyses significant adaptation and mitigation. the research done by ILRI and its partners in Ruminant Livestock and Climate Change in the Tropics 605 improving this understanding and in contribut- opportunities. The research undertaken by ILRI ing to solutions for mitigation of GHG emissions after Livestock’s Long Shadow became a major and adaptation to climate change. The focus is source of research outputs used in AR5. ILRI re- mostly on ruminants and, within this category, searchers were invited to participate in the IPCC on cattle. The chapter first reviews the scientific GHG emissions taskforce in 2009 on improving and development impacts of ILRI and partner GHG livestock emissions estimates. This work led research before suggesting research priorities on to collaboration with the International Institute climate change and tropical animal production. for Applied Systems Analysis (IIASA) and its ILRI’s predecessors did little on the global GLOBIOM model, a multi-market model with 30 environmental costs of tropical animal produc- regions covering the globe and coverage of some tion. ILRI’s research on livestock–climate change 18 or 27 commodities. This collaboration al- interactions began with the growing period mod- lowed a better disaggregation of livestock num- elling of Jones and Thornton (2000, 2003) and bers and feed sources by system, especially in the the studies of McDermott et al. (2001), Jones et al. tropics. A similar arrangement exists with the (2002) and Thornton et  al. (2002). Climate IMPACT multi-market model of IFPRI, with 158 change research at ILRI was stimulated by the regions and 60 commodities. publication of the Food and Agriculture Organiza- After a brief overview of livestock systems tion of the United Nations (FAO) book Livestock’s and their resource use, this chapter addresses Long Shadow: Environmental Issues and Options three areas: (i) climate change impacts on ru- (Steinfeld et al., 2006), which sought to ‘…assess minant livestock; (ii) adaptation of livestock sys- the full impact of the livestock sector on environ- tems to climate change; and (iii) options to mental problems, along with potential technical reduce GHG emissions. and policy approaches to mitigation’. Steinfeld et al. (2006) found that in the first decade of this century, livestock (including cattle, poultry and Livestock Production Systems pigs) contributed 40% of agricultural gross do- and Resource Use mestic product, employed 1.3 billion people and provided one-third of humanity’s protein intake. Following Seré and Steinfeld (1996) and Kruska However, livestock production also had major et al. (2003), Robinson et al. (2011) updated the negative environmental effects – polluting water most common classification for tropical livestock and altering water flows, contributing to biodiver- production. Level 1 in this classification de- sity loss and increasing air pollution as GHGs and scribed livestock production systems using land other noxious gases. Steinfeld et al. (2006, p. 112) characteristics. Level 2 linked potential to actual estimated that, in the early 2000s, livestock pro- livestock production and accounted for other en- duction accounted for some 18% of global GHG terprise options by referring to specific combin- emissions and for more than 80% of agricultural ations of crops and livestock. Level 3 addressed emissions. Extensive livestock systems contributed the intensity and scale of production by incorp- about 13% of global GHGs and intensive systems orating management practices. The resulting contributed about 5%. The major livestock sources classification has nine land-based systems and were enteric emissions (25% of the total), conver- two landless systems. The land-based systems sion of forests to pasture (34%) and manure have three climate categories – arid, humid and (about 24%) (Steinfeld et al., 2006, p. 113, Table temperate – and three agrosystem categories – 3.12). More recent estimates have revised these pastoral, mixed rainfed and mixed irrigated. The shares downwards, but livestock still is a major notation is LGA (livestock/grazing/arid), LGH contributor to global GHG emissions. (livestock/grazing/humid) and LGT (livestock/ A second impetus for new research on trop- grazing/temperate and topical); MRA (mixed/ ical livestock was the IPCC’s Fifth Assessment rainfed/arid and semi-arid), MRH (mixed/ Report (AR5; IPCC, 2014). AR5 was the first rainfed/humid) and MRT (mixed/rain-fed/tem- IPCC assessment to evaluate climate change and perate and tropical); and MIA (mixed/irrigated/ livestock interactions in some detail. It assessed arid), MIH (mixed/irrigated/humid) and MIT the literature on livestock adaptation to climate (mixed/irrigated/temperate and tropical). Map 2 change in addition to mitigation challenges and (p. xviii) shows the nine systems in Africa. 606 P. Ericksen, P. Thornton and G. Nelson Table 16.1. Livestock farming system extent and cattle numbers in Africa and Latin America, 2000. (Adapted from Robinson et al., 2011.) Area in Population in Cattle in Farming system Regiona 2000 (million km2) 2000 (million) 2000 (million TLUs) Agropastoral Central and South America 5.4 40.5 64.2 and pastoral East Asia 5.5 41.3 12.7 South Africa 0.5 19.2 6.2 South-east Asia 0.2 2.2 1.7 Sub-Saharan Africa 13.4 80.2 36.7 West Asia and North Africa 10.2 111.7 8.5 Total 35.2 295.1 129.9 Mixed extensive Central and South America 3.5 100.7 67.2 East Asia 1.7 195.4 20.3 South Africa 1.6 371.9 72.0 South-east Asia 1.2 85.3 10.2 Sub-Saharan Africa 5.1 258.7 55.5 West Asia and North Africa 0.9 87.2 5.3 Total 14.0 1099.2 230.6 Mixed intensifying Central and South America 2.4 221.2 69.4 potential East Asia 2.3 938.5 34.4 South Africa 1.8 844.6 109.5 South-east Asia 1.1 347.2 13.8 Sub-Saharan Africa 1.5 168.2 11.7 West Asia and North Africa 0.6 154.4 6.0 Total 7.3 2674.1 244.9 Other Central and South America 8.8 125.8 41.8 East Asia 1.5 104.2 9.8 South Africa 0.4 69.5 8.7 South-east Asia 1.9 40.4 7.1 Sub-Saharan Africa 4.1 109.2 6.8 West Asia and North Africa 0.2 31.3 1.4 Total 16.9 480.4 75.5 TLU, tropical livestock unit. aRegional groupings of countries are as listed in Thornton et al. (2002). Land C entral and South America and in South Asia, which would therefore have the greatest need for Table 16.1 provides statistics from Robinson et al. adaptation to climate change. (2011) for the areas of cattle-based livestock sys- An updated data set on ruminant meat and tems, estimates of the numbers of animals, and milk production by region and within region by human population by regions of Africa and Latin systems is shown in Fig. 16.5. America in 2000. The report provides similar CAFOs are part of the Seré and Steinfeld tables for pig and chicken systems in Asia. (1996) system and have been an important source Agropastoral and pastoral systems have by of production of cattle, poultry and swine in higher- far the greatest area with 35.2 million km2, of income countries for many years. FAO estimates which sub-Saharan Africa and West Asia and that 80% of growth in the livestock sector now North Africa are dominant. Mixed crop–livestock comes from these industrial production systems, systems occupy 23.8 million km2, of which sub- and this growth is likely to continue. CAFOs are Saharan Africa and Central and South America increasingly important in lower-income countries, dominate. Human and cattle population density especially for poultry, which now accounts are greatest in ‘mixed intensive’ systems in for 23 billion of the 30 billion farm a nimals, Ruminant Livestock and Climate Change in the Tropics 607 but lack of data makes it impossible to map them detail in the section below on mitigation. CAFOs, accurately outside the USA and Europe. ILRI has especially those that utilize feed concentrates done little research on CAFOs but these should be based on maize and soybean meal, generate nox- a topic for future work related to climate. ious odours that affect the quality of life in the immediate area and can be hazardous to human health. In addition, the manure generated can Water quantity be a large source of the GHGs nitrous oxide (N 2O) and CH4. One estimate is that livestock production ac- counts for almost 30% of water use in agricul- ture, most of which is water in crop production Ecosystem services for feed (Mekonnen and Hoekstra, 2010). Some research has contested this concept of Ecosystem services affected by livestock are of water use. Peden et  al., (2007) contend that two main types: (i) services provided by forests the majority of feed and fodder is rainfed, not that are lost as forested areas are converted irrigated; they propose an alternative notion, to pasture or crop production for feed; and that of livestock water productivity ‘defined as (ii) changes in grasslands that reduce a range the ratio of livestock’s beneficial outputs and of services, from water quality and quantity services to water depleted in their production’ availability to biodiversity. Quantifying defor- (Haileselassie et al., 2009). Haileselassie et al., estation is difficult because few countries col- (2009) found that livestock and water crop lect the needed data but de Sy et  al. (2015) productivity were comparable in rainfed sys- estimated that, of deforestation identified in the tems of Ethiopia. 2010 FAO Forest Resource Assessment, pasture was the dominant driver of forest area change (71.2%) and related carbon loss (71.6%) in Water quality South America, followed by commercial cropland (14% and 12.1%, respectively). Livestock reduce water quality principally by Changes in grassland ecosystem services manure runoff. Manure runoff increases both are driven by managed changes in species mix to faecal contamination of water, a major disease improve nutrient quality (see Chapter 11, this transmission mechanism where water treat- volume). Driscoll et  al. (2014) used data from ment is inadequate, and nutrient loads, which eight countries on six continents to show that have adverse human health effects and indirect few governments regulate conventionally bred effects on concentrations of harmful organisms pasture grasses to limit threats to these natural (e.g. algae blooms). Pesticides such as sheep- areas, even though these are bred with charac- dipping chemicals, and bacterial and protozoan teristics typical of invasive species and environ- contamination of soil and water are other con- mental weeds. Proença et al. (2015) reported on cerns regarding water quality (Hooda et  al., a production model that addresses some of these 2000). CAFOs present both potential benefits concerns about grassland ecosystem services. and threats to water quality. CAFOs confine live- The system of sown biodiverse permanent pas- stock waste, reducing the possibility of water tures rich in legumes has been successfully im- contamination over wide areas. However, failure plemented in Portugal on farms in Mediterranean of a containment facility can discharge large climate areas as a response to the low levels of quantities of waste in a matter of hours, over- productivity and feed quality obtained in whelming regular waste-management approaches semi-natural pastures. It consists of a mix of (Mallin and Cahoon, 2003). mostly local grasses and legumes, each mixture tailored to local environmental conditions to best cover the available environmental niches. The system combines higher pasture productiv- Air ity with soil carbon sequestration, reducing at- mospheric carbon dioxide (CO 2), providing the The most important air pollutants from livestock potential for increased farm income from pay- are emissions of GHGs. These are discussed in ments for soil carbon sequestration. 608 P. Ericksen, P. Thornton and G. Nelson Climate Change Impacts c ompared with crops (even fewer on fish and far on Ruminant Livestock fewer on pests and diseases). The major livestock-related climate impact Climate change affects livestock both directly messages from AR5 were as follows: and indirectly. The direct effects arise from • Temperature is an important limiting factor for higher temperature and humidity that slow livestock, for both meat and milk production. animal growth and increase susceptibility to dis- • Climate change will increase water stress ease. A recent study by Rose et al. (2014, p. 219) on livestock systems, affecting the water re- argued that ‘changes in climate’ may have a sources available for livestock via impacts ‘major impact on the seasonal transmission of on runoff and groundwater. gastro-intestinal nematodes in livestock’, based • Pasture response to climate change is com- on evidence from temperate and tropical condi- plex. Increases in CO2 concentration, tem- tions. Indirect effects are felt from the higher feed perature and precipitation will affect pas- prices that are likely as crop and pasture product- ture productivity and quality directly and ivity is reduced, changes in nutrient composition also have important indirect effects on of feeds and pastures occur, and climate affects plant competition, seasonal productivity livestock and wildlife pests and diseases. and plant–animal interactions. For example, AR5 highlighted that research on climate projected increases in temperature and the change impacts on livestock production systems lengthening of the growing season should was relatively limited at the time of its writing. extend forage production into late autumn ‘In comparison to crop and fish production, and early spring in temperate zones. In- c onsiderably less work has been published on creases in CO 2 will tend to benefit C3 species; observed impacts for other food production sys- however, warmer temperatures and drier tems, such as livestock or aquaculture, and to conditions will tend to favour C4 species. our knowledge nothing has been published for Often rangelands benefit from a combin- hunting or collection of wild foods other than for ation of both types of grasses as rainfall and capture fisheries’ (Porter et  al., 2014, p. 494). temperature vary throughout the year. Figure 16.1 shows one-seventh the number of • Host and pathogen systems in livestock will citations in Porter et  al. (2014) on livestock change their ranges because of climate (a) 300 (b) 300 200 200 100 100 0 0 ps ck ies es tio n tio n ge ilit y ’ o o n on ity ty r sto i i il ri C e he r s ea uc ibu a n b at at s h a c z a b i t cu Liv Fi s di ro d str cex for d llo til e d p di S s n : : y: : a f : a U d a lity ity ili t ss es s o i F o sts ab ab il ab e c ‘ Pe ai l ail va il Ac c Ac Av Av A Fig. 16.1. Livestock coverage in the ‘food security and food production systems’ chapter of AR5, Working Group II. (a) Subsectors, and pests and diseases; some citations are not mutually exclusive among categories (e.g. a few crop–livestock citations are included in both subsectors). (b) Food security dimensions. The category ‘Food security’ covers food security in general terms. (From Campbell et al., 2016.) Number of cited papers Number of cited papers Ruminant Livestock and Climate Change in the Tropics 609 change. Species diversity of some patho- role of livestock. Particularly in poor tropical gens may decrease in lowland tropical areas countries, livestock is an enormously import- as temperatures increase. For example, ant risk-management asset for hundreds of temperate regions may become more suit- millions of people. The impacts of increasing able for tropical vector-borne diseases such climate variability on downside risk and on the as Rift Valley fever and malaria. Vector-borne inter-annual stability of livestock production diseases of livestock such as African horse are not well studied. Jones and Thornton sickness and bluetongue may expand their (2009) provided some quantitative assessment range northwards to the northern hemi- of effects of climate change on livestock’s sphere. Changing frequency of extreme risk-management role. It is highly likely that weather events, particularly flooding, will the effects will be negative (Thornton and also affect diseases. Herrero, 2015). Table 16.2 gives AR5’s projected impacts of Climate change will affect all living organ- climate change on livestock in the tropics. At the isms, including livestock pests and diseases. The deadline for accepted papers for AR5 (August effects might be positive or negative for livestock 2013), detailed summaries of impacts on live- productivity depending on the biological suscep- stock systems with or without adaptation were tibility of the species to changes in temperature not available. Summaries addressing the inter- and humidity. The effects are likely to be location actions between crop and livestock enterprises specific and to vary over time as climate changes were also not available. become more pronounced. An important topic not covered in AR5 is how This research is in its infancy, but ILRI re- climate change might affect the risk-management searchers have been contributing to it since Table 16.2. AR5 livestock impacts in the tropics. (Adapted from Porter et al., 2014.) Region Subregion Climate change impacts Scenarios Africa Botswana Cost of supplying water from A2, B2, to 2050 boreholes could increase by 23% due to increased hours of pumping, under drier and warmer conditions Lowlands of Africa Reduced stocking of dairy cows, and a shift from cattle to sheep and goats, due to high temperature Highlands of East Livestock keeping could benefit Africa from increased temperature East Africa Maize stover availability per head of cattle may decrease due to water scarcity South Africa Dairy yields decrease by 10–25% A2, 2046–2065/2080–2100, ECHAM5/MPI-OM, GFDL-CM2.0/2, MRI-CGCM2.3.2 Central and Andean Mountain Beef and dairy cattle, pigs, and To 2060, hot and dry South countries chickens could decrease by scenario America between 0.9% and 3.2%, while sheep could increase by 7% Colombia, Venezuela Beef cattle choice declined To 2060, milder and wet and Ecuador scenario Argentina and Chile Beef cattle choice increased Future climate change Pernambuco, Brazil Milk production and feed intake in Future climate change cattle strongly affected 610 P. Ericksen, P. Thornton and G. Nelson the beginning of this century. McDermott Weather data in climate analyses et al. (2001) looked at the potential effects of climate change, human population growth An early step in research on climate change and and expected disease control activities on tse- its impacts was the release of MarkSim, a sto- tse distribution and trypanosomiasis risk in chastic weather generator developed at CIAT in five agroecological environments in sub- the 1990s in partnership with ILRI (Jones and Saharan Africa up to 2050. They found that Thornton, 1993, 1997, 1999, 2000). the combined effects of these changes would The livestock system classification of Seré be to contract areas under trypanosomiasis and Steinfeld (1996), further refined and devel- risk continent-wide with the greatest de- oped by Kruska et al. (2003), is driven partially crease in the impacts of animal trypanosom- by the length of growing period (LGP). The Mark- iasis in the semi-arid and subhumid zones of Sim model has since been used to refine models West Africa. of LGP. The early use of future LGP surfaces was More recently, Olwoch et al. (2008) exam- by McDermott et  al. (2001), who investigated ined effects of climate change on the range of the effects of climate, human population and the tick-borne disease East Coast fever in socio-economic changes on tsetse-transmitted sub-Saharan Africa using a species distribution trypanosomiasis to 2050. Another application model. They showed increases in East Coast fever of LGP surfaces was published as part of the suitability in the Northern Cape and Eastern study on ‘Mapping poverty and livestock in the Cape provinces of South Africa, Botswana, Ma- developing world’ (Thornton et al., 2002), which lawi, Zambia and eastern Democratic Republic had 479 Google Scholar citations to April 2020. of the Congo. The range shifts are due to changes Several projections have been made of how in temperature minima and maxima and in livestock systems might evolve by 2050 as cli- January and July rainfall. mate change affects LGP. Kristjanson et  al. Bett et al. (2017) reviewed case studies on (2004) projected LGP shifts and livestock system the epidemiology of infectious diseases. Some of changes in West Africa. They forecast declines the studies showed a positive association be- in LGP across most of West Africa, with many tween temperature and expansion of the geo- marginal cropping areas becoming even more graphical ranges of arthropod vectors, while marginal by mid-century and with rangeland others had a negative association. systems disappearing entirely in a few countries. Samy and Peterson (2016) used ecological Jones and Thornton (2009) (97th percentile in niche modelling with a comprehensive occur- Scopus citations) highlighted the possible liveli- rence data set to map the current distribution hood impacts of climate change across Africa, and explore the future potential distribution of hypothesizing that as cropping became more bluetongue virus globally under a range of cli- marginal in semi-arid zones, farmers would turn mate scenarios. Under future climate conditions, to more livestock keeping. the potential distribution of bluetongue virus MarkSim techniques were refined by was predicted to broaden, especially in Central Thornton et  al. (2006) on ‘Mapping climate Africa, the USA and western Russia. vulnerability and poverty in Africa.’ ‘Hotspots’ of climate change, identified via LGP changes projected to the middle of the 21st century for different global climate models and emissions CGIAR research on climate change scenarios, were combined with social indica- impacts in livestock systems tors to identify priority livestock systems for policies to reduce vulnerability and poverty. In the early 1990s, Philip Thornton of ILRI and The study concluded that many vulnerable Peter Jones of CIAT began two lines of research regions are likely to be adversely affected by cli- on climate change impacts and adaptation: mate change in sub-Saharan Africa, notably (i)  development of models to transform output the mixed arid–semi-arid systems in the Sahel, from climate models into weather data useful in the arid–semi-arid rangelands in eastern Africa, impact studies; and (ii) models of livestock sys- the Great Lakes and coastal regions of eastern tem performance under climate change. Africa, and all systems in southern Africa. Some Ruminant Livestock and Climate Change in the Tropics 611 of the maps and data from Thornton et  al. assessed potential priority activities for ILRI. The (2006) were used directly in the IPCC’s Fourth inventory of climate change impacts (Thornton Assessment Report (Boko et  al., 2007; IPCC, et al., 2009) listed seven topics – feeds quantity 2007) and the paper had been cited in Google and quality, heat stress, water quantity and Scholar more than 325 times to April 2020. quality, livestock diseases and disease vectors, Further analyses using MarkSim followed the biodiversity, systems and livelihoods, and indir- 2006 study. Projections of cattle trypanosomiasis ect impacts (human health effects from chan- were redone to 2030 for one of the UK Govern- ging disease burden, worsening heat-related ment’s Foresight Projects (Thornton et  al., mortality and morbidity). Table 16.3, adapted 2006). Systems impacts were analysed in Tur- from Thornton et al. (2008), summarizes gaps in kana District, Kenya (Notenbaert et  al., 2007). our understanding of the impacts of climate Impact studies were undertaken on pastoral and change and the role(s) that international re- agropastoral systems in East and West Africa for search might have in closing such gaps. the CGIAR’s Systemwide Livestock Programme Activities were ranked in relation to their (Thornton et al., 2008), and on the agricultural importance to ILRI’s mandate and the achieva- sector in East and Central Africa for the Associ- bility of outputs and outcomes. The top-ranked ation for Strengthening Agricultural Research in activities were: (i) identification of feed ‘hot- Eastern and Central Africa (ASARECA; van de spots’; (ii) improved understanding of climate Steeg et  al., 2009). Box 16.1 summarizes re- change on livestock systems and livestock keep- search utilizing MarkSim analysis and data gen- ers’ livelihoods; (iii) the development and deploy- eration. This list indicates the scope and nature ment of assessment frameworks and targeting of analyses completed, with impacts on crop and tools; and (iv) identification and dissemination livestock productivity, pest incidence, changes in of adaptation options. In the 10 years since the land use, CH emissions and poverty. Rassmann analysis was completed, considerable progress 4 and Schuetz (2017) highlighted wider studies has been made in activities (ii) and (iii). Less pro- using MarkSim, including a study on the pos- gress has been made on activities (i) and (iv), al- sible future spread of the Zika virus (Messina though new research is under way at ILRI on et al., 2016) and a projected decline in ice-skating both of these areas. days in Canada, an important recreational eco- ILRI inputs were used in the International system service in that country (Brammer et  al., Assessment of Agricultural Knowledge, Science 2015). and Technology for Development (IAASTD) A recent innovation promises to expand agricultural scenarios to 2050 (Rosegrant et al., MarkSim’s usefulness. MarkSim/GCM is a web 2009), which projected spatial livestock data. tool that uses MarkSim to generate location- Reception of the IAASTD report at its release specific weather data from GCM results used in was mixed, although it did provide an important AR5. The outputs include graphical depictions analysis of necessary changes in the global food of the data and creation of a data set that can be system. ILRI work has also contributed to the imported into the crop modelling software DSSAT. analysis of drivers of change in agricultural sys- (http://gisweb.ciat.cgiar.org/MarkSimGCM/; tems (van Vuuren et al., 2009). accessed 7 March 2020). The second version of Box 16.2 summarizes key ILRI research MarkSim will be improved over the first version outputs on climate change impacts, including as it will use 55,000 rainfall stations, compared reviews of impact and adaptation studies in with some 9000 for version 1. It will allow the mixed crop–livestock and pastoral–agro-pastoral study of novel climates – climates that will exist systems and a range of adaptation studies using in the future that currently do not exist any- different modelling approaches at varying scales where – in more detail. (e.g. household, regional, global). Tables 16.1 and 16.2 highlight the shift from livestock com- ponent impact studies to more systems-oriented Impacts on livestock systems work that attempts to understand the broader implications of climate change adaptation at dif- Thornton et  al. (2008) reviewed what was ferent scales. Much of the research of ILRI and known about climate change and livestock and its partners is tied to models of sustainable 612 P. Ericksen, P. Thornton and G. Nelson Box 16.1. Impact of the MarkSim model. MarkSim has had far-reaching impacts in: (i) modelling crop production; (ii) mapping the relationships among climate, agriculture and poverty; and (iii) modelling system effects of climate change: Modelling crop production Impacts of climate change to 2055 on maize yields in Latin Jones and Thornton (2003) (99th America and Africa percentile in Scopus) Spatial variation of crop yield response to climate change in Thornton et al. (2009) (99th East Africa percentile in Scopus) Rainfall variability, and impacts of climate change on length Thornton et al. (2007) of growing period Mapping relationships among climate, agriculture and poverty Mapping poverty and livestock in the developing world Thornton et al. (2002) Mapping climate vulnerability and poverty in Africa Thornton et al. (2006) The livestock, climate change and poverty nexus Thornton et al. (2008) Modelling system effects of climate change Effects of climate, human population and socio-economic McDermott et al. (2001) changes on tsetse-transmitted trypanosomiasis to 2050 Livestock systems changes to 2050 in West Africa Kristjanson et al. (2004) Cattle trypanosomiasis in Africa to 2030 Thornton et al. (2006) Livestock development and climate change in Turkana Notenbaert et al. (2007) District, Kenya Impacts of climate change on pastoral and agropastoral Thornton et al. (2008) systems in East and West Africa Understanding climate–land interactions in East Africa Olson et al. (2008) Spatial distribution of CH4 emissions from African domestic Herrero et al. (2008) ruminants to 2030 Influence of climate change and climate variability on the van de Steeg et al. (2009) agricultural sector of East and Central Africa Livestock system impacts in the tropics Thornton and Herrero (2010a) Climate change and crop production impacts in Thornton (2009) the Albertine Rift Possible impacts of climate change on livelihood transitions Jones and Thornton (2009) in Africa – croppers to livestock keepers? Adapting to climate change in households in East Africa at Thornton et al. (2010) the level of the household and the system Impacts of climate change on migration to 2060 New et al. (2011) Mapping hotspots of climate change and food insecurity in Ericksen et al. (2011) the global tropics Adapting to climate change in mixed crop–livestock systems Thornton et al. (2011) in developing countries Agriculture in sub-Saharan Africa in a 4°-plus world Thornton et al. (2011) Global livestock production systems Robinson et al. (2011) Consequences of climate change for pastoralism in Ericksen et al. (2012) sub-Saharan Africa Future climate change and land-use change impacts on East Moore et al. (2012) African food security MarkSim as a GCM downscaling tool: AR4 climate model Jones and Thornton (2013) ensembles Climate change adaptation in mixed crop–livestock systems Thornton and Herrero (2015) in developing countries Climate variability and vulnerability to climate change: a Thornton et al. (2014) review Continued Ruminant Livestock and Climate Change in the Tropics 613 Box 16.1. Continued. Impacts on smallholder agriculture in sub-Saharan Africa to Cooper et al. (2014) 2050 Climate change impacts on livestock Thornton et al. (2015) Carbon and biodiversity costs of converting Africa’s wet Searchinger et al. (2015) savannahs to cropland MarkSim as a GCM downscaling tool: AR5 climate model Jones and Thornton (2015) ensembles and soils data Climate change adaptation in the mixed crop–livestock Thornton and Herrero (2015) system in sub-Saharan Africa Adaptation paths for vulnerable areas Cacho et al. (2016) Pastoral farming systems and food security in sub-Saharan de Leeuw et al. (2019) Africa i ntensification (Garnett et al., 2013) and climate- variability, more information is needed concern- smart agriculture (Lipper et al., 2014). ing the nature and extent of the trade-offs among crop and livestock enterprises, and be- tween on- and off-farm income sources, as cli- Current knowledge gaps on impacts mate variability increases. This may have critical effects on food security; in addition to impacts on As initially identified by Thornton et al. (2008), food availability, variability may strongly affect identification of feed ‘hotspots’ remains a priority. the stability of food supplies and vulnerable In addition, there is much that is not well under- people’s ability to access food at affordable prices stood about the interactions of climate and cli- (Schmidhuber and Tubiello, 2007). Key to these mate variability with other drivers of change in broad issues will be the refinement of impact livestock systems and with population growth, models to assess climate variability effects on income growth and global trade. Multiple and adaptation and mitigation options at regional competing pressures are likely on tropical and and local scales, their effects on livelihoods and subtropical livestock systems in the future, to the trade-offs that arise among income, food se- produce food, to feed livestock and to produce curity and environmental objectives. energy crops, for example. While recent scien- Grace et al. (2015) identified the following tific assessments such a AR4 and AR5 (IPCC, knowledge gaps in animal disease and climate 2007, 2014) represent an accurate reflection of change: current knowledge, there remain gaps in their treatment of tropical livestock systems regard- • Information on animal diseases. The rela- ing the provision of ecosystems goods and ser- tively limited availability of epidemiological vices and the maintenance of livelihoods. (and ecological) observations on animal First, more clarity is needed concerning the disease in the tropics constrains our under- benefits of livestock, their negative impacts on standing of the climate–disease relation- GHG emissions and the environment, and the ef- ships. Current surveillance detects only a fects of climate change on livestock systems. The small proportion of livestock and wildlife regional and local variations in public costs and diseases and is not well linked to human dis- benefits associated with livestock need to be ease surveillance. understood before technology and policy op- • Disease dynamics. There are numerous tions for adaptation and mitigation can be tar- pathways – direct and indirect – through geted appropriately. Much agricultural impact which climate can influence disease. These work is reported at a continental or regional drivers are not all equal, and impacts medi- level (e.g. Lobell et al., 2008), but this aggrega- ated through changes in human popula- tion masks widespread differences. tion and behaviour may induce effects that Second, while a great deal is known about are orders of magnitude greater than those how livestock keepers manage current climate mediated through biological pathways. 614 P. Ericksen, P. Thornton and G. Nelson Table 16.3. Climate change knowledge gaps and research hypotheses. (Adapted from Thornton et al., 2008.) Feasibility of Alternative delivery Regional System Time to suppliers Feasibility of (outputs to Activity area Knowledge gaps Research outputs focus focus outputs Relative cost of outputs outputs outcomes) Feeds: What are the Localized impacts East, West MRA, LRA Short Low Very few High High (e.g. for quantity localized and hotspots and South priority and quality impacts? identified Africa setting) Rangelands: primary Rangeland net East and LRA/LRH/ Medium Medium ARIs Medium–high Medium productivity, primary South LRT species productivity Africa, distribution and distribution and North-east change due to impacts Asia CO2 and other elucidated factors; estimation of carrying capacities Crops: primary Modified crop and East, West MRA/MRH/ Long Medium–high Very few Medium Low–medium productivity, residue quality and South MRT, harvest indexes and quantity Africa, MIA/MIH/ and stover South Asia MIT production, dual purpose crops Feasibility of new New feeding East, West MRA/MRH/ Medium–long Medium NARS Medium Low–medium feeding strategies strategies and South MRT with existing developed Africa, materials South Asia Pests and diseases Hotspots identified East, West MRA/MRH/ Medium Medium OIOs Low–medium Medium of feeds of key pests, and South MRT diseases of key Africa feed crops Water Evolution of surface Understanding of East, West LRA/LRH/ Medium Medium OIOs Low–medium Medium and groundwater changes in and South LRT, supply, impacts surface and Africa, MRA/ on livestock groundwater South Asia MRH/ supply, and MRT impacts on livestock Ruminant Livestock and Climate Change in the Tropics 615 Increases in Options East, West LRA/LRT, Medium–long Medium–high Very few Low–medium Medium livestock water developed and and South MRA/ productivity tested to Africa, MRT, increase South Asia MIA/MIT livestock water productivity Animal Potential changes in Future changes in East, West All livestock Medium–long Medium–high ARIs Low–medium Medium–high health the prevalence prevalence and and South systems and intensity of intensity of Africa, epizootics in epizootics South Asia livestock predicted Impacts of diseases Impacts of East, West MRH/MRT, Medium–long Medium–high OIOs Low–medium Medium–high of intensification ‘management’ and South coast, (e.g. mastitis) diseases Africa, urban elucidated and South Asia options identified Biodiversity ‘Ecological Impacts on East, West All livestock Medium–long Medium–high GCC Low Low biodiversity’: what ecological and South systems will happen to biodiversity Africa, numbers of elucidated South Asia species as systems change? Animal breed Animal breed East, West All livestock Medium–long High OIOs Low High biodiversity: which biodiversity and South systems traits might be characterized, Africa, useful in the and a road South Asia future? map developed for future exploitation Plant biodiversity: Animal breed East, West All livestock Medium–long High OIOs Low High which traits and biodiversity and South systems hence which characterized, Africa, germplasm might and a road South Asia be useful in the map developed future? for future exploitation ARI, advanced research institute; GCC, global change community; LRA, livestock/rainfed/arid; LRH, livestock/rainfed/humid; LRT, livestock/rainfed/temperate and tropical; NARS, national agricultural research system; OIO, other international organization. 616 P. Ericksen, P. Thornton and G. Nelson Box 16.2. Impact of ILRI climate research. Climate research by ILRI and partners has had important scientific impacts on: (i) policy options; (ii) mitigation technologies; (iii) adaptation problems; and (iv) the future of tropical agriculture. Policy options Livestock production: recent trends, future prospects Thornton (2010) (99th percentile in Scopus) Discussion paper on ILRI’s research in relation to climate change Thornton et al. (2008) A review of the impacts of climate change on livestock and livestock Thornton et al. (2009) systems in developing countries, current knowledge and gaps Coping with drought and climate change in the pastoral sector in sub- Herrero et al. (2010) Saharan Africa: policy considerations Livestock and global change: emerging issues for sustainable food systems; Herrero and Thornton a brief summary of the major challenges (2013) Livestock contributions to the chapter ‘Food Security and Food Production Porter et al. (2014) Systems,’ Working Group II Livestock and the environment: what have we learnt in the last decade? Herrero et al. (2015) Impacts of climate change on the agricultural and aquatic systems and natural Thornton and Cramer resources within CGIAR’s mandate: an inventory of what is known (2012) How does climate change alter agricultural strategies to support food Thornton and Lipper security? (2014) Mitigation technologies The potential for reduced CH4 and CO2 emissions from livestock and Thornton and Herrero pasture management in the tropics; analysis based on systems (2010b) characterization in the future The impacts of climate change on livestock and livestock systems in Thornton et al. (2010) developing countries Adaptation problems Is proactive adaptation to climate change necessary in grazed rangelands? Ash et al. (2012) A study on how these systems may need to adapt Adapting smallholder mixed crop–livestock farming systems to climate Rigolot et al. (2017) variability in northern Burkina Faso with crop–livestock interactions Transitions in agro-pastoralist systems of East Africa: impacts on food Rufino et al. (2013) security and poverty. Twelve case study sites in the marginal areas, evaluating likely impacts and possible adaptations Evaluating climate-smart adaptation options in mixed crop–livestock Thornton et al. (2016) systems in developing countries: a largely qualitative approach to targeting and evaluation Climate change and pastoralism: impacts, consequences and adaptation Herrero et al. (2016) Exploring future changes in smallholder farming systems by linking Herrero et al. (2014) socio-economic scenarios with regional and household models: an early multi-scale analysis of different drives of change, including climate change The future of tropical agriculture The future of agriculture (crops and livestock) to 2050 Rosegrant et al. (2009) Drivers of change in agricultural systems to 2050 van Vuuren et al. (2009) A largely qualitative assessment of the likely effects of climate change as a Thornton and Gerber constraint to the growth of the livestock sector (2009) Kenya: climate variability and climate change and their impacts on the Herrero et al. (2010) agricultural sector Implications of future climate and atmospheric CO2 content for regional Doherty et al. (2010) biogeochemistry, biogeography and ecosystem services across East Africa Climate change and the growth of the livestock sector in developing countries Thornton and Gerber (2009) Impact of climate change on African agriculture: focus on pests and diseases Dinesh et al. (2019) Using a stakeholder and multi-model process to translate the shared Palazzo et al. (2017) socio-economic paths under climate change for the West Africa region Ruminant Livestock and Climate Change in the Tropics 617 • Multi-host diseases. The majority of climate- Improving livestock genetics is an option. sensitive diseases affect many host species Ortiz-Colón et al. (2018) reviewed work from the including livestock, wildlife and occasionally Caribbean showing that introducing a ‘slick humans. This makes them much more diffi- hair’ gene into Holstein cows by cross-breeding cult to control or eliminate than disease that with Senepols may increase heat tolerance and have only a human or livestock host (for ex- productivity. However, genetic improvements ample, when zoonotic tuberculosis is present would require substantial investments and in badgers it is much more difficult to control would involve long delays before being intro- than when it is only present in cattle). duced into production animal populations. • Joint occurrence of climate-sensitive dis- Moreover, there would be temperature limits eases. A review of risk maps reveals that a above which adaptation is not possible, even number of climate-sensitive livestock dis- with substantial genetic progress. eases occur in some common areas given that their emergence and transmission are controlled by similar ecological factors. Costs of adaptation • Lack of laboratory and epidemiology cap- acity. The lack of laboratory and epidemi- There are many possible adaptations in tropical ology capacity is a long-standing problem livestock systems for which we lack informa- in developing countries. Much effort and tion on social and private costs and benefits. expense has been spent on improving cap- Dittrich et  al. (2017) suggested techniques to acity, and best approaches exist but require assess livestock adaptations, such as cost– investment. benefit analysis, portfolio analysis, real options analysis and robust decision making, but their approach suffered from a lack of empirical data Adaptation of Livestock Systems to verify the proposed adaptations under trop- to Climate Change ical conditions. Weindl et  al. (2015) is the only study to project adaptation costs by simulating climate The AR5 text (IPCC, 2014) on adaptation re- impacts on crop and range yields productivity lies heavily on Thornton et al. (2009) for its list for ten world regions to 2045. If tropical live- of adaptation options. Adaptation options in- stock systems shift towards mixed crop–live- clude: (i) matching stocking rates with pasture stock systems and away from grazing systems, production; (ii) adjusting herd and watering adaptation costs would fall in sub-Saharan Af- point management to altered seasonal and rica and Latin America and the Caribbean, spatial patterns of forage production; (iii) man- while rising significantly in Pacific Asia and aging diet quality (using diet supplements, leg- South Asia. The Weindl model does not ac- umes, introduced pasture species and pasture count for climate change effects on livestock fertility management); (iv) more effective use disease or on animal reproductive performance of silage, pasture seeding and rotation; (v) fire and it is likely, therefore, to underestimate management to control browse encroachment; adaptation costs. (vi) using more suitable livestock breeds or spe- cies; (vii) migratory pastoralism; and (viii) bios- ecurity activities to monitor and manage pests, weeds and diseases (IPCC, 2014, p. 517). CGIAR research on climate change Research in Australia found that combin- adaptation in livestock systems ing adaptations can be more beneficial than sin- gle adaptations (Ghahramani and Moore, 2013; While there have been extensive international Moore and Ghahramani, 2013). Options in- efforts to develop options to adapt to climate clude replacing cattle with small ruminants, re- change, less has been done with producers on ducing stocking rates, better water management implementation of these options. One innov- technologies and animal health services, and ation with the potential to facilitate adaptation improving tree cover. has been agricultural insurance. 618 P. Ericksen, P. Thornton and G. Nelson The information costs and incentive prob- A challenge to any insurance approach is lems that are characteristic of agriculture have cost. While it is conceptually possible for an insur- often prevented the emergence of insurance ance scheme to self-finance, and many private in- markets in rural areas (Binswanger and Rosenz- surance programmes do so in other markets (e.g. weig, 1986). As information costs have fallen, life, automobile, health insurance) because of the use of insurance for agricultural risk man- long experience identifying actuarial risks, agri- agement has become more common in devel- cultural insurance markets have proven difficult oped countries for staple crops (e.g. maize, for the private sector to operate profitably because wheat) and to a lesser extent for other crops and of the spatial nature of agriculture. The spatial livestock. Insurance, by managing the effects of nature of farming makes it costly to monitor risks shocks, allows farms to invest more profitably in and identify losses that trigger payment. Further- non-shock periods (Alderman and Haque, 2007; more, climate change is likely to change the risk Barnett et al., 2008; Mahul and Stutley, 2010). portfolio in unknown ways, making insurance Insurance also facilitates complementary mar- management more difficult. kets, such as those for credit, inputs and produc- tion methods (Alderman, and Haque, 2007; Carter et al., 2007) by diffusing risks. Insurance can potentially help farmers to manage climate Knowledge gaps on adaptation risk by allowing them to use new adaptation strategies, while reducing the adverse effects of Thornton et  al. (2008) summarized the know- current shocks (Collier et al., 2009). ledge gaps in adaptation and followed a priority- Index insurance has emerged as a possible setting process to identify adaptation activities solution for overcoming supply-side constraints by their importance to ILRI’s mandate, the clar- to rural insurance markets and for extending ity of ILRI’s role, the presence of other providers, access to agricultural insurance. The Index- the achievability of outputs and outcomes, and based Livestock Insurance (IBLI) work is one the cost and approximate time to output. The form of that solution. By basing insurance pol- gaps were as follows icies on easily observed indices, such as precipi- tation or temperature, that are covariate with • Adequately detailed estimates of the impacts rural income and wealth risks, index insurance of climate change on livestock systems with can potentially resolve the information costs or without adaptation. and incentive problems inherent in rural finan- • The impacts of increasing climate variability. cial markets and allow provision of insurance • Information on costs and benefits of adap- coverage at a fraction of the costs of loss-based tations at given sites and seasons. This ap- polices (Chantarat et  al., 2013; Jensen and plies particularly to mixed systems, in Barrett, 2016). which the interactions between crops and There is some limited empirical evidence livestock can sometimes be managed to ad- of the effects of IBLI. Households with IBLI vantage. The challenge is to target packages coverage reduced their herd size and increased of adaptation options that are locally ap- investments that made the remaining animals propriate and amenable to scaling up. more productive (Thornton and Herrero 2010; Some of the major gaps were addressed in Gerber, et al., 2011; Jensen et al., 2017). Such the decade since Thornton et al. (2008). One ex- impacts are consistent with economic theory, ample was the impacts of climate change on whereby insurance coverage substitutes for rangeland net primary productivity (Boone et al., informal insurance mechanisms, oversized 2018). Several assessment models and targeting herds in this case. Insurance releases house- tools were developed and a special issue of Agri- holds from some risk constraints so that they cultural Systems (Volume 151, February 2017) can invest in productivity-increasing tech- was devoted to this topic. However, while these nologies, such as animal health care. In terms studies provide insights into what the impacts of of climate change adaptation, insurance re- climate change are likely to be, they do not pro- duces sensitivity to drought and lowers the vide much general guidance on priority adapta- costs of adaptation. tion activities as these are context s pecific. Ruminant Livestock and Climate Change in the Tropics 619 A review by Ash et al. (2012) gave mixed results However, the most recent epoch in which global about the need for ‘proactive adaptation’ in temperature was as high as it is now was more rangelands; while ‘incremental, autonomous than 100,000  years ago. The experiences of adaptation [would be] sufficient to deal with the pastoralists in recent millennia may therefore gradual expression of climate’ it is not known prove inadequate for adapting to current how autonomous adaptation can manage more changes in levels and variability of temperature rapid climate change in the absence of new re- and humidity. For example, Thornton and search and more supportive public policies. Herrero (2010a) simulated an increase in drought frequency to once every 3  years and Adaptation in mixed crop–livestock found that this higher frequency decreased live- systems stock densities below desirable levels. In some places, adaptation will be possible through spe- Thornton and Herrero (2015) highlight four re- cies changes, increased market orientation or search needs for appropriate adaptation options the increased ability of pastoralists to manage among mixed crop–livestock enterprises in climate risk. sub-Saharan Africa: Increasing population densities can rapidly 1. Biophysical models are needed to represent modify the accessibility to land, water and feed interactions among crops and livestock to make that makes pastoralism a viable livelihood strat- evaluations of mixed systems more robust. Most egy (Hobbs et al., 2008). Rising incomes are af- biophysical modelling has been done on the pri- fecting consumption patterns and modifying mary cereals (particularly maize, rice and expectations, with lasting impacts on traditional wheat) and legumes (groundnut and soybean), socio-cultural value systems and kinship net- but more work is needed on lesser-studied crops, works. In some places, adaptation will be pos- such as trees and other perennials. sible via farming system intensification through 2. Whole-farm models are needed because of increased market orientation and increased abil- the complex interactions of financial and phys- ity of pastoralists to manage climate-related ical resources in smallholder households. Trade- risks. In others, adaptation may need to be more offs between benefits and costs of adaptation transformative, including social innovations recommendations are inevitable and must be and changes in behaviour, institutions and cul- quantified with a whole-farm perspective. Whole- tural norms. Opportunities exist for improving farm modelling, especially in tropical A frica, is development outcomes in pastoral systems, constrained by a systemic lack of time-series through combinations of policies and institu- data. The explicit inclusion of human nutrition tional and technological alternatives that will with its appropriate metrics is also essential. vary with context and through time as the fu- 3. Use of future scenarios is needed to capture ture climate change envelope becomes less un- the nuances of smallholder systems in the con- certain (Ericksen et  al., 2012). Understanding text of larger economic and biological changes. what is possible, what is not, and where will be Some smallholder systems will intensify produc- critical for effectively improving the livelihoods tion and survive; others will become redundant of pastoralists and their rangelands (Herrero as smallholdings are aggregated into larger, more et al., 2016). intensive and more specialized systems. Research is also needed on how policy can 4. Better metrics are needed to estimate vulner- support the scaling of interventions that can ability to climate change among smallholders contribute to food and nutritional security and and to define measures of successful adaptation, poverty reduction under climate change. ILRI is such as sustainability and reduced variability of already contributing to this agenda via work on income. IBLI and cash transfers and research on effect- ive governance mechanisms that can promote adaptation. A recent collaboration with the Adaptation in pastoral systems World Agroforestry Centre called Local Govern- ance and Adaptation to Climate Change (LGACC; Pastoralists have long adapted to a highly vari- http://www.worldagroforestry.org/project/ able climate (see Chapter 15, this volume). local-governance-and-adapting-climate-change- 620 P. Ericksen, P. Thornton and G. Nelson sub-saharan-africa-lgacc; accessed 8 March agricultural emissions estimates of between 2020), for example, combined research on 4.25 and 5.25 GtCO2eq/year (Smith et al., 2014, rangeland governance with research on pro- Fig. 11.4). Estimates of emissions from enteric cesses that promote adaptation. The team was fermentation were just less than 2 GtCO2eq/year, able to draw conclusions about the comple- implying that cattle were responsible for 40–50% mentarity between governance, rangeland of agricultural emissions. Figure 16.2 reports management and climate change adaptation early 21st century estimates of anthropogenic (LGACC, 2018). GHG emissions and livestock’s share. In this fig- ure, livestock’s share of total emissions is 14.5%, with 27% from CO2, 29% from N2O and 44% Mitigation of Greenhouse Gas from CH4. Figure 16.3 shows the spatial distri- Emissions from Livestock bution of livestock GHG emissions around the turn of the century. National research on livestock emissions has The livestock sector is a major source of GHG been growing rapidly in response to the United emissions, primarily CH4, CO2 and N2O. Emis- Nations Framework Convention on Climate sions arise from five components – ruminant Change (UNFCCC) requirement of national emis- digestion, excretion of manure and urine, feed sions inventories. Patra (2012) estimated CH production, land conversion to pasture and 4and N2O emissions from Indian livestock. Svinurai transport/processing. et  al. (2018) provided estimates of enteric CH Projections from the beginning of the 21st 4emissions in Zimbabwe. What is not clear is how century to mid-century suggest that per capita comparable the country-specific results are. The meat consumption between 2010 and 2050 Standard Assessment of Agricultural Mitigation could increase by about 50% for low-income Potential and Livelihoods (SAMPLES) project countries and about 10% for higher-income coun- (http://samples.ccafs.cgiar.org; accessed 8 March tries (Nelson et al., 2018, Supplementary Fig. 4). 2020) of which ILRI researchers are a part is de- Low- and middle-income countries have a 62% signed to facilitate this cross-country comparabil- share of total global production, rising to 72% by ity (Rosenstock et al., 2016). 2050 (Thornton, 2010). The GHG mitigation challenge is how to satisfy a growing livestock product demand while reducing GHG emissions. Mitigation via supply- and demand-side options Estimates of emissions from livestock Supply-side activities to reduce GHG emissions from ruminant livestock production can be clas- Estimates of GHG emissions of livestock prod- sified as: (i) targeting reductions of enteric CH4; ucts vary considerably; emissions per unit of (ii) managing manure to reduce N2O emissions; protein are highest for beef and dairy and lower (iii) sequestering carbon in rangelands; (iv) im- for pork, chicken meat and eggs (de Vries and de plementation of better animal husbandry prac- Boer, 2010; Gerber et al., 2013) due to their dif- tices; and (v) land-use practices to sequester ferent feed and land-use intensities. Beef pro- carbon. Excluding land-use practices, Herrero duction can use up to five times more biomass to et al. (2016) found that these options have a glo- produce 1 kg of animal protein than dairy. Emis- bal mitigation potential of 2.4  GtCO2eq/year. sions intensities for the same livestock product These estimates are in the same range as those also vary largely among different regions of the proposed by Gerber et al. (2013) of 1.8 GtCO2-eq/ world (Herrero et al., 2013). Europe and North year, although strategies will vary by production America have lower emission intensities per kg system (Rivera-Ferre et al., 2016). of protein than Africa, Asia and Latin America. The AR5 review of mitigation options in Estimates of the contribution of livestock to agriculture (Smith et al., 2014) found that: GHG emissions depend on estimation methods Studies based on integrated modelling show that and data sources. AR5 reported a range of total changes in diets strongly affect future GHG Ruminant Livestock and Climate Change in the Tropics 621 Global total GHG anthropogenic emissions 27% CO2 29% N2O 44% CH4 Livestock contributes Livestock contributes Livestock contributes 5% 53% 44% 1.35% of total CO2 15% of total N2O 19% of total CH4 anthropogenic emissions anthropogenic emissions anthropogenic emissions 25 times CO 298 times CO 22 14.5% of total GHG anthropogenic emissions Fig. 16.2. Global total GHG anthropogenic emissions and livestock’s share. (From Rojas-Downing et al., 2017, based on analysis for the early 21st century in Gerber et al., 2013.) MtCO2eq/km 2/year 7.5 15 30 45 60 75 90 105120 Fig. 16.3. GHG emissions from global livestock, 1995–2005. (From Herrero et al., 2016.) emissions from food production… Technical changes in consumption was found to be mitigation options on the supply side, such as substantially higher than that of technical improved cropland or livestock management, mitigation measures. alone could reduce [emissions from 15.3 GtCO2eq/year] to 9.8 GtCO2eq/yr, whereas emissions were reduced to 4.3 GtCO eq/yr in a Supply-side options2 ‘decreased livestock product’ scenario and to 2.5 GtCO eq/yr if both technical mitigation and Supply-side efforts have focused on reducing the 2 dietary change were assumed. Hence, the GHG burden of livestock through increases in potential to reduce GHG emissions through productivity. Capper et  al. (2009) showed that 622 P. Ericksen, P. Thornton and G. Nelson US dairy production in 2007 used only 21% of achieved if reduced consumption of animal- the animals, 23% of the feed, 35% of the water based products is combined with sustained and 10% of the milk that had been required in productivity gains in plant production, but the 1944 to produce 1 billion kg of milk. Emissions economic feasibility of the latter is uncertain. from dairy cattle fell in consequence, with CH4 Scherer and Verburg (2017) compared emissions only 43% and 56% of N2O emissions supply- and demand-side options under the in 2007 relative to 1944. Overall, the carbon- label of ‘climate-smart agriculture’. Adaptation equivalent footprint of 1 billion kg of milk in the measures under climate-smart agriculture can USA in 2007 was 34% of that in 1944. Similar involve technological advances, new farming evidence was found by Gerber et al. (2011) who practices, and changes in food origin and supply identified four reasons for the reduction in emis- chain management. Unlike Weindl et al. (2017), sions from dairy systems as they intensify: (i) Scherer and Verburg (2017) did not use an inte- higher-quality diets; (ii) higher proportions of grated global model, so their findings are weaker feed energy and protein used for production and with regard to demand-side measures. Their not maintenance; (iii) higher nitrogen efficiency; findings were that: (i) emissions reductions are and (iv) a concentration approach to reducing possible with demand measures, such as a vegan unit emissions through genetics and animal diet or local sourcing, but their economics are health. very uncertain and site-specific; and (ii) supply- Gerber et al. (2013, p. xiii) provided a global side measures can also have mitigation effects, review of mitigation potentials to reduce GHG but the latter are probably less effective than emissions from ruminant and non-ruminant demand measures. livestock. They found that a ‘30% reduction of Ripple et al. (2013) argued for both supply- GHG emissions would be possible, for example, if and demand-side options, citing modelling of a producers in a given system, region and climate food tax proportional to the mean GHG emis- adopted the technologies and practice currently sions per unit of food sold. Shields and Orme- used by the 10% of producers with the lowest Evans (2015) argued that a mitigation strategy emission intensity’. of intensifying production would not be socially The technical changes modelled in Gerber sustainable because of its adverse effects on ani- et al. (2013) are due to productivity gains from mal welfare. higher digestibility feeds, herd health interven- Valin et  al. (2013) reported results from tions and genetic selection for animals with GLOBIOM modelling of productivity increases in higher milk productivity. It is not clear whether crops and livestock. They found that closing the technologies producing these gains are yield gaps by 50% for crops and 25% for live- profitable. stock by 2050 would decrease agriculture and Weindl et al. (2017) compared supply- and land-use change emissions by 8% overall, and by demand-side scenarios in carbon dynamics to 12% per calorie produced. However, the out- 2050. They mapped the results of two demand come is sensitive to the technological path and scenarios – a continuation of trends in global which factor benefits from productivity gains: diets, including levels of animal products, and a sustainable land intensification would increase gradual change in diet projections to lower GHG savings by one-third when compared with shares of animal-based calories in diets, with a fertilizer-intensive pathway. Improvements in 15% as the upper limit in 2050 for calories from the crop or livestock sector have different out- livestock and fish – and four supply scenarios, comes: crop yield gains would bring the largest ranging from current levels of productivity in food provision benefits, whereas livestock yield low-productivity animal systems to slight to low gains would bring the largest cuts in GHGs. to moderate productivity gains. Changes in diet N 2O is a powerful GHG that is emitted as a would produce substantial reductions in CO 2 consequence of the use of both organic and in- burden at all levels of productivity change, ran- organic nitrogenous fertilizers. Some quantity ging from –40% to –57%. Changes in productiv- of applied nitrogen is not taken up by the plants ity without changes in diet would increase the and is lost to ground water and the atmosphere. CO2 burden substantially. The highest abate- In the early part of the 21st century, it was dis- ment of carbon emissions (63–78%) can be covered that the roots of many plants release Ruminant Livestock and Climate Change in the Tropics 623 substances that inhibit nitrogen release. The i ncrease to meet demand and using data from process is called ‘biological nitrification inhib- FAO on livestock species and diet composition ition’. Early research focused on the tropical pas- and Food and Agriculture Organization Corpor- ture grass, Brachiaria spp., and researchers have ate Statistical Database (FAOSTAT) population since looked into the possibility of enhancing projections. For the latter, Africa was divided biological nitrification inhibition in wheat, bar- into regions to be more specific about diets by ley and rye (Subbarao et al., 2009; Moreta et al., production system and season variation, along 2014; Byrnes et al., 2017; Karwat et al., 2017; with the level of intensification. To move from Subbarao et  al., 2017; Nuñez et  al., 2018; diets to CH 4 emissions, they used the RUMIN- Teutscherova et al., 2019). ANT model. The results showed the importance of the assumptions about population growth and Mitigation research in livestock systems changes in densities, as these drove the pro- jected increase in total CH4 emissions, estimated A key contribution to both ILRI and other re- to be 42% between 2000 and 2030. Emissions searchers in the study of livestock emissions was intensities differed between production systems, the development of the RUMINANT model, as but all were estimated to increase by 2030. first described by Herrero (1997) and subse- Total emissions varied by region, with the Horn quently by Herrero et al. (2013), with reference of Africa estimated to be the largest emitting re- to Sniffen et  al. (1992) and AFRC (1993). It is gion. Cattle contributed over 80% of emissions used to predict feed intake, nutrient supply and across the continent. These findings were in line CH emissions. These numbers are then aggre- with other studies. Steinfeld et  al. (2006) esti-4 gated to systems, countries, regions and contin- mated emissions from Africa to be about 13% of ents using the population projections. the global total of enteric CH4; Herrero et  al. The main mitigation research at ILRI began (2008) estimated the contribution to be about just after the publication of Livestock’s Long 10%. The differences are due largely to assump- Shadow (Steinfeld et al., 2006), building on earl- tions and inherent uncertainties in emissions ier research at ILRI on five factors – digestion, factor estimates, which suggested the need for manure, feed production, land conversion and more research to have better CH4 emissions esti- transport/processing – that contribute to rumi- mates and targeting of interventions to reduce nant-related GHG emissions. The goal was to emissions. develop spatially disaggregated livestock system In 2009, the IPCC GHG emissions taskforce data and better information on differential im- invited ILRI’s contribution to the emissions fac- pacts and emissions by system, species, region, tor database. ILRI’s contributions included both technology and country. biological research into emissions from cattle Herrero et  al. (2008) published the first production systems (e.g. Pelster et al., 2016) and study estimating emissions from African domes- a long-term collaboration with the GLOBIOM tic ruminants. This study combined country- model at IIASA. level calculations of changes in livestock Thornton and Herrero (2010b) estimated production due to population densities and cli- the potential for four interventions to reduce mate change with spatially explicit distributions GHG emissions from livestock: (i) adoption of of CH 4 emissions. The classification system built improved pastures; (ii) intensification of rumin- upon earlier ILRI efforts to better classify and ant diets; (iii) changes in land-use practices; and map livestock production systems (Kruska et al., (iv) changing breeds of ruminants. They esti- 2003; Kruska, 2006), accounting for differences mated reductions in emissions intensities, per in land areas, population densities, numbers of unit of milk or meat, and reductions in numbers livestock and diets for ruminants. Climate of animals (e.g. from improved productivity), as change was modelled as changes in LGP using well as carbon sequestered through the land the MarkSim model, which resulted in changes management options. Restoration of degraded in area under different production systems rangelands had the highest mitigation potential, (Thornton et al., 2006). For animal population followed by agroforestry, which both sequesters changes, national projections were made as- carbon and improves diet quality (and hence suming that production and productivity animal productivity). Improving breeds and 624 P. Ericksen, P. Thornton and G. Nelson grain supplementation had the lowest mitiga- The special issue of Proceedings of the Na- tion potentials. The total of all interventions tional Academy of Sciences USA used the first combined was a range of 6–12% reduction in biologically consistent, spatially disaggregated current livestock-related emissions (depending global data set of the main biophysical inter- on assumptions about adoption rates). actions among feed use, animal production and Herrero et  al. (2016) assessed three inter- GHG emissions. It highlighted three points: (i) ventions: (i) technical and management inter- feed-use efficiencies are a key driver of productiv- ventions; (ii) intensification and the associated ity and therefore of GHG emissions per unit of structural changes of livestock systems; and output; (ii) grasslands are a critical resource, (iii) moderation of demand for livestock products. which provide almost 50% of plant biomass for All such interventions have the technical poten- animals; and (iii) mixed crop–livestock systems tial to mitigate emissions from livestock, but produce over 60% of animal production across their economic potential may be far smaller due the world. to adoption costs on the supply side and a lack of CH 4 from enteric fermentation is the largest effective policies for promoting healthy levels of source of non-CO2 emissions, with cattle ac- consumption of livestock products (Fig. 16.4). counting for 77%. Developing world regions In 2013, a special issue of Proceedings of contribute 75% of the global emissions from the National Academy of Sciences USA was pub- livestock, and sub-Saharan Africa is a global lished, representing several years of intensive hotspot for high emissions intensities, driven by work to improve the modelling of heterogeneity low animal productivity per unit of land and in livestock system characteristics and their low-quality feeds, which extend the growing evolution, using spatially explicit data sets and periods of animals raised on grasslands or crop different assumptions by region about future residues. growth. Herrero et al. (2013) focused on differ- Herrero et  al. (2016) updated the 2013 ences among systems in land-use intensities analysis with new data on livestock production and GHG emissions. They concluded that these systems and on differences between technical differences showed potential for improvements mitigation potential and economic potential. in all tropical livestock systems, given their low First, they reviewed the major studies of GHG productivity. This study produced an innovative emissions from livestock, including both IPCC data set on biomass use, production, feed effi- emissions guidelines as well as life cycle assess- ciency, excretion and GHG emissions for 28 re- ments, focusing on uncertainties in the esti- gions, eight livestock production systems, four mates. They estimated that over the period animal species and three livestock products 1995–2005, annual global GHG direct and (Figs. 16.5–16.7). indirect emissions from livestock were 5.6–7.5 0.8 0.7 0.6 0.5 0.4 0.4 0.2 0.1 0 Carbon sequestration Improved feed Use of feed Avoided LUC due Animal Rangeland Carbon sequestration Manure due to improved digestibility additives to intensification management rehabilitation due to legume sowing management grazing management of ruminant systems Fig. 16.4. Mitigation potentials of supply-side measures. Red represents the range for each practice, where available. LUC, land-use change. (From Herrero et al., 2016.) Technical mitigation potential (GtCO2eq) Ruminant Livestock and Climate Change in the Tropics 625 (a) Bovine milk (b) Bovine meat EUR EUR OCE LGA OCE NAM LGH NAM LAM LGT LAM EAS MXA EAS SEA MXH SEA SAS MXTUrban SAS MNA Other MNA SSA SSA 0 50 100 150 200 250 0 5 10 15 Million t Million t (c) Small ruminant milk (d) Small ruminant meat EUR EUR OCE OCE NAM NAM LAM LAM EAS EAS SEA SEA SAS SAS MNA MNA SSA SSA 0 1 2 3 4 5 6 7 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Million t Million t Fig. 16.5. Bovine meat (a) and milk (b) production and small ruminant milk (c) and meat (d) by region. EUR, Europe; OCE, Oceania; NAM, North America; LAM, Latin America and the Caribbean; EAS, Eastern Asia; SEA, South-East Asia; SAS, South Asia; MNA, Middle East and North Africa; SSA, sub-Saharan Africa; LGA, livestock/grazing/arid; LGH, livestock/grazing/humid; LGT, livestock/grazing/ temperate; MXA, mixed/arid; MXH, mixed/humid; MXT, mixed/temperate; Other, other systems; Urban, urban systems. (From Herrero et al., 2013.) GtCO2eq. They then estimated emissions reduc- be 175, 200 and 225 tCO2eq. For measures tar- tions potential for several supply options, con- geting soil carbon sequestration in grazing cluding that these practices could help mitigate lands, higher mitigation levels of 250, 375 and between 0.01 and 0.5  GtCO2eq/year or about 750 tCO2eq/year were found. two-thirds of livestock emissions. The supply Most emissions results to date have been options included feed additives, improved feed derived from studies of temperate livestock and digestibility, manure management, soil carbon extrapolated to the tropics. Without more accur- sequestration in grasslands, animal productivity ate data, existing models used to calculate emis- and health, and avoided deforestation due to sions from smallholdings are more likely to give intensification unreliable estimates and, in turn, are less useful Demand-side options, discussed in greater for policy in the tropics (Rufino et al., 2014; Kim detail below, comprise a new agenda that has and Kirschbaum, 2015). In 2013, ILRI began gained traction in Europe and North America in collaboration with the Karlsruhe Institute of response to concerns that livestock production Technology in Germany to measure the global uses a disproportionate amount of land, emits environmental impacts of livestock production, significant GHGs and can have negative health in particular GHG emissions, in order to derive effects. The study assessed the potential for miti- better estimates under tropical conditions gation over the range of GHG taxes of US$20, (e.g. Zhou et  al., 2014a, near Lake Victoria; US$50 and US$100/tCO eq. The 2030 mitiga- Zhou et al., 2014b, for a wheat–maize rotation 2 tion potentials for these taxes were projected to in subtropical China; Pelster et  al., 2017, for 626 P. Ericksen, P. Thornton and G. Nelson (a) Grains (b) Grass EUR EUR OCE OCE NAM NAM LAM LAM EAS EAS SEA SEA SAS SAS MNA MNA SSA SSA 0 100 200 300 400 0 100 200 300 400 500 Million t Million t (c) Occasional feeds (d) Stover Dairy cattle EUR EUR Dairy small ruminant OCE OCE Other cattle NAM NAM Other small ruminant LAM LAM Pigs EAS EAS Poultry SEA SEA SAS SAS MNA MNA SSA SSA 0 50 100 150 200 0 50 100 150 200 250 300 Million t Million t Fig. 16.6. Regional estimates of feed production for grains (a), grass (b), occasional feeds (c) and stover (d). See Fig. 16.5 for abbreviations. (From Herrero et al., 2013.) c attle and tree fodder in Kenya; Rosenstock et al., The Mazingira facility responds to several 2016, for croplands in Kenya and Tanzania). analytical challenges. Little is known about cur- In 2012, ILRI researchers engaged with the rent baselines (e.g. Hickman et al., 2014 found CGIAR Research Programme on Climate Change, only 20 studies on N2O and NO fluxes from agri- Agriculture and Food Security (CCAFS) flagship culture in sub-Saharan Africa). Second, existing on Low Emissions Agriculture began collaborat- models have not accounted for all of the processes ing on SAMPLES. Rosenstock et  al. (2013) de- through which livestock emit GHGs (Rufino et al., scribed the SAMPLES protocol for improving data 2014). Third, time is needed for measurements quality and quantity from tropical smallholdings. to start and for data quality to be evaluated. An This protocol was based on five innovations: (i) initial SAMPLES study provided ‘the most com- systematic data collection; (ii) informed sampling prehensive study in Africa to date’ of annual from emissions hotspots; (iii) quantifying emis- in situ CO 2, CH4 and N2O emissions from soils in a sions at several spatial scales, including whole mixed crop–livestock system in western Kenya farm and landscape; (iv) using a multi-criteria ap- (Pelster et  al., 2017). The authors found that proach to link GHG emissions reductions with land classes did not make much difference in productivity gains; and (v) offering cost-differen- fluxes, nor did management because input use tiated measurements, depending on user needs. was so low. The lack of a management effect is ILRI established a modern environmental probably representative of most smallholdings laboratory in 2014. The Mazingira (Kiswahili in Africa, but the land class effect has not been for environment) Centre is the only facility in widely tested. A second study found that land Africa with the capacity for accurate measure- use and soil texture influenced GHG fluxes, al- ments of GHG emissions from soils, manure and though this study measured fluxes in the labora- ruminant digestion, using field and laboratory tory rather than in situ (Wanyama et al., 2018). measurements and analysis (https://mazingira. Pelster et  al. (2016) measured emissions from ilri.org; accessed 8 March 2020). excreta from cattle fed diets representative of Ruminant Livestock and Climate Change in the Tropics 627 (a) Ruminant GHG emissions (b) EUR LGA LGH OCE LGT MXA NAM MXHMXT LAM URBANOTHER EAS SEA SAS MNA SSA kg CO2eq/kg protein 10 25 50 1002505001000 0 100 200 300 400 500 Mt CO eq 2 (c) (d) Milk Meat 10000 10000 5000 LGA EUR 5000 LGH OCE LGT NAM MRA LAM MRH EAS 1000 MRT SEA 1000 SAS 500 MNA 500 SSA 100 100 50 50 10 10 8 9 10 11 12 8 9 10 11 12 ME (MJ/kg DM feed) ME (MJ/kg DM feed) Fig. 16.7. Global non-CO2 GHG emissions from ruminant livestock. (a) Non-CO2 GHG emissions from global ruminant livestock (cattle, sheep, and goats) by production system and region. (b) Spatial distribution of non-CO2 GHG emissions from ruminants (kg CO2eq/kg edible animal protein). (c, d) Relationship between diet quality of ruminants and the non-CO2 GHG emission intensity for edible animal protein from ruminant milk (c) and meat (d). DM, dry matter; see Fig. 16.5 for other abbreviations. (From Herrero et al., 2013.) East African conditions and found that CH4 and variation in live weight, feed sources and feed N2O emissions were lower than current IPCC es- availability. timates. The lower emissions were apparently Previous studies have shown that improv- due to the low nitrogen content of the excreta, ing dietary quality and quantity results in live reflecting the low nitrogen content of animal weight gains, which reduce emissions intensities diets in the sample. per unit of live weight. Feed quality is the key Another problem in establishing tropical factor influencing CH4 production from rumin- emissions baselines is seasonal variability in feed ant digestion as shown in a meta-analysis of ani- quality and supply. Goopy et al. (2018) defined a mal experimentation data (Hristov et al., 2013)1. method based on animal energy requirements, Blümmel et al. (2009, 2013) studied the poten- derived from field measurements of live weight, tial to reduce GHG emissions in India. Although milk production, locomotion and feed digestibil- the research emphasis was on use of crop res- ity. Emissions factors for annual CH 4 production idues to improve productivity, the India work were produced for three locations in western found that a collateral benefit could be reduc- Kenya (Ndung’u et al., 2018; Onyango, 2018). tions in GHG emissions intensities per unit of In all locations, the emissions factors per unit of output, and possibly a reduction in total emis- live weight by type of animal and agroecology sions per herd, if productivity gains allowed a re- differed from the current IPCC estimates due to duction in herd sizes. kg CO2eq/kg protein kg CO2eq/kg protein 628 P. Ericksen, P. Thornton and G. Nelson Demand-side options ‘total abatement calorie cost’ – than policies tar- geting emissions from livestock only. Much has recently been written about de- Revell (2015) used a partial equilibrium mand-side interventions (Garnett 2009; Smith model of beef, poultry, pig and sheep meats for et  al., 2013; Valin et  al., 2013). Springmann the major regions of the world to explore scen- et al. (2017) estimated the mitigation benefits of arios that might reduce meat consumption and a tax on foods whose production is GHG inten- GHG emissions. He concluded that economic sive and where current consumption levels in and population growth to 2050 without any some countries have negative health effects mitigation measures would lead to a 21% in- (Fig. 16.8). They found a double benefit from crease in per capita meat consumption and a this policy approach – a substantial reduction in 63% increase in total consumption and GHG GHG emissions, and health-promoting out- emissions by 2050. However, the mitigation pro- comes in middle- and high-income countries. jections from the scenarios generated only a Average GHG taxes on food commodities (based 14% reduction in cumulative emissions from the on an emissions tax of US$52/tCO eq) were baseline 2050 projections, insufficient to meet 2 highest for animal-sourced foods, such as beef the 2050 target of a 50% reduction in global (US$2.8/kg), lamb (US$1.3/kg), and pork and GHG emissions. poultry (US$0.3/kg each), which corresponded Schader et al. (2015) explored the scope for to 40%, 15%, 7% and 9% of the mean global sustainable livestock production by modelling producer prices of these commodities. the effects of a third strategy in which animal Springmann et  al. (2018) showed that be- feeds that compete with food production are re- tween 2010 and 2050, as a result of expected duced, and in an extreme scenario, animals are changes in population and income levels, the fed only from grasslands and by-products from environmental effects of the food system could food production. While the extreme scenario increase by 50–90% in the absence of techno- largely reduces animal protein per capita by logical changes and dedicated mitigation meas- some 70%, it could provide adequate energy ures. The same study also found that no single and proteins and reduce environmental impacts measure is enough to keep these effects within compared with a 2050 reference scenario as all planetary boundaries simultaneously, and f ollows: GHG emissions −18%, arable land oc- that a combination of measures is needed to cupation −26%, nitrogen surplus −46%, phos- sufficiently mitigate the projected increase in phorus surplus −40%, non-renewable energy e nvironmental pressures. use −36%, pesticide-use −22%, and freshwater Havlik et al. (2014) found that sustainable use −21%. intensification of livestock production systems White and Hall (2017) used the total re- might become a key climate-mitigation tech- moval of animals as the extreme boundary to nology. However, livestock production systems potential mitigation options and required the vary widely, making the implementation of cli- fewest assumptions to model the yearly nutri- mate-mitigation policies a costly challenge. They tional and GHG impacts of eliminating ani- projected that by 2030 autonomous transitions mals from US agriculture. Although modelled towards more efficient systems would de- plants-only agriculture produced 23% more crease emissions by 0.74 GtCO 2eq/year, mainly food, it met fewer of the US population's require- through avoided emissions from the conversion of ments for essential nutrients. When nutritional 162 million ha of natural land. A moderate miti- adequacy was evaluated by using least-cost diets gation policy targeting emissions from both the produced from the foods available, more nutri- agricultural and land-use change sectors with a ent deficiencies, a greater excess of energy and a carbon price of US$10/tCO2eq could lead to an need to consume a greater amount of food solids abatement of 3.22 GtCO2eq/year. Livestock sys- were encountered in plants-only diets. In the tem transitions would contribute 21% of the simulated system with no animals, estimated total abatement, intra- and interregional reloca- agricultural GHG decreased (28%) but did not tion of livestock production another 40% and all fully counterbalance the animal contribution of other mechanisms would add 39%. Mitigation GHG (49% in this model). This assessment sug- policies targeting emissions from land-use change gests that removing animals from US agriculture are five to ten times more efficient – measured in would reduce agricultural GHG emissions but Ruminant Livestock and Climate Change in the Tropics 629 (a) 12 (b) 45 (c) Price GHG tax Price Consumption 10 35 0.0 8 25 –0.2 6 15 Beef Milk 4 5 –0.4 Oils Lamb 2 –5 Rice –0.6 Poultry 0 –15 Pork Wheat –0.8 Vegetables Eggs Other –1.0 Fig. 16.8. Impacts of GHG taxes on food prices, consumption and GHG emissions. (a) Prices and GHG taxes by food commodity. (b) Percentage changes in price and consumption by food commodity. (c) Change in GHG emissions by food commodity and region. Regions include high-income countries (HICs) and the low- and middle-income countries of Africa (AFR), the USA (AMR),the Eastern Mediterranean (EMR), Europe (EUR), South-east Asia (SEA) and the Western Pacific (WPR), and an aggregate of all regions (World). Impacts are for a tax scenario in which GHG taxes are levied on all food commodities. (From Springmann et al., 2018.) P r i c e ( U S $ / k g ) L a m b B e e f P o r k P o u l t r y F r u i t s ( t e m p . ) L e g u m e s V e g e t a b l e s F r u i t s ( t r o p . ) R i c e O i l s M i l k O i l c r o p s S u g a r R o o t s W h e a t O t h e r g r a i n s M a i z e C h a n g e ( % ) B e e f O i l s M i l k L a m b P o u l t r y O t h e r g r a i n s R i c e W h e a t P o r k M a i z e E g g s O i l c r o p s V e g e t a b l e s S u g a r L e g u m e s R o o t s F r u i t s ( t r o p . ) F r u i t s ( t e m p . ) C h a n g e i n G H G e m i s s i o n s ( G t C O e q ) 2 W o r l d H I C A F R A M R E M R E U R S E A W P R 630 P. Ericksen, P. Thornton and G. Nelson would also create a food supply incapable of from animal production and animal product supporting the US population’s nutritional re- consumption; and (vii) extend field tests under quirements. tropical conditions of actual emission levels and possible reductions. Examples of the latter are pilot projects for Low Emissions Development The Future options (Ericksen and Crane, 2018; Kashangaki and Ericksen 2018). Future climate research priorities for tropical There has been less research on climate livestock have three components – mitigation, adaptation in tropical livestock than there has adaptation and policy. been on mitigation. Additional adaptation Research has established the mitigation research requires a broader view of adaptation potential of technical changes in the systems beyond technical change, involving changes in responsible for most GHG emissions from pro- behaviour, institutions and culture. Priorities for duction animals. The best-understood systems adaptation studies in tropical livestock systems are dairy and beef, which account for about include the following: 70% of GHG emissions from world livestock supply chains (Gerber et  al., 2013, p. 18). • More effort on the specific tropical problems Other work by Gerber et  al. (2011), on inten- of heat stress and animal performance, on sive dairying in a temperate climate, has estab- the genetics of reproduction under greater lished ranges of possible mitigation gains and heat stress, and on pests and diseases that the components – feed, genetics, health and do not exist in temperate climates. management – of such gains and the output • Improved capacity for surveillance of climate- costs of those changes. The lessons of this sensitive diseases, coupled with new diag- work are applicable as first approximations to nostics for these diseases (see Chapters 2, 3 mitigation paths for low-productivity dairying and 5–10, this volume). in the tropics, but more in situ measurements • An expanded programme of characteriz- from tropical systems are needed to sharpen ing, testing and disseminating perennial estimates of potential gains. Mitigation work forage species adapted to hotter, drier and on the supply side must rely less on the as- more variable climates (see Chapters 12 sumption that temperate data and models are and 13, this volume). directly transferable to tropical conditions and • Decision support tools to target adaptation instead will require greater focus on new find- programmes and to monitor their effects, ings under tropical conditions. including new measures of adaptation at The future of mitigation research is to: the household level. (i) estimate potential GHG reductions from less The models underpinning policy recom- well-studied tropical systems, such as extensive mendations for climate are inherently complex beef on pastures, intensive fattening on small- because of the number and scale of the climate, holdings and nutrient cycling in mixed crop– biological and behavioural relationships in- livestock farms; (ii) identify the components – feed volved. Policy recommendations from climate re- quality and management, animal genetics, health, search involving animals, in particular, require a overall herd management, and demand reduc- closer integration of supply- and demand- side tion – of potential GHG reductions; (iii) refine modelling because of the interactions between estimates of success probabilities from investiga- the two sides: tions of feed-use efficiency in the tropics; (iv) iden- tify profitability constraints, including policies, to • More research is needed on the policy in- adoption of potential technical changes; (v) ‘back- centives to promote broad adoption of cast’ projections from published models, not- mitigation and adaptation practices in the ably MarkSim and LGP-based work, into actual tropics, given the externality problems in- data to test the validity of these projections; volved in both. (vi) strengthen demand-side mitigation research • The literature comparing supply and de- in comparison with supply-side efforts to esti- mand measures is limited. Future modelling mate least-cost paths for emissions reductions by ILRI and partners must involve closer Ruminant Livestock and Climate Change in the Tropics 631 integration between supply and demand a ssistance to such risk-management components (e.g. Weindl et al., 2017). interventions as IBLI. • The dependence of arid rangelands on • In mixed crop–livestock systems, we also livestock demands an extended research have not assessed the impacts of production and policy effort that recognizes that shifts away from ruminants towards poultry technical options are limited (Ericksen on livelihoods and food security. et  al., 2012; Herrero et  al., 2016) for • Countries also need support to develop GHG mitigation, for adaptation and for protocols and data to monitor and report on raising productivity even (see Chapter their commitments to UNFCCC and to pre- 11, this volume, on the difficulties of pare credible investment plans. raising productivity from arid range- lands). Improved technical and institu- tional support for rangeland manage- Acknowledgements ment is needed to identify sustainable land management practices and to pro- The authors acknowledge the comments of John mote them. Related to this would be McIntire and Caroline Bosire on earlier drafts. Note 1 Most of the studies cited in the meta-analysis of Hristov et al. (2013) were conducted on ruminants con- suming significant shares of grain in their diets, which is still uncommon in African livestock production compared with other regions. References AFRC (1993) The Nutrient Requirments of Dairy Cattle. CAB International, Wallingford, UK. Alderman, H. and Haque, T. (2007) Insurance against covariate shocks: the role of index-based insurance in social protection in low-income countries of Africa. World Bank Working Paper No. 95. World Bank, Washington, D.C. Ash, A., Thornton, P., Stokes, C. and Togotohyn, C. (2012) Is proactive adaptation to climate change neces- sary in grazed rangelands? Rangeland Ecology & Management 65, 563–568. Barnett, B.J., Barrett, C.B. and Skees, J.R. (2008) Poverty traps and index-based risk transfer products. World Development 36, 1766–1785. Bett, B. Kiunga, P., Gachohi, J., Sindato, C., Mbotha, D., et al. (2017) Effects of climate change on the oc- currence and distribution of livestock diseases. Preventive Veterinary Medicine 137, 119–129. Binswanger, H.P. and Rosenzweig, M.R. (1986) Behavioural and material determinants of production relations in agriculture. Journal of Development Studies 22, 503–539. Blümmel, M., Tui, S.H.K., Valbuena, D., Duncan, A. and Herrero, M. (2013) Biomass in crop–livestock systems in the context of the livestock revolution. Secheresse 24, 330–339. Blümmel, M., Anandan, S. and Prasad, C.S. (2009) Potential and limitations of by-product based feeding systems to mitigate greenhouse gases for improved livestock productivity. In: Diversification of Animal Nutrition Research in the Changing Scenario: 13th Biennial Animal Nutrition Conference of the Ani- mal Nutrition Society of India, Vol. 1: Lead Papers. Animal Nutrition Society of India, Bangalore, India, p. 68–74. Boko, M., Niang, I., Nyong, A., Vogel, A., Githeko, A., et al. (2007) Africa Climate Change 2007: Impacts, Adaptation and Vulnerability: Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Boone, R.B., Conant, R.T., Sircely, J., Thornton, P.K. and Herrero, M. (2018) Climate change impacts on selected global rangeland ecosystem services. Global Change Biology 24, 1382–1393. Brammer, J.R., Samson, J. and Humphries, M.M. (2015) Declining availability of outdoor skating in Canada. Nature Climate Change 5, 2–4. 632 P. Ericksen, P. Thornton and G. Nelson Byrnes, R.C., Nùñez, J., Arenas, L., Rao, I., Trujillo, C., et al. (2017) Biological nitrification inhibition by Bra- chiaria grasses mitigates soil nitrous oxide emissions from bovine urine patches. Soil Biology and Biochemistry 107, 156–163. Cacho, O., Moss, J., Thornton, P.K., Herrero, M., Henderson, B. and Bodirsky, B. (2016) Adaptation path- ways for vulnerable areas. Background report for the State of Food and Agriculture 2016 (SOFA): climate change, agriculture and food security. FAO, Rome. Campbell, B.M., Vermeulen, S.J., Aggarwal, P.K., Corner-Dolloff, C., Girvetz, E., et al. (2016) Reducing risks to food security from climate change. Global Food Security 11, 34–43. Capper, J.L., Cady, R.A. and Bauman, D.E. (2009) The environmental impact of dairy production: 1944 compared with 2007. Journal of Animal Science 87, 2160–2167. Carter, M.R., Galarza, F.B. and Boucher, S.R. (2007) Underwriting area-based yield insurance to crowd-in credit supply and demand. Savings and Development 31, 335–362. Chantarat, S., Mude, A.G., Barrett, C.B. and Carter, M.R. (2013) Designing index-based livestock insurance for managing asset risk in northern Kenya. Journal of Risk and Insurance 80, 205–237. CIAT (1973) External review team: beef production systems program. CIAT, Cali, Colombia. Collier B., Skees J. and Barnett B. (2009) Weather index insurance and climate change: opportunities and challenges in lower income countries. The Geneva Papers on Risk and Insurance – Issues and Prac- tice 34, 401–424. Cooper, P., Vermeulen, S., Hansen, J., Thornton, P.K., Ramirez-Villegas, J., et al. (2014) Smallholder agri- culture and climate variability and change in sub-Saharan Africa: looking forward to 2050. In: Africa Agriculture Status Report 2014, Climate Change and Smallholder Agriculture in Sub-Saharan Africa. Alliance for a Green Revolution in Africa (AGRA). AGRA, Nairobi, pp. 18–50. de Leeuw, J., Osano, P., Said, M.Y., Ayantunde, A.A., Dube, S., et al. (2019) Pastoral farming systems and food security in Sub-Saharan Africa. Priorities for science and policy. In: Dixon, J., Garrity, D.P, Boffa, J.-M., Williams, T.O., Amede, T. et al. (eds) Farming Systems and Food Security in Africa: Priorities for Policies and Science under Global Change. EarthScan, Routledge, Oxford. de Sy, V., Herold, M., Achard, F., Beuchle, R., Clevers, J.G., et al. (2015) Land use patterns and related car- bon losses following deforestation in South America. Environmental Research Letters 10, 124004. de Vries, M. and de Boer, I.J.M. (2010) Comparing environmental impacts for livestock products: a review of life cycle assessments. Livestock Science 128, 1–11. Dinesh, D., Bett, B., Boone, R., Grace, D., Kinyangi, J., et al. (2019) Impact of climate change on African agriculture: focus on pests and diseases. CCAFS Info Note. CCAFS, Copenhagen. Dittrich, R., Werford, A., Topp, C.F.E., Eory, V. and Moran, D. (2017) A guide towards climate change adap- tation in the livestock sector: adaptation options and the role of robust decision-making tools for their economic appraisal. Regional Environmental Change 17, 1701–1712. Doherty, R., Sitch, S., Smith, B., Lewis, S. and Thornton, P. (2010) Implications of future climate and atmos- pheric CO2 content for regional biogeochemistry, biogeography and ecosystem services across East Africa. Global Change Biology 16, 617–640. Driscoll, D.A., Catford, J.A., Barney, J.N., Hulme, P.E., Inderjit, et al. (2014) New pasture plants intensify invasive species risk. Proceedings of the National Academy of Sciences USA 111, 16622–16627. Ericksen, P. and Crane, T. (2018) The feasibility of low emissions development interventions for the East African livestock sector: lessons from Kenya and Ethiopia. ILRI, Nairobi. Ericksen, P., de Leeuw, J., Thornton, P., Notenbaert, A., Cramer, L., et al. (2012). Climate change in sub-Saharan Africa: what consequences for pastoralism? In: Catley, A., Lind, J., and Scoones, I. (eds) Pastoralism and Development in Africa: Dynamic Change at the Margins. EarthScan, London, pp. 71–81. Ericksen, P., Thornton, P., Notenbaert, A., Cramer, L., Jones, P. and Herrero, M. (2011) Mapping hotspots of climate change and food insecurity in the global tropics. CCAFS Report No. 5. CCAFS, Copenhagen. Garnett, T., Appleby, M.C., Balmford, A., Bateman, I.J., Benton, T.G. et al. (2013) Sustainable intensification in agriculture: premises and policies. Science 341, 33–34. Garnett, T. (2009) Livestock-related greenhouse gas emissions: impacts and options for policy makers. Environmental Science and Policy 12, 491–503. Gerber, P., Vellinga, T., Opio, C. and Steinfeld, H. (2011) Productivity gains and greenhouse gas emissions intensity in dairy systems. Livestock Science, 139, 100–108. Gerber, P.J., Steinfeld, H. and Henderson, B. (2013) Tackling Climate Change Through Livestock: A Global Assessment of Emissions and Mitigation Opportunities. FAO, Rome. Ghahramani, A. and Moore, A.D. (2013) Climate change and broadacre livestock production across south- ern Australia. 2. Adaptation options via grassland management. Crop and Pasture Science 64, 615. Ruminant Livestock and Climate Change in the Tropics 633 Goopy, J.P., Onyango, A.A., Dickhoefer, U. and Butterbach-Bahl, K. (2018) A new approach for improving emission factors for enteric methane emissions of cattle in smallholder systems of East Africa – results for Nyando, Western Kenya. Agricultural Systems 161, 72–80. Grace, D. Bett, B., Lindahl, J. and Robinson, T. (2015) Climate and livestock disease: assessing the vulner- ability of agricultural systems to livestock pests under climate change scenarios. CCAFS Working Paper No. 116. CCAFS, Copenhagen, Denmark. Haileselassie, A., Peden, D., Gebreselassie, S., Amede, T. and Descheemaeker, K. (2009) Livestock water productivity in mixed crop–livestock farming systems of the Blue Nile basin: Assessing variability and prospects for improvement. Agricultural Systems 102, 33–40. Havlik, P., Valin, H., Herrero, M., Obersteiner, M., Schmid, E. et al. (2014) Climate change mitigation through livestock system transitions. Proceedings of the National Academy of Sciences USA 111, 3709–3714. Henricksen, B.L. and Durkin, J.W. (1986) Growing period and drought early warning in Africa using satellite data. International Journal of Remote Sensing 7, 1583–1608. Herrero, M. (1997) Modelling dairy grazing systems: an integrated approach. PhD thesis, University of Edinburgh, Edinburgh, UK. Herrero, M. and Thornton, P.K. (2013) Livestock and global change: Emerging issues for sustainable food systems. Proceedings of the National Academy of Sciences USA 110, 20878–20881. Herrero, M., Thornton, P.K. Kruska, R. and Reid, R.S. (2008) Systems dynamics and the spatial distribu- tion of methane emissions from African domestic ruminants to 2030. Agriculture, Ecosystems and Environment 126, 122–137. Herrero, M., Thornton, P.K., Nouala, S., Notenbaert, A., Ericksen, P., et al. (2010) Coping with drought and climate change in the pastoral sector in Sub-Saharan Africa: policy considerations. Paper presented to the 8th Meeting of Ministers responsible for Animal Resources in Africa, 10–11 May, Entebbe, Uganda. ILRI, Nairobi. Herrero, M., Havlik, P., Valin, H., Notenbaert, A., Rufino, M.C., et al. (2013) Biomass use, production, feed efficiencies, and greenhouse gas emissions from global livestock systems. Proceedings of the National Academy of Sciences USA 110, 20888–20893. Herrero, M., Thornton, P.K., Bernués, A., Baltenweck, I., Vervoort, J., et al. (2014) Exploring future changes in smallholder farming systems by linking socio-economic scenarios with regional and household models. Global Environmental Change 24, 165–182. Herrero, M.T., Wirsenius, S., Henderson, B.B., Rigolot, C., Thornton, P.K., et al. (2015) Livestock and the environment: What have we learned in the past decade? Annual Review of Environment and Resources 40, 177–202. Herrero, M., Henderson, B., Havlik, P., Thornton, P.K., Conant, R.T., et al. (2016) Greenhouse gas mitigation potentials in the livestock sector. Nature Climate Change 6, 452–461. Hickman, J.E., Scholes, R.J., Rosenstock, T.S., Pérez García-Pando, C. and Nyamangara, J. (2014) Assessing non-CO2 climate-forcing emissions and mitigation in sub-Saharan Africa. Current Opinion in Environmental Sustainability 9–10, 65–72. Hobbs, N.T., Galvin, K.A., Stokes, C.J., Lackett, J.M., Ash, A.J., et al. (2008) Fragmentation of rangelands: implications for humans, animals, and landscapes. Global Environmental Change 18, 776–785. Hooda, P.S., Edwards, A.C., Anderson, H.A. and Miller, A. (2000) A review of water quality concerns in live- stock farming areas. Science of the Total Environment 250 143–167. Hristov, A.N., Oh, J., Firkins, J.L., Dijkstra, J., Kebreab, E., et al. (2013) Special topics – mitigation of me- thane and nitrous oxide emissions from animal operations: I. A review of enteric methane mitigation options. Journal of Animal Science 91, 5045–5069. IPCC (2007) Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, New York. IPCC (2014) Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth As- sessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, New York. Jensen, N.D. and Barrett, C.B. (2016) Agricultural index insurance for development. Applied Economic Perspectives and Policy 39, ppw022. Jensen, N.D., Barrett, C.B. and Mude, A.G. (2017) Cash transfers and index insurance: A comparative impact analysis from northern Kenya. Journal of Development Economics 129, 14–28. Jones, P. and Thornton, P. (1993) A rainfall generator for agricultural applications in the tropics. Agricultural and Forest Meteorology 63, 1–19. 634 P. Ericksen, P. Thornton and G. Nelson Jones, P. and Thornton, P. (1997) Spatial and temporal variability of rainfall related to a third-order Markov model. Agricultural and Forest Meteorology 86, 127–138. Jones, P. and Thornton, P. (1999) Fitting a third-order Markov rainfall model to interpolated climate surfaces. Agricultural and Forest Meteorology 97, 213–231. Jones, P. and Thornton, P. (2000) MarkSim: Software to Generate Daily Weather Data for Latin America and Africa. Agronomy Journal 92, 445–453. Jones, P. and Thornton, P. (2002) Spatial modeling of risk in natural resource management. Conservation Ecology 5, 27. Jones, P. and Thornton, P. (2003) The potential impacts of climate change on maize production in Africa and Latin America in 2055. Global Environmental Change 13, 51–59. Jones, P.G. and Thornton, P.K. (2009) Croppers to livestock keepers: livelihood transitions to 2050 in Africa due to climate change. Environmental Science & Policy 12, 427–437. Jones, P.G. and Thornton, P.K. (2013) Generating downscaled weather data from a suite of climate models for agricultural modelling applications. Agricultural Systems 114, 1–5. Jones, P.G. and Thornton, P.K. (2015) Representative soil profiles for the Harmonized World Soil Database at different spatial resolutions for agricultural modelling applications. Agricultural Systems 139, 93–99. Jones, P.G., Thornton, P.K., Díaz, W., Wilkens, P.W., Jones, A.L. (eds) (2002) MarkSim: a computer tool that generates simulated weather data for crop modeling and risk assessment. CD-ROM. CIAT, Cali, Columbia. Karwat, H., Moreta, D., Arango, J., Núñez, J., Rao, I., et al. (2017) Residual effect of BNI by Brachiaria humid- icola pasture on nitrogen recovery and grain yield of subsequent maize. Plant and Soil 420, 389–406. Kashangaki, J. and Ericksen, P. (2018) Cost–benefit analysis of fodder production as a low emissions de- velopment strategy for the Kenyan dairy sector. ILRI, Nairobi. Kim, D.G. and Kirschbaum, M.U. (2015) The effect of land-use change on the net exchange rates of green- house gases: a compilation of estimates. Agriculture, Ecosystems & Environment 208, 114–126. King, J.M. (1983) Livestock water needs in pastoral Africa in relation to climate and forage. ILCA Research Report No 7. ILCA, Addis Ababa. Kristjanson, P.M., Thornton, P.K., Kruska, R.L., Reid, R., Henninger, N., et al. (2004) Mapping livestock systems and changes to 2050: implications for West Africa. In: Williams, T.O., Tarawali, S.A., Hiernaux, P. and Fernandez-Rivera, S. (eds) Sustainable Crop–Livestock Production for Improved Livelihoods and Natural Resource Management in West Africa. Proceedings of an International Conference held at the International Institute of Tropical Agriculture (IITA), 19–22 November 2001, Ibadan, Nigeria. ILRI, Nairobi, and CTA, Wageningen, The Netherlands, pp. 28–44. Kruska, R.L. (2006) Mapped Seré & Steinfeld livestock production systems for Africa, version 3. CD-ROM. ILRI, Nairobi. Kruska, R.L., Reid, R.S., Thornton, P.K., Henninger, N. and Kristjanson, P.M. (2003) Mapping livestock- oriented agricultural production systems for the developing world. Agricultural Systems 77, 39–63. LGACC (2018) Multilevel dialogue and planning for improving rangelands in northern Kenya. LGACC Policy Brief No. 1. ILRI, Nairobi. Lipper, L., Thornton, P., Campbell, B.M., Baedeker, T., Braimoh, A., et al. (2014) Climate-smart agriculture for food security. Nature Climate Change 4, 1068–1072. Lobell, D.B., Burke, M.B., Tebaldi, C., Mastrandrea, M.D., Falcon, W.P. and Naylor, R.L. (2008) Prioritizing climate change adaptation needs for food security in 2030. Science 319, 607–610. Mahul, O. and Stutley, C. (2010) Government Support to Agricultural Insurance: Challenges and Options for Developing Countries. World Bank, Washington, D.C. Mallin, M.A. and Cahoon, L.B. (2003) Industrialized animal production – a major source of nutrient and mi- crobial pollution to aquatic ecosystems. Population and Environment 24, 369–385. McDermott, J.J., Kristjanson, P.M., Kruska, R.L., Reid, R.S., Robinson T.P., et al. (2001) Effects of climate, human population and socio-economic changes on tsetse-transmitted trypanosomosis to 2050. In: Black, S.J. and Seed, J.R. (eds) World Class Parasites, Vol. 1. The African Trypanosomes. Kluwer Aca- demic Publishers, Boston, pp. 25–38. Mekonnen, M.M. and Hoekstra, A.Y. (2010) The green, blue and grey water footprint of farm animals and animal products. Unesco-IHE Institute for Water Education, Delft, Netherlands. Messina, J.P., Kraemer, M.U., Brady, O.J., Pigott, D.M., Shearer, F.M., et al. (2016) Mapping global environ- mental suitability for Zika virus. Elife 5, e15272. Moore, A.D. and Ghahramani, A. (2013) Climate change and broadacre livestock production across south- ern Australia. 1. Impacts of climate change on pasture and livestock productivity, and on sustainable levels of profitability. Global Change Biology 19, 1440–1455. Ruminant Livestock and Climate Change in the Tropics 635 Moore, N., Alagarswamy, G., Pijanowski, B.C., Thornton, P.K., Lofgren, B., et al. (2012) East African food security as influenced by future climate change and land use change at local to regional scales. Climatic Change 110, 823–844. Moreta, D.E., Arango, J., Sotelo, M., Vergara, D., Rincón, A. et al. (2014) Biological nitrification inhibition (BNI) in Brachiaria pastures: A novel strategy to improve eco-efficiency of crop-livestock systems and to mitigate climate change. Tropical Grasslands-Forrajes Tropicales 2, 88–91. Ndung’u, P., Bebe, B.O., Ondiek, J.O., Butterbach-Bahl, K., Merbold, L. and Goopy, J.P. (2018) Improved region-specific emission factors for enteric methane emissions from cattle in smallholder mixed crop:livestock systems of Nandi County, Kenya. Animal Production Science 59, 1136–1146. Nelson, G., Bogard, J., Lividini, K., Arsenault, J., Riley, M., et al. (2018) Income growth and climate change effects on global nutrition security to mid-century. Nature Sustainability 1, 773–781. New, M., Anderson, K., Fung, F. and Thornton, P. (2011) Migration and global environmental change: SR8: the possible impacts of high levels of climate change in 2060 and implications for migration. Report for the UK Government Foresight Project on Migration and Global Environmental Change. Government Office for Science, London. Notenbaert, A., Thornton, P. and Herrero, M. (2007) Livestock development and climate change in Turkana District, Kenya. ILRI Targeting and Innovation Discussion Paper No. 7. ILRI, Nairobi. Nuñez, J., Arevalo, A., Karwat, H., Egenolf, K., Miles, J., et al. (2018) Biological nitrification inhibition activity in a soil-grown biparental population of the forage grass, Brachiaria humidicola. Plant and Soil 426, 401–411. Olson, J.M., Alagarswamy, G., Andresen, J.A., Campbell, D.J., Davis, A.Y., et al. (2008) Integrating diverse methods to understand climate–land interactions in East Africa. Geoforum 39, 898–911. Olwoch, J.M., Reyers, B., Engelbrecht, F.A. and Erasmus, B.F.N. (2008) Climate change and the tick- borne disease, Theileriosis (East Coast fever) in sub-Saharan Africa. Journal of Arid Environments 2, 108–120. Onyango, A.A. (2018) Contribution of smallholder ruminant livestock farming to enteric methane emissions in Lower Nyando, Western Kenya. PhD thesis, University of Hohenheim, Stuttgart, Germany. Ortiz-Colón, G., Fain, S.J., Parés, I.K., Curbelo-Rodríguez, J., Jiménez-Cabán, E., et al. (2018) Assessing climate vulnerabilities and adaptive strategies for resilient beef and dairy operations in the tropics. Climatic Change 146, 47–58. Palazzo, A., Vervoot, J., Mason-d’Croz, D., Rutting, L., Havlik, P., et al. (2017) Linking regional stakeholder scenarios and shared socioeconomic pathways: quantified West African food and climate futures in a global context. Global Environmental Change 45, 227–242. Patra, A.K. (2012) Estimation of methane and nitrous oxide emissions from Indian livestock. Journal of En- vironmental Monitoring 14, 2673. Peden, D., Tadesse, G., Misra, A.K., Awad Amed, F., Astatke, A., et al. (2007) Water and livestock for human development. In: Molden, D. (ed) Water for Food, Water for Life: A Comprehensive Assessment of Water Management in Agriculture. Oxford University Press, Oxford, pp. 485–514. Pelster, D.E., Gisore, B., Goopy, J., Korir, D., Koske, J.K., et al. (2016) Methane and nitrous oxide emissions from cattle excreta on an East African grassland. Journal of Environment Quality 45, 1531. Pelster, D.E., Rufino, M., Rosenstock, T., Mango, J., Saiz, G., et al. (2017) Smallholder farms in eastern African tropical highlands have low soil greenhouse gas fluxes. Biogeosciences 14, 187–202. Porter, J.R., Xie, L., Challinor, A.J., Cochrane, K., Howden, S.M., et al. (2014) Food security and food pro- duction systems. In: Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J., Mastrandrea, M.D., et al. (eds) Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge/New York, pp. 485–533. Proença, V., Aguiar, C. and Domingos, T. (2015) Highly productive sown biodiverse pastures with low inva- sion risk. Proceedings of the National Academy of Sciences USA 112, E1695. Rassmann, K. and Schuetz, T. (2017) An assessment of the influence of CCAFS’ climate data and tools on out- comes achieved 2010–2016. Report from an Outcome Harvesting evaluation. CCAFS, Eurasburg, Germany. Revell, B.J. (2015) One man’s meat … 2050? Ruminations on future meat demand in the context of global warming. Journal of Agricultural Economics 66, 573–614. Rigolot, C., de Voil, P., Douxchamps, S., Prestwidge, D., van Wijk, M., et al. (2017) Interactions between intervention packages, climatic risk, climate change and food security in mixed crop–livestock sys- tems in Burkina Faso. Agricultural Systems 151, 217–224. Ripple, W.J., Smith, P., Baberl, H., Montzka, S.A., McAlpine, C. and Boucher, D.H. (2013) Ruminants, climate change and climate policy. Nature Climate Change 4, 2–5. 636 P. Ericksen, P. Thornton and G. Nelson Rivera-Ferre, M.G., López-i-Gelats, F., Howden, M., Smith, P., Morton, J.F. and Herrero, M. (2016) Re-framing the climate change debate in the livestock sector: mitigation and adaptation options. WIREs Climate Change 7, 869–892. Robinson, T.P., Thornton, P.K., Franceschini, G., Kruska, R.L., Chiozza, F., et al. (2011) Global livestock production systems. FAO, Rome, and ILRI, Nairobi. Rojas-Downing, M.M., Nejadhashemi, A.P., Harrigan, T. and Woznicki, S.A. (2017) Climate change and live- stock: Impacts, adaptation, and mitigation. Climate Risk Management 16, 145–163. Rose, H., Hoar, B., Kutz, S.J. and Morgan, E.R. (2014) Exploiting parallels between livestock and wildlife: predicting the impact of climate change on gastrointestinal nematodes in ruminants. International Journal for Parasitology: Parasites and Wildlife 3, 209–219. Rosegrant, M.W., Fernandez, M., Sinha, A., Alder, J., Ahammad, H., et al. (2009) Looking into the future for agriculture and KST. In: McIntyre, B.D., Herren, H.R., Wakhungu, J. and Watson, R.T. (eds) Inter- national Assessment of Agricultural Knowledge, Science and Technology for Development (IAASTD): Agriculture at a Crossroads, Global Report. Island Press, Washington, D.C., pp. 307–376. Rosenstock, T.S., Rufino, M.C., Butterbach-Bahl, K. and Wollenberg, E. (2013) Toward a protocol for quan- tifying the greenhouse gas balance and identifying mitigation options in smallholder farming systems. Environmental Research Letters 8, 021003. Rosenstock, T.S., Ruffino, T.S., Butterbach-Bahl, K., Wollenberg, E. and Richars, M. (eds) (2016) Methods for Measuring Greenhouse Gas Balances and Evaluating Mitigation Options in Smallholder Agricul- ture. Springer, Cham, Switzerland. Rufino, M.C., Thornton, P.K., Ng’ang’a, S.K., Mutie, I., Jones, P.G., et al. (2013) Transitions in agro-pastoralist systems of East Africa: Impacts on food security and poverty. Agriculture, Ecosystems & Environment 179, 215–230. Rufino, M.C., Brandt, P., Herrero, M. and Butterbach-Bahl, K. (2014) Reducing uncertainty in nitrogen budgets for African livestock systems. Environmental Research Letters 9, 105008. Samy, A.M. and Peterson, A.T. (2016) Climate change influences on the global potential distribution of blue- tongue virus. PLoS One 11, e0150489. Sandford, S. (1983) Organization and management of water supplies in tropical Africa. ILCA Research Report No. 8. ILCA, Addis Ababa. Schader, C., Muller, A., Scialabba, N.-H., Hecht, J., Isensee, A., et  al. (2015) Impacts of feeding less food-competing feedstuffs to livestock on global food system sustainability. Journal of The Royal Soci- ety Interface 12, 20150891. Scherer, L. and Verburg, P.H. (2017) Mapping and linking supply- and demand-side measures in climate- smart agriculture. A review. Agronomy for Sustainable Development 37, 66. Schmidhuber, J. and Tubiello, F.N. (2007) Global food security under climate change. Proceedings of the National Academy of Sciences USA 104, 19703–19708. Searchinger, T.D., Estes, L., Thornton, P.K., Beringer, T., Notenbaert, A., et al. (2015) High carbon and bio- diversity costs from converting Africa’s wt savannahs to cropland.Nature Climate Change 5, 481–486. Seré, C. and Steinfeld, H. (1996) World livestock production systems: current status, issues and trends. FAO Animal Production and Health Paper No. 127. FAO, Rome. Shields, S. and Orme-Evans, G. (2015) The impacts of climate change mitigation strategies on animal welfare. Animals 5, 361–394. Smith, P., Haberl, H., Popp, A., Erb, K.H., Lauk, C., et al. (2013) How much land-based greenhouse gas mitigation can be achieved without compromising food security and environmental goals? Global Change Biology 19, 2285–2302. Smith P., Bustamante, M., Ahammad, H., Clark, H., Dong, H. et al. (2014) Agriculture, forestry and other land use (AFOLU). In: Edenhofer, O., Pichs-Madruga, R., Sokona, Y., Farahani, E., Kadner, S., et al. (eds) Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge/New York, pp. 811–922. Sniffen, C.J., O’Connor, J.D., van Soest, P.J., Fox, D.G. and Russell, J.B. (1992) A net carbohydrate and protein system for evaluating cattle diets: II. Carbohydrate and protein availability. Journal of Animal Science 70, 3562–3577. Springmann, M., Mason-d’Croz, D., Robinson, S., Wiebe, K., Godfray, H.C.J., et al. (2017) Mitigation potential and global health impacts from emissions pricing of food commodities. Nature Climate Change 7, 69–74. Springmann, M., Clark, M., Mason-d’Croz, D., Wiebe, K., Bodirsky, B.L., et al. (2018) Options for keeping the food system within environmental limits. Nature 562, 519–525. Ruminant Livestock and Climate Change in the Tropics 637 Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M. and de Haan, C. (2006) Livestock’s Long Shadow: Environmental Issues and Options. FAO, Rome. Subbarao, G.V., Kishii,M., Nakahara, K., Ishikawa, T., Ban, T., et al. (2009) Biological nitrification inhibition (BNI) – is there potential for genetic interventions in the Triticeae? Breeding Science 59, 529–545. Subbarao, G.V., Arango, J., Masahiro, K., Hooper, A.M., Yoshihashi, T., et al. (2017) Genetic mitigation strat- egies to tackle agricultural GHG emissions: the case for biological nitrification inhibition technology. Plant Science 262, 165–168. Svinurai, W., Mapanda, F., Sithole, D., Moyo, E.N., Ndidzano, K., et al. (2018) Enteric methane emissions and their response to agro-ecological and livestock production systems dynamics in Zimbabwe. Sci- ence of the Total Environment 616–617, 710–719. Teutscherova, N., Vazquez, E., Arango, J., Arevalo, A., Benito, M. and Pulleman, M. (2019) Native arbuscu- lar mycorrhizal fungi increase the abundance of ammonia-oxidizing bacteria but suppress nitrous oxide emissions shortly after urea application. Geoderma 338, 493–501. Thornton, P.K. (2009) Climate change assessment in the Albertine Rift: climate model data downscaling, and crop model simulation. Consultant’s Report to the Wildlife Conservation Society, New York. Thornton, P.K. (2010) Livestock production: recent trends, future prospects. Philosophical Transactions of the Royal Society B: Biological Sciences 365, 2853–2867. Thornton, P.K. and Cramer, L. (eds) (2012) Impacts of climate change on the agricultural and aquatic sys- tems and natural resources within the CGIAR’s mandate. CCAFS Working Paper No. 23. CCAFS, Copenhagen. Thornton, P.K. and Gerber, P. (2009) Will climate change constrain growth of the livestock sector? A back- ground paper for the SOFA 2009. ILRI, Nairobi, and FAO, Rome. Thornton, P.K. and Herrero, M. (2010a) The inter-linkages between rapid growth in livestock production, climate change, and the impacts on water resources, land use, and deforestation. Background paper for the 2010 World Development Report. World Bank Policy Research Working Paper WPS 5178. ILRI, Nairobi. Thornton, P.K. and Herrero, M. (2010b) Potential for reduced methane and carbon dioxide emissions from livestock and pasture management in the tropics. Proceedings of the National Academy of Sciences USA 107, 19667–19672. Thornton, P.K. and Herrero, M. (2015) Adapting to climate change in the mixed crop and livestock farming systems in sub-Saharan Africa. Nature Climate Change 5, 830–836. Thornton, P.K. and Lipper, L. (2014) How does climate change alter agricultural strategies to support food security? IFPRI Discussion Paper 01340. IFPRI, Washington, D.C. Thornton, P.K., Kruska, R.L., Henninger, N., Kristjanson, P.M., Reid, R.S., et al. (2002) Mapping poverty and livestock in the developing world. ILRI, Nairobi. Thornton, P.K., Jones, P.G., Owiyo, T.M., Kruska, R.L., Herrero, M., et al. (2006) Mapping climate change vulnerability and poverty in Africa. ILRI, Nairobi. Thornton, P.K., Herrero, M., Freeman, A., Okeyo, M., Rege, R.E.O., et  al. (2007) Vulnerability, climate change and livestock – research opportunities and challenges for poverty alleviation. Journal of Semi-Arid Tropical Agricultural Research 4, 26520771. Thornton, P.K., van de Steeg, J., Notenbaert, A. and Herrero, M. (2008) The livestock–climate–poverty nexus: a discussion paper on ILRI research in relation to climate change. ILRI Discussion Paper 11. ILRI, Nairobi. Thornton, P.K., van de Steeg, J., Notenbaert, A. and Herrero, M. (2009) The impacts of climate change on livestock and livestock systems in developing countries: a review of what we know and what we need to know. Agricultural Systems 101, 113–127. Thornton, P.K., Jones, P.G., Alagarswamy, G., Andresen, J. and Herrero, M. (2010) Adapting to cli- mate change: agricultural system and household impacts in East Africa. Agricultural Systems 103, 73–82. Thornton, P.K., Jones, P.G., Ericksen, P.J. and Challinor, A.J. (2011) Agriculture and food systems in sub-Saharan Africa in a 4°C+ world. Philosophical Transactions of the Royal Society A: Mathematical Physical and Engineering Sciences 369, 117–136. Thornton, P.K., Ericksen, P.J., Herrero, M. and Challinor, A.J. (2014) Climate variability and vulnerability to climate change: a review. Global Change Biology 20, 3313–3328. Thornton, P.K., Boone, R.B. and Ramirez-Villegas, J. (2015) Climate change impacts on livestock. CCAFS Working Paper No. 120. Thornton, P.K., Rosenstock, T., Lamanna, C., Bell, P., Förch, W., et al. (2016) Evaluating climate smart adaptation options in mixed crop–livestock systems in developing countries. In: Lipper, L., McCarthy, 638 P. Ericksen, P. Thornton and G. Nelson N., Zilberman, D., Asfaw, S. and Branca, G. (eds) Climate Smart Agriculture: Building Resilience to Climate Change. Springer, New York. Valin, H., Havlík, P, Mosnier, A., Herrero, M., Schmid, E. and Obersteiner, M. (2013) Agricultural product- ivity and greenhouse gas emissions: trade-offs or synergies between mitigation and food security? Environmental Research Letters 8, 035019. van de Steeg, J.A., Herrero, M., Kinyangi, J., Thornton, P.K., Rao, K.P.C., et al. (2009) The influence of climate variability and climate change on the agricultural sector in East and Central Africa – sensitiz- ing the ASARECA strategic plan to climate change. Research Report No. 22. ILRI, Nairobi. van Vuuren, D.P., Ochola, W.O., Riha, S., Giampietro, M., Ginzo, H., et al. (2009) Outlook on agricultural change and its drivers. In: McIntyre, B.D., Herren, H.R., Wakhungu, J. and Watson, R.T. (eds) Agri- culture at a Crossroads. Island Press, Washington, D.C, pp. 255–305. Wanyama, I., Rufino, M.C., Pelster, D.E., Wanyama, G., Atzberger, C., et al. (2018) Land use, land use history, and soil type affect soil greenhouse gas fluxes from agricultural landscapes of the East Afri- can highlands. Journal of Geophysical Research: Biogeosciences 123, 976–990. Weindl, I., Lotze-Campen, H., Popp, A., Müller, C., Havlik, P., et al. (2015) Livestock in a changing climate: production system transitions as an adaptation strategy for agriculture. Environmental Research Let- ters 10, 094021. Weindl, I., Popp, A., Bodirsky, B.L., Rolinski, S., Lotze-Campen, H., et al. (2017) Livestock and human use of land: productivity trends and dietary choices as drivers of future land and carbon dynamics. Global and Planetary Change 159, 1–10. White, R.R. and Hall, M.B. (2017) Nutritional and greenhouse gas impacts of removing animals from US agriculture. Proceedings of the National Academy of Sciences USA 114, E10301–E10308. Zhou, M., Brandt, P., Pelster, D., Rufino, M.C., Robinson, T. and Butterbach-Bahl, K. (2014a) Regional nitrogen budget of the Lake Victoria Basin, East Africa: syntheses, uncertainties and perspectives. Environmental Research Letters 9, 105009. Zhou, M., Zhu, B., Brüggemann, N., Bergmann, J., Wang, Y. and Butterbach-Bahl, K. (2014b) N2O and CH4 emissions, and NO − 3 leaching on a crop-yield basis from a subtropical rain-fed wheat–maize rotation in response to different types of nitrogen fertilizer. Ecosystems 17, 286–301. 17 Economics and Policy Research at ILRI, 1975–2018 Mohammad Jabbar1, Steve Staal2, John McIntire3 and Simeon Ehui4 1Dhaka, Bangladesh; 2International Livestock Research Institute, Nairobi, Kenya; 3Santa Barbara, California, USA; 4World Bank, Washington, DC, USA Contents Executive Summary 640 Goals of economics and policy research 640 Research spending 640 Scientific impacts 640 Development impacts 641 Introduction 641 Policy Problems 642 The historical problem of supply response 642 Markets, institutions and competitiveness 648 Ruminants 648 Pigs and poultry 650 Animal health services and productivity 653 Responding to the ‘Livestock Revolution’ 655 Policy and technical barriers to smallholder dairy development 656 Dairy reform in Ethiopia 656 Dairy reform in Kenya 658 Comparisons of dairying in South Asia and East Africa 660 Land rights 661 Pastoral systems 662 Land tenure, resource allocation and productivity 663 Collective action for common resource management 664 Land tenure and fodder trees 666 Livestock and poverty 666 Food security and nutrition 667 IBLI in the arid rangelands of Kenya and Ethiopia 668 Impact 669 Policy lessons 669 Livestock sector analyses and master plans as part of development policies 669 The Future 670 References 671 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 639 640 M. Jabbar et al. Executive Summary • Definition of a path from research to devel- opment impact known as ‘Livestock – A Goals of economics and policy Pathway out of Poverty’. research • Contributions to global knowledge about distortions in agricultural incentives via a The goals of livestock policy and economics re- wide range of livestock system, value-chain, search at the International Livestock Research trade and incentive studies. Institute (ILRI) have been to increase small- • The notable scientific achievement compris- holder returns from animal agriculture by: (i) ing the joint effort of ILRI, the International analysing the productivity and targeting of live- Food Policy Research Institute (IFPRI) and stock-based technologies; (ii) identifying policy the Food and Agriculture Organization of barriers that lower farm prices, raise input costs the United Nations (FAO) to produce a study or lower the financial, information, and risk of global livestock trends: ‘Livestock to costs of new agricultural innovations; (iii) sup- 2020: the next food revolution’ (Delgado porting institutions that improve productivity, et al., 1999). create assets and improve the performance of • Informed pro-poor regulatory systems for value chains; and (iv) creating a policy and improved dairying in Kenya and Ethiopia, regulatory environment that allows animal for live-animal marketing in Sudan, and for agriculture to contribute to growth and poverty better structuring and financing of animal reduction. health in various livestock systems. • Using remote sensing and household sur- vey data, development of an index-based livestock insurance product, achieving Research spending wide scientific impact through the many well-cited papers produced from that data International Livestock Centre for Africa in Kenya and Ethiopia. (ILCA) spending on economics and policy re- • Developing new methods: incorporating search was some US$48 million from 1975 to geographic information systems (GIS)- 1994, or 13% of the ILCA total of US$374 mil- derived variables in econometric analysis, lion. The sum of livestock systems research, integrating models through farm, sector which often had a policy component, plus eco- and global levels, including participatory nomics and policy research during the ILCA approaches in field investigations and in the era was US$120 million or 32% of the 1975– derivation of policy recommendations. 1994 total. ILRI lifetime spending on economics and The second scientific impact was to develop policy research has been about US$198 million technical and economic models derived from ori- since 1975, or 11% of the 1975–2018 total of ginal field data that were applicable to policy US$1.75 billion. The sum of economics and pol- problems: icy work plus livestock systems research, which nearly always made some policy recommenda- • Collected and analysed data needed to tions, was about US$375 million, or 21% of the support policy measures, including field 1975–2018 total. surveys, modelling with GIS, bioeconomic models and surveys, randomized controlled trials. • Property rights studies of grazing land, Scientific impacts arable land and tree tenure, leading to site-specific policy recommendations in The scientific impact of economics and policy Ethiopia, Kenya and some countries in West research at ILRI and its predecessors, ILCA Africa. This included demonstrations of the and the International Laboratory for Re- scope and impact of collective action in search on Animal Diseases (ILRAD), has been land rights, livestock value chains and nat- substantial: ural resource management. Economics and Policy Research at ILRI, 1975–2018 641 • Quantification of environmental con- crop–livestock farming systems in Ethiopia straints to growth, including soil erosion after overthrow of the Derg in 1991. constraining higher land productivity in • Evaluation of the effects of property rights, Ethiopia. including traditional informal tenure, on • Identification of barriers to women’s par- land use, land management and tree tenure ticipation in technology and product mar- as it affected incentives to adopt alley farm- kets, notably in Nigeria and Kenya. ing practices. • The landmark work on dairy reform in • Contribution to reforms of animal health Kenya, which used a combination of an in- services in sub-Saharan Africa. novative data set, modelling and communi- • A national programme to control classical cations strategy to achieve substantial eco- swine fever launched in India following an nomic benefits. ILRI study disclosing the economic impacts of the disease in three states in the coun- try’s north-eastern region. • Livestock master plans: the major achieve- Development impacts ment in influencing public expenditure was production of livestock master plans in ILRI’s policy and economics research, with the Ethiopia, Rwanda and Tanzania and in the exception of the Kenya dairy policy effort, had state of Bihar, India. limited identifiable development impact on out- • Capacity development: the main achieve- put or equity. Some indication of development ment was to link economists and scien- impact can be seen in the following: tists in other programmes and in other institutions through the African Livestock • Analytical and organizational contribu- Policy Analysis Network (ALPAN), the Afri- tions to dairy market reforms in Kenya that can Trypanotolerant Livestock Network benefited small producers and poor con- (ATLN) and the epidemiology-economics sumers. The estimated net present value programme started at ILRAD in 1987. The (NPV) of these reforms, in terms of add- epidemiology-economics work has con- itional producer and consumer benefits tinued for 30 years (see Chapters 5 and 6, minus reform programme costs, was this volume) and has had a strong scientific US$230 million. impact in proposing solutions to animal • Purchases of index-based livestock insur- health problems. ance (IBLI) policies in Kenya at an approxi- mate amount of US$25 million and pay- outs to herders of US$10 million; purchases of IBLI policies in Ethiopia of US$2.5 mil- Introduction lion and pay-outs of US$370,000; and adoption by the government of Kenya of a This chapter defines policy as the actions of national IBLI policy, based on the ILRI/IBLI governments and public agencies. Policy actions model, which now provides insurance to can include, inter alia, the writing and enforce- 80,000 herder beneficiaries in the risky ment of laws and regulations; the funding arid and semi-arid areas of Kenya. or conducting of research that investigates • Identification of measures to promote public actions; the provision of information the supply response after the ‘Livestock and other inputs; and the building, operating Revolution’ including supporting infra- or funding of productive infrastructure structure, targeted producer transfers, (e.g. laboratories for disease diagnosis) and definition of the conditions for success- market infrastructure. In their policy actions, ful collective action, and modernization governments typically make use of fiscal in- of sanitary and phytosanitary (SPS) struments such as taxes, subsidies, and wage standards. and price controls, as well as institutional con- • Contribution to the design and analysis structs and measures (organizations, rules of reforms of property rights in mixed and regulations). 642 M. Jabbar et al. Two factors drive demand for public policy. Policy Problems The first consists of market failures (typically characterized by inefficient distribution of The overall goal of economics and policy re- goods and services), negative externalities search at ILRI and its predecessor ILCA has been (when production and/or consumption im- to increase smallholder returns from farming poses external costs on third parties, causing with animals. This research has sought to: social costs to exceed private costs) or of (i) analyse the targeting and productivity of live- ‘free-riders’ (where some individuals consume stock-based technologies; (ii) identify policy bar- more than their fair share or pay less than their riers that lower farm prices, raise input costs or fair share of the cost of a shared resource). To increase the financial, information and risk address such market failures, governments costs of new livestock-related innovations; and may intervene to correct market distortions to (iii) support institutions that improve livestock improve efficiency and equity. The second productivity, create livestock assets and improve driver of demand for public policy is ambitions the performance of livestock value chains. to achieve social objectives of efficiency and The following problems have been the focus equity through public goods that maximize so- of ILCA and ILRI economics and policy research: ciety’s return and ensure the equitable distribu- (i) the historical problem of supply response; tion of returns. (ii) animal health services and productivity; (iii) Zilberman and Heiman (2004) placed stud- responding to the ‘Livestock Revolution’; (iv) ies of public policy into three classes: policy and technical barriers to smallholder • Class I: advancing scientific understand- dairying; (v) livestock and poverty; (vi) markets, ing in describing systems and markets, institutions and competitiveness; (vii) land ten- assessing performance of markets, mak- ure; and (viii) livestock master plans. Economic ing sectoral accounts and estimating and policy analyses linked to specific technolo- productivity. gies are discussed for animal health in Chapters • Class II: contributing to technical change 1–9 and for feed production in Chapters 11–14 that lifts productivity, such as breeding new (this volume). stock or forage plants. • Class III: advising on policies to improve efficiency or equity, such as improving ex- The historical problem ternal terms of trade, building institutions, of supply response investing in public goods, eliminating mar- ket distortions and calculating the costs of An initial problem of policy research at the bad policies. founding of ILCA in the mid-1970s was how to Studies of scientific understanding and stimulate supply through higher productivity. technical change are of a ‘diagnostic’ nature: This led to policy research on the structure of they provide information to agents – individuals, production and trade and on the role of incen- firms and sectors – with or without specific tives in raising production. The research on this policy implications. Such studies may indicate problem had very limited impact and has largely market failures, negative externalities, or free- been forgotten, mainly because it failed to use rider problems and their underlying causes, ILCA’s comparative advantage in farm data ac- and suggest interventions as corrective meas- cess and analysis. ures. The focus of this chapter is on Class I and The initial ILCA policy papers analysed sub- III types of policy studies, ranging from diag- continental trends in production, demand and nostic market performance studies to analyses trade, and their drivers (Montgolfier-Kouevi and of the impacts of interventions and polices. Vlavonou, 1981; Addis Anteneh, 1984; Sand- Table 17.1 summarizes policy research im- ford, 1985). These studies highlighted that Afri- pacts by policy class and gives specific ca’s low livestock productivity and growth rates problems addressed. Table 17.2 gives selected contrasted sharply with the rapidly growing de- applications of bioeconomic models to policy mand for animal-source foods, which implied an problems. increasing dependence on imports. National Economics and Policy Research at ILRI, 1975–2018 643 Table 17.1. Livestock policy research impacts by class and problem. Policy Class II: Policy Class III: Policy Class I: providing contributing to technical improving efficiency Themes information change and equity The problem of Scenario modelling after 1995: McIntire et al. (1992); Perry and Randolph supply response Thornton (2010), Herrero Pingali et al. (1987) (1999) on animal et al. (2010); major impact disease; Jones and for ILCA before 1995; Addis Thornton (2009) on Anteneh on public services crop–livestock (1983, 1984, 1991); interactions as Thornton et al. (2011), affected by climate Robinson et al. (2011) on change mapping global livestock Animal health Addis Anteneh (1983, 1984, Perry et al. (2002) on Perry and Randolph services 1991) animal health and (1999) poverty; Kristjanson et al. (1999) on trypanosomiasis vaccine; McDermott and Arimi (2002) on brucellosis, McDermott and Coleman (2001) on trypanosomiasis control strategies Responding to the Livestock Revolution Livestock Thornton et al. (2002) on Thornton et al. (2002) assets, poverty analytics on poverty poverty, and analytics; Giller financial et al. (2011) on markets targeting technical change Barriers to Holloway et al. (2000a) smallholder on dairying in dairying Ethiopia Land tenure, Verburg et al. (2009), Fritz Reid et al. (2000b) on Coughenor et al. property et al. (2015) on cropland and land use in Ethiopia, (1985) on pastoral rights, and field size Berhanu models, Lawry institutions Gebremedhin on land et al. (1994); conservation in McCarthy et al. Ethiopia, Franzel and (1999) Wambugu (2007) on fodder shrubs Livestock master plans Public investment Randolph et al. (2007) on role Minor impact Coughenor et al. of livestock, Thornton et al. (1985) on pastoral (2011) on adaptation to models, Randolph climate change et al. (2007) on role of livestock, Havlik et al. (2013) on adaptation to climate change; Shapiro et al. (2015) Table 17.2. Selected applications of bioeconomic models, various years. Theme and chapter Region/country/ Policy Technical Problem/study in this volume climate Method Objective function Treatments recommendations recommendations ECF Gettinby et al. Acaricide resistance/ Sub-Saharan Process Minimize acaricide Acaricides Regulate acaricide Elucidate genetic (1988) Ticks (Chapter 10) Africa; simulation resistance in use basis of acaricide subhumid treated tick resistance in ticks populations Mukhebi et al. ECF (Chapters 5 Kenya; Partial budgeting Financial and ITM and acaricides Monitor benefits and Joint use of (1992) and 6) subhumid economic costs of dipping acaricides and returns to ITM frequency carefully ITM and tick control; IRR 48% Nyangito ECF immunization Kenya; Whole-farm Financial returns to ITM and acaricides Monitor benefits and Joint use of et al. and tick control/ subhumid simulation ITM and tick costs of dipping acaricides and (1996a,b) ECF, ticks control frequency carefully ITM (Chapters 5,6, and, 10) Ticks and their East Africa; Literature review Financial and Acaricides in Buffer zones between Strategic application control (Chapter subhumid biological rotation and domestic stock and of acaricides 10) and humid sustainability of combination alternate hosts long-term (wildlife) acaricides Cattle productivity Cartwright Beef production/ Botswana; arid Herd simulation Output and profit Weaning ages of NA Manage weaning et al. (1982) livestock systems maximization calves ages and milk (Chapter 15) offtake policy Konandreas Beef and milk Botswana; arid Herd simulation Profit maximization Feed Supplementation Feed et al. (1983) production/ of product and supplementation economically supplementation livestock systems input choice in cross-bred and superior in (Chapter 15) indigenous cows cross-bred cows; no public policy implications Thornton Dual-purpose beef Colombia; Simulation model Optimize beef Improved grass- None (pending more Not specific (1987)a and milk humid and from station production legume pastures results on pasture production/ subhumid trials and field in dual-purpose outcomes) livestock systems surveys systems (Chapter 15) Feeds and forages Powell et al. Soil fertility Several Survey, station Yield and financial Mulching, manuring, (1995) management/ sub-Saharan trials, tabular optimization of mineral fertilizers, livestock systems African models, organic and crop-residue (Chapter 15) countries; deterministic mineral soil management subhumid simulation amendments and semi-arid Elbasha et al. Planted forages West and Survey, herd Ex post economic Planted forages, Most recommendations (1999) (Chapter 13) Central simulation surplus; ‘fodder bank’ technical; policies African model, tabular ‘research paid (usually include providing countries; model of forage for itself’ Stylosanthes extension subhumid production spp.) information, credit and humid and subsidized seed Kristjanson Dual-purpose Central and Survey and Ex ante economic Improved dual- No technical No policy et al. (2001) cowpea/planted northern tabular model surplus; cowpea purpose cowpea recommendations recommendations forages (Chapter Nigeria; after station grain 128–154% cultivars 13) subhumid trials of NPV gross and benefits; cowpea semi-arid hay from −28% to −54% Kristjanson Dual-purpose pearl 105 districts Ex ante Ex ante economic Higher straw quality Value incremental No policy and Zerbini millet and within nine deterministic surplus; meat in pearl millet and traction and recommendations (1999) sorghum/ states of simulation and milk sorghum manure output, multidimensional semi-arid productivity (IRR value effects in new crops (Chapter 14) India 26–43%) areas (e.g. north-east Brazil) Valbuena Conservation 12 sites in nine Literature and Optimize crop- Mulching versus Value output and soil Favour conservation et al. (2012) agriculture countries in data review residue uses feeding quality agriculture in sites sub-Saharan among feed and with higher mean Africa mulch biomass production Continued Table 17.2. Continued. Theme/part and chapter in this Region/country/ Policy Technical Problem/study volume climate Method Objective function Treatments recommendations recommendations Trypanosomiasis ILCA/ILRAD Trypanosomiasis/ East Africa/ Survey and Optimize use of Trypanocidal drugs, Choice method of (1988) livestock systems Kenya coast; tabular model trypanotolerant sprays, traps, trypanosomiasis (Chapters 2 and 3) humid stock trypanotolerant control as functions stock of challenge, livestock system and susceptibility of animals Itty (1992), Itty Trypanotolerant Several Surveys, herd/ Optimize use of Vector control, Policy choice among et al. (1988) livestock and sub-Saharan flock trypanotolerant trypanocidal treatments as trypanocides African deterministic stock drugs, function of land countries; simulations trypanotolerant potential and stock subhumid stock density Agyemang N’Dama cattle and The Gambia; Herd Profit maximization Milk offtake of Private management et al. (1997) milk offtake subhumid measurements, trypanotolerant recommendation of deterministic stock partial milk offtake simulations Kristjanson Trypanosomiasis Sub-Saharan Ex ante simulation NPV of net benefits Vaccine None et al. (1999) vaccine/ Africa; to vaccine trypanosomiasis subhumid (Chapters 2 and 3) and humid McDermott Trypanosomiasis Sub-Saharan Deterministic Rate of Curative drugs, None and (Chapters 2 and 3) Africa; humid epidemiological trypanosomiasis vector control, Coleman and model without prevalence trypanotolerant (2001) subhumid prices or costs cattle, vaccine Reid et al. Environmental effects Sub-Saharan Survey and Rate of change of Implicit treatments None None (2000a) of trypanosomiasis Africa; humid time-series tsetse infestation of population control/tick-borne and projections as function of growth and land disease subhumid population growth use intensity Dairying Nicholson Dairying Coastal Kenya; Household Impact of dairying Breed, feed, Promoting access of Not specific et al. (1999) humid surveys on income and management, smallholders to employment disease control dairy technology through financial and institutional measures Kaitibie et al. Reform of dairy Rural Kenya, Household Impact of dairy Liberalization of Promoting direct Not specific (2010a) marketing policy mainly surveys, market market reforms raw-milk trade sales of raw milk highlands surveys income and by smallholders employment and traders Other diseases McDermott and Brucellosis/ Arimi (2002), veterinary McDermott epidemiology et al. (2013) (Chapter 5) Kimani et al. RVF/ECF (Chapter 6) Kenya; Simulation of RVF Reduce human and Vaccination and Enhanced Enhanced (2016) subhumid epidemics animal health disease surveillance and surveillance and effects of RVF surveillance earlier vaccination earlier vaccination measured in disability-adjusted life years Vertisol technology and land management Gryseels (1988); Improved drainage Ethiopia; Farm budget Profit/ha Enhanced Vertisol Extend the Target technology to Gryseels and as part of highlands analysis and drainage and technology; reduce poorly drained Anderson broad-bed linear tillage quality cost of the BBM areas (1983); management/ programming Getachew livestock systems model Asamenew (Chapter 15) et al. (1993) Rutherford Improved plough for Ethiopia; Farm and market Estimate return to Enhanced Vertisol Expand credit supply, Expand use of BBM (2008) broad-bed highlands surveys, analysis research in the drainage with reduce cost of the for pond management/ of research and BBM BBM BBM construction and livestock systems development field irrigation (Chapter 15) costs BBM, broad-bed maker; ECF, East Coast fever; ITM, infection-and treatment method; IRR, internal rate of return; NA, not applicable; NPV, net present value; RVF, Rift Valley fever. aThornton (1987) was part of a CIAT effort on modelling beef production, beginning roughly with CIAT (1975). 648 M. Jabbar et al. trends differed significantly and there was signifi- reducing protection to uncompetitive public cant cross-border trade in live animals among agencies, such as parastatals, which were pro- many countries, especially in West and East Af- ducing, trading or processing livestock products. rica. In West Africa, the large cities in the coastal Related studies of import competition in dairy region in the south were major centres of meat products were inconclusive in terms of policy re- and milk consumption, while livestock produc- commendations (von Massow, 1989). tion was concentrated in the arid, semi-arid Two trade studies were conducted around and subhumid zones in the Sahel to the north. the time of the ILCA/ILRI merger in 1995. The A similar pattern – production in the pastoral first compared the effects of livestock policies on and agro-pastoral zones and consumption in output, consumption, trade and government the cities – held in East Africa. This policy work revenues in Côte d’Ivoire, Mali, Nigeria, Sudan established that demand was not a constraint and Zimbabwe (Williams, 1993). The study to growth of African livestock and agriculture found that inflation and exchange rates were key more generally. variables in livestock pricing policies, and that Several early studies, notably Jahnke (1982), some success had been seen in stabilizing real observed the poor incentives for all agricultural domestic prices. As a result of policy changes in production in sub-Saharan Africa due to low of- exchange rates and domestic prices since the ficial prices and overvalued exchange rates as 1980s, there had been a shift away from taxing domestic factors, and as a result of dumping and livestock producers. These changes had positive food aid as external factors. The 1970s and impacts on beef production and consumption in 1980s generated extensive study of price incen- all study countries except Côte d’Ivoire, with tives in global and sub-Saharan African agricul- mixed effects on exports. ture culminating in the landmark work of Schiff The second study reviewed macroeco- and Valdés (1992) to which ILCA and ILRI made nomic, sectoral and trade policies based on infor- little contribution. mation developed for international farm trade ILCA contributions on incentives problems negotiations (Williams et  al., 1995). The study were microeconomic. An ILCA paper argued focused on ruminant livestock in Africa, Asia that international trade and pricing policies and Latin America and on pork and poultry in were particularly important for livestock given West Asia and North Africa. It found that, along the widespread trade in live animals in sub- with macroeconomic, sectoral and trade pol- Saharan Africa (Solomon Bekure and Macdon- icies, investments in infrastructure, animal ald, 1985). First, many people derive income health, and processing and marketing facilities from livestock production and these incomes are were required to promote efficient resource use directly affected by livestock prices. Second, con- and growth in production, consumption and sumers spend an important share of their in- trade. It projected positive effects on livestock come on livestock products and this share will product supply from better incentives through rise with economic growth. Third, livestock pri- trade liberalization, though the effects varied by cing policies are important to governments be- region, country and commodity. cause of their implications for incentives, public spending and revenue. The study further argued that ‘positive’ policies – enhancing effective Markets, institutions measures – improved efficiency in live-animal and competitiveness markets where transport costs were high. Ex- amples of such ‘positive’ policies were gazetting Ruminants animal trekking routes to reduce costs of crop damage along the routes and establishing water Market participation by smallholders can allow and feeding facilities to reduce weight loss and them to expand crop and livestock supply. To mortality in transit. ‘Negative’ policies – elimin- identify barriers to wider market participation, ating ineffective measures – could also improve ILRI policy research has focused on the struc- the efficiency of livestock markets, such as by ture, conduct and performance of product and eliminating arbitrary or discriminatory licens- input markets with a particular emphasis on ing of traders and other intermediaries and by market access. Economics and Policy Research at ILRI, 1975–2018 649 Kebede Andargachew and Brokken (1993) to market opportunities, so better transmission analysed sheep price patterns in the central of demand and price information will benefit highlands of Ethiopia to determine the effects of them. animal and market characteristics and season A series of studies of livestock trade in Ni- on price. Weekly price variations were evident in geria found that markets were quite responsive redistribution, intermediate and terminal mar- to incentives. Regular supply–demand imbal- kets. Animal characteristics (weight, age, body ances, determined by regional comparative ad- condition, sex and colour), as well as purpose for vantage, were corrected by live-animal trade buying and buying season, were important in from other regions of the country (Jabbar, explaining price variation. The findings indi- 1993, 1995, 1998). Seasonal excess demand cated that producers targeted market strategies for small ruminants in southern Nigeria was to gain from coordination of fattening, breeding met by supplies from the north. These findings and trading operations. The study did not find suggested that production technologies that any subtle market inefficiencies that might be contribute to a regular increase in supply might addressed by policy beyond the obvious meas- be appropriate for smallholders, while seasonal ures of increasing market density and lowering commercial production might be geared to peak transaction costs by providing infrastructure. season markets. ILRI collaborated in a study led by the Ethi- A series of studies on cattle breed prefer- opian Agricultural Research Organization ences applying hedonic analysis of cattle prices (EARO) in which hedonic price models were in south-west Nigeria found small but signifi- used to determine seasonal and intermarket dif- cant price differentials by breed (Jabbar et  al., ferences in prices of sheep in Ethiopia (Ayele 1995, 1997; Jabbar and Diedhiou, 2001). In Solomon et al., 2003). There were significant dif- these studies, Muturu, a locally adapted West ferences in prices among seasons and markets. African Shorthorn breed, illustrated the rela- Seasons in which farmers faced severe cash tionship between farmer preferences and market shortages exhibited the lowest adjusted prices for values of cattle breeds. Even though Muturu is animals they sold, indicating that, although live- known for its superior abilities to resist diseases, stock may provide a fallback position for cash in particularly trypanosomiasis, and is productive times of crisis, terms of trade may be worst when under high humidity, heat stress, water restric- farmers need cash the most. In general, there tion and poor-quality feed, the Muturu was rated was no clear progression in price of sheep along the least desirable for market value and mobility. the primary to terminal market chain ending in For producers, the perceived limited marketabil- Addis Ababa as would normally be expected, ex- ity, low market value and the need for mobility to cept that the farthest market had the lowest market are important components of the re- price. In the case of goats, the price differences turns to raising Muturu. For these reasons, between markets followed to some extent the ex- farmers’ aversion to Muturu as an investment pected differences between primary, secondary imply little scope for its in vivo conservation. and terminal markets. One possible reason for Eliciting farmers’ knowledge about traits of the different price progression along the supply breeds and their relationships to prices can be chains of sheep and goats is that, in general, the useful for designing breeding policy and strategy Ethiopian highlands are not a major production for breed development programmes (Jabbar or consumption area for goats, so supplies come et al., 1999). mainly from the lowlands, with the result that Another study on the spatial integration the price movement followed the market chain of livestock markets in Niger found that live- from primary markets in pastoral areas to the stock markets were ‘related, but not closely terminal market in Addis Ababa. On the other integrated’ (Fafchamps and Gavian, 1996). The hand, most of the higher-quality sheep originat- authors suggested policy options for improving ing in both highlands and lowlands are bought livestock market efficiency by investing in ani- by exporters and processors in the intermediate mal transport and lifting official restrictions on markets so both average quality and price are animal trade. lower at the Addis Ababa terminal market. The Ayele Solomon et al. (2003) surveyed live- implication is that producers and traders respond stock markets and traders in the highlands of 650 M. Jabbar et al. the Amhara, Oromiya and Tigray regions in unit distance marketing costs than other ap- 2000 and 2001. They showed that the numbers proaches and hence indicate priorities for public of agents (wholesalers, brokers and retailers) investments to reduce those costs. who trade many species had increased signifi- Bahta and Malope (2014) examined the cantly since the collapse of the Derg regime competitiveness of smallholder beef farmers in 1991. This expansion in the numbers of agents, Botswana using data from randomly selected combined with small numbers of inspectors, producers. There was significant inefficiency, led to many unlicensed traders, irregular inspec- with about 74% of the variation in actual profit tions and a scarcity of fencing, feed and water from maximum profit between farms arising troughs. from differences in farmer practices. The mean Jabbar et  al. (2008) found that Ethiopian profit-efficiency level of 0.58 suggested scope to livestock markets were characterized by non- improve beef profitability with current technol- standardized products and lack of public infor- ogy. Profit drivers included education, distance mation about quantities and prices. Consequently, to market, herd size, access to information and livestock trading was largely a personalized crop income. The main policy lesson was to im- business among brokers with regular buyers prove market access so as to raise profits among and sellers; this regularity is a form of social cap- smaller producers. ital used for gathering information, searching buyers/sellers, negotiating prices and enforcing Pigs and poultry contracts. Business relationships were based principally on trust developed over time, without With the merger of ILCA and ILRAD into ILRI in strong ethnic, religious or family ties. Although 1995, the mandate of the new institute ex- most transactions were conducted in the phys- panded to include pigs and poultry. A series of ical presence of parties, contract disputes were ILRI-led studies in Asia beginning in the 1990s common and were typically settled mainly identified constraints to smallholder production through informal means as formal legal systems in non-ruminant livestock. were absent or expensive. It was argued that pol- Lapar et al. (2003a) studied a cross-section icies to reduce transactions costs and multiple of smallholders in northern Luzon, in the Philip- taxes while increasing access to market informa- pines. They investigated factors motivating tion would improve trader margins and market smallholders’ decisions to sell products or con- performance. sume them as functions of transaction costs, Examining the factors that affect market labour mobility, capital formation and indebted- participation and sales by households, Ehui et al. ness. The strong effect of animal numbers on the (2003) showed that physical capital (ownership participation and selling decisions of farmers of different species of livestock and landhold- suggested that policy interventions may be ings) and financial capital (crop and non-farm needed to support smallholder access to input income) are the main factors influencing market and output markets. The availability of alternative participation and sales, rather than the distance occupation opportunities, however, significantly to markets and towns. These results suggested affected the viability of social and economic that constraints to production of livestock and prescriptions, and policy makers must be cogni- livestock products (e.g. capital to purchase ani- zant of these results when targeting objectives mals, feeds and processing equipment) were the for smallholders. Remittances had a positive in- main factors limiting market participation. fluence on market participation, suggesting the Baltenweck and Staal (2007) explored spa- importance of financial security in enabling tial measures of market access for milk and smallholders to manage risks and subsistence beans in the Kenyan highlands. Measures of requirements. market access were used to create spatial price Costales et al. (2006) studied how scale af- decay functions in relation to transaction costs. fected access to markets by hog producers in The effects of market access differed significantly southern Luzon, a major hog-producing area in depending on the traded good. The analysis also the Philippines. Regional data indicated that be- demonstrated that spatial price formation could tween the 1990s and 2000s there had been an be used to generate more accurate measures of expansion of larger hog farms and displacement Economics and Policy Research at ILRI, 1975–2018 651 of small farms. The study applied a probit model The study on pig contract farming con- to identify factors that determine participation ducted in four provinces in northern Vietnam in hog production and applied a profit-efficiency showed that there was limited scope for small- model to identify the role of transaction cost bar- holder pig producers to participate in formal riers in smallholder performance. The model re- contracts; however, smallholders were found to sults showed that the decision to participate in participate in informal contracts with coopera- hog production was positively influenced by the tives and with input/output traders that facili- availability of family labour and by the cap- tated their access to pig markets. To under- acity to deal with fixed transaction costs for ac- stand the drivers of these smallholders to cess to financial resources. Market participation participate in these types of contractual ar- was negatively influenced by higher opportunity rangements for pig and piglet production, a costs of family labour due to access to off-farm multinomial logit model was applied. The results job markets and by distance to hog markets suggested that the significant determinants of outside the village. Comparison of contract and smallholders’ participation in contractual ar- independent growers among hog producers rangements are age, proportion of time spent showed that most contract growers tended to in pig-raising, location, distance to veterinary specialize in fattening pigs to slaughter. In con- shops and access to animal health services. trast, independent producers tended to combine A study on livestock development in Viet- the production of weaners (piglets) with slaugh- nam identified barriers to input and output mar- ter hogs or specialized in weaner piglet produc- kets for smallholders (Lapar et  al., 2003b). The tion. In general, contract growers had larger uncertain quality and high prices of animal levels of operations than independent growers. feeds, including raw materials for feed process- Contract growers exhibited better access to out- ing, the variable quality and high cost of more put markets and to good-quality feeds and stock, productive animal breeds, and the high costs of feed credit, veterinary health services and credit veterinary inputs were found to be the principal for expansion purposes. barriers in livestock input markets. Constraints One study analysed contract farming in to reaching output markets included poor-quality Bangladesh poultry (Jabbar et al., 2007), while and unsafe meat and meat products, lack of a another covered pig farming in Vietnam legal framework and standards, bottlenecks in (Tiongco et  al., 2009). The Bangladesh study the distribution channel and limited access to in- found three emerging contracts – production formation. In addition, the prevailing marketing marketing, formal input marketing and infor- system and channels for each type of commodity mal output marketing. The profitability of from farm to market have evolved into a mul- broiler farms did not differ significantly be- ti-stage system that is characterized by high tween contract and independent farms, but it transaction costs and lack of market integration. differed between the two sample districts. Con- Another study in Vietnam addressed whether tract layer farms performed much better than national livestock production can remain com- independent layer farms. These differences petitive under rapid demand and import growth were due to differences in the feed conversion (Akter et  al., 2004). The study applied a policy ratio, fattening days and sale weight for broil- analysis matrix to assess the competitiveness of ers, and egg production per bird per laying poultry and pig production based on 1999 data period and length of laying period for layers. from a sample of 2213 farms. Poultry and egg Based on a sample of independent poultry production from cross-bred and exotic breeds farms in five districts, the key reasons for busi- were competitive in the north, while egg produc- ness failure after 1 or more years of operation tion with local breeds was uncompetitive in the were identified as high input prices, irregular south due to low productivity and high per-unit supply of day-old chicks and poor-quality vet- cost. Economies of scale in poultry production erinary drugs. Some input market problems existed in the north but were not so clear in the have been solved by contract farming in other south. Domestic prices of outputs and inputs contexts, but formal contract farming seems to were higher than world prices due to trade have offered few opportunities for commercial protection. In the long run, small poultry poultry farmers in Bangladesh. farmers might not be able to compete in a more 652 M. Jabbar et al. liberalized economic environment with low-pro- of product and input market domains and ductive local breeds and higher per-unit cost; pol- household characteristics that may improve ac- icy support such as access to credit and inputs for cess to information, technology and manage- smallholders would be needed to maintain their ment decisions. competitiveness. In the case of pigs, there are significant dif- Pig production under existing technologies ferences in production behaviour and efficiency and market conditions was highly competitive, levels between the north and the south among especially with local and cross-breeds in the farms producing different breeds, between mixed north and exotic breeds in the south. At the time and specialized farms, between household and of the survey, the producers in the south were ap- commercial farms, and among producers lo- parently benefiting due to a greater role of formal cated in different agroecological regions. Access market conditions, which favoured cross-breeds to better output markets, land size, herd size and and exotic breeds, while in the north, policy education of the household head significantly interventions made input costs higher and out- reduced inefficiency, while access to govern- put prices lower. Some economies of scale were ment-supplied inputs, age of the household demonstrated in pig production, in that medi- head, female-headed households and family- um-sized farms were more cost-effective and supplied crude feeds significantly increased small farms were least competitive. inefficiency in both regions. The direction of Using the same data set for Vietnam, sto- influence on efficiency differs between the two chastic frontier production functions were used regions for access to credit, proportion of output to assess the effects of market and other factors sold at market rather than at the farm gate and on technical efficiency, respectively, in poultry family labour supply. Generally, market-related (Jabbar and Akter, 2006) and pig production (Jab- factors had a more consistent influence on pro- bar and Akter, 2008) In the developing-c ountry duction efficiency in the south of Vietnam, production environment, farm production effi- where the experience of market economics is ciency is often measured in terms of on-farm longer than in the north. Policy actions on pro- resources and producer characteristics. In these viding better extension, more timely access to studies, it was postulated that input and output better-quality inputs through the private sector, market-related factors also influence farm pro- making credit more easily accessible to small- duction decisions and hence farm efficiency. holders and providing opportunity to sell output In the case of poultry, in general there are at better-priced secondary markets are expected significant differences in the production behav- to increase productivity and reduce inefficiency iour and efficiency levels between the north and among producers located in different agroeco- the south among farms producing different logical regions. breeds of poultry, between mixed and specialized In some countries, local products are shield- poultry farms, between household and commer- ed from international competition by ‘natural’ cial farms, and among producers located in factors influencing the purchase of products, different agroecological regions. Sale at market- such as strong local tastes (or preferences) that place rather than at the farm gate, market favour the local product and the absence (or distance and flock size significantly reduced inef- relative absence) of complementary retail out- ficiency in both regions. Contract farming or lets or home appliances suitable for storing and sale, the number of visits by extension staff, fam- preparing potential imported substitutes (Tis- ily labour supply, land size and education levels dell, 2009). The desire for fresh meat rather than of households had significantly reduced ineffi- chilled or frozen meat, the absence of supermar- ciency in the north but had no significant effect ket outlets and limited refrigeration possibilities in the south. The direction and significance of in homes can limit imports into developing influence on efficiency differ between the two re- countries of meat supplied by developed coun- gions for credit use, inputs from government, tries. This study gave some simple economic ratio of home-produced crude feed, producer analysis of how local producers of livestock age and gender of the household head. There- benefit from natural protection. Drawing on the fore, opportunities exist for improving average results of research completed in Vietnam and efficiency through interventions in a number other sources, factors that provide natural Economics and Policy Research at ILRI, 1975–2018 653 protection to Vietnam’s pork industry were to the detriment of non-staff variable costs, such identified, with particular attention given to as drugs and fuel; (iii) cost recovery was their implications for small-scale household pig limited as a share of livestock service budgets; producers compared with larger-scale commer- and (iv) private service delivery was weak, partly cial pig producers. It was noted that the protec- because of resistance from the public veterinary tion of Vietnam’s pig industry was not based on sector. a preference for pork from local breeds but arose Mohammed Mussa and Gavian (1994) re- for other reasons. viewed the privatization of animal health ser- vices in Ethiopia. They argued that vaccination and vector control are public goods because the benefits extend to the whole economy, while Animal health services and productivity curative services (diagnosis and treatment) of non-transmittable diseases are primarily private Policy aspects of animal health research at ILCA goods (although some effects of repeated cura- concentrated initially on institutions – who pro- tive treatments, such as induced resistance to vides animal health services, how can services trypanocides or other antibiotics, would become provision be more efficient and what is the scope public goods). The policy lesson is that preventa- for private provision? Public agencies were tive services work better when managed by the the main providers of animal health services state, while privatization is feasible for curative throughout sub-Saharan Africa before 1980. treatments. After the introduction of structural adjustment The question of payment for public services programmes in the late 1970s and early 1980s was studied by Swallow and Woudyalew (1994) and given the failure of veterinary services to who investigated whether communities in reach many livestock farmers, the form and pri- south-west Ethiopia would pay in cash and/or cing of veterinary services became important labour for trypanosomiasis control. When asked policy issues. The debate centred on the justifica- about the maximum amounts of money and/or tion for public financing of animal health ser- labour that they would be willing to pay, 59% of vices, especially those that could be considered households volunteered both money and labour private goods, such as curative services. and only 3% volunteered neither. Willingness to Addis Anteneh (1983, 1985) reviewed na- contribute money was related to the gender of tional information on the financing of livestock the household head, the number of cattle held services in selected African countries. He found by the household and the participation of the that: (i) services were underfunded with respect household in a monitoring exercise being con- to the share of livestock in agricultural gross- ducted by the research organization. Willing- domestic product (GDP); (ii) costs were mostly ness to contribute labour was related negatively operating budgets and hence capital funding to off-farm employment status of the head of the per staff was generally inadequate; (iii) services household, and positively to the information were largely funded with public resources and available to the respondent about the pro- foreign aid, not with user charges, which dis- gramme. Apart from direct applicability of these couraged user participation in service manage- results to increase local involvement of the af- ment; (iv) there was potential for more public fected population in the control programme, the spending because user charges – head taxes, study stressed that the methodology used here, slaughterhouse fees and taxes – were sometimes when integrated into a participatory research greater than public spending on livestock ser- approach, can generate practical results for vices, indicating that livestock revenue had been evaluating the prospects for local participation diverted to other sectors; and (v) the quality of in the provision of public goods. veterinary services was low because of lack of Hall et al. (2004) evaluated the welfare ef- funding, especially for non-staff variable costs. fects of herd health programmes on smallholder A later study on financing livestock services dairies in central Thailand. Dairy farmers had (Addis Anteneh, 1991, Part 7) arrived at similar appropriate incentives to adopt herd health conclusions: (i) livestock services were again measures; following adoption of control meas- underfunded; (ii) staff costs dominated services ures, there was an improvement in the efficiency 654 M. Jabbar et al. of policy support to dairying. Following a reduc- Eastern markets, but the binding constraint was tion in disease incidence on adopters’ farms, the domestic input costs rather than the costs of study found an increase in farm profits. compliance. Sensitivity analyses revealed that, Outbreaks of Rift Valley fever (RVF) in East while investments in feed efficiency and animal Africa in the past prompted a ban by Middle productivity would enhance Ethiopia’s export Eastern countries on imports of live animals competitiveness, the highly competitive nature from that region. Nin-Pratt et al. (2003a) evalu- of international beef markets may still prevent ated the certification of exported live animals market access by Ethiopia’s beef producers (Rich granted in an RVF-free zone, as a case for Ethi- et al., 2008; Rich and Perry, 2009). opia, as one way of handling RVF and complying Somalia was a traditional supplier of live with international regulations. The study also sheep, goats and camels to Middle Eastern mar- examined policies (export tax, sales tax and in- kets. Following the collapse of the Somali state creased transaction costs) to make producers in 1991, a rapid appraisal identified the institu- bear programme costs. The study concluded tions active in livestock exports (Mugunieri et al., that implementing an animal health programme 2008) to understand informal policies. A more in Ethiopia’s Somali region was economically detailed study was then conducted with export- feasible and would benefit poor livestock produ- ers to understand how the Somali origin satis- cers. It suggested that increasing taxes on live- fied import requirements for product quality and stock sales offered the best way to fund the cost (Asfaw Negassa et  al., 2008). Constraints health certification plan. This option, in which a along the export chains were mapped, and ap- transfer from middle and better-off producers to propriate steps for addressing the constraints poor livestock producers is implicit, reduces were recommended. The recommendations in- harms to exports and welfare and increases cluded: a certification system for health and benefits to the poor. quality; provision of market information, train- While Ethiopia has been a major supplier of ing in market opportunities for traders; forma- animals to Middle Eastern markets, its market tion of trade associations and other collective share has varied (Asfaw Negassa and Jabbar, action forums to share information, capital, and 2008). The reasons included an inadequate strengthen ability to negotiate; and harmoniza- supply of good-quality live animals to the export tion of informal and formal taxes and fees in abattoirs in Ethiopia, with some abattoirs operat- marketing chains. ing at less than half their capacity, which raised Sudan (including South Sudan, which be- the average fixed costs per animal and reduced came independent in 2011) was an historical competitiveness in export and domestic markets. exporter of live sheep and sheep meat to the Mid- Livestock census data revealed very low com- dle East, but its market share has fallen over time mercial offtake rates of cattle and shoats from because Sudanese supply failed to meet Middle Ethiopia’s smallholder farmers and pastoralists. Eastern standards. A study analysing supply- This limited market participation by Ethiopia’s chain constraints for Sudanese exports of sheep smallholders and herders implied that, under and mutton found that they faced long costly the prevailing production and marketing condi- journeys by trekking or trucking, which reduced tions, small-scale farms and pastoral systems were the animals’ health and quality (El Dirani et al., not supplying sufficient numbers of good-quality 2009). A high incidence of disease and mortal- live animals at competitive prices to make efficient ity and low offtake rates among traditional produ- use of the country’s meat-processing capacity in cers limited the supply of high-quality animals. its export abattoirs, lowering the competitive- There were elaborate systems of inspection, test- ness of Ethiopia’s domestic and export livestock ing and screening for diseases and other SPS markets. standards in the supply chains, but, because of In addition, SPS barriers and animal dis- poor enforcement, too many unacceptable ani- eases have traditionally constrained market ac- mals remained in the export lots, which led to cess. A system dynamics model was applied to rejections at destination. Although major mar- examine a proposed SPS certification system. kets in Sudan’s hinterlands were integrated with The model’s results indicated that the system the terminal market in Khartoum, as indicated may not be viable for beef exports to Middle by price cointegration, responses to price shocks Economics and Policy Research at ILRI, 1975–2018 655 were variable among markets, with some come from imports and would therefore pose markets more responsive than others and with both a threat and an opportunity to domestic supply markets responding more quickly and in- producers. tensely to shocks than terminal markets. Policy The ‘Livestock Revolution’ study warned recommendations to increase supply of export that smallholders might not benefit from rapid quality animals were to: invest in health, exten- growth in international trade. It recommended sion and higher-quality inputs leading to in- new measures to defend smallholder interests1: creased offtake rate especially from larger flocks; to reduce rejection rate throughout the long • Removing policy distortions that promote supply chain, increase investment in proper la- artificial economies of scale, such as credit boratory facilities including equipment and and tax breaks and other subsidies and trained manpower; rigorously enforce screening trade protection or support to large-scale procedures and standards, and enhance coord- producers. ination among various agencies involved in sup- • Building institutions to link smallholders to porting export oriented production and trading. markets, for example by facilitating vertical integration of smallholders, cooperatives and other forms of collective action. • Creating public goods through the provi- Responding to the ‘Livestock Revolution’ sion of services for animal health and live- stock extension, research and education. A third policy problem has been how trade and • Regulating the environmental and public globalization affected incentives for livestock pro- health costs of livestock production and duction in poor countries and what policy could consumption, such as water pollution and do to mitigate adverse effects or to exploit favour- land degradation on the one hand and, on able effects. This problem became more acute at the other, the obesity epidemic and the the turn of the century as growing populations, emergence of new human diseases origin- urbanization and rising incomes in developing ating in livestock. countries fuelled a rapid increase in demand for Nin-Pratt et al. (2001) analysed the role of animal-source foods, a phenomenon that be- China as an importer of livestock and other farm come known as the ‘Livestock Revolution’. products. They analysed productivity growth A 1999 collaborative study involving IFPRI, in China’s pig and poultry sectors and projected FAO and ILRI, ‘Livestock to 2020: the next food China’s meat trade to 2010 in a general equilib- revolution’ (Delgado et  al., 1999) – became rium model of the Chinese economy. China’s net known as the Livestock Revolution study, used a trade position was projected to be sensitive to global model to analyse changing supply and de- posited growth of its GDP and its non-ruminant mand as they affected poverty, nutrition and productivity, implying uncertainty in the policy health, and the environment. From 1971 to contexts of other developing countries. 1995, meat consumption in developing coun- A similar study (Nin-Pratt et  al., 2003b) tries grew almost three times as fast as in devel- examined the implications of trade liberalization oped countries. It was projected that by 2020, on Vietnam’s smallholders, including the developing countries would be consuming 100 consequences for poverty alleviation. While the million t more meat and 223 million t more milk impact of trade liberalization on Vietnam’s live- than they had in 1993, dwarfing the projected stock production tended to be small, a more increases in meat and milk consumption in de- open Vietnamese economy would increase com- veloped countries. Again from 1993 to 2020, petitive pressure on domestic producers. per-capita consumption of meat and milk in de- New non-tariff barriers have emerged in veloping countries was projected to increase, re- the form of more stringent SPS standards. The spectively, by 42% and 55%, in contrast to devel- implications of such barriers were studied for ex- oped countries, where meat consumption was ports of live animals from the Somali region of projected to increase by 9% and milk consump- Ethiopia to the Middle East, where the exporters tion to decrease by 2%. Much of the increased faced major high costs of compliance with the consumption in developing countries would standards required by the importers (Nin-Pratt 656 M. Jabbar et al. et al., 2003a). Moreover, a ban on livestock ex- Policy and technical barriers to ports from the Horn of Africa imposed by Saudi smallholder dairy development Arabia in 1998 and 2000 following an outbreak of RVF severely affected trade. The cost to the So- Dairying was an important theme in the early mali economy of the ban was estimated to be at ILCA policy work because of its potential for ex- least US$21.8 million, with the total reaching pansion in sub-Saharan Africa, its potential up to US$36 million under some scenarios. The benefits to smallholders as a source of economic estimated loss in regional value added was growth and the growing political problem of ris- US$195 million, almost equal to the value added ing imports. Early stage studies were conducted produced in an average year. In the short run, in Nigeria, Ethiopia and Kenya. middle- and higher-income households could Research in Nigeria has examined demand, manage the negative effects of the Saudi import price determinants, policy reforms, and market ban by increasing consumption. Poor pastoral- development. Jabbar and di Domenico (1990, ists with limited production capacity, however, 1993) and Jansen (1992) described dairy con- would lose income because increased consump- sumption and its determinants in northern and tion of their own production was insufficient to southern Nigeria. In both regions of Nigeria, the compensate for export losses. type of product consumed, and the frequency of An important concern from the Livestock consumption differed markedly among ethnic Revolution was that large producers would dis- groups and between urban and rural popula- place smallholders as markets opened because tions. In the south, per-capita income of they could exploit economies of scale in production dairy-consuming households did not differ sig- and in finance. A study in Bangladesh analysed nificantly. Among the consumers, the income the effects of policy and scale on the efficiency of elasticity of dairy consumption was higher for dairy and poultry farms (Jabbar et al., 2005). For rural households in the south-east. In northern dairy, they showed that breed, management, feed Nigeria, dairy product demand was found to be cost, choice of markets and access to credit for income inelastic, and larger households tended liquidity and to extension contact at times of real to consume relatively fewer dairy products per need to solve a production constraint were sig- household member than smaller households. nificant variables affecting the profitability and The strongest conclusion from these studies was efficiency of dairy farms. Policy interventions – that pricing structures and local consumer pref- infrastructure, waste management, access to fi- erences for traditional products argued for devel- nance and creation of producers’ organizations – opment of traditional production systems using favouring small farms would increase the overall indigenous cattle breeds. Production increase efficiency of the dairy sector. would require provision of breeding and health Baker and Enahoro (2014), in an overview services and better feeds. Support for better of six studies, argued that information from a processing, storage and transportation of trad- large number of household studies on livestock itional products would be required to access was not fully utilized by aggregate models, which higher-income urban consumers. failed to recognize heterogeneity, dynamics and exogenous forces on livestock systems. House- Dairy reform in Ethiopia hold-level studies are not standardized and rarely identify and characterize key drivers and The question of dairying potential was particu- mechanisms for exploiting heterogeneity in pol- larly important in Ethiopia, where bad policies icy analysis. The analysis defined and addressed had limited growth in dairying. A sequence of the dichotomy in approaches to policy analysis studies – Mbogoh (1984), von Massow (1989) for developing countries’ livestock sectors and and Brokken and Senait Seyoum (1992), and the gap in analytical approaches and identifies later Mbogoh (1992) and Mbogoh and Ochuonyo aspects of the way forward. Evidence was pre- (1992) – identified policy and technical barriers sented of inconsistencies and practicalities that to dairying in Ethiopia. A later review of dairy emphasized the gap, but all studies presented development in Ethiopia over 50 years: (i) identi- evidence of integrative progress and listed op- fied trends in production, consumption, policy portunities for accelerating it. and development interventions; (ii) provided Economics and Policy Research at ILRI, 1975–2018 657 evidence of the potential impact of improved in what appeared to be a single market (e.g. fluid dairy cattle; (iii) examined factors that promote milk in Addis Ababa) explained why produ- smallholder dairying; and (iv) identified policy cers accepted widely different prices for a homo- and technology issues for public interventions geneous product in the same markets; and (Ahmed et al., 2003). (iii) contracts between producer and buyer co- Ethiopian dairying has passed through operatives played a central role in reducing three phases, matching shifts in national eco- transaction costs. nomic policies. Since the early 1990s, the transi- Subsequent work explained the impact of tion to a market economy has taken place and transaction costs and the choice of production the dairy sector has been growing. Milk produc- techniques on decisions to sell fluid milk to Ethi- tion during the 1990s expanded at an annual opian cooperatives (Holloway et  al., 2000a,b). rate of 3.0% compared with 1.6–1.7% during Creating local markets to minimize the time re- the preceding three decades. Some 60% of the quired to sell milk increases the number of produ- growth in milk production was due to herd cers and amounts sold. Institutional investments, growth; only 25% was due to higher productiv- such as the formation of milk groups, provided a ity per animal. less costly mechanism for increasing market There were institutional reasons for lower participation. Although milk groups are a sim- productivity in Ethiopia. Although dairy co- ple institutional innovation, they appear to be a operatives in Ethiopia were not as strong as in necessary first step in developing more sophisti- Kenya, cooperatives induced increased partici- cated cooperatives. pation of smallholders in fluid milk markets in Farmer cooperatives have been identified as the Ethiopian highlands. The survival of the catalysts to market participation. Analysis of milk groups that supplied inputs and processed data from the Ethiopian highlands where farm- and marketed dairy products depended on their ers organized themselves in a dairy cooperative continued ability to capture value-added dairy showed that male household heads and exten- processing and to return those value-added sion visitations affected cross-bred cow adoption benefits to their members. positively, while credit use and the number of Contrasts between Kenyan and Ethiopian local-breed cows currently milked affected adop- dairying were investigated to elucidate the roles tion negatively (Holloway et  al., 2000b). Male of cooperatives in reducing transaction costs. heads of household, extension visits and the The similarities of the highland agroclimates in number of local-breed cows affected output Kenya and in Ethiopia imply that dairy develop- positively, while credit use affected output nega- ment in Ethiopia would benefit from the Kenyan tively, as did distance to market. This study also experience, yet Ethiopia’s dairy system was for suggested that extension is a potentially import- many years less productive than Kenya’s. Part of ant catalyst for market expansion. Consequently, the difference was attributed to informality – in several important questions arise concerning the early 1990s, Ethiopia had not developed a the actual impacts of extension on participation, formal dairy system and some 88% of urban the number of extension-requesting households milk supply passed through informal markets willing to pay for services if it was privatized, the (Staal, 1995). Another reason was policy tax- corresponding demand schedule for extension ation – Kenya and Ethiopia both had an inter- services and the requisite conditions for the national comparative advantage, but Ethiopian existence of a private market for the service. supply was restrained by an overvalued cur- One study addressed these transactional rency causing low domestic producer prices. The issues (Holloway and Ehui, 2001). For each unit devaluation of the Ethiopian birr (ETB) in the increase in extension, the transaction cost was early 1990s greatly improved the potential of lowered by ETB0.62. Hence, extension was agricultural production for import competition shown as a promising market-entry catalyst. and for export markets. Furthermore, the willingness to pay for one add- Staal et  al. (1997) argued that: (i) the itional extension visit ranged from ETB0.6 to growth in smallholder dairying was limited by ETB6.7. The study estimated the marginal cost of high transaction costs for both production and each extension visit at ETB2.1, based on the an- marketing; (ii) transaction costs across producers nual extension budget of the local administrative 658 M. Jabbar et al. units and the estimated number of extension which was further demonstrated by predictions visits made during the year. The willingness to of technology uptake changing with a shift in pay estimates showed that some 39% of partici- infrastructure policy. Although requiring large pating households would purchase extension geo-referenced data sets and high-resolution GIS services. layers, the methodology demonstrated the po- Reinforcing the findings of Holloway et al. tential to better unravel the multiple effects of (2000a), other studies found that households location on farmers’ decisions on technology with a higher education level, a larger number and land use. of cows and a greater non-farm income were The above study was done within the frame- positively associated with value of sales of dairy work of ILRI’s Smallholder Dairy Project, a joint products. This suggests that income from the initiative with the Kenya Agricultural Research sale of milk, butter and cheese can be increased Institute (KARI) and the Kenya Ministry of Live- through education and training, especially tar- stock Development, which began in 1997 to geting women (Holloway et  al., 2000b; Ehui address farming practices, marketing and exten- et al., 2003; Lapar and Ehui 2004). sion. The policy aim of the Smallholder Dairy Project was to achieve a better policy environ- Dairy reform in Kenya ment for raw-milk trading to raise producer prices and to improve supply from smallholders. The most productive policy research at ILRI was A policy-change strategy was developed, which the long-term engagement in Kenyan dairying. included generating evidence about raw-milk Kenya was an attractive site for policy research markets and working with civil-society organ- in that its dairy sector was highly productive, it izations who were voices in policy advocacy and had high unrealized potential and major policy had connections to public agencies. barriers to achieving that potential, and it had a Until 1992, the Kenyan Dairy Board (KDB) base of technical and economic research to in- officially controlled dairy pricing and marketing. form policy recommendations. During the early 1990s, as input prices paid by In a study of adoption of improved dairy producers increased at a higher rate than the cattle and related technologies, a methodo- KDB-controlled prices of milk, producers began logical innovation was generated by applying to divert sales to the informal market. Conse- GIS-derived variables in econometric analysis. quently, supply to Kenya Co-operative Creamer- This study by Staal et  al. (2002) demonstrated ies (KCC) fell substantially, causing shortages of the usefulness of integrating GIS-measures into processed milk in the formal market. To stimu- analysis of technology uptake for better differ- late supply, the Kenyan government announced entiating and understanding locational effects. the liberalization of dairy prices and the lifting of A set of GIS-derived measures of market access the KCC monopoly on processed milk sales to and agroclimate were included in a standard urban areas. The market response was an in- household model of technology uptake, applied crease in raw-milk supply to the KCC and, conse- to smallholder dairy farms in Kenya, using a quently, in supply of processed milk to retailers. sample of 3330 geo-referenced farm house- The benefits of policy reform were limited, holds. The three technologies examined were however, because few dairy traders entered the keeping dairy cattle, planting specialized fodder market owing to the dominance of KCC, which and using concentrated dairy feed. Logit es- obstructed price liberalization. The raw-milk timations were conducted that significantly sales policy, however, did not change. differentiated the effects of individual household The path from policy research to policy characteristics from those related to location. change in Kenya dairying has been well chron- The predicted values of the locational variables icled (Leksmono et  al., 2006; Kaitibie et  al., were then used to make spatial predictions of 2010b). A first step was to investigate dairy mar- technology potential. Comparisons were made ket liberalization with a policy analysis matrix with estimations based only on survey data, which (Staal and Shapiro, 1994). Following output demonstrated that, while overall explanatory price liberalization, Kenya continued reducing power may not improve with GIS-derived variables, government support and intervention within the latter yielded more practical interpretations, the livestock sector, specifically for veterinary Economics and Policy Research at ILRI, 1975–2018 659 and artificial insemination services. Policy ana- concerns were likely to lower milk consumption lysis by ILRI measured the changes between in Kenya, would reduce health benefits to the 1990 and 1995 in milk marketing and service country’s low-income consumers, and would provision by the dairy farmer cooperative soci- destroy the livelihoods of hundreds of thou- eties, which played a central role in meeting the sands of small producers and vendors. needs of dairy production (Owango et al., 1998). The KDB and milk processors repeatedly Most notable were the changes in the unregu- challenged the public statements made by the lated raw-milk market, which helped increase civil-society organizations but were unable to real market prices paid to producers by up to produce evidence to back up their anti-raw-milk 50%. Large increases were also observed in the claims, whereas the robust evidence from the provision of veterinary and artificial insemin- Smallholder Dairy Project strongly supported ation services by the dairy cooperatives, whose the arguments of the civil-society organizations. producer base and credit facilities allowed them Some change in perceptions by the KDB oc- to compete with independent private traders. curred when it visited a group of project-trained A contentious policy issue following market milk vendors and saw that the vendors demon- liberalization was regulation of the informal strated good milk-safety practices. milk market, a complex network of farmers and A Dairy Policy Forum was held in May groups selling raw milk directly or through 2004 at the close of the Smallholder Dairy Pro- venders to consumers or shops. The price liberal- ject, where farmer advocates and senior officials ization of 1992 allowed other private milk pro- were prominent. At the forum, the minister of cessors to enter the market, causing the near livestock committed the government to passing collapse of KCC. The liberalization was also in- the stalled Kenya Dairy Bill and to take into ac- terpreted as allowing the sale of raw milk in count the mass of evidence and stakeholder urban areas, which was technically illegal, and opinion presented. In time, the KDB came to the raw-milk market quickly expanded through view the training and certification of raw-milk small vendors (Staal and Shapiro, 1994). By traders as an intermediate step towards formal- 2000, this market was estimated to control 80– izing the country’s small-scale milk trade rather 90% of the total liquid milk market, even though than as a means to promote raw-milk trading. it was fiercely opposed by the KDB, and officials An external evaluation used a ‘Research retained the authority to confiscate illegal vend- and Policy in Development Outcome Assessment’ ors’ milk and equipment. Research by the Small- approach to document how the Smallholder holder Dairy Project found that this authority Dairy Project research and policy engagement imposed constraints on the markets for milk strategy led to the policy change (Leksmono from smallholders as the price for milk paid by et  al., 2006). The study concluded that the the vendor decreased with the quality of milk Smallholder Dairy Project was the principal bought, even though a larger volume of sales driver of the policy changes, notably because of would be expected to impose lower unit transac- the following: tion costs. This result was thought to be due to the fact that vendors were restricted to handling • Private processors changed their marketing relatively small quantities (e.g. 30  litres/day) strategy to focus on the value and safety of due to risks of confiscation (Staal et al., 2002). processed, packaged milk without overtly Larger producers opposed the reform. The attacking small-scale milk vendors. Some Kenya Dairy Processors Association launched a processors also encouraged small-scale milk high-profile ‘Safe Milk’ campaign against raw- vendors to trade in processed products. milk marketing. The Smallholder Dairy Project • Virtually all subsequent projects in Ken- and its civil-society organization partners re- ya’s dairy subsector have used the research sponded by publishing some of their results in results of the Smallholder Dairy Project, the local news media. In addition, the civil-society and many have also linked with the pro- organization partners held a press conference to ject’s implementing institutions in other contest this campaign’s anti-raw-milk messages dairy- related activities, both in service- using evidence from the Smallholder Dairy delivery and policy-related areas (Leksmono Project showing that unsubstantiated health et al., 2006). 660 M. Jabbar et al. • One point of opposition to reform was by other projects led by the national agricultural health risks from drinking raw milk. Prod- research system, ministry and non-governmen- uct quality analysis showed that if milk is tal organizations in Kenya and elsewhere in East boiled, a near-universal practice in Kenya, Africa. The project created greater regional it is almost entirely safe. To further improve awareness among policy makers in Ethiopia, hygiene in the informal milk sector, the Tanzania and Uganda of pro-poor policy impli- Smallholder Dairy Project developed a cations of small-scale milk markets. The Associ- training programme for informal vendors ation for Strengthening Agricultural Research to teach them improved practices for hand- in Eastern and Central Africa (ASARECA) and ling milk. its policy programme, the Eastern and Central Africa Programme for Agricultural Policy Ana- Eventually, the work of the Smallholder lysis (ECAPAPA), built on the Smallholder Dairy Dairy Project achieved an agreement to train Project and ILRI policy recommendations to and certify small traders of raw milks, with the seek harmonized pro-poor dairy policies in the KDB taking up the training and licensing of region. Through ECAPAPA, dairy policy makers traders using guidelines and training materials and regulators in Rwanda, Tanzania and Uganda developed by the Smallholder Dairy Project. adopted new institutional approaches and ap- There were further revisions of the draft Dairy propriate technologies to harmonize stand- Industry Act, stalled since 1997, to recognize ards and improve informal milk markets across and formalize the role of small-scale raw-milk the region. traders and to increase the number of groups ILRI’s results were used to: (i) promulgate representing poor farmers. common dairy industry standards in East Africa; The change in Kenyan dairy policy to allow (ii) advance a regional agreement to promote greater market participation by small producers the movement of certified milk traders across had the principal effect of lowering transaction borders; (iii) publish training materials for milk costs, thereby raising prices to producers and standards, and provide certification of milk lowering prices to consumers. An economic sur- traders and accreditation of their trainers; and plus model was used to compute economic bene- (iv) train and provide certification of informal fits of these price shifts, with movements in the milk traders by involving private trainers. milk-supply curve being attributed to the policy In 2016, ILRI began working with the state changes affecting the informal market. Kaitibie government of Assam, India, where the infor- et al. (2010b)2 reported a best estimate of the net mal market suppled the great majority of local benefits of the reform to have an NPV of US$230 milk. Working closely with Assam’s Dairy Devel- million over 1997–2039. opment Department, the training and certifica- A study following the Kenya dairy reform tion approach of the Smallholder Dairy Project clarified the reasons for its political support. Infor- was adapted for Assam and was piloted locally, mal milk markets created more employment per again with a local non-governmental organiza- unit of product than did formal markets. A study tion as the main training service provider. Stud- by Omore (2004) of employment in milk markets ies showed that trained vendors sold safer milk in Kenya, Bangladesh and Ghana found that in- and demonstrated better knowledge of hygiene. formal milk markets employed up to five times as This is thought to be the first time in India that many people per 100 litres of milk handled. public funds have been devoted to improving the The Smallholder Dairy Project in Kenya not informal (or unorganized) milk market (Lindahl only benefited the dairy sector and the wider et al., 2017). economy in Kenya, its experiences created posi- tive externalities (international public goods) in the East Africa region and beyond. The partner- ship and communication strategy in this reform Comparisons of dairying in South led to the Smallholder Dairy Project receiving Asia and East Africa the 2004 CGIAR Communications Award. The project’s extension materials and market agent ILRI and FAO’s Pro-Poor Livestock Policy Initia- training materials and methods were taken up tive studied dairy development in East Africa and Economics and Policy Research at ILRI, 1975–2018 661 South Asia to assess the roles of policies and in- The study found that demand factors ex- stitutions and their impact on the poor (Staal plain much of dairy development in East Africa, et al., 2008a,b,c). The dairy sector in South Asia as shown by the rapid growth of milk production followed a different path to that of East Africa. in Kenya, Sudan and Uganda. Development of Consumption of dairy products is higher on formal milk markets, input markets, technology average in South Asia than in East Africa owing and policy do not explain the differences between to demand factors. Differences in growth in fast-growing countries and the rest, which may South Asia are more related to the possibility of imply that much of the increased production in expanding supply to match the growing demand response to demand came from herd growth ra- for dairy products. Multivariate econometric ther than from productivity growth. models incorporating technology and policy fac- This finding of lack of significance of input tors showed that India and Pakistan were able to market and technology on output growth was link the agricultural transformation originating supported by an earlier study by Freeman et al. in the Green Revolution to successfully expand (1998) that analysed the impact of credit on milk production; this is reflected in the contribu- milk output by smallholder dairy producers in tion of input markets and technology to growth Ethiopia and Kenya. It found no consistent rela- in milk production. In the case of countries with tionship between farmers’ credit constraints and slow growth in milk production, such as Bangla- their borrowing. However, farms that were credit desh and Nepal, development of cereal production constrained increased output more when given and feed markets and a growing demand did not access to credit than farms that were not credit induce a technical change in dairying, as was the constrained, indicating that the credit constraint case in India and Pakistan. As in East Africa, de- did limit the supply response. This finding indi- velopment of formal milk markets in South Asia is cated that demand for credit become important not associated with increased growth rates. to acquire inputs to increase output for market. Detailed analysis of the drivers and impacts These results suggest that adjusting supply of dairy sector growth on employment, income to type and quality of products demanded, ex- and nutrition was done for Ethiopia and Kenya panding demand by reducing consumer prices in East Africa and for India and Pakistan in and reducing transaction costs will contribute to South Asia. Although informal and commercial expand the dairy sector in East Africa. dairying coexist in both regions, informal production still dominates and is generally com- petitive. For example, the study conducted an Land rights analysis by district across India, which sought to find evidence that the presence and success of Land rights are legally or socially enforceable cooperatives was associated with greater dairy claims. They can be permanent via ownership or development. However, there was little evidence temporary via rental or other fixed-term con- towards this, which suggests that India’s well- tracts. The analysis of land rights in animal pro- known Operation Flood, which used revenues duction is important for two reasons. First, land from donated imported milk powder to fund is the primary factor in both grazing and mixed dairy cooperative development, did not play any systems. Second, nearly all modern efforts to ex- significant role in driving dairy development in pand agricultural production necessarily involve India, where to this day the dairy cooperative more intensive land use, lowering the amount of sector retains a relatively small market share. land used per unit of output, whether that out- The evidence suggests that relatively efficient put is forage, an arable crop such as rice or a per- informal milk markets played the key role in manent crop such as coffee. linking producers to growing consumer de- ILRI research on land rights has had four mand. In fact, the dairy industry grew more themes: (i) early work on pastoral systems, quickly in Pakistan, where cooperatives played focusing on group ranches and enclosures; almost no role, than in India. Policies that build (ii) studies on land tenure, resource allocation on traditional production systems, with new and productivity beginning in the early 1990s; focus on employment, food safety and quality, (iii) collective action for common resource man- are expected to be pro-poor. agement; and (iv) land rights and fodder trees. 662 M. Jabbar et al. Pastoral systems farming, falling range productivity outside the enclosures and falling costs of enclosures. Land tenure differs markedly between pastoral The Maasailand study (Solomon Bekure and mixed systems. Pastoral/agro-pastoral sys- et al., 1991) conducted in Kenya was exceptional tems typically use land held in common. Mixed in that it made detailed policy recommendations systems depend on private or social land tenure about the land rights of pastoralists. that recognizes certain rights that can be appro- The Borana study (summarized by Coppock, priated by individuals or communities. 1994) in southern Ethiopia found that the Borana A common historical view of land rights in system was moving from traditional pastoralism African pastoralism was that such rights did not to a semi-sedentary system with more reliance exist or were not enforced. It was argued that on crops and private grazing. This study high- this market failure made land use inefficient by lighted the need for a strengthening of trad- weakening incentives to improve it. A common itional authority in resource management. It political and development perspective on pas- further concluded that the agroecological diver- toral land tenure was that it could be ignored as sity of the Borana rangelands called for selective a subject of scientific evaluation; for example, policies that supported crop–livestock integra- the first major ILCA book was entitled Evaluation tion and extensive livestock production as neces- and Mapping of Tropical African Rangelands but sary rather than a ‘one policy fits all’ approach said nothing of land tenure or of changes in to the entire area. traditional forms of land management (ILCA, In the early 1990s, ILRI sponsored two lit- 1975) as possible remedies on overgrazing. Pratt erature reviews on land tenure and property and Gwynne (1977) mentioned pastoral tenure rights. The reviews of 18 studies on land tenure in East Africa as a barrier to be dismantled on in Africa (Swallow, 1994; Swallow and Bromley, the path to stopping overgrazing, to sedentariz- 1995) asked the following questions: ing pastoralists and ultimately to a generalized How do property institutions affect the use extension of ranching. The historical work of • and management of resources? Gallais, who had meticulously characterized the How do property institutions create or deny land administration of the Peulh herders of cen- • opportunities for the adoption of new tech- tral Mali in the 1950s (Gallais, 1967) and later nologies and expansion of agricultural pro- proposed a modern legal codification of that sys- duction? tem (Gallais and Boudet, 1980), had little effect3. How does the structure of government af- Land rights of common property range- • fect property institutions? lands was, on the other hand, of immediate How do changes in economic and tech- interest to ILCA researchers. Sandford’s magis- • nical conditions affect resource use and terial book, Management of Pastoral Develop- property institutions? ment in the Third World, which is still the most often cited work in the history of ILCA/ILRAD/ The review by Swallow and Bromley (1994) ILRI, presented a scheme for allocating land indicated that groups of livestock owners could ‘among uses and among users’ (Sandford, 1983, manage common property rangelands without pp. 135–136). formal organizations or institutions if the group Early research published by ILCA on Soma- was relatively small, if entry into the group was lia and Sudan found spontaneous range enclos- costly and if the members of the group did not dis- ures – the assertion of private property rights in count the future heavily. Thus, a local rangeland grazing and the defence of those rights by fen- management regime could only be effective if its cing (Behnke, 1986). Such spontaneous en- institutions were governed locally. It was argued closures were influenced by density-dependent that it was more effective for governments to en- factors – commercial animal husbandry, com- force boundaries among groups than to seek to mercial fodder markets and ‘heavy stocking of establish the internal group conditions for efficient pastures’ – and by density-independent factors – resource management (McCarthy et al., 1999). drought, water development and official land- A collaborative project titled ‘Property rights, tenure policies. Behnke found that spontaneous risk, and livestock development’ implemented by enclosures responded to increasing profits from ILRI, IFPRI and the Göttingen Research Institute Economics and Policy Research at ILRI, 1975–2018 663 for Rural Development during 1996–1999 Derg) took power. Under the monarchy, most of sought to support reforms of property institu- the land had been held by the aristocracy and tions and land policies in the semi-arid areas of the church. The Derg nationalized all land and sub-Saharan Africa. The specific objectives were: redistributed it to farmers on a per-capita basis (i) a better understanding of how environmental using local norms, giving them usufruct with no risk affects the use and management of re- right to sell, rent or transfer. There was provision sources under alternative property rights re- for periodic redistribution when new families gimes; (ii) identifying circumstances under were formed or when some families abandoned which land use and property rights change; and farming. The period of the Derg was one of sus- (iii) identifying how policy and other external pended animation as far as ILCA research on interventions can assist communities to achieve land issues was concerned. desirable pathways and mitigate negative im- Property institutions changed slowly after pacts of undesirable pathways (ILRI, 2000). the overthrow of the Derg in 1991. The new Part of the ILRI/IPFRI/Göttingen study fo- government began to tolerate decollectivization, cused on the Borana rangelands in semi-arid labour mobility and informal renting within southern Ethiopia (Kamara, 2001), where IL- extended families. Changes in land use and the CA’s historic 1994 study of the pastoral system resurfacing of rural factor markets following de- in the Borana area was characterized by exten- collectivization provided evidence that emerging sive livestock production and was a valuable factor markets brought better land use. There source of young stock for power and for export were observed changes in adoption of soil con- (Coppock, 1994; see Chapter 15, this volume). servation, tree planting, crop rotation and fallow Development in Borana was limited by aridity, practices, and increased use of organic and inor- causing low plant biomass productivity, and by ganic fertilizers associated with the new land periodic droughts, causing herd deaths. policy. Selling, hiring, renting, and trade of land, The Kamara (2001) study focused on the ef- labour and draught animals also grew (Omiti fects of environmental risk, market variables and et al., 1999, 2000). population pressure on land use and property The study of land rights in Ethiopia after rights. The results largely conformed to the prin- the collapse of the Derg was an important part cipal hypotheses about institutional change. of ILRI’s policy research in the 1990s. One study Community cooperation in resource manage- identified factors influencing the evolution of ment was determined by demography, wealth, land-tenure institutions to determine the effect off-farm income and social capital. Rainfall vari- of land tenure on investment, productivity and ability affected stock densities only in areas of efficiency in crop–livestock systems and to assess high rainfall variability. Market variables did not the impact of tenure on household access to determine stock densities or community level co- feed. For instance, the issue of land access by pri- operation but did affect land allocation to crops. vate commercial investors and land tenure and Changes in property rights were explained by a farming practices in the highlands of Ethiopia ban on wildfires, the creation of peasant associ- was presented in one paper (Gavian and Amare ations, sedentarization programmes and devel- Teklu, 1996) and a second paper presented evi- opment interventions (Kamara, 2001). A related dence on the nature of access to land by farmers study in the same area had examined the evolu- in one region of the Ethiopian highlands (Gavi- tion of land rights (Kamara, 2000). Kamara an and Ehui, 1999). found substantial privatization of land, related to Two studies dealt with the efficiency of land the change in national policies after the fall of the tenure contracts (Gavian and Ehui, 1999; Derg in 1991 and, chiefly, as function of rapid Ahmed et al., 2002). The first indicated that, al- growth in population density and cultivation. though the informally contracted lands were farmed 10–16% less efficiently, the hypothesis Land tenure, resource allocation that land tenure is a constraint to agricultural and productivity productivity was rejected. The second found higher technical efficiency between owner- An armed revolution overthrew the Ethiopian cultivated or rented plots and sharecropped, or monarchy in 1974 and a military regime (the borrowed plots. This difference was attributed to 664 M. Jabbar et al. restrictions imposed on the tenant in the share- tenure insecurity in western Niger would not cropping and borrowing contracts, which some- justify a major change in the tenure system. times involved labour and animal power supply by the tenant. A mild policy recommendation Collective action for common resource was to ‘facilitate more efficient transactions’ management Benin and Pender (2001) found that crop yields in the Amhara region were significantly A subset of the land tenure–productivity prob- higher, particularly in villages where the last lem is that of collective action in common re- major land redistribution took place in 1997– source management. A long-term area of study 1998. The authors also found that plots on by ILCA and later by ILRI was traditional agri- which households felt more secure (i.e. expect- culture in highland central Ethiopia which was ing to operate the plot for the next 5 years) were long constrained by lack of modern inputs, a associated with higher crop yield, suggesting variable environment and a severe risk of soil that security of tenure may be associated with erosion. Collective watershed management had other yield-enhancing management practices. been proposed as a model to manage modern Together, these results suggest that improving inputs while controlling soil erosion. The policy tenure security can bring about substantial in- question was whether collective watershed crements in crop productivity. management, as a form of tenure, could achieve Examining the evolution since 1991 in these policy objectives. land rental markets of the highlands of nor- A public-goods problem of watershed devel- thern Ethiopia, Benin et al. (2005) showed that opment in Ginchi, Oromia, was presented in a changes in the production environment and nat- game-theoretical model to study the logic of vol- ural resource endowments, changes in human untary contributions to an indivisible public capital, access to credit, commercialization of good: namely, a central drainage channel to cereal production and tenure security are the solve a waterlogging problem that constrained major forces contributing to the changes in land early planting of a high yielding wheat variety rental arrangements. Reduction in production (Gaspart et al., 1998). The most striking result of risk, through increased availability of moisture this study was that there is indeed a clear posi- or reduced degradation of soil, has reduced the tive relationship between the magnitude of per- need for risk-pooling arrangements associated sonal stakes and the effort spent on building the with sharecropping in favour of fixed-rent drainage channel. In other words, in the equilib- leases. Furthermore, increasing commercializa- rium selection process, a social norm of the kind tion of cereals caused an increase in land ren- ‘from each according to his expected gains’ tals, while an increase in credit supply caused seems to have been at work to favour coordin- an increase in fixed-rent leases. The same work ation of individual efforts. Out of 33 members of showed that alternative land rentals had a posi- the community who contributed to drain con- tive impact on cereal yields, suggesting that ten- struction, five had additional leadership roles. ure innovations after the eviction of the Derg Even though, taken singly, the leadership factor had evolved to reduce production inefficiencies. was the most statistically significant independ- The most widely cited paper on land tenure and ent variable, taken together it was self-interested investment in Ethiopia (Deininger and Jin, considerations that played the major role. 2006) found that tenure security could en- A bioeconomic model was developed to hance agricultural productivity and that public evaluate watershed management in central policy to improve tenure security would there- Ethiopia. The baseline in the model showed that, fore be justified. without technological or policy intervention, in- Related evidence from semi-arid Niger gave come and nutrition could not be sustainably im- evidence on ‘traditional land tenure [as] an im- proved in the watershed without serious soil pediment to allocative efficiency’ on millet farms losses (Okumu et al., 2004). Although cash in- (Gavian and Fafchamps, 1996). Gavian and Faf- comes could rise by more than 40% over a 12- champs found that land security was important year period, average soil losses could be as high for input allocation decisions, such as the use as 31  t/ha. With the adoption of a package of of labour and manure, but that the degree of new land-management technologies, however, Economics and Policy Research at ILRI, 1975–2018 665 the model projected, on average a 10% increase region revealed different causal factors for soil in cash incomes and a 28% decline in aggregate conservation adoption versus intensity of use erosion. The policy implications were: (i) the (Berhanu Gebremedhin and Swinton, 2001). need for more secure tenure to promote new Farmers’ reasons for adopting soil conservation technology; (ii) a shift from subsistence livestock measures varied sharply between stone terraces management to commercial; and (iii) a site- and soil bunds. Long-term investments in stone specific approach to land management within terraces were associated with more secure land the watershed (Okumu et al., 2004). tenure, more labour availability, proximity to the A study in Tigray region, in northern Ethi- household and learning opportunities via local opia, investigated the determinants of collective food-for-work projects. By contrast, short-term action and its effectiveness in managing com- investments in soil bunds were strongly linked to munity woodlots (Berhanu Gebremedhin et  al., insecure land tenure and the absence of local 2003, 2004). The studies suggest that collective food-for-work projects. actions may be more beneficial and more effect- In Ethiopia, particularly in the Amhara re- ive when managed at the village level rather gion, one source of tenure insecurity was land than at a county (wereda) level. Collective actions redistribution, which had been ongoing since were more productive when external interven- 1974 to equalize land holdings and quality tions were demand driven rather than imposed. across households. However, its short- and Population density and market access affected long-term effects may have mixed impacts on the probability of successful interventions. Collect- farmer land management and productivity. Ex- ive actions are more successful in intermediate- pectations of future land redistribution may population-density communities with poorer undermine farmers’ incentives to invest in land market access. At higher population densities improvements and soil fertility, as the farmers’ and with better market access, private approaches ability to reap the benefits of such investments is were more effective. undermined. Redistribution might, however, im- Berhanu Gebremedhin and Swinton (2003) prove access to land of households that have examined the relationships among public and relative surpluses of other important factors of private conservation investments. Public con- production, such as labour, oxen or cash to pur- servation campaigns on private land reduced chase inputs, particularly in the context of pro- adoption of stone terraces and soil bunds. hibited land sales and restricted lease markets Whereas capacity factors largely influenced the that exist in Ethiopia. Thus, land redistribution adoption decision, expected returns carried may increase the intensity of land management more influence for the intensity of stone terrace and use of purchased inputs, which may in turn adoption (measured as metres of terrace per increase productivity. hectare). More stone terracing was built where A research project in the Amhara region of fertile but erosion-prone silty soils in higher Ethiopia looked at land degradation and identi- rainfall areas offered valuable yield benefits. The fied options (Pender et  al., 2001). That project intensity of terracing was also greater in remote classified geographical units into various devel- villages where limited off-farm employment op- opment domains defined by combining produc- portunities reduced construction costs. These tion potential or ecology, population pressure results highlighted the importance of appropriate (high versus low) and market access (high ver- public interventions. Direct public involvement sus low). It has been found that there are signifi- in constructing soil conservation structures on cant differences in the extent of degradation and private lands appeared to undermine incentives its causes across the various development do- for private conservation. When done on public mains. Therefore, there are no one-size-fits-all lands, however, public conservation activities solutions to the problems across the domains. can encourage private soil conservation. Secure Technology and institutional options suitable for land tenure rights clearly reinforced private in- different domains to increase productivity and centives to make long-term investments in soil reduce degradation need to be introduced. conservation. A review article by Williams (1998) covered A related issue was how tenure security in- common property issues in semi-arid West fluenced investment in land. A study in Tigray Africa, specifically the problems created by 666 M. Jabbar et al. population growth, land pressure on water and alley farming in the West and Central African grazing, the lack of participation in governance land markets studied, though Adesina et  al. by resource users, and the role of the state in re- (2000) showed that relieving the specific land solving non-market conflicts. tenure constraints faced by women farmers would be necessary to raise their share of bene- Land tenure and fodder trees fits from alley farming or from other fallow substitutes. The International Institute for Tropical Agri- culture (IITA) had for many years studied alley farming, a system in which leguminous trees were planted between rows of food crops, Livestock and poverty such as maize or cassava. Nitrogen fixed by the trees could be returned to the soil as mulch Poverty was not a theme of ILCA/ILRAD/ILRI for uptake by crops, or the leaves could be cut research before the mid-1990s and the words and fed to livestock. Long-term collaboration ‘poor’ or ‘poverty’ as keywords in published among IITA, ILCA/ILRI and national pro- work rarely appear before 2000. There was some grammes in West Africa investigated agro- analysis of wealth disparities in Maasailand in nomic and economic aspects of leguminous the 1980s by King et  al. (1984) and Grandin tree farming. (1988) but no systematic or even sporadic effort Given the long-term character of tree in- to relate ILRI’s work to poverty in Africa, or any- vestments, adequate land tenure was thought to where else, before 2000. be needed to provide incentives to plant and ILRI adopted the theme of livestock as a maintain trees. One study included results from ‘pathway out of poverty’ for its 2002–2010 a sample of 248 farms in southern Nigeria be- strategy (ILRI, 2002). Two landmark studies – tween 1984 and 1991. While that study did not Perry et al. (2002) and Thornton et al. (2002) – collect tenure data, it did show that high turn- examined welfare among livestock keepers and over in plot ownership had no effect on tree paths by which they might escape poverty. Sub- farming (Lawry et al., 1994, p. 3). sequent work identified three paths along which A wider study in humid West and Central research might assist by: (i) securing the assets Africa tested the land tenure argument. This of the poor; (ii) improving the productivity of as- work characterized land and tree tenure prac- sets; and (iii) encouraging market participation tices and their implications for tree management by the poor. in Cameroon, Nigeria and Togo. The review Thornton et  al. (2002) produced the first found that 66%, 50% and 56% of the land, re- set of maps to locate poor livestock keepers by spectively, in Cameroon, Nigeria, and Togo was country, region and production system. They under tenure that provided long-term security estimated that out of nearly 1 billion poor and was, therefore, favourable for adoption of people living in the developing world, about 550 alley farming (Lawry and Stienbarger, 1991, million depended on livestock for their liveli- p. 62). Tenure had a significant role in the adop- hoods, most of them located in sub-Saharan Af- tion, continuation and discontinuation of alley rica and South Asia. Some 366 million and 103 farming. Because a significant proportion of the million livestock-dependent poor people live, land in the three countries was under a favour- respectively, in rain-fed and irrigated mixed sys- able tenure system, it was concluded that land tems, another 30 million in rangelands, and the tenure was not a major constraint to the adop- remaining 50 million or so in highlands and tion of alley farming, if other favourable factors other areas. were present (Lawry and Stienbarger, 1991; Subsequent microeconomic studies assessed Lawry et  al., 1994). In a study of southwest poverty dynamics and its relation to livestock. Cameroon, Adesina et al. (2000) found no statis- Kristjanson et  al. (2004) followed over 1700 tically significant effect of land tenure security households in 20 communities in western on the probability of adopting alley farming. Kenya. The communities differed in population An aggressive policy of tenure reform would density, farm size, agricultural potential, pov- therefore not be generally necessary to promote erty rate and human immunodeficiency virus Economics and Policy Research at ILRI, 1975–2018 667 prevalence. As they emerged from poverty, are wholly or partially dependent on the live- households typically first acquired food, then stock economy. clothes, shelter, primary education and small Little et al. (2008) examined poverty among animals, including chickens, sheep and goats. Kenya pastoralists. They argued that external ob- The results showed movement by households servers tended to ‘homogenize’ the concept of into and out of poverty over the 25-year period. ‘pastoralist’ by failing to acknowledge the diverse Of the households that had escaped poverty, livelihoods, wealth and income in pastoral areas. 73% mentioned diversifying income into cash The study concludes that what is not needed is crops and/or selling food crops when a house- another development label (stereotype) that hold member obtained an off-farm job, 57% equates pastoralism with poverty, thereby em- mentioned cash crop production and 42% men- powering outside interests to transform rather tioned that they diversified their on-farm incomes than strengthen pastoral livelihoods. through livestock, ranging from poultry to dairy. Radeny et al. (2007) showed that education On-farm diversification of income sources away among Tanzanian pastoralists influenced liveli- from a sole reliance on crops through invest- hood choices and improved the viability of pas- ment in chickens, sheep, goats and/or cattle toralism by diversifying it with crop production. helped many of the households in the study to escape poverty. Poor health, health-related ex- penses and funerals were the principal reasons cited by households for having fallen into poverty. Food security and nutrition The slaughter of livestock to meet emergency needs was mentioned by 63% of households as a A fundamental policy question under the head- reason for falling into poverty. ing of ‘livestock and poverty’ is how the benefits Kristjanson et  al. (2007) replicated the of technical change accrue to the rich and the community approach in some 3800 households poor and between women and men. The ques- in two regions of highland Peru, based on 10-year tion is especially relevant when technical change and 25-year recall. The reasons for movements involves a cash good such as milk or meat, com- into or out of poverty were identified at commu- modities not consumed in large quantities by nity and household levels, as was the role of live- most poor households, raising the possibility stock in the different paths. Diversification of that producing cash goods can worsen the nutri- income through livestock and intensification of tion of the poor (Pinstrup-Andersen, 2000). livestock activities through improved breeds Studies in Ethiopia on which ILRI collaborated helped many households escape poverty, but tended to reject that adverse possibility. these results varied across households. A study was started in 1997 in collabor- Ouma et  al. (2003), in a study in Kenya, ation with national institutions near Holetta in used data from a survey of cattle-keeping house- the highlands of Ethiopia. The work involved holds in intensive, semi-intensive and extensive an on-farm trial of cross-bred dairy cows and systems. This work assessed the contribution of animal draught power to assess the nutritional non-market benefits of cattle to the competitive- impacts of market-oriented dairying. A first ness and survival of smallholder enterprises. analysis evaluated the nutritional and health Some 50–70% of the benefits from smallholder status of women and children in households cattle are non-cash and smallholder cattle pro- with and without cross-bred cows (Shapiro duction systems are relatively competitive and et al., 2000; Ahmed et al., 2000). Malnutrition, efficient in the utilization of household produc- as measured in pre-school children by stunting, tion factors when non-market benefits are taken wasting and underweight, and as measured by into consideration. This is especially so for exten- body mass index in adult women, was lower in sive systems, which are non-market-oriented. households with cross-bred cows than in those The study concluded by emphasizing the import- with local cows. Calorie, protein and nutrient ance of the non-market roles of cattle in evalu- intake were significantly higher in the cross- ations of smallholder cattle production systems, bred cow group. as this will have a bearing on any policy-related The analysis further assessed the effects of interventions whose target are households that milk and income on decision making (Haider 668 M. Jabbar et al. et al., 2000). Women in households with cross- health and nutrition. Anthropometrical indica- bred cows contributed over 80% of household tors – stunting (height for age), underweight expenditure on food. A second extension of the (weight for age) and wasting (weight for height) – study revealed that steady increases in income are generally used as means of assessing from dairy in Ethiopia translated directly into in- prevalence of malnutrition among pre-school creases in expenditure on purchased food, non- children or children under 5  years. A study in food and farm inputs (Ahmed et al., 2000, 2003). highland Ethiopia tested the hypothesis that ac- Tangka et al. (2002) analysed the food se- cess to animal-source foods affected nutrition in curity and supply effects of smallholder dairying pre-school children (Okike et al., 2005). A child’s in peri-urban Ethiopia. Econometric analysis of nutritional and health status are jointly deter- panel data was used to evaluate the effects of mined by dietary intake, maternal wellbeing and dairying on food consumption, calorie intake the state of the physical environment as it influ- and marketed surplus in a treatment group of enced agricultural production and health status. households in contrast to a control group with- Presence of dairy cows in the household signifi- out the dairy technology. There were substantive cantly contributed to the health of children. The and statistically significant improvements in findings implied the need for multi- or transdisci- food security and marketed surplus with improved plinary approaches to research and develop- cattle. These impacts were reflected mainly ment incorporating heath, nutrition, sanitation, through the effects of income and wealth, meas- and farming practices for improving the health ured by animal value and land area. Household and nutrition of rural households. income had a positive and significant effect on A study in Selale District, in the Ethiopian food consumption. Regression estimates show highlands, examined the relationship between that elasticity of expenditure on food with re- smallholder dairying, time allocation by gender spect to income, animal value and cropland and income receipts by gender (Lenjiso et  al., area at the mean levels was respectively 0.29, 2016). In market participant households, in- 0.18 and 0.26. The largest share (63%) of the come from milk was higher because of higher difference in calorie intake between the cross- output and marketed surplus, but control of bred and local-breed cattle households was at- income shifted from women to men compared tributed to differences in the explanatory vari- with non-participant households. Policy lessons ables, while the estimated parameter differences from this work were inconclusive. between the two groups accounted for 37% of the difference. The value of animal assets had a positive and statistically significant impact on calorie intake in both the combined and IBLI in the arid rangelands cross-bred cattle regressions. The increase in of Kenya and Ethiopia animal values for the cross-bred cattle house- holds was estimated to increase their caloric in- An IBLI project developed a market tool for risk take by 12.7% relative to the local-breed cattle management by pastoralists in arid and semi- households. The value of food marketed by the arid Kenya. Following the inception of house- cross-bred cattle group was 82% higher than hold surveys in Marsabit, Kenya, in 2009 and that in the local-breed cattle group. A total of the launch of the IBLI’s insurance product in 76% of the increase in the value of marketed January 2010 (Jensen et al., 2015), the IBLI model surplus food for the cross-bred cattle over the lo- has combined biological, economic and institu- cal-breed cattle groups was accounted for by the tional research involving scientists, herders, difference in household characteristics, while private firms and regulators to: (i) protect pas- only 24% of the increase could be attributed to toralists from livestock losses by assessing forage differences in the estimated parameters. House- availability during the rainy season(s), as an holds in market-oriented dairying increased their index of production risk among a sample of 924 income and animal values significantly com- herding households in Marsabit county of arid pared with households in traditional dairying. northern Kenya; (ii) measure household demog- Nutritional status of children under 5 years raphy, income and wealth in that sample over of age is often a good indicator of community a survey period of 5  years; (iii) define and sell Economics and Policy Research at ILRI, 1975–2018 669 insurance policies against covariate risks caused policy impact of IBLI was that the insurance by drought; (iv) measure an index of vegetation product had a favourable benefit:cost ratio com- to define a trigger for insurance payments, using pared with other social protection programmes remote sensing data; (v) analyse the consump- in Kenya (Janzen and Carter, 2019). tion and investment behaviour of pastoralists of IBLI expanded into southern Ethiopia in varying herd sizes to estimate the impact of in- 2012 and has since generated pay-outs to herd- surance on sales, income, consumption and ers of approximately US$370,000 (Matsuda herd viability; (vi) identify effective institutional et al., 2019). A major finding from the studies of and extension models for the uptake of the prod- IBLI in Ethiopia is that index insurance is a com- uct; and (vii) work with herders and private bro- plement, not a substitute, to traditional risk- kers to monitor demand for index insurance, to sharing mechanisms (Takahashi et  al., 2019). continue adaptive testing of the insurance instru- Since 2019, an IBLI product has been integrated ments and to analyse the development impacts into the Africa Risk Capacity (ARC) to offer index of IBLI (a recent summary of IBLI is given by insurance to national partners targeting pas- Fava and Jensen, 2020). toral regions. To date, more than 86,000 policies have been sold with the ARC micro-insurance Impact scheme and more than 25,000 pastoralists are protected through the macro-level programmes. There was widespread adoption of index insur- ance in Marsabit, although many herders did Policy lessons not renew their policies after seasons of low pay- outs (Jensen et  al., 2015, p. 3). Insurance had Successful policy is impossible without a base of three broad impacts in Kenya, First, insurance, data collection and analysis whether payments were triggered or not, had a Index insurance for livestock leaves sub- positive impact in maintaining consumption stantial idiosyncratic risks (Jensen et al., 2016), and in preserving livestock wealth (Janzen and with roughly 60–75% of risk uncovered. These Carter, 2019), through IBLI’s generation of idiosyncratic risks have to be managed by trad- roughly US$10 million in pay-outs to Kenya itional risk-sharing mechanisms or by associ- herders. During the drought of 2011, house- ated public policies such as social funds. holds in Marsabit county with IBLI coverage had The arid and semi-arid counties of Kenya higher incomes and milk production; (Jensen are poor enough and risky enough that commer- et  al., 2015), were 27–36% less likely to skip cial livestock insurance will need public finan- meals and were 22–36% less likely to make dis- cial support for some time. tress sales of livestock (Janzen and Carter, Market agents – insurance brokers, 2019). Jensen et al. (2017) found, over 3 years regulators, and extension and research of IBLI coverage, that average veterinary ex- collaborators – had insufficient capacity at the penditures doubled and livestock sales in non- onset of IBLI. The commercial and regulatory drought years increased by an average of 46% of capacities of Kenya have grown since the incep- the mean. tion of IBLI, but international research support Impact on the wider policy environment in will be needed for some time to maintain a flow Kenya is a second category in which the pro- of information and analysis on programme gramme had a strong effect measure through operations and outcomes. the expansion of IBLI by the government of Kenya as the Kenya Livestock Insurance Project (KLIP). KLIP now provides subsidized insurance Livestock sector analyses and master to 18,000 pastoral households, representing plans as part of development policies over 80,000 beneficiaries, across eight counties of northern Kenya, and plans to serve 100,000 ILRI has pioneered the use of system dynamics households across 16 counties by 2021. The models in agri-food and livestock value chains. 2016/17 drought was among the worst in One application was in Rich et  al. (2009) who Kenya in the past 20 years, and KLIP paid out assessed the viability of a two-stage export certi- $7 million to pastoralists. One indicator of the fication system in Ethiopia using quarantine 670 M. Jabbar et al. stations and feedlots to ensure disease-free and (http://projects.worldbank.org/P159382? higher-quality beef for export to Middle Eastern lang=en; accessed 8 March 2020), new donor markets.The model found that the costs of com- project financing of US$75 million and new pri- plying with SPS regulations did not constrain vate-sector investments of US$200 million. The competitiveness but that high feed costs would higher livestock productivity and income levels do so. Later models at ILRI evaluated sheep and resulting from the plan’s investment interven- goat marketing in Mozambique (Hamza et  al., tions are projected to lift more than 2.3 million 2014), reforms to improve competitiveness in of Ethiopia’s 11 million livestock-keeping house- the beef sector in Botswana (Dizyee et al., 2017) holds out of poverty. and assessments of animal disease and food safety (Grace et al., 2017; Rich et al., 2018). Lie et  al. (2017, 2018) used spatial tech- niques in a model of the dairy value chain in The Future Nicaragua to quantify the market effects of feed quality. The goal of ILCA/ILRAD/ILRI policy research A growing area of ILRI policy support has was to increase smallholder returns by: (i) improv- been the development of ‘livestock master ing the productivity of technologies through plans’. Such plans set priorities within livestock technical, economic and financial analysis; development strategies to generate public and (ii) identifying policy barriers that lower farm private investments. The government of Ethiopia prices, raise input costs or increase the financial, has developed a Growth and Transformation information and risk costs of new methods; and Plan II 2015–2020, which prioritizes agricul- (iii) building institutions to raise productivity, ture and livestock investments to reduce poverty, create assets and reduce the external costs of to raise national income, to increase exports and animal agriculture. to improve food and nutritional security. The We are unable to estimate most of the de- Growth and Transformation Plan includes a live- velopment benefits of policy research at ILRI stock master plan, based on an analytical tool and partners, for several reasons. One is that known as the Livestock Sector Investment Policy many policy studies made little or no effort to Toolkit (www.au-ibar.org/2012-10-01-13-08- calculate impact research on the policy process 42/news/171-au-ibar/451-the-alive-livestock- or on outcomes of policy changes. This pattern sector-investment-policy-toolkit-lispt; accessed began with early ILCA work, such as Addis An- 9 March 2020). teneh (1983, 1984, 1985, 1991) and the dairy The Ethiopia livestock master plan (Shapiro policy studies of Brokken and Senait Seyoum et  al., 2015) was based on a 15-year sectoral (1992), continuing with McCarthy et al. (1999) model of potential outcomes of livestock invest- and the contemporary livestock and poverty ments in terms of increased production and (Thornton et  al., 2002), animal health (Perry value added for technology and service invest- et al., 2002), and climate change investigations ments under associated policy scenarios. The (Thornton and Herrero, 2010). One recent in- modelling incorporated the red meat and dairy novation is the preparation of ‘livestock master value chains subject to constraints in animal plans’ (e.g. Shapiro et  al., 2015, for Ethiopia, health, feeds and genetics. The livestock master which developed benefit–cost results for specific plan, as derived from the sector model, com- policy measures). The book of Herrero et  al. prises a 5-year investment roadmap and assess- (2014) on African livestock futures proposed ments of potential medium-term impacts of specific policy measures, but there has been no combined technology and policy interventions, effort to cost those recommendations and to see and informed the Ethiopian government’s if they have been implemented. Growth and Transformation Plan II livestock A second reason is broader – the inter- targets for 2015–2020. national system has neglected the assessment of Since 2016, the plan has served as the basis policy research with the exception of IPFRI’s for new funding and projects for the country’s work. The reasons for this failure include the time livestock sector. This includes livestock invest- lag between research output and policy changes, ments of US$132 million by the World Bank the difficulty of attributing policy changes to Economics and Policy Research at ILRI, 1975–2018 671 research products, and the futile and counter- the work at ILRI and other centres. The ICRISAT productive demands by donors for simple an- village-level studies had the greatest scientific swers to complex questions in unrealistically impact of economics and policy work across the short periods, which leads to hasty and incon- international agricultural research institutions clusive studies. Exceptions to these generaliza- because the work was sustained for many years tions were ILRI’s research on the Kenyan dairy and was specifically linked to technology gener- policy, which had significant economic and cap- ation. Future field investigations of livestock acity development benefits (Kaitibie et  al., systems should renew the ICRISAT village-level 2010a; Leksmono et al., 2006), development of studies model over a sufficiently long period in the East Coast fever vaccine (see Chapter 6, this African and in other developing country situ- volume) and the public–private partnerships in ations. Latin America that led to the planting of large areas of the forage grass Brachiaria spp. There is an important contrast between the Acknowledgements extended data collection and analysis done as part of the ICRISAT village-level studies (Walker The authors thank Iain Wright and Isabelle and Ryan, 1990) in semi-arid central India and Baltenweck for comments. Notes 1 These were the same classes of measures recommended by Schiff and Valdés (1992) in a global review of agricultural incentives. 2 There are several versions of the Kaitibie et al. (2010b) paper. The version published in a major policy journal had only 37 citations to May, 2020. 3 Gallais was a member of the 1981 Quinquennial Review of ILCA (CGIAR/TAC, 1982). His deep knowledge of Sahelian pastoralism and land tenure is only faintly apparent in the findings of that Review. References Adesina, A.A., Mbila, D., Nkamleu, G.B. and Endamana, D. (2000) Econometric analysis of the determin- ants of adoption of alley farming by farmers in the forest zone of southwest Cameroon. Agriculture, Ecosystems & Environment 80, 255–265. Addis Anteneh. (1983) Financing animal health services in some African countries. Livestock Policy Unit Working Paper No. 1. ILCA, Addis Ababa. Addis Anteneh. (1984) Trends in sub-Saharan Africa’s livestock industries. ILCA Bulletin No. 18. ILCA, Addis Ababa. Addis Anteneh, A. (1985) Financing livestock services: do Directors of Veterinary Services have a distinct role? ALPAN network Paper No. 6. ILCA, Addis Ababa. Addis Anteneh. (1991) The financing and staffing of livestock services in sub-Saharan Africa: a cross-country analysis. Livestock Economics Division. Working Document No. 16. ILCA, Addis Ababa. Agyemang, K., Dwinger, R.H., Little, D.A. and Rowlands, G.J. (1997) Village N’Dama cattle production in West Africa: Six years of research in The Gambia. ILRI, Nairobi. Ahmed, M.M., Jabbar, M. and Ehui, S. (2000) Household-level economic and nutritional impacts of mar- ket-oriented dairy production in the Ethiopian highlands. Food and Nutrition Bulletin 21, 460–465. Ahmed, M.M., Berhanu Gebremehdin, Benin, S. and Ehui, S. (2002) Measurement and sources of technical efficiency of land tenure contracts in Ethiopia. Environment and Development Economics 7, 507–527. Ahmed, M.A., Emana, B., Jabbar, M.A., Tangka, F. and Ehui, S. (2003) Economic and nutritional impacts of market-oriented dairy production in the Ethiopian highlands. ILRI Socio-Economics and Policy Research Working Paper 51. ILRI, Nairobi. Akter, S., Jabbar, M., Ehui, S. and Goletti, F. (2004) Competitiveness of poultry and pig production in Vietnam: an application of policy analysis matrix. Quarterly Journal of International Agriculture 43, 1962–1979. 672 M. Jabbar et al. Asfaw Negassa and Jabbar, M A (2008) Livestock ownership, commercial off take and related reasons in Ethiopia. ILRI Research Report No. 8. ILRI, Nairobi. Asfaw Negassa, Costalgi, R., Metete, G., Jabbar, M., Oyieke, S.O., et al. (2008) Towards improving live- stock export marketing support services in the Somali context: survey findings and implications. ILRI, Nairobi. Ayele Solomon, Assegid Workalemahu, Jabbar, M.A., Ahmed, M.A. and Belachew Hurissa (2003) Livestock marketing in Ethiopia: a review of structure, performance and development initiatives. Socio-economics and Policy Research Working Paper 52. ILRI, Nairobi. Bahta, S. and Malope, P. (2014) Measurement of competitiveness in smallholder livestock systems and emerging policy advocacy: an application to Botswana. Food Policy 49, 408–417. Baker, D. and Enahoro, D. (2014) Policy analysis and advocacy for livestock-based development: the gap between household-level analysis and higher-level models. Food Policy 49, 361–364. Baltenweck, I. and Staal, S. (2007) Beyond one-size-fits-all: differentiating market access measures for commodities in the Kenyan highlands. Journal of Agricultural Economics 58, 536–548. Behnke, R.H. (1986) The implications of spontaneous range enclosure for African livestock development policy. African Livestock Policy Analysis Network Paper No. 12. ILCA, Addis Ababa. Benin, S. and Pender, J. (2001) Impacts of land redistribution on land management and productivity in the Ethiopian highlands. Land Degradation and Development 12, 555–568. Benin, S., Ahmed, M., Pender, J. and Ehui, S. (2005) Development of land rental markets and agricultural productivity growth: the case of northern Ethiopia. Journal of African Economies 14, 21–54. Benin, S., Ehui, S. and Pender, J. (2003a) Policies for livestock development in the Ethiopian highlands. Environment, Development and Sustainability 5, 491–510. Benin, S., Jabbar, M. and Ehui, S. (2003b) Livestock marketing in the Ethiopian highlands: changes in structure and conduct since market liberalization in 1991. Presented at the Workshop on Harnessing Markets for Agricultural Growth in Ethiopia: Bridging the Opportunities and Challenges, 7–8 July, ILRI, Addis Ababa, Ethiopia. Berhanu Gebremedhin and Swinton, S. (2001) Reconciling food-for-work, project feasibility with food aid targeting in Tigray, Ethiopia. Food Policy 26, 85–95. Berhanu Gebremedhin and Swinton, S. (2003) Investment in soil conservation in northern Ethiopia: The role of land tenure security and public programmes. Agricultural Economics 29, 69–84. Berhanu Gebremedhin, Pender, J. and Tesfaye, G. (2003) Community natural resource management: the case of woodlots in Northern Ethiopia. Environment and Development Economics 8, 129–148. Berhanu Gebremedhin, Pender, J. and Tesfaye, G. (2004) Collective action for grazing land management in crop–livestock mixed systems in the highlands of northern Ethiopia. Agricultural Systems 82, 273–290. Brokken, R.F. and Senait Seyoum. eds) (1992) Dairy Marketing in Sub-Saharan Africa. Proceedings of a Symposium Held at ILCA, Addis Ababa, Ethiopia, 26–30 November 1990. ILCA, Addis Ababa. Cartwright, T.C., Anderson, F.M., Buck, N.G., Nelsen, T.C., Trail, J.C.M., et al. (1982) Systems modelling in cattle production-an application in Botswana. World Animal Review 41, 40–45. CGIAR/TAC (1982) Report of the TAC quinquennial review of the International Livestock Centre for Africa (ILCA). TAC Secretariat, FAO, Rome. CIAT (1975) Beef production systems: reprint from Annual Report of CIAT, 1975. CIAT, Cali, Colombia. Coppock, D.L. (1994) The Borana plateau of southern Ethiopia: synthesis of pastoral research, develop- ment and change, 1980–91. Systems Study No. 5. ILCA, Addis Ababa. Costales, A., Delgado, C.L., Catelo, M.A., Lapar, M., Tiongco, M., et al. (2006) Scale and access issues affecting smallholder hog producers in an expanding peri-urban market: Southern Luzon, Philippines. IFPRI Research Report 151. IFPRI, Washington, D.C. Coughenor, M.B., Ellis, J.E., Swift, D.M., Coppock, D.L., Galvin, K.M., et al. (1985) Energy extraction and use in a nomadic pastoral ecosystem. Science 230, 619–625. Deininger, K. and Jin, S. (2006) Tenure security and land-related investment: evidence from Ethiopia. European Economic Review 50, 1245–1277. Delgado, C.L., Rosegrant, M., Steinfeld, H., Ehui, S. and Courbois, B. (1999) Livestock to 2020: the next food revolution. IFPRI, Washington, DC, FAO, Rome and ILRI, Nairobi. Delgado, C.L., Rosegrant, M.W. and Meijer, S. (2001) Livestock to 2020: the livestock revolution con- tinues. Paper presented at the annual meeting of the International Agricultural Trade Research Con- sortium (IATRC), 18–19 January Auckland, New Zealand. Dizyee, K., Baker, A.D. and Rich, K.M. (2017) A quantitative value chain analysis of policy options for the beef sector in Botswana. Agricultural Systems 156, 13–24. Economics and Policy Research at ILRI, 1975–2018 673 Ehui S.K., Ahmed M.M., Berhanu Gebremehdin, Benin S.E., Nin-Pratt, A. and Lapar, M.L. (2003) 10 Years of Livestock Policy Analysis. Policies for improving productivity, competitiveness and sustainable live- lihoods of smallholder livestock producers. ILRI, Nairobi. El Dirani, O.H., Jabbar, M. A., and Babikar, B. I. (2009) Constraints in the market chains for export of Sudanese sheep and sheep meat to the Middle East. ILRI Research Report No. 16. ILRI, Nairobi. Elbasha, E., Thornton, P.K., and Tarawali, G. (1999) An ex post economic impact assessment of planted forages in West Africa. ILRI Impact Assessment Series No. 2. ILRI, Nairobi. Fafchamps, M. and Gavian, S. (1996) The spatial integration of livestock markets in Niger. Journal of Afri- can Economies 5, 366–405. Fava, F. and Jensen, N. (2020) Lessons learnt from the implementation of index-insurance in the African drylands: toward early response and regional scaling. Available at: https://cgspace.cgiar.org/bitstream/ handle/10568/107356/PosterAGU2019.pdf?sequence=1&isAllowed=y (accessed 13 May 2020). Franzel, S. and Wambugu, C. (2007) The uptake of fodder shrubs among smallholders in East Africa: Key elements that facilitate widespread adoption. In: Hare, M.D. and Wongpichet, K. (eds) Forages: A Pathway to Prosperity for Smallholder Farmers. Proceedings of an International Symposium, Faculty of Agriculture, Ubon Ratchathani University, Thailand, pp. 203–222. Freeman, H.A., Ehui, S.K. and Jabbar, M.A. (1998) Credit constraints and smallholder dairy production in the East African highlands: application of a switching regression model. Agricultural Economics 19, 33–44. Fritz, S., See, L., McCallum, I., You, L., Bun, A., et al. (2015) Mapping global cropland and field size. Global Change Biology 21, 1980–1992. Gallais, J. (1967) Le Delta Intérieur du Niger: Étude de Géographie Régionale. IFAN, Dakar. Gallais, J. and Boudet, G. (1980) Projet de code pastoral concernant plus spécialement la région du Delta Central du Niger au Mali. GERDAT-IEMVT, Maisons-Alfort, France. Gaspart, F., Jabbar, M., Mélard, C. and Platteau, J.-P. (1998) Participation in the construction of a local public good with indivisibilities: an application to watershed development in Ethiopia. Journal of Afri- can Economies 7, 157–184. Gavian, S. and Amare Teklu (1996) Land tenure and farming practices: The Case of Tiyo Woreda, Arsi, Ethiopia. In: Mulat, D., Wolday, A., Tesfaye, Z., Solomon, B. and Ehui, S. (eds) Sustainable Intensifi- cation of Agriculture in Ethiopia. Proceedings of the second annual conference of the Ethiopian Agri- cultural Economics Society, 3–4 October, Addis Ababa, Ethiopia, pp. 74–97. Gavian, S. and Ehui, S. (1999) Measuring the production efficiency of alternative land tenure contracts in a mixed crop–livestock system in Ethiopia. Agricultural Economics 20, 37–49. Gavian, S. and Fafchamps, M. (1996) Land tenure and allocative efficiency in Niger. American Journal of Agricultural Economics 78, 460–471. Getachew Asamenew., Hailu Beyene., Negatu, W. and Ayele, G. (1993) A survey of the farming systems of Vertisol areas of the Ethiopian highlands. ILCA, Addis Ababa. Gettinby, G., Newson, R.M., Calpin, M.M.J. and Paton, G. (1988) A simulation model for genetic resistance to acaricides in the African brown ear tick, Rhipicephalus appendiculatus. Preventive Veterinary Medicine 6, 183–197. Giller, K.E., Tittonell, P., Rufino, M.C., van Wijk, M.T., Zingore, S., et al. (2011) Communicating complexity: integrated assessment of trade-offs concerning soil fertility management within African farming sys- tems to support innovation and development. Agricultural Systems 104, 191–203. Grandin, B.E. (1988) Wealth and pastoral dairy production: a case study from Maasailand. Human Ecology 16, 1–21. Gryseels, G. (1988) Role of livestock on mixed smallholder farms in the Ethiopian highlands. PhD thesis, State Agricultural University, Wageningen, The Netherlands. Gryseels, G. and Anderson, F.M. (1983) Research on farm and livestock productivity in the central Ethiop- ian highlands: initial results, 1977–1980. ILCA Research Report No. 4. ILCA, Addis Ababa. Haider, J., Shapiro, B.I., Tsegaye, D., and G/Wold, A. (2000) The nutritional and health status of women and children in households with and without cross-bred cows in Holetta, Wereda, Ethiopia. In: Jabbar, M.A., Peden, D.G., Mohamed Saleem, M.A. and Li Pun, H. (eds) Agro-ecosystem, Natural Resource Management and Human Health Related Research in East Africa. Proceedings of an IDRC/ILRI inter- national workshop held at ILRI, Addis Ababa, Ethiopia, 11–15 May 1998, pp. 124–136. Hall, D.C., Ehui, S.K. and Shapiro, B.I. (2004) Economic analysis of the impact of adopting herd health con- trol programs on smallholder dairy farms in Central Thailand. Agricultural Economics 31, 335–342. Hamza, K., Rich, K.M., Baker, D. and Hendrickx, S. (2014) Commercializing smallholder value chains for goats in Mozambique: a system dynamics approach. In: Proceedings in System Dynamics and 674 M. Jabbar et al. Innovation in Food Networks. Proceedings of the 8th International European Forum on System Dy- namics and Innovation in Food Networks, 17–21 February. International Center for Food Chain and Network Research, University of Bonn, Germany, pp. 117–134. Havlik, P., Valin, H., Mosnier, A., Obersteiner, M., Baker, J.S., et al. (2013) Crop productivity and the global livestock sector: implications for land use change and greenhouse gas emissions. American Journal of Agricultural Economics 95, 442–448. Herrero, M., Havlik, P., McIntire, J., Palazzo, A. and Valin, H. (2014) African livestock futures: realizing the potential of livestock for food security, poverty reduction and the environment in sub-Saharan Africa. Office of the Special Representative of the UN Secretary General for Food Security and Nutrition and the United Nations System Influenza Coordination, Geneva, Switzerland. Herrero, M., Thornton, P.K., Notenbaert, A.M., Wood, S., Freeman, H.A., et al. (2010) Smart Investments in Sustainable Food Production: Revisiting Mixed Crop-Livestock Systems. Science 327, 822–825. Holloway, G.J. and Ehui, S.K. (2001) Demand, supply and willingness-to-pay for extension services in an emerging-market setting. American Journal of Agricultural Economics 83, 764–768. Holloway, G., Nicholson, C., Delgado, C., Staal, S. and Ehui, S. (2000a) How to make a milk market: a case study from the Ethiopian highlands. Working Paper. ILRI, Addis Ababa. Holloway, G.J., Barrett, C.B. and Ehui, S. (2000b) Crossbred-cow adoption and milk-market participation in a multivariate count data framework. In: Bayesian Methods with Applications to Science, Policy and Official Statistics: Eurostat Special Issue. Eurostat, Luxembourg, pp. 233–242. ILCA (1975) Evaluation and Mapping of Tropical African Rangelands. Proceedings of a Seminar, 3–8 March, Bamako, Mali. ILCA, Addis Ababa. ILCA/ILRAD (1988) Livestock Production in Tsetse Affected Areas of Africa. Proceedings of a Meeting Held in Nairobi, Kenya, 23–27 November 1987. ILCA, Addis Ababa, and ILRAD, Nairobi. ILRI (2000) ILRI medium-term plan 2001–2003: making the livestock revolution work for the poor. ILRI, Nairobi. ILRI (2002) Livestock – a pathway out of poverty. ILRI's strategy to 2010. ILRI, Nairobi. Itty, P. (1992) Economics of Village Cattle Production in Tsetse Affected Areas of Africa: A Study of Tryp- anosomiasis Controlling Trypanotolerant Cattle and Chemotherapy in Ethiopia, Kenya, Côte d’Ivoire, the Gambia, Zaire and Togo. Hartung-Gorre Verlag, Konstanz, Germany. Itty, P., Chema, S., d’Ieteren, G.D.M., Durkin, J.W., Leak, S.G.A., et al. (1988) Economic aspects of cattle production and of chemoprophylaxis for control of trypanosomiasis in village East African Zebu Cattle in Kenya. In: Livestock Production in Tsetse Affected Areas of Africa. Proceedings of a Meeting Held in Nairobi, Kenya, 23–27 November 1987. ILCA, Addis Ababa, and ILRAD, Nairobi, pp. 360–376. Jabbar, M.A. (1993) Market potential for increased small ruminant production in southwest Nigeria. In: Ayeni A.O. and Bosman H.G. (eds) Goat Production Systems in the Humid Tropics. Proceedings of an International Workshop, 6–9 July 1992, Obafemi Awolowo University, Ile-Lfe, Nigeria, pp. 97–107. Jabbar, M.A. (1995) Market niches for increased small ruminant production in southern Nigeria. Oxford Agrarian Studies 23, 85–96. Jabbar, M.A. (1998) Buyer preferences for sheep and goats in southern Nigeria: a hedonic price analysis. Agricultural Economics 18, 21–30. Jabbar, M.A. and Akter, S. (2006) Market and other factors affecting farm-specific production efficiency in poultry production in Vietnam. Journal of Food Products Marketing 12, 99–113. Jabbar, M.A. and Akter, S. (2008) Market and other factors affecting farm specific production efficiency in pig production in Vietnam. Journal of International Food & Agribusiness Marketing 20, 29–53. Jabbar, M.A. and Diedhiou, M.L. (2001) Does breed matter to cattle farmers and buyers? Evidence from West Africa. Ecological Economics 45, 461–472. Jabbar, M.A. and di Domenico, C. (1990) Dairy consumption patterns in southern Nigeria. In: Dairy Marketing in sub-Saharan Africa. Proceedings of a Symposium, 26–30 November, ILCA, Addis Ababa, Ethiopia, pp. 185–202. Jabbar, M.A. and di Domenico, C. (1993) Demand for dairy products among the indigenous population of southern Nigeria. Indian Journal of Dairy Science 46, 363–370. Jabbar, M.A., Benin, S., Eleni Gabremedhin and Paulos, Z. (2008) Market institutions and transaction costs influencing trader performance in live animal marketing in rural Ethiopian markets. Journal of African Economies 17, 747–764. Jabbar, M.A., Swallow, B.M. and d’Ieteren, G.D.M. (1995) The demand for trypanotolerant cattle: Evidence on farmers breed preferences and breeding practices in southern Nigeria. Presented at the 23rd meeting of the International Scientific Council for Trypanosomiasis Research and Control, 11–15 September, Banjul, The Gambia. Jabbar, M.A., Swallow, B.M., d’Iteren, G.D. and Busari, A. A. (1997) Farmer preferences and market val- ues of cattle breeds of West and Central Africa. Socioeconomic and Policy Research Working Paper No 21. LPAP, ILRI, Addis Ababa. Economics and Policy Research at ILRI, 1975–2018 675 Jabbar, M., Swallow, B. and Rege, E. (1999) Incorporation of farmer knowledge and preferences in designing breeding policy and conservation strategy for domestic animals. Outlook on Agriculture 28, 239–243. Jabbar, M.A., Islam, S.M.F., Delgado, C., Ehui, S., Akanada, M.A.I., et al. (2005) Policy and scale factors influencing efficiency in dairy and poultry production in Bangladesh. ILRI, Nairobi. Jabbar, M.A., Rahman, M.H., Talukder, R.K. and Raha, S.K. (2007) Alternative institutional arrangements for contract farming in poultry production in Bangladesh and their impacts on equity. ILRI Research Report 7. ILRI, Nairobi. Jabbar, M.A., Baker, D. and Fadiga, M. (2010) Demand for livestock products in developing countries with a focus on quality and safety attributes: evidence from Asia and Africa. ILRI Research Report 24. ILRI, Addis Ababa. Jahnke, H.E. (1982) Livestock Production Systems and Livestock Development in Tropical Africa. Kieler Wissenschaftsverlag Vauk, Kiel, Germany. Jansen, H.G. (1992) Dairy consumption in northern Nigeria: implications for development policies. Food Policy 17, 214–226. Janzen, S.A., and Carter. M.R. (2019) After the drought: the impact of microinsurance on consumption smoothing and asset protection. American Journal of Agricultural Economics 101, 651–671. Jensen, N.D., Barrett, C.B. and Mude, A.G. (2015) The favourable impacts of index-based livestock insur- ance: evaluation results from Ethiopia and Kenya. ILRI Research Brief No. 52. ILRI, Nairobi. Jensen, N.D., Barrett, C.B. and Mude, A.G. (2016) Index insurance quality and basis risk: evidence from northern Kenya. American Journal of Agricultural Economics 98, 1450–1469. Jensen, N.D., Barrett, C.B. and Mude, A.G. (2017) Cash transfers and index insurance: a comparative impact analysis from northern Kenya. Journal of Development Economics 129, 14–28. Jones, P.G. and Thornton, P.K. (2009) Croppers to livestock keepers: livelihood transitions to 2050 in Af- rica due to climate change. Environmental Science and Policy 12, 427–437. Kaitibie, S., Omore, A., Rich, K. and Kristjanson, P. (2010a) Kenya dairy policy change: influence path- ways and economic impacts. World Development 38, 1494–1505. Kaitibie, S., Omore, A., Rich, K., Salaysa, B., Hooton, N., et al. (2010b) Influence pathways and economic impacts of policy change in the Kenyan dairy sector. ILRI Research Report. ILRI, Nairobi. Kamara, A. (2000) Ethiopian case study. In: McCarthy, N., Swallow, B., Kirk, M. and Hazell, P. (eds) Property Rights, Risk and Livestock Development in Africa. ILRI, Nairobi, and IFPRI, Washington, D.C., pp. 396–426. Kamara, A.B. (2001) Property rights, risk and livestock development in Ethiopia. Socioeconomic Studies on Rural Development 123. Kieler Wissenschaftsverlag Vauk, Kiel, Germany. Kebede Andargachew and Brokken, R.F. (1993) Intra-annual sheep price patterns and factors underlying price variations in the central highlands of Ethiopia. Agricultural Economics 8, 125–138. Kimani, T., Schelling, E., Bett, B., Ngigi, M., Randolph, T. and Fuhrimann, S. (2016) Public health benefits from livestock Rift Valley fever control: a simulation of two epidemics in Kenya. EcoHealth 13, 729–742. King, J.M., Sayers, A.R., Peacock, C.P. and Kontrohr, E. (1984) Maasai herd and flock structures in rela- tion to livestock wealth. Climate and Development 13, 21–56. Konandreas, P.A., Anderson, F.M. and Trail, J.C.M. (1983) Economic trade-offs between milk and meat production under various supplementation levels in Botswana. ILCA, Addis Ababa. Kristjanson, P.M. and Zerbini, E. (1999) Genetic enhancement of sorghum and millet residues fed to ruminants. An ex ante assessment of returns to research. ILRI Impact Assessment Series No. 3. ILRI, Nairobi. Kristjanson, P.M., Swallow, B.M., Rowlands, G.J., Kruska, R.L. and de Leeuw, P. (1999) Measuring the costs of African animal trypanosomosis, the potential benefits of control and returns to research. Agri- cultural Systems 59, 79–98. Kristjanson, P.M., Tarawali, S., Okiki, I., Singh, B.B., Thornton, P.K., et  al. (2001) Genetically improved dual-purpose cowpea: assessment of adoption and impact in the dry savanna regions of West Africa. ILRI Impact Assessment Series 9. ILRI, Nairobi. Kristjanson, P., Krishna, A., Radeny, M. and Nindo, W. (2004) Pathways out of poverty in western Kenya and the role of livestock. Pro-Poor Livestock Policy Initiative Working Paper No. 14. FAO, Rome. Kristjanson, P., Krishna, A., Radeny, M., Kuan, J., Quilca, G., et al. (2007) Poverty dynamics and the role of livestock in the Peruvian Andes. Agricultural Systems 94, 294–308. Lenjiso, B., Smits, J. and Ruben, R. (2016) Transforming gender relations through the market: the impact of smallholder milk market participation on women’s intra-household bargaining position in Ethiopia. Journal of Development Studies 52, 1002–1018. Lapar, M.L.A. and Ehui, S.K. (2004) Factors affecting adoption of dual-purpose forages in the Philippine uplands. Agricultural Systems 81, 95–114. 676 M. Jabbar et al. Lapar, M.L., Holloway, G. and Ehui, S. (2003a) Policy options promoting market participation of small- holder livestock producers: a case study from the Philippines. Food Policy 28, 187–211. Lapar, M.L., Vu Trong Binh and Ehui, S. (2003b) Identifying barriers to entry to livestock input and output markets in South-East Asia: the case of Vietnam. Socio-economics and Policy Research Working Paper No. 56. ILRI, Nairobi. Lawry, S.W. and Stienbarger, D.M. (1991) Tenure and Alley Farming in the Humid Zone of West Africa: Final Report of Research in Cameroon, Nigeria and Togo. LTC, Madison, Wisconsin. Lawry, S., Stienbarger, D. and Jabbar, M.A. (1994) Land tenure and the potential for the adoption of alley farming in West Africa. Outlook on Agriculture 23, 183–187. Leksmono, C., Young, J., Hooton, N., Muriuki, H. and Romney, D. (2006) Informal traders lock horns with the formal milk industry: the role of research in pro-poor dairy policy shift in Kenya. ODI Working Paper 266. ODI, London, and ILRI, Nairobi. Lie, H., Rich, K.M. and Burkart, S. (2017) Participatory system dynamics modelling for dairy value chain development in Nicaragua. Development in Practice 27, 785–800. Lie, H., Rich, K.M., van der Hoek, R. and Dizyee, K. (2018) Quantifying and evaluating policy options for inclusive dairy value chain development in Nicaragua: a system dynamics approach. Agricultural Systems 164, 193–222. Lindahl, J., Deka, R.P. and Grace, D. (2017) Improving the traditional dairy sector in Assam: policy, prac- tice and incentives. ILRI Research Brief 81. ILRI, Nairobi. Little, P.D., McPeak, J., Barrett, C.B. and Kristjanson, P. (2008) Challenging orthodoxies: understanding poverty in pastoral areas of East Africa. Development and Change 39, 587–611. Matsuda, A., Takahashi, K. and Ikegami, M. (2019) Direct and indirect impact of index-based livestock insurance in southern Ethiopia. The Geneva Papers on Risk and Insurance – Issues and Practice 44, 481–502. Mbogoh, S.G. (1984) Dairy development and internal dairy marketing in sub-Saharan Africa: performance, policies and options. LPU Working Paper No. 5. ILCA, Addis Ababa. Mbogoh, S.G. (1992) Marketing efficiency, pricing and policy implications: a case study of milk marketing in Addis Ababa, Ethiopia. In: Brokken, R.F. and Senait, S. (eds) Dairy Marketing in sub-Saharan Af- rica. Proceedings of a symposium, 26–30 November 1990, Addis Ababa. ILCA, Addis Ababa, pp. 239–256. Mbogoh, S.G. and Ochuonyo, J.B. (1992) Kenya’s dairy industry: the marketing system and the marketing and pricing policies for fresh milk. In: Brokken, R.F. and Senait, S. (eds) Dairy Marketing in sub-Saharan Africa. Proceedings of a symposium, 26–30 November 1990, Addis Ababa. ILCA, Addis Ababa, pp. 269–278. McCarthy, N., Swallow, B., Kirk, M. and Hazell, P. (eds) (1999) Property Rights, Risk, and Livestock De- velopment in Africa. ILRI, Nairobi, and IPFRI, Washington, D.C. McDermott, J.J. and Arimi, S. (2002) Brucellosis in sub-Saharan Africa: epidemiology, control and impact. Veterinary Microbiology 90, 111–134. McDermott, J.J. and Coleman, P.G. (2001) Comparing apples and oranges – model-based assessment of different tsetse-transmitted trypanosomiasis control strategies. International Journal for Parasitology 31, 603–609. McDermott, J., Grace, D. and Zinsstag, J. (2013) Economics of brucellosis impact and control in low-income countries. Scientific and Technical Review 32, 249–261. McIntire, J., Bourzat, D. and Pingali, P. (1992) Crop–Livestock Interaction in Sub-Saharan Africa. World Bank, Washington, D.C. Mohammed Mussa and Gavian, S. (1994) Privatization of animal health services: policy issues for Ethiopia. Paper prepared for the Fourth Annual Conference on the Ethiopian Economy, Hor Ras Hotel, Debre Zeit, Ethiopia. Montgolfier-Kouevi, C. and Vlavonou, A. (1981) Trends and prospects for livestock and crop production in Tropical Africa. ILCA Working Paper No. 5. ILCA, Addis Ababa. Mugunieri, L., Costagli, R., Osman, I.O., Oyieke, S.O., Jabbar, M., et al. (2008) A rapid appraisal of institu- tions supporting Somali livestock export. Improvement and diversification of Somali livestock trade and marketing project. ILRI Discussion Paper No. 14. ILRI, Nairobi. Mukhebi, A., Perry, B.D. and Kruska, R. (1992) Estimated economics of theileriosis control in Africa. Preventive Veterinary Medicine 12, 73–85. Nicholson, C.F., Thornton, P.K., Mohammed, L., Muinga, R.W., Mwamachi, D.M., et al. (1999) Smallholder dairy technology in coastal Kenya. An adoption and impact study. ILRI Impact Assessment Series. ILRI, Nairobi. Economics and Policy Research at ILRI, 1975–2018 677 Nin-Pratt, A., Hertel, T.W., Foster, K.A. and Rae, A.N. (2001) Productivity growth and catching-up: implica- tions for China’s trade in livestock products. Paper presented at the 2001 AAEA Annual Meeting, 5–8 August, Chicago, Illinois. Nin-Pratt, A., Bonnet, P., Ehui, S., Jabbar, M. and de Haan, C. (2003a) Benefits and costs of compliance of SPS rules: the case of Rift Valley Fever in Ethiopia. ILRI, Nairobi. Nin-Pratt, A., Lapar, M.L. and Ehui, S. (2003b) Globalization, trade liberalization and poverty alleviation in Southeast Asia: the case of the livestock sector in Vietnam. Paper presented at the Sixth Annual Con- ference on Global Economic Analysis, Scheveningen, Netherlands, 12–14 June. ILRI, Nairobi. Nyangito, H.O., Richardson, J.W., Mukhebi, A.W., Zimmel, P., Namken, J. and Berry, B.P. (1996a) Whole farm simulation analysis of economic impacts of East Coast Fever immunization strategies on mixed crop–livestock farms in Kenya. Agricultural Systems 51, 1–27. Nyangito, H.O., Richardson, J.W., Mundy, D.S., Mukhebi, A.W., Zimmel, P. and Namken, J. (1996b) Eco- nomic impacts of East Coast Fever immunization on smallholder farms, Kenya: a simulation analysis. Agricultural Economics 13, 163–177. Okike, I, Jabbar, M.A, Gugsa Abate and Lema Ketema (2005) Household and environmental factors influ- encing anthropometric outcomes among pre-school children in a rural Ethiopian community. Ecology of Food and Nutrition 44, 167–187. Okumu, B.N., Russell, N., Jabbar, M.A., Colman, D., Mohamed Saleem, M.A. and Pender, J. (2004) Eco- nomic impacts of technology, population growth and soil erosion at watershed level: the case of the Ginchi in Ethiopia. Journal of Agricultural Economics 53, 503–523. Omiti, J.M., Parton, K.A., Sinden, J.A. and Ehui, S.K. (1999) Monitoring changes in land-use practices follow- ing agrarian de-collectivisation in Ethiopia. Agriculture, Ecosystems and Environment 72, 111–118. Omiti, J.M., Parton, K.A., Ehui, S.K. and Sinden, J.A. (2000) Some policy implications of the resurfacing of rural factor markets following agrarian de-collectivization in Ethiopia. Human Ecology 28, 585–603. Omore, A.O. (2004) Employment generation through small-scale dairy marketing and processing: experi- ences from Kenya, Bangladesh and Ghana. A Joint Study by the ILRI Market-oriented Smallholder Dairy Project and the FAO Animal Production and Health Division, FAO, Rome. Ouma, E.A., Obare, G.A. and Staal, S.J. (2003) Cattle as assets: assessment of non-market benefits from cattle in smallholder Kenyan crop-livestock systems. In: Proceedings of the 25th International Confer- ence of Agricultural Economists (IAAE), 16–22 August 2003, Durban, South Africa, pp. 328–334. Owango, M.O., Lukuyu, B., Staal, S.J., Kenyanjui, M., Njubi, D. and Thorpe, W. (1998) Dairy co-operatives and policy reform in Kenya: effects of livestock service and milk market liberalization. Food Policy 23, 173–185. Pender, J., Berhanu Gebremedhin, Benin, S. and Ehui, S. (2001) Strategies for sustainable agricultural development in the Ethiopian highlands. American Journal of Agricultural Economics 83, 1231–1240. Perry, B.D. and Randolph, T.F. (1999) Improving the assessment of the economic impact of parasitic dis- eases and of their control in production animals. Veterinary Parasitology 84, 145–168. Perry, B.D., Randolph, T.F., McDermott, J., Sones, K. and Thornton, P. (2002) Investing in Animal Health Research to Alleviate Poverty. ILRI, Nairobi. Pingali, P., Bigot, Y. and Binswanger, H.P. (1987) Agricultural Mechanization and the Evolution of Farm- ing Systems in Sub-Saharan Africa. Johns Hopkins University Press, Baltimore/London. Pinstrup-Andersen, P. (2000) 25 Years of IFPRI Policy Research: Reflections. IFPRI, Washington, D.C. Powell, J., Fernandez-Rivera, S., Williams, T. and Renard, C. (eds) (1995) Livestock and sustainable nutri- ent cycling in mixed farming systems of sub-Saharan Africa, Vol. II: Technical Papers. Proceedings of an international conference, 22–26 November 1993, Addis Ababa, Ethiopia. ILCA, Addis Ababa. Pratt, D.J. and Gwynne, M.D. (1977) Rangeland Ecology and Management in East Africa. Robert E. Krieg- er, Huntington, New York. Radeny M., Nkedianye, D., Kristjanson, P., Herrero, M. (2007) Livelihood choices and returns among pas- toralists: evidence from southern Kenya. Nomadic Peoples 11, 31–55. Randolph, T.F., Schelling, E., Grace, D., Nicholson, C.F., Leroy, J.L., et al. (2007) Invited review: role of live- stock in human nutrition and health for poverty reduction in developing countries. Journal of Animal Science 85, 2788–2800. Reid, R.S., Kruska, R.L., Deichmann, U., Thornton, P.K. and Leak, S.G.A. (2000a) Human population growth and the extinction of the tsetse fly. Agriculture, Ecosystems and Environment 77, 227–236. Reid, R.S., Kruska, R.L., Muthui, N., Taye, A., Wotton, S., et al. (2000b) Land-use and land-cover dynamics in response to changes in climatic, biological and socio-political forces: The case of southwestern Ethiopia. Landscape Ecology 15, 339–355. 678 M. Jabbar et al. Rich, K.M. and Perry, B.D. (2009) The impact of changing global animal health trading standards on market access for livestock products by developing countries; assessment of commodity based trade. Commis- sioned report to the Department for International Development (DFID) of the Government of the UK. Rich, K.M., Perry, B.D. and Kaitibie, S. (2009) Commodity-based trade and market access for developing country livestock products: the case of beef exports from Ethiopia. International Food and Agribusi- ness Management Review 12, 1–22. Rich, K.M., Perry, B.D., Kaitibie, S., Gobana, M. and Tewolde, N. (2008) Enabling livestock product ex- ports from Ethiopia: understanding the costs, sustainability and poverty reduction implications of sani- tary and phytosanitary compliance. Final report for the Texas Agricultural Experiment Station, Texas A&M University Sanitary and Phytosanitary Livestock and Meat Marketing Program. Rich, K.M., Rich, M. and Dizyee, K. (2018) Participatory system approaches for urban and peri-urban agri- culture planning: the role of system dynamics and spatial group model building. Agricultural Systems 160, 110–123. Robinson, T., Thornton, P., Franceschini, G., Kruska, R., Chiozza, F., et al. (2011) Global livestock produc- tion systems. FAO, Rome, and ILRI, Nairobi. Rutherford, A. (2008) Broad bed maker technology package innovations in Ethiopian farming systems: an ex post impact assessment. ILRI Research Report No. 20. ILRI, Nairobi. Sandford, S. (1983) Management of Pastoral Development in the Third World. Wiley, Chichester, UK. Sandford, S. (1985) Better livestock policies for Africa. ALPAN Network paper No. 1. ILCA, Addis Ababa. Schiff, M. and Valdés, A. (1992) The plundering of agriculture in developing countries. The World Bank. Washington, DC. Shapiro, B.I., Haider, J., G/Wold, A. and Misgina, A. (2000) The intra-household economic and nutrition im- pacts of market-oriented dairy production: evidence from the Ethiopian highlands. In: Jabbar, M.A., Peden, D.G., Mohamed-Saleem, M.A. and Li Pun, H. (eds) Agro-ecosystems, Natural Resources Man- agement and Human Health Related Research in East Africa. Proceedings of an IDRC/ILRI International Workshop, 11–15 May 1998, ILRI, Addis Ababa, Ethiopia. ILRI, Addis Ababa, pp. 109–123. Shapiro, B.I., Getachew Gebru, Solomon Desta, Asfaw Negassa, Kidus Negussie, et al. (2015) Ethiopia livestock master plan. ILRI Project Report. ILRI, Nairobi. Solomon Bekure and Macdonald, I. (1985) Some policy issues of livestock marketing in Africa. ALPAN Network Paper No, 2. ILCA, Addis Ababa. Solomon Bekure, de Leeuw, P.N., Grandin, B.E. and Neate, P.J.H. (eds) (1991) Maasai Herding: An Analysis of the Livestock Production System of Maasai Pastoralists in Eastern Kajiado District, Kenya. ILCA, Addis Ababa. Staal, S.J. (1995) Periurban dairying and public policy in Ethiopia and Kenya: a comparative economic and institutional analysis. PhD thesis, University of Florida, Gainesville, Florida. Staal S.J. and Shapiro, B.I. (1994) The effects of recent price liberalization on Kenyan peri-urban dairy. Food Policy 19, 533–549. Staal, S.J., Delgado, C. and Nicholson, C. (1997) Smallholder dairying under transactions costs in East Africa. World Development 25, 779–794. Staal, S.J., Owango, M., Muriuki, H., Kenyanjui, M., Lukuyu, B.A., et al. (2001a) Dairy systems character- isation of the greater Nairobi milk shed. Smallholder Dairy (R&D) Project report. ILRI, Nairobi. Staal, S.J., Ehui, S. and Tanner, J. (2001b) Livestock–environment interactions under intensifying produc- tion. In: Lee, D.R. and Barrett, C.B. (eds) Tradeoffs or Synergies? Agricultural Intensification, Eco- nomic Development and the Environment. CAB International, Wallingford, UK, pp. 345–364. Staal, S.J., Wolff, T., Baltenweck, I., Waithaka, M., Njoroge, L. and Kruska, R. (2001c) Spatial econometric analysis of farmer strategies: an application to feeding practices on smallholder Kenyan dairy farms. Paper presented at the International Workshop on Forage Demand and Adoption, 18–20 June, ILRI, Addis Ababa, Ethiopia. Staal, S.J., Baltenweck, I., Waithaka, M., DeWolff, T. and Njoroge, L. (2002) Location and uptake: inte- grated household and GIS analysis of technology adoption and land use, with application to small- holder dairy farms in Kenya. Agricultural Economics 27, 295–315. Staal, S.J., Nin-Pratt, A. and Jabbar, M.A. (2008a) Dairy development for the resource poor. Part 1: A compari- son of dairy policies in South Asia and East Africa. FAO/PPLPI Working Paper, No. 44-1. FAO, Rome. Staal, S.J., Nin-Pratt, A. and Jabbar, M.A. (2008b) Dairy development for the resource poor. Part 2: Paki- stan and India dairy development case studies. FAO/PPLPI Working Paper, No. 44-2. FAO, Rome. Staal, S.J., Nin-Pratt, A. and Jabbar, M.A. (2008c) Dairy development for the resource poor. Part 3: Kenya and Ethiopia Dairy Development Case Studies. FAO/PPLPI Working Paper, No. 44-3. FAO, Rome. Stienbarger, D.M. (1990) Tenure and alley farming: a literature review with particular reference to the West African Humid Zone. LTC Paper No. 138. Land Tenure Center, University of Wisconsin-Madison, Madison, Wisconsin. Economics and Policy Research at ILRI, 1975–2018 679 Swallow, B.M. (1994) Evaluating the relationships between property rights, risk, technology and productiv- ity in sub-Saharan Africa. Socio-economic and Policy Research Working Document No. 18. ILCA, Addis Ababa. Swallow, B.M. and Bromley, D.W. (1994) Co-management or no management: the prospects for internal governance of common property regimes. Oxford Agrarian Studies 22, 3–16. Swallow, B.M. and Bromley, D.W. (1995) Institutions, governance and incentives in common property regimes for African rangelands. Environment and Resource Economics 6, 99–118. Swallow, B.M. and Woudyalew, M. (1994) Evaluating willingness to contribute to a local public good: appli- cation of contingent valuation to tsetse control in Ethiopia. Ecological Economics 11, 153–161. Takahashi, K., Barrett, C.B. and Ikegami, M. (2019) Does index insurance crowd in or crowd out in- formal risk sharing? Evidence from rural Ethiopia. American Journal of Agricultural Economics 101, 672–691. Tangka, F.K., Emerson, R.D. and Jabbar, M.A. (2002) Food security effects of intensified dairying: evi- dence from the Ethiopian highlands. ILRI, Nairobi. Thornton, P.K. (1987) A beef production model for the savannas of Colombia: model description and user notes. CIAT. Cali, Columbia. Thornton, P.K. (2010) Livestock production: recent trends, future prospects. Philosophical Transcripts of the Royal Society B: Biological Sciences 365, 2853–2867. Thornton, P.K. and Herrero, M. (2010) The potential for reduced methane and carbon dioxide emissions from livestock and pasture management in the tropics. Proceedings of the National Academy of Sci- ences USA 107, 19667–19672. Thornton, P.K., Jones, P.G., Ericksen, P.J. and Challinor, A.J. (2011) Agriculture and food systems in sub-Saharan Africa in a 4°C+ world. Philosophical Transactions of the Royal Society A: Mathematical Physical and Engineering Sciences 369, 117–136. Thornton, P.K., Kruska, R.L., Henninger, N., Kristjanson, P.M., Reid, R.S., et al. (2002) Mapping Poverty and Livestock in the Developing World. ILRI, Nairobi. Tiongco, M., Lapar, M.L., Costales, A., Son, N.T., Jabbar, M. and Staal, S. (2009) Is contract farming really pro-poor? Empirical evidence from northern Vietnam. Contributed paper presented at the 27th Con- ference of the International Association of Agricultural Economics, 16–22 August, Beijing, China. ILRI, Nairobi. Tisdell, C.A. (2009) The survival of small-scale agricultural producers in Asia, particularly Vietnam: general issues illustrated by Vietnam’s agricultural sector, especially its pig production. In: Salazar, A. and Rios, I. (eds) Sustainable Agriculture: Technology Planning and Management. Nova Science Publishers, Hauppauge, New York, pp. 315–328. Valbuena, D., Erenstein, O., Homann-kee, T.S., Abdoulaye, T., Claessens, I., et al. (2012) Conservation agriculture in mixed crop–livestock systems: scoping crop residue trade-offs in sub-Saharan Africa and South Asia. Field Crops Research 132, 175–184. Verburg, P.H., van de Steeg, J., Veldkamp, A. and Willemen, L. (2009) From land cover change to land function dynamics: a major challenge to improve land characterization. Journal of Environmental Management 90, 1327–1335. von Massow, V.H. (1989) Dairy imports into sub-Saharan Africa: problems, policies and prospects. ILCA Research Report No. 17. ILCA, Addis Ababa. Walker, T.S. and Ryan, J.G. (1990) Village and Household Economics in India’s Semi-arid Tropics. Johns Hopkins University Press, Baltimore, Maryland. Williams, T.O. (1993) Trade and pricing policies in the context of sustainable livestock production in sub-Saharan Africa. In: Ehui S.K. and Lipner M. (eds) Livestock and Resource Management Policy: Issues and Priorities for Research. Proceedings of the Research Planning Workshop, 24–27 March, ILCA, Addis Ababa, pp. 37–42. Williams, T.O. (1998) Multiple uses of common pool resources in semi-arid West Africa: a survey of existing practices and options for sustainable resource management. ODI Natural Resource Perspectives No. 38. ODI, London. Williams, T.O., de Rosa, D.A. and Badiane, O. (1995) Macroeconomic, international trade and sectoral pol- icies in livestock development: an analysis with particular reference to low income countries. In: Wil- son, R.T., Ehui S. and Mack, S. (eds) Livestock development strategies for low income countries. Proceedings of the joint FAO/ILRI round table on livestock development strategies for low income countries, 27 February–2 March, ILRI, Nairobi, pp. 47–69. Zilberman, D. and Heiman, A. (2004) The value of economics research. In: Pardey, Philip G., and Smith, Vincent H. (eds) What’s Economics Worth? Valuing Policy Research. IFPRI, Washington D.C., pp. 275–299. 18 The Impact of ILRI Research on Gender Catherine Hill1, Nicoline de Haan2, Alessandra Galiè2 and Nelly Njiru2 1Vancouver, Canada; 2International Livestock Research Institute, Nairobi, Kenya Contents Executive Summary 681 Research spending and Altmetrics on gender questions 681 Scientific impact 681 Development impacts 681 Evolution of Gender Research at ILRI and its Predecessors 682 1970s–1980s 682 1990s–early 2000s: the merger and clash of cultures 682 2005–2010: institutionalizing a gender research agenda 683 2010 onward: from an ILRI gender strategy to CRP and beyond 684 Impacts of Gender Research at ILRI 685 Scientific perspectives, methods and level of analysis 685 Influence on scientific perspectives 685 Methods 685 Gender at the landscape level 686 Farming systems and technologies 687 Institutions 687 Fodder and forages 687 Animal health and food safety 688 Genetics 689 Environment 690 Nutrition and food security 690 Markets and value chains 691 Empowerment 692 Capacity development and partnerships 692 Policy impacts 693 Improving data and statistics for policy 693 The Future 693 References 694 Annex 1. Gender Research in CGIAR 698 © International Livestock Research Institute 2020. The Impact of the International 680 Livestock Research Institute (eds J. McIntire and D. Grace) C. Hill et al. 681 Executive Summary accessed 10 March 2020), is limited. The ILRI institutional database on the keyword ‘gender’ This chapter discusses the evolution of gender within Altmetrics has 1.0% of the Altmetrics research at the International Livestock Research global database; the rate is even lower for the Institute (ILRI) and its predecessors1, and in the keyword ‘women’. The only major papers (over context of CGIAR. It then reviews the impact of 100 citations) with specific notice of the work ILRI’s gender research in a number of areas in- and output of women are the systems studies for cluding development, science, capacity and policy. Maasailand (Solomon Bekure et  al., 1991) and Discrimination against women in access Borana (Coppock, 1994). There has been a recent to skills, assets, employment, education and health- increase in gender-specific articles produced by care is costly in foregone output and in heightened ILRI staff and partners, such as Galiè et  al. inequality. Research at international agricultural (2019a) on the Women’s Empowerment in Live- research centres, although often ad hoc, has long stock Index (WELI), which had a download sought to identify technical and policy measures index of 1.9k by April 2019. to eliminate or reduce bias against women in agri- culture. Specifically, research at ILRI has focused on gendered access to assets, such as livestock and Scientific impact land, and to technology needed to raise livestock and crop production. As livestock often provide a Research on gender at ILCA began in the 1970s significant share of women’s employment and in- in the principal systems studies conducted by come, and are often an asset they have control ILCA, notably those in Kaduna in Nigeria, Bora- over, identifying gender-based biases through re- na in Ethiopia, Maasailand in Kenya, and Niono search can be a powerful tool to improve the con- and Macina in Mali. The main gender-related dition of women and improve the sector as a impacts of the systems studies were: (i) to iden- whole. Gender research in livestock also enhances tify gender bias in ownership of livestock, land the effectiveness of interventions by increasing and other assets; (ii) to identify gender bias in the relevance of livestock technologies and insti- a ccess to technology and advisory services; and tutions to local communities by addressing needs, (iii) to give methodological guidance on avoiding preferences, constraints and challenges of all gender bias in the design and conduct of field farmers. Recent work also looks at how livestock i nvestigations. can empower women by revealing social relation- Recent scientific impacts of gender work at ships and power dynamics in decision making ILRI have included: (i) institutionalizing a stra- that affect livestock interventions and how it is tegic approach to avoid bias in experimental de- possible to build upon livestock as an asset for sign and conduct; (ii) refining field methods to empowerment. show potential gains from greater gender equity in the generation and application of field results; (iii) a paradigm shift among technical scientists Research spending and Altmetrics to understand that women, as well as men, are on gender questions members of their client groups and have their own demands and needs in terms of animal health services, feeds and forages, environment The accounting of ILRI and its predecessors is and genetics, as well as facing different con- insufficient to estimate the gender share of re- straints and challenges related to livestock; and search and development spending at the three (iv) developing indicators and an understanding institutions. During the CGIAR Research Pro- of women’s empowerment through livestock. gramme (CRP) on Livestock and Fish (2011– 2017), this changed with the mandate of the CRP to spend 10% of its budget on gender. At the Development impacts same time, there has been an increase in direct funding to projects on gender2, but it is prema- The following development impacts were identified: ture to estimate the impact of this funding. ILRI’s presence in the gender literature, as • Defining beneficiary populations for technical indicated by Altmetric (www.altmetric.com/; changes in livestock investments, including 682 The Impact of ILRI Research on Gender in dairy, in vaccination campaigns and in in the drier areas of Nigeria and Niger (Dupire, plant-breeding programmes. 1960, 2018; Stenning, 1994; Hopen, 2018). • Adapting advisory services to the poten- Waters-Bayer (1985), in the Kaduna study, tially different needs and constraints of fe- looked at resource control and decision making male and male farmers in crop and livestock in Fulani households, specifically at Fulani production. women’s processing and marketing of milk • Strengthening personal and institutional products, and highlighted the ways in which capacities to enable ILRI, other CGIAR these agro-pastoral women understood market centres and several partners to do gender forces and recognized the social and local polit- analysis in both pastoral and mixed farm- ical functions of their work. The research also ing systems. explored the limits of Fulani women’s know- • More efficient identification of target groups ledge of connections between the local, national for campaigns to improve food safety. and international economies. The study argued • Identifying livestock assets as a means of in favour of the participatory research to build women’s empowerment. on rural people’s knowledge to enable them to understand and cope better with external influ- ences on their activities and help them ‘better Evolution of Gender Research at ILRI defend their own interests against the mac- and its Predecessors roplanning State’(Waters-Bayer, 1985). In an effort to bring gender more to the forefront, in 1984, ILCA hosted a workshop on 1970s–1980s women in agriculture in West Africa, sponsored by the Ford Foundation. The conference covered When ILCA opened in 1974, its aim was ‘to inte- a wide range of topics related to women in devel- grate sociological, economic, and biological re- opment and included an often-cited paper by search and development related to livestock in Okali and Sumberg (1985), which focused on Africa’ (Waters-Bayer and Bayer, 2014). Despite ownership patterns between women and men its development objectives, ILCA did not system- in small-ruminant production systems and the atically include gender in its early work. A review intra-h ousehold processes therein. of gender-related impacts of CGIAR research By the late 1980s, there had been a shift in criticized the methods and results of inter- ILCA’s focus, and it began to emphasize more national agricultural research on gender issues discipline-based research in animal health, nu- more than a decade after most of the inter- trition and genetics, and stressed ‘precise’ meas- national agricultural research centres had been urement. This approach conflicted with ILCA’s established (Jiggins, 1986). previous innovations in livestock systems re- ILCA social scientists working in field con- search and development and its investments ditions (see Chapter 15, this volume) recognized in social scientists. This new approach led to the important roles of women in agriculture and the  removal of ILCA’s social scientists, apart sought to understand these roles within com- from economists (Romney and Minjauw, 2006; plex farming systems. What became the early, Waters- Bayer and Bayer, 2014), implying that innovative, gender research focused on intra- some of the innovation and momentum on gen- household decision making, indigenous know- der disappeared. ledge, farming systems, the roles of women and even the power and privilege aspects of extract- ive research (Waters-Bayer, 1985). The ILCA Subhumid Zone Programme at 1990s–early 2000s: the merger Kaduna, Nigeria, was a pioneering effort in I LRI’s and clash of cultures evolution (von Kaufmann et al., 1986). It stud- ied the production systems of settled and season- The merger in 1995 of ILCA and the Inter- ally transhumant agro-pastoralists to find ways national Laboratory for Research on Animal of increasing crop and livestock production. Diseases (ILRAD)3 into ILRI altered livestock sys- Milk sellers were exclusively women in the rural tems research. ILRAD had little tradition of field- markets in central Nigeria, as had been described based research, apart from epidemiology, and C. Hill et al. 683 the influence of ILCA’s systems work weakened economic activities and the potential gains that after the merger. ILRI’s nascent institutional cul- can be realized by reducing bias against women ture tended to see the social sciences, which had in access to inputs and services. been significant in the systems studies, as ‘soft sci- ences’4 and this included gender research at ILRI. ILRI’s work in the 1990s on smallholder live- 2005–2010: institutionalizing a gender stock development later did open opportunities research agenda for gender work. CGIAR’s Systemwide Livestock Programme was one such initiative, drawing at- In 2005, ILRI undertook a gender audit, which tention to the importance of investing in mixed reviewed its understanding of gender analysis in crop–livestock systems and indigenous livestock research, its gender equity in the organization breeds and in community-based management of and its mainstreaming of a gender-based ap- animal genetic resources (ILRI, 2014). proach. The audit found a good-faith effort to Again, in 1996, ILRI realigned its agenda to improve gender equity and diversity in the work- emphasize the importance of considering people place among staff and that management backed as part of the livestock systems, which included this effort. It noted that this effort had helped women. The gender perspective of ILRI’s re- create a supportive environment for main- search at the time focused on the roles of, and streaming gender analysis in the ILRI’s research constraints to, productivity faced by women as programmes, as an understanding of gender in agricultural producers. Specific studies of small- the workplace is known to facilitate the integra- holder dairying on the Kenya Coast indicated tion of gender in research. that women operators were more productive In terms of gender analysis in research, the than men, even where men owned cows where audit found that there was no policy on gender the woman operator received the bulk of the analysis in setting research priorities. It was additional earnings. The Kenya Coast research noted that gender analysis in ILRI was discussed also looked at the importance of targeting exten- more than practised and that ILRI’s strategy did sion to women as well as to men to make tech- not mention gender issues. The gender audit nical advice more effective. Other research in also found that, although there was an under- Kenya focused on the importance of understand- standing of the importance of gender to this ing women’s and men’s roles on trypanosomia- point, there had been little training of scientific sis; this yielded a recognition of the importance staff, managers and students in gender analysis of understanding gender-differentiated willing- or integration. The institution had not yet insti- ness to adopt disease-control strategies (Echessah tuted a unit or focal point to systematize gender et al., 1997). Research in Ethiopia involving the in its programme, although some staff were con- Ethiopian Institute of Agricultural Research sidered to have expertise in gender analysis (EIAR), the Ethiopian Health and Nutrition (Roothaert et al., 2006). The audit therefore con- Research Institute (EHNRI) and ILRI looked at cluded that, while there was good understand- the impact of cross-bred cattle on men’s and ing of what gender analysis is in the research women’s decision making around dairying in- capacity of ILRI staff, undertaking such analysis come (Nicholson et al., 1999). was limited (Roothaert et al., 2006). In 1999, a study of 54 households in a As a result of the audit, ILRI formed a task semi-arid subregion of western Niger highlighted force in 2006 to develop a research agenda on shifts in livestock ownership related to long-term women and livestock issues, but it was only in economic and environmental changes (Turner, 2008 that the task force began to have meaning- 1999). Turner found significant shifts away from ful dialogue with experts and partners (Njuki cattle owned by men and towards sheep and et al., 2011). This was initiated through a global goats owned by women over the period 1984– e-consultation, the Global Challenge Dialogue 1994. To some extent, these shifts have gone un- on Women and Livestock (Gonsalves, 2013). noticed in the gender literature, yet they confirm The consultation brought together major live- the point, often made in the same literature, that stock players and proposed: (i) the production of separate survey and analytical approaches are a landmark document providing evidence of the needed to capture the importance of women’s feminization of the livestock sector throughout 684 The Impact of ILRI Research on Gender the world; (ii) a plan for revitalizing a global The authors hypothesized that livestock pathways women’s and livestock alliance; (iii) a review of out of poverty: (i) secure current and future as- strategies used by research and development or- sets; (ii) sustain and improve the productivity of ganizations to reach women; and (iv) plans for agricultural systems in which livestock are im- scaling out those strategies that have been suc- portant; and (iii) facilitate greater participation cessful in reaching women with livestock inter- of the poor in livestock-related markets. Each of ventions (ILRI, 2012). these brought a new attention to gender in its While gender analysis was not systematic- own right and to the importance of livestock as ally integrated in ILRI’s research, several pro- an asset for women. jects included gender outcomes. Most projects were development oriented and included women as beneficiaries of the technologies without ana- 2010 onward: from an ILRI gender lysing the actual needs of the women involved. strategy to CRP and beyond For example, the broad-bed maker tool in some mixed farming systems of the Ethiopian high- In 2010, ILRI developed a common set of gender, lands (Rutherford, 2008), the Improving Prod- livestock and livelihood indicators to help the uctivity of Market Success (IPMS) of Ethiopian centre measure the impacts that projects and Farmers project (2008–2013) and the East other livestock interventions, such as markets Africa Dairy Development (EADD) were efforts and biotechnology, had on poverty, gender and to introduce interventions that included women equity (Njuki et al., 2011). These indicators were as beneficiaries. developed for household-level surveys with the The EADD Phase 1 project in Kenya, Rwanda potential for adaptation for community-level and Uganda (2008–2013) set out to double the focus group discussions. dairy income of 179,000 smallholder families in Another turning point for institutionalizing 10  years. Its entry point was women as benefi- gender approaches at ILRI came when the Insti- ciaries of training and as producers. In 2009, tute produced its ‘Strategy and plan of action to EADD set out to address this gap by developing a mainstream gender in ILRI’ (ILRI, 2012). The gender strategy, hiring a gender and youth coord- strategy recognized ILRI’s need to guide and inator in 2010, developing gender disaggregated design the consolidation of ILRI’s expertise and data templates and a gender work plan and gender resources, to engage stakeholders, and to performance targets, and outlining strategies to ensure that men and women participate in and include women in project activities. Although a benefit from ILRI’s research. It also emphasized development project, it did open up a new under- the need for commitment from ILRI’s board, standing of gender and the need to focus specific- management and staff and from its many other ally on women (Baltenweck and Mutinda, 2013). partners. ILRI’s gender strategy represented a In 2009, ILRI established a new theme on true shift over time from research that looked at ‘Poverty, Gender and Impact’. This demonstrated women as components in farming systems re- a shift in its commitment to ensuring gender- search to a full gender and agricultural research responsive research by focusing on two compo- theme including production, processing, mar- nents: (i) investigations where the research kets, value chains and strategic gender research. agenda has been set by scientists, such as forages In 2014, ILRI introduced a new theme – or genetics, and gender considerations are inte- Enabling Innovation – which focused on adap- grated to study the subject more effectively; and tive capacity and increased attention to gender. (ii) strategic research where the subject is gender. This continued the research of the Innovations Important in the 2000s was ILRI’s work Work Unit established in 2007, which, in part, developing a conceptual framework on livestock also generated information and learning to em- as a pathway out of poverty that had, at its core, power women in livestock innovation (Waters- the importance of assets, markets and other in- Bayer and Bayer, 2014). The Innovation Works stitutions (Kristjanson et  al., 2004). ILRI used Unit recognized women’s key roles in livestock this framework in a seminal literature review to production, nutrition and health, noting that discuss women and livestock as a pathway out of most resource-poor livestock keepers are women, poverty for women (Kristjanson et  al., 2010). and campaigning to keep gender issues at the C. Hill et al. 685 forefront of livestock research and development. dynamics in fodder seed innovation systems, This included a greater emphasis on the impact and gender dynamics in forage conservations of technologies and policies on women and a systems; (iii) genetics, through gender-sensitive greater awareness of gender issues overall. community breeding of small ruminants (Marshall A later shift in ILRI’s gender research fol- et al., 2019); and (iv) the environment, through lowed the development of the CRPs in 2010 and gender and land tenure for reduced land degrad- 2011. With the CRP on Livestock and Fish (2012– ation, increased intensification, labour dynam- 2016), ILRI recognized the need to consolidate the ics, gender norms, and gender and pastoralism centre’s gender expertise and resources to ensure (de Haan and Mulema, 2018). that men and women participated in and bene- fited from CRP work. Gender was one of the pro- gramme’s six themes along with animal health, Impacts of Gender Research at ILRI genetics, feeds and forages, sustainable interven- tions and value-chain development (Galiè and The impacts of ILRI’s gender work can be grouped Kantor, 2016; CGIAR, 2013). by influences on: (i) scientific perspectives, The CRP on Livestock and Fish focused on methods and levels of analysis; (ii) farming sys- gender relations and dynamics, access to and tems and technologies; and (iii) empowerment. control of productive resources, and gender- transformative approaches. The CRP explored local meanings of livestock ownership across Scientific perspectives, methods three CRP value-chain countries (Tanzania, and level of analysis Ethiopia and Nicaragua) (Galiè, 2015). The CRP developed an article reviewing tools developed in livestock and fish value chains (Farnworth Influence on scientific perspectives et al., 2015) and a policy brief looking at gender Research by Waters-Bayer (1985) on agro- relationships and farmers’ capacity to mitigate pastoral Fulani women in Nigeria and by Okali and c limate change (Gumucio and Rueda, 2015). To Sumberg (1985) on women and small- ruminant enhance the capacity of scientists to integrate production in the subhumid areas of southern gender in their work, the CRP on Livestock and Nigeria provided some early understanding on the Fish engaged the Royal Tropical Institute in the intersection of gender and livestock production. Netherlands, which, together with the ILRI gen- Recently, gender research has gained ground in der scientists, coached them; this work led to the ILRI’s work and moved from an issue relevant for publication of findings from 14 gender-integrated studying other subjects such as breeding and livestock and fish research studies (Pyburn and animal health (integrated gender analysis) to a van Eerdewijk, 2016). research topic in its own right (strategic gender In 2015, ILRI and Emory University in research) and is thus part and parcel of under- Georgia, USA, identified a mismatch between the standing how rural households, value chains and limited attention to livestock issues in the WELI, livestock systems work, with its own research which focused on agriculture in general (includ- agenda. Similarly, gender equity through livestock ing livestock, crops and fish) and the importance is increasingly accepted as a goal of ILRI’s work in of livestock in East Africa. The WELI was subse- its own right and a driver of change, rather than a quently developed to explore how women’s em- secondary outcome of livestock interventions. powerment can be supported through livestock and to assess women’s empowerment quantita- Methods tively, particularly in a case study of Tanzania (Galiè, 2018a). Mulema et  al. (2019) undertook research in Currently, ILRI’s gender research work is Ethiopia that demonstrated the significance of focusing on: (i) animal health, through enhan- several empowerment indicators (e.g. cultural cing gendered capabilities to address threats norms and women’s inputs into production deci- through a gendered lens and engaging women sions; autonomy in plot management; member- in health services; (ii) feed and forages, through ship of farmer groups; ability to speak in public, gender-sensitive forage interventions, gender enhancing their participation in different stages; 686 The Impact of ILRI Research on Gender access to information and extension services, outlines a Theory of Change and research cycle education and land size) in influencing women’s approach as well as gender-responsive goals, participation in different stages of agricultural o bjectives, research questions, activities and out- research. This work contributed to the literature comes that can inform research and interventions on women’s empowerment in relation to agri- in livestock health. The WELI has been found to be cultural research and to promoting the integra- particularly useful for measuring the impact of tion of proactive, holistic gender perspectives in livestock projects on women’s empowerment over research strategies. time. ILRI, together with Emory University, devel- ILRI has also developed indicators on gen- oped the WELI and piloted it in Tanzania in der, livestock and livelihoods to measure the 2015 (Galiè et  al., 2019a). The WELI helps re- impacts of livestock interventions at household searchers and decision makers better understand and community levels (Njuki et al., 2011). The which interventions work best for empowering women’s empowerment in livestock-focused rural women. Such evidence is important to agriculture – the IMMANA project (2015–2018) – fine-tune interventions and provide better em- in Kenya, Uganda and Tanzania developed new powering opportunities for rural women. The ac- metrics for women’s empowerment and animal- tual discussions on empowerment between rural source food intakes that are sensitive to mater- women and men also provide value, opening nal and child nutrition, and are relevant to spaces for individuals, communities and house- different livestock value chains, including pork, holds to think about what empowerment means, dairy cattle and poultry. who has access to more opportunities for em- The CRP on Livestock and Fish developed a powerment, and how social and gender norms af- set of tools for social and gender analysis for fect the ability of individuals to succeed. value chains (Kruijssen et  al., 2016). These Tavenner et al. (2018) analysed resources, tools, adapted from existing tools from other or- decision making and labour dynamics in dairy ganizations, help users to explore gender rela- farm households in western Kenya. This study tionships and the underlying causes of inequities. found statistically significant differences in prac- One helps users undertake a supplementary tices based on gender. The most divergent re- gender and social analysis when there is already sponses between men and women were decision an existing value-chain analysis, while the other making around the morning and evening milk helps users undertake a full value-chain analysis sales. The authors argued that the choice of including underlying causes of gender inequality. interviewee affected research findings because In 2016, the CRP on Livestock and Fish in- survey respondents may have different percep- tegrated gender into the Feed Assessment tool tions or valuation about ‘who does what’. (FEAST), a participatory tool focused on feeds Galiè et  al. (2019b) discussed some of the and forage and developed by scientists at ILRI, difficulties encountered in adopting a mixed- the Centro Internacional de Agricultura Tropical method approach that results in contradictory (CIAT) and the International Center for Agricul- quantitative and qualitative findings. The article tural Research in the Dry Areas (ICARDA). This discusses some reasons for this discrepancy in- supported researchers and practitioners in their cluding the different definitions, domains and research to surface the issues of gender relation- indicators adopted by the two approaches when ships and how they affect livestock farming, studying ‘food security’ ‘nutrition security’ and particularly feeding practices and innovations ‘women’s empowerment’. In addition, the quali- (Lukuyu et al., 2016). The resulting app has now tative study may have given space to a discussion been gendered into G-FEAST, which specifically on ‘aspirational’ versus ‘actual’ gender roles in looks at gendered preferences for forages. guaranteeing food and nutrition security that Waithanji and Grace (2014) also developed a quantitative and closed research questions may gender strategy to support mycotoxin control have not provided. given that, in many regions, women are respon- sible for producing food for home consumption Gender at the landscape level and may also have roles in feeding and caring for livestock. The strategy is an important tool for Gender work has typically been done at the r esearchers working on mycotoxin control as it household level and has studied intra-household C. Hill et al. 687 dynamics. To widen the impact of gender re- women construct new pathways for women’s search, in 2017, the CRP on Livestock began participation because of the ways that various work on gender at the landscape level through participatory strategies relate to one another, ra- the development of national livestock master ther than due to the efficacy of one strategy over plans. New versions of such national master another. plans will guide investment towards women in the livestock sector (Shapiro et  al., 2015 for Fodder and forages Ethiopia). There is also a move to integrate gen- der in modelling work and livestock sector ana- Work at ILRI and the Kenya Agricultural Re- lytics that underpin the national master plans. search Institute (KARI) on smallholder dairying An ongoing project is developing a methodology based on a fodder cut-and-carry system found to scale gender dynamics from the household that an integrated dairy development package and community levels to higher national and re- had limited acceptance among farmers. A subse- gional levels in the context of the feminization of quent study looked at women’s roles and labour, agriculture (Galiè et al., 2019d). and found that women were more likely than men to adopt more of the package and demon- strated higher milk yields per lactating cow Farming systems and technologies (11.5  litres/day) than male contact farmers (6.8 litres/day) (Mullins et al., 1996). Although Institutions women faced increased workloads as dairy oper- ators, they also perceived improvements in the Farnworth and Colverson (2015) found that welfare and long-term development benefits of rural advisory services operate in environments their households through women’s income structured by gender relationships. In other going to school fees, books, and food purchases words, women often have less-effective partici- (Mullins et al., 1996). pation in community decisions, in value-chain A study on a traditional Maasai forage networks and in innovation platforms. Because conservation system (ololili) in Tanzania (Galiè, women are reached less often by advisory ser- 2018a) found that the system relied heavily on vices, it is more costly for them to adopt new women’s labour when it was in use during the methods. The study concluded that advisory ser- dry season, whereas livestock management in- vices should be seen as a facilitation system to volved both women and men. Women’s and tackle underlying gender relationships that con- men’s groups were found to have similar know- strain access and implementation rather than as ledge of local forage plants but ranked their im- a supposedly gender-neutral service. portance differently. They also showed the same Omondi et al. (2014) found that women were level of interest in intensifying forage growing in reluctant to participate in dairy hubs in Kenya the ololili. At the same time, gender norms and because of their loss of control of income from dynamics were found to strongly affect the abil- milk sales, underscoring the importance of intra- ity of women – mostly poor women, and widows household income distribution. The findings im- in particular – to manage ololili. These social plied the need for evidence-based interventions constraints in the governance of the ololili, if not and changes in structures that encourage addressed at the inset of any intensification w omen’s participation, promote more equitable intervention, were found to be likely to decrease i ncome distribution from dairying and/or com- the success of forage technology interventions pensate women’s loss of income, without negative because they limited the sustainability of the impacts on the stability of gender relationships system. within the households. Galiè (2018b) showed how a forage breed- Basu et  al. (2019) analysed approaches to ing intervention can enhance the empowerment women’s participation adopted by the EADD by of female farmers. The author demonstrated looking at how participation actually emerges in practical challenges faced by a breeding pro- specific contexts through gendered negotiations gramme that aims to include gender consider- with participatory development policies. The ations in its activities and showed how a lack of authors discussed how initiatives that include access to seed because of gender-discriminating 688 The Impact of ILRI Research on Gender norms and practices at local and national levels disease-control measures. The study noted that can hinder progress towards empowerment. Ul- to control the disease, farmers require informa- timately, the article challenges assumptions that tion and money for disinfectant as well as the gender considerations be integrated in breeding agency to make decisions. It found that women programmes to enhance their effectiveness only, work closely with livestock, often detecting the by showing the empowering potential of a gen- disease or symptoms. However, men typically der-responsive programme to progress towards make decisions, control household income, and gender equality. The article also shows the im- have access to training and veterinary services. portance of taking into account the wider con- Elsewhere, Kiama et al. (2016) conducted a text (e.g. socio-cultural, policy and seed systems) qualitative study with male and female dairy in which a breeding programme is implemented, farmers in Kenya on their awareness and per- to ensure its benefits reach both female and male ceptions of mycotoxins and how their risk of farmers. dietary exposure of mycotoxins is influenced by these. The gender analysis found that those re- Animal health and food safety sponsible for mitigating risk of exposure are not always those with the knowledge of how to do Galiè (2017) studied smallholder livestock keep- so. It also pointed to the importance of extension ers in Tanzania and found that while men and services targeting women as they are the main women were both involved in animal health handlers of food. The study found that farmers management and had similar knowledge of dis- had a high level of awareness of the harm of eat- eases, women faced more constraints than men ing mouldy food even though risk categories, in accessing livestock services, disease informa- awareness of mycotoxicosis and carcinogenic tion and veterinary drugs because of restrictive effects were generally low. Typically, women norms on both their movement and their inter- were more careful than men not to feed cattle actions with unrelated men, because of biases spoilt maize and they were key decision makers about their reliability in identifying diseases and in dairy cow diets and disposal of mouldy foods. paying for services, and because they had limited Furthermore, while farmers agreed that hy- control over the household resources. The study gienic handling was the most important method suggested supporting women’s groups as a way to enhance meat and milk safety, it was women of enhancing women’s control over livestock who took more care in ensuring that this hap- and revenues, and access to animal health infor- pened, while men were more likely to treat sick mation and income-generating opportunities. animals. The study recommended enhancing the capacity Kimani et al. (2012) investigated the gender of service providers in gender-responsive ap- and social determinants of the risk of exposure proaches and organizing community outreach of Cryptosporidium spp. from urban dairying in activities that highlight the benefits of shared Dagoretti, Nairobi. The study found that gender, intra-household decision making. It recom- age and household roles are all determinants in mended that research institutes include gender exposure to Cryptosporidium spp. For example, considerations when identifying priority species farm labourers and people aged 50–65  years and diseases for research on animal medicines had the most contact with cattle, while women and assess which format (e.g. size or tempera- had greater contact with raw milk and children ture sensitivity) increases the accessibility of had relatively higher consumption of raw milk. animal medicines at local level. Women had more contact than men with cattle Dione et al. (2016) explored how gender re- faeces. Age also played a factor, as older women lationships affect African swine fever control had more contact than older men. Socio- protocols and how current male-centred ap- economics was a partial factor, with those living proaches often disregard women’s roles in pig in poverty consuming less milk than others, al- husbandry. Specifically, the research looked at though their exposure to cattle was not affected. how women and men in Uganda perceive African There was no significant gender difference in swine fever and the factors affecting how they knowledge of cryptosporidiosis symptoms or respond to it in efforts to encourage farmers other zoonotic diseases in the dairying sector; to adopt improved husbandry practices and however, the level of education was a determining C. Hill et al. 689 factor in awareness, with those with higher It measured gender capacities at organizational levels of education more aware of the disease and staff levels of national and regional re- and factors affecting its transmission. search institutes and assessed them in relation Jumba et al. (2016) illustrated that vaccines to the institutional and policy environment that against East Coast fever, a major tick-borne dis- enables or disables other capacities. On a scale of ease of cattle and buffalo, can increase overall 1–5, the study found that core gender capacities household productivity while making it more are insufficiently to partially developed (2.4–2.9), unequal. This resulted from an increase in wom- pointing to the need to substantially improve the en’s labour on livestock at the same time as their gender capacity of these organizations to sup- husbands controlled income from increased live- port genetics research. stock sales. As a consequence, women were Other research by Mora Benard et al. (2016) sometimes reluctant to buy vaccines. in Nicaragua demonstrated gender disparities in Working on contagious bovine pleuropneu- milk production and breeding technologies (arti- monia (CBPP) in Kenya, Muindi et  al. (2015) ficial insemination). Ramaswamy and Galiè (2018) noted that women and men perceived the effects studied gender trait preferences in poultry in of, and were affected differently by, CBPP occur- Ethiopia and showed that women valued traits rence because of prevailing gender norms. While related to behaviour and feathers that breeding women perceived cattle mortality to be the programmes usually neglect by focusing on greatest effect of CBPP because it caused food meat yield and taste only. Women’s preferred shortages and a decline in income from milk traits affected whether a breed was adopted by a sales, men perceived reduced participation in household or not. The same study also showed cattle markets to be the greatest effect of CBPP that men respondents preferred traits related to occurrence. The findings pointed to the need to productivity, health and marketing of chickens incorporate gender in animal health research to with a view to scaling up their poultry keeping to develop appropriate interventions to prevent or an intensive system for business. Women re- mitigate small-ruminant diseases. A related ex- sponders, in contrast, aspired to increase the ample is given by Wieland et al. (2016) in Ethi- scale of their poultry keeping within their opia in a Participatory Epidemiology and Gender household level only and therefore valued traits Project. This project provided insights about the to increase productivity in extensive systems. differential veterinary knowledge of women and Women were not interested in upgrading poultry men in households keeping sheep and goats re- to a business because of the high labour require- lated to their gender-specific roles and about the ments (mostly their responsibility); their lack of need to target interventions, such as deworm- land to keep chickens intensively or assets to ing, accordingly. make financial investments needed for intensifi- Research on 20 livestock and fish value cation; and their loss of control over the benefits chains found that the influence of gender on risk provided by chickens when, with intensification, exposure and management is essential for im- men took on the marketing of the birds. The au- proving food safety in informal markets (Grace thors, therefore, recommend that, to increase et  al., 2015). Socially constructed gender roles adoption, poultry breeding programmes include were more important determinants of health gendered preferences for both traits and chicken- risk than biological differences between men and raising systems. women; variations in risk exposure were mainly ILRI participates in the CGIAR Gender and due to gender-based differences in occupational Breeding Initiative, which seeks to build an ap- exposure. proach that incorporates gender perspectives from the beginning of a breeding programme Genetics through implementation and impact assess- ment. The initiative is currently working on a Gender inequalities can affect the orientation toolbox that helps such incorporation. The tool- and outcomes of programmes to improve live- box will be used to assure the gender relevance stock genetics. A study by Rijke (2017) focused of tools in the CGIAR Excellence in Breeding on the gender capacities of national partners Platform (Liljander et al., 2015), while support- in the African Chicken Genetic Gains project. ing national agricultural research institutes and 690 The Impact of ILRI Research on Gender other breeding programmes. As part of this ini- gender relationships affect the capacity of live- tiative, Galiè et al. (2019a) analysed approaches stock producers to mitigate climate change. The to see what is effective in making a plant-breeding study also demonstrated that women face cer- programme gender responsive. The authors ar- tain limitations as agents of change compared gued that a programme needs to: (i) adapt its cri- with men due to gaps in access to, and control teria to select farmers to host and evaluate trials over, productive resources. Related work by to ensure women (who own smaller parcels of Gumucio and Rueda (2015), derived from a re- land than men or none at all) are involved; view of 105 national policy documents in seven (ii)  adapt its process to evaluate trials in ways Latin American countries, concluded that devel- that women can express their preferences (e.g. opment and environmental policies often failed by using scoring systems that require little liter- to recognize women’s roles as producers in the acy or, for example, by creating a safe space for national economy. women to assess the crops and express openly their preferences; in a community with strong Nutrition and food security purdah practices, this may entail a women-only field trip and domestic space to discuss and score Galiè and de Haan (2019) highlighted the rele- trials); (iii) expand the traits it considered for fur- vance of gender in policy pathways for food and ther breeding to include traits preferred by differ- nutrition security. Price et  al. (2018) explored ent groups of women and men; (iv) expand the the linkages between women’s empowerment crops it included in its portfolio (to include crops and household nutrition in relation to livestock of interest to women and men); and (v) include knowledge and looked at perceptions of women’s both oral and visual information-sharing ap- empowerment from the perspective of female proaches to reach women who are often more farmers in Tanzania. The study found that illiterate than men. However, for a gender- women perceived an increased ability to provide responsive breeding programme to result in ac- nutrition for their families if they had more con- tual gender-equitable outcomes (e.g. producing trol over livestock, income and agricultural seed that benefits both women and men), a co- r esources. However, women were reluctant to herent and comprehensive package of techno- describe the direct links between empowerment logical (e.g. improved seed) and institutional in livestock work and household nutrition, in (e.g. policy and governance) solutions needs to part because they could not imagine that it be developed by multidisciplinary teams. would be possible to gain significant power over livestock within their societal constraints. Women Environment frequently described opportunities for becoming empowered outside the livestock sector (i.e. in The EADD programme in Kenya examined sus- new crop agriculture or business) where gender tainable milk intensification, climate change norms were less entrenched than in livestock be- mitigation and gender dynamics in determinants cause they are less constrained by tradition. of participation and in distribution of benefits Similarly, Galiè et  al. (2019b) presented a (Tavenner et  al., 2018). Household surveys mixed-methods study that examined the rela- covered decision making, resources and labour tionship between women’s empowerment, house- dynamics in cattle-keeping households in Bomet, hold food insecurity, and maternal and child diet Nandi, Uasin Gishu and Kericho counties. While in two regions of Tanzania. Indicators across women and men reported similarly on some three domains of women’s empowerment were issues, they contested others. The research dem- scored and matched to a household food inse- onstrated the challenges of interpreting gender curity access scale. Qualitative research helped dynamics and addressing challenges in the dairy appreciate the gender dynamics affecting the sector methodologically and programmatically. empowerment–food security nexus in a forage Gumucio et  al. (2015) looked at capacity conservation system. In cluster-adjusted regres- and gender relationships in the context of miti- sion analyses, scores from each domain were gating climate change. Based on a review of silv- significantly associated with women’s dietary opastoral production systems in Costa Rica, diversity but not with household food security. Colombia and Nicaragua, the work found that All three empowerment domains were positively C. Hill et al. 691 associated with food security and nutrition in development initiatives. It helped to explore the qualitative analysis. The authors discussed challenges and identify opportunities for pro- some of the methodological challenges encoun- moting gender equality and women’s empower- tered when combining quantitative and qualita- ment through increasing women’s access to tive methods and the implications of the findings. skills, knowledge and assets and by increasing Other research in rural Kenya examined women’s participation in market-oriented agri- how ‘women’s time use and decision-making cultural production and their control over the patterns related to dairy income and consump- benefits. tion are associated with intensification’ and Recent research on cattle and dairy market found that ‘children in high-intensity house- participation in Kenya demonstrated the ad- holds r eceived more milk than children in vances in gender research in recent years to in- medium-i ntensity households’ and that women clude attention to the gendered nature of market in high-intensity households also spent less time participation and privilege over dairy income on dairy activities than women in medium- (Tavenner and Crane, 2018). This research dem- intensity households. Although women seemed onstrated the importance of considering the to be gaining control over evening milk sales, social trade-offs and the gendered costs of dairy men appeared to be increasingly controlling commercialization in interventions aimed at total dairy income, a trend countered by the in- redressing gender power imbalances. Elsewhere, crease in reported joint decision making (Njuki recent work on milk trading in peri-urban et al., 2015). Galiè et al. (2019b) confirmed this N airobi revealed strong gender-based constraints in their recent article on milk production. faced by women milk traders that result in milk business being more lucrative for men than for Markets and value chains women (Galiè et al., 2019c). Similarly, market-oriented smallholder de- Farnworth et  al. (2015) examined current velopment in the dairy sector in Holetta in research to develop analytical frameworks and E thiopia and Kiambu in Kenya contributed to i mplementation guidelines to support gender the question of whether smallholder research analysis in livestock and fish value chains. Njuki results in women were losing control over in- and Sanginga (2013) carried out research on come in the East African highlands and sug- women and livestock and provided empirical gested the need for more robust understanding evidence from different production systems in of the context in which gender roles and rela- Kenya, Tanzania and Mozambique of the im- tionships exist and the subsequent impact on portance of livestock as an asset to women and women’s time use, participation in market-related their participation in livestock product markets. livestock activities and benefits (Tangka et  al., They explored intra-household income manage- 1999). McPeak and Doss (2006) also high- ment and the economic benefits of livestock lighted the importance of understanding gen- markets to women, focusing on how markets, dered roles and relationships in producing and products and women’s participation in markets marketing dairy products through their re- influence their livestock income. search on mobile pastoralists in Kenya. They From experience in the IPMS project in found that women had the right to sell milk, but Ethiopia, Aregu et al. (2010) demonstrated that men were responsible for the whole herd and site-specific commodity-based gender analysis is where they would camp. If women’s marketing essential for understanding the different roles of objectives conflicted with men’s herd-management women and men in the production of specific objectives, then men used location to restrict commodities, marketing and decision making women’s access to markets. and their share in the benefits; in identifying po- A study on pork consumption in Uganda tential barriers for women’s and men’s partici- studied the reasons why pork consumption is pation in market-led development initiatives and lower for women than for men at pork joints technology adoption; and in identifying what (Mabwire, 2018). The study focused on two main actions may be required by the project in order to possible reasons: the attributes of retailer outlets overcome some of these barriers that limit women’s and gendered perceptions associated with pork participation in these particular commodities consumption at joints. The study found hygiene 692 The Impact of ILRI Research on Gender (of the outlet environment and the waiters and Capacity development waitresses) to be the main attribute that women and partnerships consider when eating at joints. It also found that the communities usually negatively label un- Working closely with scientists and partners has accompanied female pork consumers as ‘lonely’, been the best approach to develop capacity on ‘single’ or ‘prostitutes’. gender and to leverage that capacity for a larger impact. In 2013, in collaboration with Transi- tion International, ILRI produced a gender cap- Empowerment acity assessment tool to evaluate existing skills and gaps in partners’ gender capacities and Empowerment through livestock is a new area identify measures to address them. In 2015, the of work for ILRI. It has meant a move away from tool was implemented in four of the value-chain simply ensuring that women can benefit from countries (Ethiopia, Nicaragua, Tanzania and technologies developed by ILRI to one where the Uganda) in the CRP on Livestock and Fish (ILRI, research is on how women can benefit from live- 2017). The ILRI gender team has engaged in stock based on their own needs and aspirations, addressing some of the identified gaps during and how they can potentially be empowered by capacity development workshops and through livestock. Three initial areas of work have been the use of a training manual. The team has also understanding: (i) the concept of livestock own- undertaken capacity development at national ership as part of empowerment; (ii) the links levels through, for example, close collaboration among food security, nutrition security and with the Ethiopia Institute of Agriculture Re- empowerment; and (iii) livestock as an asset for search (EIAR) and with the Food and Agricul- empowerment (Galiè and de Haan, 2018). The ture Organization of the United Nations (FAO) to team has also engaged in developing new concep- develop an approach to building capacity at tualizations of empowerment based on fieldwork p olicy levels. with livestock keepers (Galiè and Farnworth, Under the CRP on Livestock and Fish, the 2019). gender team was embedded under different As empowerment is often also within a flagships to provide coaching to individual context, it has also meant increased engage- scientists and technicians. Doing so resulted ment and research of gender-transformative in a cadre of researchers who had a more approaches. Gender-transformative approaches in-depth understanding of gender in their aim to deepen social change by addressing some subject area and in the development of a book of the norms that constrict a particular group on how to integrate gender within different by determining, for example, what behaviour is areas of livestock development (Pyburn and acceptable for women and men (e.g. of a given van Eerdewijk, 2016). It also resulted in the ethic group, social status or age) or what re- investment of a gender strategy for the Afri- sources and opportunities they are entitled to or can Chicken Genetic Gains project and the can claim (Galiè and Kantor, 2016). Gender- placement of a gender expert in the project to transformative approaches are often contrasted provide support in Ethiopia, Tanzania and to ‘ accommodative approaches’. Accommoda- Nigeria (Rijke, 2017). tive approaches recognize and respond to the Working together, EIAR, ILRI and ICARDA, specific needs and realities of men and women together with the Ethiopian Agricultural Trans- based on their existing roles and responsibilities formation Agency, began to integrate gender in as shaped by existing social and economic struc- agricultural programmes by sharing the gender tures; they do not question the systemic bar- capacity assessment methodology and tools de- riers put up by the social context of people’s veloped by the CRP on Livestock and Fish (ILRI/ lives (Cornwall and Edwards, 2010). Using both ICARDA, 2017). The results and experiences empowerment as an entry point and gen- from gender capacity assessment of the small- der-transformative approaches is a new area of ruminant value-chain partners were also distrib- work for ILRI but an important one in under- uted through the Agricultural Transformation standing the potential for livestock to improve Agency to stimulate interest in and appreciation livelihoods. of the methodology and tools. C. Hill et al. 693 Policy impacts Improving data and statistics for policy A regional dairy development project was im- Of note is the contribution to science impacts of plemented in Kenya, Tanzania and Uganda by the development of the WELI, which helps quan- Heifer International with ILRI and other partners. tify empowerment in a way whereby scientists One project objective was to increase women’s can measure its changes over time. The WELI participation in producers’ organizations and in provides a common framework for determining the dairy value chain (Pyburn and van Eerdewi- the effectiveness of various interventions and jk, 2016; Basu et al., 2019). The project included can support decision makers and policy makers two studies, one to assess women’s roles in the in measuring progress against the investments dairy value chain beyond production, and the they make and can strengthen the integration of second to analyse the inclusion of women and empowerment in development policies and pro- youth in producer organizations. Both studies grammes. illustrated that in all three countries, participa- The Rural Household Multi-Indicator Sur- tion of women was higher at the production vey (RHoMIS) framework produces standardized, links of the value chains and weaker at higher coherent, cost-effective, quantitative, decision- links. Participation of women in leadership posi- relevant information to support efficient and im- tions in producer organizations, cooperatives pactful development programming for planning and credit agencies was insignificant. and monitoring investments in sustainable in- A study of a sheep value chain in Ethiopia tensification across a range of rural contexts. identified gender-specific constraints for par- RHoMIS captures information on farm product- ticipation in the value chain (Wieland et  al., ivity and practices, nutrition, food security, gen- 2016). The results showed that men and der equity, climate and poverty (van Wijk et al., women both faced constraints in terms of 2016). The core set of data feeds into a global capital – social, financial, human, natural, pol- discussion on the success of sustainable intensi- itical, cultural and physical – but women faced fication. RHoMIS includes a gender equity more severe constraints than men. Projects to indicator, ‘Gendered income over assets and support pro-poor value chains would therefore foodstuffs’ (van Wijk and Hammond, 2018), and need to devise mechanisms to release women’s since its inception in 2015, RHoMIS has been capital constraints. a pplied in 22 countries. Another study in Tanzania showed that women and men had similar knowledge of ani- mal disease management and its possible impact The Future on food security (Galiè et  al., 2017). However, women faced more constraints than men in Based on ongoing work, the future of research gaining access to veterinary services, informa- on gender and livestock covers two different but tion on diseases and animal drugs. The implica- equally important agendas (for further elabor- tions are that veterinary and extension services ation, see chapter on ‘Conclusion: The Future of should give proper attention to different service Research at ILRI’, this volume): constraints faced by men and women. • Improving the productivity and efficiency Quisumbing et  al. (2013) assessed the im- of the livestock sector by increasing the op- pact of dairy value-chain interventions on gen- portunities for women and men to engage der issues, including ownership of assets, asset in the livestock sector. control and decision making, and time alloca- • Strengthening (economic) empowerment tion. The study results indicated that value-chain of women through livestock. interventions increased joint household assets of men and women. Value-chain interventions Research to inform and support this will be did not alter production decision making, al- undertaken under the following themes: though they did have an impact on intra-house- hold decisions. The value-chain interventions • Conceptual framing of gender and livestock. also increased the time allocated to dairying, • Increasing options to engage equitably most of which was provided by women. in the livestock sector and identifying 694 The Impact of ILRI Research on Gender gender-specific interventions in ILRI’s re- than the household scale. From the per- search for development. spective of modelling, the RHoMIS and • (Economic) empowerment through livestock GENNOVATE (a global comparative research and livestock as a business for women. initiative that addresses the question of how • Gendered empowerment and nutrition. gender norms and agency influence men, • Gender in livestock policy and at the land- women, and youth to adopt innovation in scape scale. This includes investment plans in agriculture: https://gender.cgiar.org/themes/ the livestock sector (in Ethiopia and Namibia) gennovate/; accessed 14 April 2020) initia- that focus on gender at a broad scale rather tives provide potential for impact. Notes 1 ILRI refers to the International Livestock Centre for Africa (ILCA, 1974–1994) and the International Laboratory for Research on Animal Diseases (ILRAD, 1973–1994) unless specified otherwise. ILCA and ILRAD were merged to form ILRI in 1995. 2 In terms of gender in the workplace at ILRI and the institution’s inclusion of women among its scientists, the share of female scientists rose from an average of 22% in 1980–1986 to 32% in 2006–2011 (the most recent period before the arrival of the CRPs.). 3 ILRAD’s work on gender was limited to some aspects of its field epidemiology studies in East Africa after 1986. 4 One experienced ILRI ecologist was said to have referred to gender as the ‘really hard science.’ References Aregu, L., Bishop-Sambrook, C., Puskur, R. and Tesema, E. (2010) Opportunities for promoting gender equality in rural Ethiopia through the commercialization of agriculture. IPMS Working Paper 18. ILRI, Nairobi. Baltenweck, I. and Mutinda, G. (2013) Gender in the East Africa Dairy Development Project. Presented at the Livestock and Fish Gender Working Group Workshop and Planning Meeting, 14–18 October, Addis Ababa, Ethiopia. ILRI, Nairobi. Basu, P., Galiè, A. and Baltenweck, I. (2019) Presence and property: gendered perspectives on participa- tion in a dairy development project in Kenya and Uganda. Women’s Studies International Forum 74, 68–76. CGIAR (2013) Gender strategy for Livestock and Fish. CGIAR Research Plan 3.7. Available at: https:// genderinsite.net/sites/default/files/CGIAR%20Research%20Program%20on%20Livestock%20 and%20Fish%20-%20Gender%20strategy.pdf; accessed 10 March 2020). CGIAR/IEA (2017) Evaluation of gender in research and in CGIAR workplace. Available at: www.cgiar.org/ research/publication/evaluation-gender-research-cgiar-workplace/ (accessed 26 March 2020). Coppock, D. L. (1994) The Borana plateau of southern Ethiopia: synthesis of pastoral research, develop- ment and change. Systems Study No. 5. ILCA, Addis Ababa. Cornwall, A. and Edwards, J. (2010) Introduction: negotiating empowerment. IDS Bulletin 41, 1–9. de Haan, N. and Mulema, A.A. (2018) Moving up the livestock ladder: gender and equity. Presented at A Stakeholder Consultative Workshop, 16 October, ILRI, Addis Ababa. ILRI, Nairobi. Dione, M.M., Ochago, R., Ouma, E.A., Lule, P. and Birungi, R. (2016) The gender dimensions of a pig dis- ease: African swine fever in Uganda. In: Pyburn, R. and van Eerdewijk, A. (eds) A Different Kettle of Fish? Gender Integration in Livestock and Fish Research. LM Publishers, Volendam, Netherlands, pp. 69–76. Dupire, M. (1960) Situation de la femme dans une société pastorale (Peul Wodaabe, nomades du Niger). In: Paulme, D (ed.) Femmes d’Afrique Noire. La Haye, Mouton, Paris, pp. 51–91. Dupire, M. (2018) Exploitation du sol, communautés résidentielles et organisation lignagère des pasteurs WoDaaBe (Niger). In: Pastoralism in Tropical Africa. Routledge, London, pp. 322–337. Echessah, P.N., Swallow, B., Kamara, D.W. and Curry, J. J. (1997) Willingness to contribute labor and money to tsetse control: application of contingent valuation in Busia District, Kenya. World Development 25, 239–253. C. Hill et al. 695 Farnworth, C.R. and Colverson, K.E. (2015) Building a gender-transformative extension and advisory facili- tation system in sub-Saharan Africa. Journal of Gender, Agriculture and Food Security 1, 20–39. Farnworth, C. R., Kantor, P., Kruijssen, F., Longley, C. and Colverson, K. E. (2015) Gender integration in livestock and fisheries value chains: emerging good practices from analysis to action. International Journal of Agricultural Resources, Governance and Ecology 11, 262–279. Galiè, A. (2015) Exploring gender perceptions of resource ownership and their implications for food security among rural livestock owners in Tanzania, Ethiopia, and Nicaragua. Agriculture and Food Security 4, 21–29. Galiè, A. (2017) Gendered perspectives on smallholder cattle production and health management in three sites in Tanzania. Journal of Gender, Agriculture and Food Security 2, 43–65. Galiè, A. (2018a) Gender-responsive forage intensification in the ololili system of Tanzania. In: Tufan, H.A., Grando, S. and Meola, C. (eds) State of the Knowledge for Gender in Breeding: Case Studies for Practitioners. CGIAR Gender and Breeding Initiative, Lima, pp. 44–54. Galiè, A. (2018b) Gendered seed: from variety selection to control over seed. In: Tufan, H.A., Grando, S. and Meola, C. (eds) State of the Knowledge for Gender in Breeding: Case Studies for Practitioners. CGIAR Gender and Breeding Initiative, Lima, pp. 109–121. Galiè, A. and N. de Haan (2019) ‘Leveraging Gender for Food and Nutrition Security Through Agriculture.’ In: Ferranti, P., Berry, E.M., Anderson, J.R. (Eds.), Encyclopedia of Food Security and Sustainability, vol. 3., pp. 426–431. Elsevier. https://doi.org/10.1016/B978-0-08-100596-5.21568-0 Galiè, A. and Farnworth, C. (2019) Power through: a new concept in the empowerment discourse. Global Food Security 21, 13–17. Galiè, A. and Kantor, P. (2016) From gender analysis to transforming gender norms: using empower- ment pathways to enhance gender equity and food security in Tanzania. In: Njuki, J., Parkins, J. and Kaler, A. (eds) Transforming Gender and Food Security in the Global South. Routledge, London, pp. 189–216. Galiè, A., Distefano, F., Kangogo, D., Mattioli, R.C., Wieland, B. and Baltenweck, I. (2017) Gendered per- spectives on smallholder cattle production and health management in three sites in Tanzania. Journal of Gender, Agriculture and Food Security 2, 43–65. Galiè, A., Teufel, N., Korir, L., Baltenweck, I., Girard, A.W., et al. (2019a) The Women’s Empowerment in Livestock Index. Social Indicators Research 142, 799–825. Galiè, A., Teufel, N., Girard, A.W., Baltenweck, I., Dominguez-Salas, P. et al. (2019b) Women’s empowerment, food security and nutrition of pastoral communities in Tanzania. Global Food Security 23, 125–134. Galiè, A., Njiru, N., Heckert, J., Myers, E. and Alonso, S. (2019c) Gendered barriers and opportunities among milk traders in the informal sector in peri-urban Nairobi. Poster prepared for the 2019 Agricul- ture, Nutrition and Health (ANH) Academy Week, 24–29 June, Hyderabad, India. ILRI, Nairobi. Galiè, A., Oloo, S., and Pfeifer, C. (2019d) Exploring feminization of agriculture through gender dynamics across scales. Presented at the Seeds of Change Conference, 2–4 April, University of Canberra, Aus- tralia. ILRI, Nairobi. Gonsalves, J. (2013) A new relevance and better prospects for wider uptake of social learning within CGIAR. CCAFS Working Paper No. 37. CCAFS, Copenhagen. Grace, D., Roesel, K., Kang’ethe, E., Bonfoh, B. and Theis, S. (2015) Gender roles and food safety in 20 informal livestock and fish value chains. IFPRI, Washington, D.C. Gumucio, T. and Rueda, M.T. (2015) Influencing gender-inclusive climate change policies in Latin America. Journal of Gender, Agriculture and Food Security 1, 42–61. Gumucio, T., Mora Benard, M.A., Clavijo, M., Hernández, M.C. and Tafur, M. (2015) Silvopastoral Systems in Latin America: Mitigation Opportunities for Men and Women Livestock Producers. CCAFS Policy Brief. CCAFS, Copenhagen. Gurung, B. and Mentor, H. (2002) Mainstreaming gender-sensitive participatory approaches: The CIAT case study. CGIAR, Systemwide Program on Participatory Research and Gender Analysis (PRGA), Cali, Colombia. Hopen, C.E. (2018) The Pastoral Fulbe Family in Gwandu. Routledge, London. ILRI (2012) Strategy and plan of action to mainstream gender in ILRI. ILRI Policy Brief 6. ILRI, Nairobi. ILRI (2014) One Health, EcoHealth and agriculture associated diseases – report of a regional dialogue, New Delhi, India, 25 November 2013. ILRI, Nairobi. ILRI (2017) CGIAR Research Programme on Livestock and Fish: 2016 performance monitoring report. ILRI, Nairobi. ILRI/ICARDA (2017) Gender capacity development training module 3: Gender responsive organizations. Pre- sented at the Gender Capacity Development Training, 23–27 October, ILRI, Addis Ababa. ILRI, Nairobi. 696 The Impact of ILRI Research on Gender Jiggins, J. (1986) Gender-related impacts and the work of the International Agricultural Research Centers. CGIAR Study Paper No. 17. World Bank, Washington, D.C. Jumba, H., Teufel, N., Kiara, H. and Baltenweck, I. (2016) The use of the Infection and Treatment Method vaccine in controlling East Coast Fever in Kenya: does gender matter for adoption and impact? ILRI Research Brief 71. ILRI, Nairobi. Kiama, T.N., Lindahl, J.F., Sirma, A.J., Senerwa, D.M., Waithanji, E.M., et al. (2016) Kenya dairy farmer perception of moulds and mycotoxins and implications for exposure to aflatoxins: a gendered ana- lysis. African Journal of Food, Agriculture, Nutrition and Development 16, 11106–11125. Kimani, V.N., Mitoko, G., McDermott, B., Grace, D., Ambia, J., et al. (2012) Social and gender determin- ants of risk of cryptosporidiosis, an emerging zoonosis, in Dagoretti, Nairobi, Kenya. Tropical Animal Health and Production 44 (Suppl. 1), S17–S23. Kristjanson, P., Krishna, A., Radeny, M. and Nindo, W. (2004) Pathways out of poverty in western Kenya and the role of livestock. Pro-Poor Livestock Policy Initiative Working Paper No.14. FAO, Rome. Kristjanson, P., Waters-Bayer, A., Johnson, N., Tipilda, A., Njuki, J., et al. (2010) Livestock and women’s livelihoods: a review of the recent evidence. Discussion Paper No. 20. ILRI, Nairobi. Kruijssen, F., Kantor, P., Galiè, A. and Farnworth, C.R. (2016) Adding gender transformation into value chain analysis. In: Pyburn, R. and van Eerdewijk, A. (eds) A Different Kettle of Fish? Gender Integra- tion in Livestock and Fish Research. LM Publishers, Volendam, Netherlands, pp. 45–53. Liljander, A., Yu, M., O’Brien, E., Heller, M., Nepper, J.F., et al. (2015) Field-applicable recombinase poly- merase amplification assay for rapid detection of Mycoplasma capricolum subsp. capripneumoniae. Journal of Clinical Microbiology 53, 2810–2815. Lukuyu, B.A., Kinati, W., Sultana, N. and Mulema, A.A. (2016) A FEAST for women and men: genderizing a feed-assessment tool. In: Pyburn, R. and van Eerdewijk, A. (eds) A Different Kettle of Fish? Gender Integration in Livestock and Fish Research. LM Publishers, Volendam, Netherlands, pp. 36–44. Mabwire, M.I. (2018) Social-cultural values that influence gendered decisions for pork consumption: A case study of pork retail outlets in Kampala District. MSc thesis, Makerere University, Kampala, Uganda. Marshall, K., de Haan, N. and Galiè, A. ( 2019) Ensuring gender-responsive livestock genetic improvement in low to middle income countries. Proceedings of the Association for the Advancement of Animal Breeding And Genetics 23, 171–174. McPeak, J. and Doss, C. R. (2006) Are household production decisions cooperative? Evidence on pas- toral migration and milk sales from northern Kenya. American Journal of Agricultural Economics 88, 525–541. Mora Benard, M.A., Mena Urbina, M.A., Corrales, R., van der Hoek, R. and Ojango, J.M. (2016) The silent cattle breeders in central Nicaragua. In: Pyburn, R. and van Eerdewijk, A. (eds) A Different Kettle of Fish? Gender Integration in Livestock and Fish Research. LM Publishers, Volendam, Netherlands, pp. 85–92. Muindi, P., Waithanji, E. and Bukachi, S. (2015) Gendered effects of contagious bovine pleuropneumonia (CBPP) occurrence and control in a pastoral community in Ijara sub county, northeastern Kenya. ILRI Discussion Paper 32. ILRI, Nairobi. Mulema, A.A., Jogo, W., Damtew, E., Mekonnen, K. and Thorne, P. (2019) Women farmers’ participation in the agricultural research process: Implications for agricultural sustainability in Ethiopia. Inter- national Journal of Agricultural Sustainability 17, 127–145. Mullins, G., Waborne, L., Tsangari, P. and Maarse, L. (1996) Impacts of intensive dairy production on small holder farm women in coastal Kenya. Human Ecology 24, 14–15. Nicholson, C.F., Thornton, P.K., Mohammed, L., Muinga, R.W., Mwamachi, D.M., et al. (1999) Smallholder dairy technology in coastal Kenya. An adoption and impact study. ILRI Impact Assessment Series No. 5. ILRI, Nairobi. Njuki, J. and Sanginga, P.C. (eds) (2013) Women, Livestock Ownership and Markets: Bridging the Gen- der Gap in Eastern and Southern Africa. Taylor & Francis, London. Njuki, J., Poole, J., Johnson, N., Baltenweck, I., Pali, P., et al. (2011) Gender, livestock and livelihood in- dicators. ILRI, Nairobi. Njuki, J.M., Wyatt, A., Baltenweck, I., Yount, K., Null, C., et al. (2015) An exploratory study of dairying in- tensification, women’s decision making, and time use and implications for child nutrition in Kenya. European Journal of Development Research 28, 722–740. Okali, C. and Sumberg, J.E. (1985) Sheep and goats, men and women: household relations and small ruminant development in southwest Nigeria. Agricultural Systems 18, 39–56. Omondi, I., Zander, K., Bauer, S. and Baltenweck, I. (2014) Using dairy hubs to improve farmers’ access to milk markets in Kenya: gender implications. Presented at Tropentag 2014: Bridging the Gap be- C. Hill et al. 697 tween Increasing Knowledge and Decreasing Resources Workshop, 17–19 September, Prague, Czech Republic. ILRI, Nairobi. Price, M., Galiè, A., Marshall, J. and Agu, N. (2018) Elucidating the linkages between women’s empower- ment in livestock and nutrition: a qualitative study of smallholder livestock raisers in Tanzania. Devel- opment in Practice 28, 510–524. Pyburn, R. and van Eerdewijk, A. (eds) (2016) A Different Kettle of Fish? Gender Integration in Livestock and Fish Research. LM Publishers, Volendam, Netherlands. Quisumbing, A.R., Roy, S., Njuki, J., Tanvin, K. and Waithanji, E. (2013) Can dairy value-chain projects change gender norms in rural Bangladesh? Impacts on assets gender norms and time use. IFPRI Discussion Paper 01311. IFPRI, Washington, D.C. Ramaswamy, M. and Galiè, A. (2018) Poultry trait preferences and gender in Ethiopia. In: Tufan, H.A., Grando, S. and Meola, C. (eds) State of the Knowledge for Gender in Breeding: Case Studies for Practi- tioners. CGIAR Gender and Breeding Initiative, Lima, pp. 55–65. Rijke, E. (2017) Gender capacity assessment of the African Chicken Genetic Gains project partners in Ethiopia. ILRI Project Report. ILRI, Nairobi. Romney, D. and Minjauw, B. (2006) Why the ‘Livestock Revolution’ requires research on people. In: Cer- nea, M.M. and Kassam, A.H. (eds) Researching the Culture in Agri-culture: Social Research for International Development. CAB International, Wallingford, UK, pp. 194–208. Roothaert, R., Chiche, Y. and Mulindi, M. (2006) Gender audit and action plan for mainstreaming gender analysis in ILRI. ILRI, Nairobi. Rutherford, A. S. (2008) Broad bed maker technology package innovations in Ethiopian farming systems: an ex post impact assessment. LRI Research Report 20. ILRI, Nairobi. Shapiro, B.I., Getachew Gebru, Solomon Desta, Asfaw Negassa, Kidus Negussie, et al. (2015) Ethiopia livestock master plan. ILRI Project Report. ILRI, Nairobi. Solomon Bekure, de Leeuw, P.N., Grandin, B.E. and Neate, P.J.H. (1991) Maasai Herding: An Analysis of the Livestock Production System of Maasai Pastoralists in Eastern Kajiado District. Kenya. ILCA, Addis Ababa. Stenning, D.J. (1994) Savannah Nomads: A Study of the Wodaabe Pastoral Fulani of Western Bornu Province Northern Region, Nigeria. LIT Verlag Münster, Germany. Tangka, F., Ouma, E.A. and Staal, S.J. (1999) Women and the sustainable development of market-orient- ed dairying: evidence from the highlands of East Africa. Paper presented at the International Sustain- able Development Research Conference, University of Leeds, 25–26 March. ILRI, Nairobi. Tavenner, K. and Crane, T.A. (2018) Gender power in Kenyan dairy: cows, commodities, and commercial- ization. Agriculture and Human Values 35, 701–715. Tavenner, K., Fraval, S., Omondi, I. and Crane, T. A. (2018) Gendered reporting of household dynamics in the Kenyan dairy sector: trends and implications for low emissions dairy development. Gender, Tech- nology and Development 22, 1–19. Turner, M.D. (1999) Merging local and regional analyses of land-use change: the case of livestock in the Sahel. Annals of the Association of American Geographers 89, 192–219. van Wijk, M. and Hammond, J. (2018) The Rural Household Multiple Indicator Survey (RHOMIS): evaluat- ing where and for whom Sustainable Intensification works. Feed the Future Innovation Lab for Col- laborative Research on Sustainable Intensification. Kansas State University, Manhattan, Kansas. van Wijk, M.T.V., Hammond, J., Etten, J. V., Pagella, T., Ritzema, R.S., et al. (2016). The Rural Household Multi-Indicator Survey (RHoMIS): a rapid, cost-effective and flexible tool for farm household charac- terisation, targeting interventions and monitoring progress towards climate-smart agriculture. CCAFS, Copenhagen. von Kaufmann, R., Chater, S. and Blench, R.(eds) (1986) Livestock systems research in Nigeria’s subhu- mid zone. Proceedings of the Second ILCA/NAPRI Symposium, 29 October–2 November 1984, Kaduna, Nigeria. ILCA, Addis Ababa. Waithanji, E. and Grace, D. (2014) Tools and concepts for mainstreaming gender in aflatoxin research at the International Livestock Research Institute. ILRI Project Report. ILRI, Nairobi. Waters-Bayer, A. (1985) Dairying by settled Fulani women in Central Nigeria and some implications for dairy development. Pastoral Development Unit, ODI, London. Waters-Bayer, A. and Bayer, W. (2014) Supporting development processes through livestock innovation systems research. Presented at Tropentag 2014: Bridging the Gap between Increasing Knowledge and Decreasing Resources Workshop, 17–19 September, Prague, Czech Republic. ILRI, Nairobi. Wieland, B., Kinati, W. and Mulema, A.A. (2016) Sheep are like fast-growing cabbage: Gender dimensions of small ruminant health in Ethiopia. In: Pyburn, R. and van Eerdewijk, A. (eds) A Different Kettle of Fish? Gender Integration in Livestock and Fish Research. LM Publishers, Volendam, Netherlands, pp. 61–68. 698 The Impact of ILRI Research on Gender Annex 1: Gender Research in CGIAR CIAT programme; and (iv) the Gender and Diversity Programme (1999–2012) hosted by CGIAR’s efforts to address gender in international the International Centre for Research on Agrofor- agricultural research began in the 1970s and estry (ICRAF) (CGIAR/IEA, 2017). The Participa- have evolved across time and institutions. A re- tory Research and Gender Analysis Programme cent evaluation of gender in CGIAR identified essentially moved gender analysis out of the three phases of gender mainstreaming: a first Gender and Diversity Programme (Gurung and phase in the 1990s, a second from the 1990s to Menter, 2002) and focused on gender research 2011, and a third after 2010 (CGIAR/IEA, primarily on crop and natural resource manage- 2017). The recent CGIAR 2010–2015 Strategy ment research. and Results Framework provided the foundation ILRI and its predecessors played an active for the first round of CRP proposals and identi- part in the Consortium, although the focus and fied gender inequality as a critical area directly intensity of gender research changed over time. affecting CGIAR’s likelihood of success in Gender efforts began with researcher-led in- achieving its four system-level outcomes of re- tra-household approaches to farming systems ducing rural poverty, increasing food security, research in the 1970s and 1980s. Reinvigorated improving nutrition and health, and the sus- efforts on gender in the 1990s were in part due tainable management of natural resources. This to the influence of CGIAR’s Gender and Diversity was a crucial step in acknowledging the import- Programme and of the Systemwide Programme ance of gender equity to the effectiveness of on Participatory Approaches and Gender Ana- CGIAR research. The Consortium developed and lysis. In the early 2000s, gender found growing adopted its first explicit Consortium Level Gen- attention, with a focus on poverty reduction and der Strategy in 2011 and implemented this in renewed interest in social sciences. The hiring of 2012 alongside the first-generation CRPs, cover- a Programme Leader in the latter part of the ing both gender mainstreaming in research and 2000s for a theme that included gender (Liveli- at the CGIAR workplace (CGIAR/IEA, 2017). hoods, Gender and Institutions) and a new Gender Gender Research Coordinators were appointed Strategy helped institutionalize gender in ILRI. in each CRP to lead the gender strategies, sup- These efforts strengthened under the inter-c entre ported by a Senior Gender Adviser at the Consor- research collaboration on the CRP on Livestock tium, and the wider Gender Network has provided and Fish, followed by the more recent CRP on the capacity to advance the process (CGIAR/IEA, Livestock and were supported by gender strat- 2017). egies to guide more strategic research as well as Several CGIAR programmes have focused gender-mainstreamed research and capacity on gender and have had wide influence across the strengthening. centres. These include: (i) the Intra-h ousehold ILCA had played an important role in early Research Programme (1992–2003), led by IFPRI; CGIAR research by highlighting women’s roles (ii) the CGIAR Gender Programme (1991–1999), in farming systems research. Notably, ILCA’s led by CIAT, focusing in part on gender staffing r esearch drew attention to the importance of as well as on gender analysis in research; (iii) the women in pastoral livelihoods in East and West Participatory Research and Gender Analysis Africa and, importantly, contributed to the dis- Programme (1997–2011), which was a system- course on participatory versus extractive know- wide programme until 2010 when it became a ledge systems. Conclusion: The Future of Research at ILRI Jimmy Smith1, Iain Wright1 Delia Grace2 and Brian Perry3 with Isabelle Baltenweck1, Bernard Bett1, the late Michael Blümmel4, Nicoline de Haan1, Alan Duncan5, Polly Ericksen1, Olivier Hanotte6, Jean Hanson4, Chris Jones1, Steve Kemp1, Okeyo Mwai1, Gerald C. Nelson7, Vish Nene1, Lance Robinson8, Steve Staal1, Philip K. Thornton9 and Barbara Wieland4 1International Livestock Research Institute, Nairobi, Kenya; 2International Livestock Research Institute, Nairobi, Kenya and University of Greenwich, UK; 3University of Oxford and University of Edinburgh, UK; 4International Livestock Research Institute, Addis Ababa, Ethiopia; 5International Livestock Research Institute, Addis Ababa, Ethiopia, and Global Academy of Agriculture and Food Security, University of Edinburgh, UK; 6International Livestock Research Institute, Addis Ababa, Ethiopia, and School of Life Science, University of Nottingham, Nottingham, UK; 7University of Illinois, Urbana-Champaign, Champaign, Illinois, USA; 8Ruwaza Sustainable Development, Stratford, Ontario, Canada; 9CGIAR Research Programme on Climate Change, Agriculture and Food Security (CCAFS) and International Livestock Research Institute, Nairobi, Kenya; University of Edinburgh, Edinburgh, UK Contents Introduction 700 The Contributions of Livestock 700 Livestock and food and nutrition security 700 Nutritional benefits of animal-source foods (ASFs) 701 Overconsumption of ASFs 701 Livestock and prosperity 701 Livestock and natural resources 701 Livestock and climate change 702 The Research Context: Demand and Supply of Livestock Products 702 Demand 702 Supply 702 Imports 704 Industrial livestock production 705 Supply from small- and medium-sized producers 706 Livestock and resource use 707 Labour 707 Land 708 Environmental costs 710 Livestock and the Global Development Goals 710 Priorities for ILRI’s Research in the Next 10 Years 710 Livestock and food and nutrition security 710 Livestock health 712 Herd health 713 © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 699 700 Jimmy Smith et al. Transboundary animal diseases 713 Diagnostics and vaccines 713 Animal welfare 714 Livestock genetics 714 Livestock feed and forages 715 Plant improvement of feed and forages 716 Reduced cost of feed and more efficient feed value chains 717 Optimized feed strategies and sourcing 717 Livestock and human health 717 Food-borne diseases 718 Emerging zoonoses 718 Neglected zoonoses 719 Antimicrobial resistance 719 Livestock and climate change 719 Mitigation 719 Adaptation 720 Policy modelling 720 Livestock and natural resources 720 Livestock and prosperity: Growth with equity 721 The future of smallholders 722 Market links 722 Trade in livestock products 723 Modelling livestock systems and economies 724 Livestock and gender equity 725 ILRI Operations in a Competitive Global Environment 726 Changed funding environment 726 Collaboration with national agricultural research systems (NARS) 727 Impact at scale 727 References 728 Introduction • The research context for livestock in terms of global Sustainable Development Goals. This book has presented the achievements and • Research challenges in the principal domains development impacts of livestock research con- covered in this book – animal health and genet- ducted by the International Livestock Research ics, including livestock and human health; pri- Institute (ILRI), its predecessors the International mary production, focusing on animal feed and Livestock Centre for Africa (ILCA) and the Inter- forages; livestock systems, including policy and national Laboratory for Research on Animal economics, climate change, and gender. Diseases (ILRAD), and selected partners over a • The potential research contribution to the 45-year period. These achievements and im- Sustainable Development Goals. pacts are summarized in the Introduction to this • Future operational issues for ILRI and volume and in the Executive Summaries of each partners. chapter. This final chapter looks at the future – what should ILRI do in the next 10 years? – by considering the challenges facing livestock The Contributions of Livestock research and livestock development and sug- gesting priorities for work in those areas. Livestock and food and nutrition security The chapter considers future priorities for ILRI by reviewing the following: An estimated 820 million of the world’s people • The contributions livestock make to human are food insecure. Some 2 billion are undernour- welfare. ished because they do not consume enough • Global projections of animal product con- protein, vitamins and minerals to lead healthy sumption, production and resource use. lives; of these people, perhaps 1.5 billion may The Future of Research at ILRI 701 suffer from micronutrient deficiencies. The major- are now overweight and a further 600 million are ity of food-insecure and malnourished people obese (HLPE, 2016, p. 15). It has been contended live in the tropics, which is ILRI’s principal man- for some time that high levels of meat consump- date zone (FAO/IFAD/WFP, 2018). tion in particular may increase risks of heart The welfare effects of food and nutrition disease, diabetes and certain cancers. However, insecurity are significant. Some African and recently the evidence for this has been questioned Asian nations may be losing as much as 10% of (Leroy and Cofnas, 2019). Part of the global research their annual gross domestic product (GDP) to agenda will be to reduce overconsumption of undernutrition (World Bank, 2019). A leading ASFs in high- and middle- income countries example of these effects is childhood stunting, where it has adverse health effects. manifesting as low height for age and often causing delayed cognitive development and durable harm to individual welfare. The Food and Agriculture Livestock and prosperity Organization of the United Nations (FAO) has estimated that about 150 million children are The demand for livestock foods is increasing stunted in low- and middle-income countries as both the global population and household (FAO/IFAD/WFP, 2018). incomes increase. With growth of the world population projected to increase from 7.8 billion Nutritional benefits of animal-source in March 2020 to 9.7 billion by 2050, and with foods (ASFs) growing incomes and urbanization, demand for ASF is projected to continue to increase; the ASFs – milk, meat, offal and eggs – provide scarce highest growth in total and per capita consump- nutrients in much of the world. These foods pro- tion of ASFs is expected to occur in low- and duce approximately 26% of the protein (often of middle-income countries. higher biological value than plant protein) and Some 750 million people in the world live in 13% of the calories consumed globally. They extreme poverty (World Bank, 2019). Many poor also provide essential micronutrients, including people keep livestock as their only productive asset calcium, iron, phosphorus and zinc, as well as and as one of their few sources of income. The world vitamins A, B12 and D. Many of these micronu- livestock sector provides employment to 1.3 bil- trients tend to be more bioavailable in ASFs than lion people (about one in six people on the planet at in other foods. Some micronutrients, such as the end of 2019) and some form of livelihood to vitamin B12, are found naturally only in ASFs. 1 billion poor people. Meeting the growing demand For many groups, it is physiologically impossible for ASF can therefore be a source of economic to maintain bodily health without consumption growth (incomes and employment) of the poor. of ASFs. The international scientific community Ten times more women own livestock than agrees that the first 1000 days of life are critical own land, and a majority of poor livestock pro- for child development; many nutritionists assert ducers are women (HLPE, 2016). Empirical work, that the nutritional requirements of this period to which ILRI has contributed, suggests that in- cannot be achieved without ASFs in the diet creasing women’s control over assets, including (Adu‐Afarwuah et al., 2017; Grace et al., 2018). livestock, improves welfare, food security, house- Part of the global effort to improve the food and hold nutrition, education and health (see Chapter nutritional status of millions of people will be in- 18, this volume). Research to improve livestock creasing their consumption of ASFs from their productivity is therefore likely to improve wom- existing low levels of such consumption. en’s welfare in direct ways. Overconsumption of ASFs Livestock and natural resources One cost of overconsuming milk, meat and eggs, as well as other foods, is that it can contribute to Livestock can both benefit and harm natural re- increased health risks, including obesity. Now sources. On the positive side, livestock convert considered the world’s number one health prob- large amounts of biomass and waste material, lem, obesity kills three times more people than for which there is no alternative use, into valu- undernutrition; approximately 1 billion adults able products. Some 26% of global ice-free land, 702 Jimmy Smith et al. much of which cannot be used for crops, is used other environmental variables, however, and for grazing animals. Globally, livestock is estimated practices that have beneficial rather than detri- to use 3.73 billion ha of land: 3.38 billion ha for mental co-effects should be favoured. grazing and 0.35 billion ha for feed production Climate change will force tropical farmers to (HLPE, 2016, p. 35). In addition to using land that adapt and this is particularly true of animal agri- is not suitable for crops, animals are beneficial to culture, where stresses from heat, varying growing mixed crop-and-livestock farming by providing periods, and pests and disease are expected to become manure for soils and power for cultivation and more severe. Climate adaptation (see Chapter 16, by consuming crop residues. Research to inten- this volume) has recently been an important part sify animal production using resources that are of ILRI’s global research portfolio and should be- not usable for direct consumption by people can come more important in the next 20 years. therefore produce substantial benefits. On the negative side, livestock are an import- ant user of natural resources (land, water and soil The Research Context: Demand and nutrients) that have competing uses. Some 33% of Supply of Livestock Products cropland is used for cultivating feed crops, which can compete with food crops in many regions. Some 15% of agricultural water use is linked to Demand livestock production. Animal production can also be a cause of deforestation and biodiversity de- Demand for ASF has grown rapidly in low- and struction (HLPE, 2016). The local resource effects middle-income countries, driven by rising house- of livestock, including damage to soils and water, hold incomes and population growth (Delgado can also be adverse. Research to mitigate the re- et al., 1999; Rosegrant et al., 2009; Alexandratos and source costs of animal production should there- Bruinsma, 2012). FAO (2011, p. viii) notes that the fore be an explicit global priority, with attention to ‘vast majority of growth in [demand from 2000 to the differing intensities of ASF and protein crops, 2030] is caused by increasing per capita consump- such as soybeans and grain legumes, in terms of tion rates rather than by increasing population water footprint and marginal land use. levels’; this income effect will continue to dominate through to 2050. Projections of consumption show continued demand growth in all regions, as illus- Livestock and climate change trated in Fig. F.1 for meat and Fig. F. 3 for milk with a general slowing of demand growth for all regions The life cycle of livestock production and con- except South Asia and sub-Saharan Africa until sumption contributes an estimated 14.5% of 2050. Growth rates for meat and milk commodities global greenhouse gas (GHG) emissions. After are projected to remain higher in developing the first global estimates of livestock-generated than in developed countries, notably in South Asia GHGs were published (Steinfeld et al., 2006), and in sub-Saharan Africa, suggesting that the much has been done to measure animal contri- derived demand for research on those goods will re- butions to climate change and to explore mitiga- main strong for decades. tion options (Gerber et al., 2013). It is now ac- cepted that significant potential exists for cutting emissions from tropical livestock through effi- Supply ciency gains at the level of the individual ani- mal, as has been done in temperate areas. Wider Livestock populations have grown rapidly across adoption of best production and marketing developing countries, notably in China and other practices and technologies in feed, health, hus- East Asian nations, and most rapidly in poultry. Es- bandry and manure management – as well as timates for the period 1971–2007 (Fig. F.2) showed greater use of technologies such as biogas gener- consistent trends: (i) meat production in developed ators, renewable energy, and energy-saving countries grew more slowly than in developing methods – could help the global livestock sector countries; and (ii) monogastrics grew more rapidly reduce its outputs of GHGs by as much as 30% than ruminants in all country groupings. (HLPE, 2016). GHG mitigation technologies Projections for the periods 2007–2030 and may come at a cost to animal welfare and 2007–2050 (Fig. F.2) indicated that: (i) meat The Future of Research at ILRI 703 1971–2007 2007–2030 2007–2050 6 4 2 0 Period World East Asia LAC SSA Region Developing South Asia MENA Developed Fig. F.1. Aggregate meat consumption growth rates by period and region, 1971–2007, 2007–2030 and 2007–2050. Rates for 1971–2007 are estimated from historical data; rates for 2007–2030 and 2007–2050 are projected from a world agricultural model. LAC, Latin America and the Caribbean; MENA, Middle East and North Africa; SSA, Sub-Saharan Africa. (Data from Alexandratos and Bruinsma, 2012.) production growth rates for all species are ex- commodities – poultry, beef, milk, mutton, pork pected to decline until 2030 and to remain and eggs. While the livestock number growth roughly constant from 2030 through 2050; (ii) rates of sub-Saharan Africa will rise somewhat numbers for developing countries are projected more rapidly than those of other regions (Fig. F.4), to grow more rapidly than those of developed the share of sub-Saharan Africa in world live- countries (Fig. F.4); (iii) growth rates for poultry stock numbers is projected to rise only from 3% exceed those for the other species in all regions to 5%. This pattern of growth to 2050, although to 2050 (Alexandratos and Bruinsma, 2012, slower than observed in the past 50 years, will Table 4.19); (iv) the high annual rate of growth in offer important production and trade opportun- pig production from 1971 to 2007 will fall to ities for African and Asian producers. barely 1.0% between 2030 and 2050; and (v) the Animal supply growth has two components: high historical rate of growth in poultry produc- (i) numbers of animals; and (ii) productivity per tion is also expected to fall from 2030 to 2050. animal, such as carcass weight or milk yield. In At the rates of growth applied here, by 2050 the earlier period of 1961–2007 (Fig. F.5), growth it is projected that developing countries will pro- of numbers was roughly three times as high in duce about 70% of the world’s meat (ruminant developing countries as in developed. Growth and monogastric) and about 63% of the world’s rates of animal numbers in both groups of coun- milk, which are substantial gains over the esti- tries are projected to fall from 2007 to 2050 mated 2007 values (58% for meat and 46% although the rate in developing countries will for milk). This trend towards an increasing share remain about twice as high as that in developed of production and consumption in low- and countries. The lower projected growth of animal middle-i ncome countries holds for all livestock numbers to 2050 will be general across continents Growth rates in annual percentages 704 Jimmy Smith et al. 1971–2007 2007–2030 2007–2050 6 4 2 0 Period Product All meat Poultry meat Pig meat Bovine meat Ovine meat Fig. F.2. Aggregate meat production growth rates by period and product, 1971–2007 to 2050. Rates for 1971–2007 are estimated from historical data; rates for 2007–2030 and 2007–2050 are projected from a world agricultural model. LAC, Latin America and the Caribbean; MENA, Middle East and North Africa; SSA, Sub-Saharan Africa. (Data from Alexandratos and Bruinsma, 2012.) and species, although growth in aggregate ani- Africa will dominate global productivity growth mal numbers in sub-Saharan Africa will continue in all species save poultry in the period ending in to be strong through 2050. 2050. Vigorous projected productivity growth in Productivity growth per animal, as measured sub-Saharan Africa will depend on development by rates of carcass weight growth, rose dramat- and application of new methods and on reduc- ically in developing countries for pigs and poultry tion of intermediation costs through investment in the period 1962–2006 (Fig. F.6). Estimated in infrastructure. aggregate rates of carcass weight growth across An important question for ILRI’s research is species and regions averaged more than 6% in how supply of ASF in its mandate areas can meet that earlier period. Productivity growth rates are growing demand in economically, socially and projected to fall over the period 2006–2050. environmentally sustainable ways while improv- The projected patterns of growth in animal ing competitiveness from those mandate areas. numbers and carcass weight will adjust the im- Supply for ILRI’s principal mandate region of portance of the two factors in output growth sub-Saharan Africa has three components: through over time. As shown in Figs F.6 and F.7, growth imports, industrial production or smallholder of animal numbers explained more of production production. growth in the historical period of 1962–2006 than did growth of animal weights. In the pro- Imports jection period of 2006–2050, the share of growth in animal weights is projected to become more Imports now meet much of the demand for ASF important for cattle and poultry. Sub-Saharan in sub-Saharan Africa. Sub-Saharan Africa’s Growth rates in annual percentages The Future of Research at ILRI 705 1971–2007 2007–2030 2007–2050 6 4 2 0 World East Asia LAC SSA Region Developing South Asia MENA Developed Fig. F.3. Aggregate milk consumption growth rates by period and region, 1971–2007 to 2050. Rates for 1971–2007 are estimated from historical data; rates for 2007–2030 and 2007–2050 are projected from a world agricultural model. LAC, Latin America and the Caribbean; MENA, Middle East and North Africa; SSA, sub-Saharan Africa. (Data from Alexandratos and Bruinsma, 2012.) average 2008–2014 annual food import bill was food imports, given projected exports. Sub- about US$46.1 billion, representing about 3.2% Saharan Africa has some potential for additional of regional GDP. Milk, meat and eggs accounted exports of meat and milk, but any increase for 14% of food imports, or 0.4% of regional will require reducing yield gaps with competing GDP. Projections using the International Model regions. for Policy Analysis of Agricultural Commodities and Trade (IMPACT) (http://impact-model.ifpri. Industrial livestock production org/; accessed 11 March 2020) showed that the food import bill could reach US$147 billion an- A second component is industrial livestock pro- nually by 2030 (US$142 billion in 2015 terms), duction, which has been essential in Latin or around 2.9% of projected GDP, with milk, America and East Asia. The potential for indus- meat and eggs accounting for US$36 billion, or trial supply in Africa is high but confronts 0.7% of GDP (i.e. the relative share of net imports of barriers of feed transformation efficiency, infra- milk, meat and eggs as a share of GDP would more structure and regulation of externalities. Where than double from 2016 to 2030). While ASF im- concentrate feeds for monogastrics are accessible ports are likely to grow in sub-Saharan Africa, at competitive prices, industrial systems have they are not projected to pose a significantly grown; where feed costs are high relative to larger burden on the region’s capacity to finance product prices, industrial systems will grow more Growth rates in annual percentages 706 Jimmy Smith et al. 1961–1963 2005–2007 2050 100 50 9.0 40 75 6.0 30 50 20 3.0 25 10 0.0 0 0 y ltr ig s ts o y u P oa ffa l ltru Pi gs ts lo ry gs ts lo o g u o go a ffa ult Pi oa ffa P p& bd P p& bu o g bu e n e nd P ep & nd Sh e a le Sh e le a e a Sh le Ca tt Ca tt tt Ca Species Developed East Asia MENA SSA Region Developing LAC South Asia World 1961–1963 and 2005–2007 numbers estimated from historical data; 2050 numbers projected from world agricultural model. LAC, Latin A,erica and the Caribbean;MENA, Middle East and North Africa; SSA, sub-Saharan Africa. + (Data from Alexandratos and Bruinsma, 2012). Fig. F.4. Animal numbers by period, region and species, 1961–1963, 2005–2007 and 2050. Numbers for 1961–1963 and 2005–2007 are estimated from historical data; numbers for 2050 are projected from a world agricultural model. LAC, Latin America and the Caribbean; MENA, Middle East and North Africa; SSA, sub-Saharan Africa. (Data from Alexandratos and Bruinsma, 2012.) slowly than extensive ruminant systems. Conse- Supply from small- and medium-sized quently, ruminant subsystems, apart from dairy- producers ing, will probably remain extensive into the fore- seeable future because their feed-conversion A third element is expanding supply from small- ratios are much higher than those of monogas- and medium-sized producers in Africa and in trics. Another factor maintaining extensive ani- parts of Asia. Although imports of ASF are mal production is labour costs, where low wages growing in developing countries, over 85% of favour smallholder agriculture over larger enter- the total consumption of ASF is now met by prises. Wages in Africa are likely to remain low local supply, much of it from smallholders. With for some time compared with Latin America and respect to sub-Saharan Africa, United Nations East Asia, based on projections of per capita eco- projections are that the rural population share nomic growth rates. Furthermore, industrial in Africa will remain high for decades, provid- livestock production systems can have high en- ing a labour base for the smallholder sector to vironmental costs owing to waste disposal, dominate agriculture; this is unlike much of water nutrient loading and antimicrobial resist- Asia where the rural population has peaked, ance. The projected costs of production factors – and farm sizes have already increased substan- feed costs, infrastructure and environmental tially due to consolidation. Africa’s evolving externalities – suggest that smallholders will demography and land:labour ratios therefore continue to dominate ruminant production in imply both falling farm sizes (Masters et al., sub-Saharan Africa, and that larger units will 2013) and a long-term opportunity for small- gradually dominate monogastrics. holder agriculture. Animal numbers in millions The Future of Research at ILRI 707 1961–2007 2007–2050 10 8 5 2 0 Region Species Poultry Pigs Sheep and goats Cattle and buffalo Fig. F.5. Growth rates of animal numbers by period, region and species, 1961–2007 to 2007–2050. Rates for 1961–2007 are estimated from historical data; numbers for 2007–2050 are projected from a world agricultural model. LAC, Latin America and the Caribbean; MENA, Middle East and North Africa; SSA, sub-Saharan Africa. (Data from Alexandratos and Bruinsma, 2012.) Meeting growing ASF demand in sub- generally, given the growing integration of crop Saharan Africa and achieving the region’s ASF and livestock production at all scales. A fourth export potential implies several research object- objective is to reduce the environmental costs of ives for ILRI. The first is to develop genetic, health livestock production, processing and marketing. and feeding improvements to lower production costs in ruminants for small and medium farmers; Chapters 1–9 (this volume) show the important Livestock and resource use progress in animal health and genetics made at ILRI to date. The second is to develop genetic and Labour health improvements in monogastrics for all enterprise scales, given the likely continued glo- Employment is a major policy problem in Africa, bal trend towards larger production units in particularly among the young, where the 15– monogastrics. The third is to identify policy and 24-year age cohort represents 19% of the popu- infrastructure barriers to agricultural production lation (United Nations Economic Commission Growth rates in annual percentages W D oe rlv delopin E gast A S so iu ath Asia LAC MENA S D Se Aveloped W o D re ldvelopin E gast S Ao su iath Asia LAC MENA S D Se Aveloped 708 Jimmy Smith et al. 1962–2006 2006–2050 10 5 0 Region Species Poultry Pigs Sheep and goats Cattle and buffalo Fig. F.6. Growth rates of animal carcass weights by period, region and species, 1962–2006 to 2006– 2050. Rates for 1962–2006 are estimated from historical data; rates for 2006–2050 are projected from a world agricultural model. LAC, Latin America and the Caribbean; MENA, Middle East and North Africa; SSA, Sub-Saharan Africa. (Data from Alexandratos and Bruinsma, 2012.) for Africa, 2016, pp. 17–18). Labour-intensive is largely complete, and hence additional land agricultural production, including livestock, has for pasture in these regions is unavailable. New substantial rural job creation potential in Africa pasture is available for livestock in Latin Amer- where demand for new employment is high for ica and Central Africa but mainly from forest demographic reasons. To the extent that growth land conversion, which has unbearable costs of livestock production occurs in more labour- in GHG emissions and would therefore not be intensive segments – dairy, pigs and poultry – as sustainable. is projected to be the case in sub-Saharan Africa, The historical phase of cropland expansion then animal agriculture, including processing is nearly complete. Alexandratos and Bruinsma and marketing, will promote labour use. (2012, Tables 4.8 and 4.9 and Fig. 4.3) pro- jected slow global growth in arable land and in Land its rainfed and irrigated components for 2007– 2050. Arable areas of all types will contract in Livestock use land directly, as pasture, and indir- developed nations, expand slowly in sub-Saharan ectly, by consuming grain and crop residues. The Africa and in Latin America and the Caribbean, historical phase of pastureland expansion in and remain roughly constant in the Middle East Africa, the Middle East, South Asia and East Asia and North Africa and in South Asia and East Growth rates in annual percentages W D oe rv ldelopin E gast A S so iauth Asia LAC MENA D Se Sve Aloped W o D re ldvelopin E gast A S so iauth Asia LAC MENA S D Se Aveloped The Future of Research at ILRI 709 1962–2006 2006–2050 150 100 50 0 –50 Region Species Poultry Pigs Sheep and goats Cattle and buffalo Fig. F.7. Shares of growth rates in animal numbers by period, region and species, 1962–2006 and 2006–2050. Rates for 1962–2006 are estimated from historical data; rates for 2006–2050 are projected from a world agricultural model. LAC, Latin America and the Caribbean; MENA, Middle East and North Africa; SSA, Sub-Saharan Africa. (Data from Alexandratos and Bruinsma, 2012.) Asia. Arable land per capita will become par- of the dry feed consumed by livestock is now ticularly scarce in developing countries, falling inedible by humans, this is projected to change from an estimated 0.19 ha per person in 2017 over time. Long-term projections of grain dis- to a projected 0.14 in 2050. The quantity of appearance imply a shift in developing countries cropland for livestock use will therefore con- to more feed-grain use (Alexandratos and Bru- tract globally in per capita terms, implying that insma, 2012, Fig. 3.5) and to less use of pastures more intensive use of that land will be necessary. and crop residues. Therefore, while ruminants Such increasing land scarcity will, of course, will remain important converters of inedible bio- drive an increase in cropping intensity in all re- mass into high-quality protein for human con- gions, but this growth will be limited by rising sumption, the share of edible biomass (i.e. grain water scarcity and environmental constraints and soy) fed to animals will increase in develop- to irrigation. ing nations, where agricultural land is scarcest. One argument in favour of livestock pro- Research on intensification of animal produc- duction for food is that it uses resources, such as tion – in essence, research to raise feed:product pastures and crop residues, that do not compete conversion ratios – will correspondingly become with direct human use. While it is true that most more important. Growth rates in annual percentages W D oe rv ldelopin E gast A S so iauth Asia LAC MENA S D Se Aveloped W o D re ldvelopin E gast A S so iauth Asia LAC MENA S D Se Aveloped 710 Jimmy Smith et al. Environmental costs ILRAD were created, in part because of new expectations from the world community. The The principal global environmental costs of live- original CGIAR mandate was to produce more stock are GHGs and zoonoses. Assessments across food through generating technologies as inter- the production cycle of animals indicate that some national and regional public goods. That man- 14.5% of global GHG emissions are attributable to date subsequently expanded to include poverty livestock (see Chapter 16, this volume). The devel- reduction, embodied in ILRI’s theme in the early oping world contributes 75% of global GHG emis- 2000s of ‘Pathways out of Poverty’. The ‘new sions from ruminants and 56% of emissions from expectations’ include participation in achiev- monogastrics (Herrero et al., 2013). The shares ing the global Sustainable Development Goals. attributed to the developing world will rise if prod- Figure F.8 indicates the potential contribution of uctivity per animal in the tropics does not rise, as it livestock research and development to the 17 has done in the developed world (Capper et al., Sustainable Development Goals (www.un.org/ 2009, for dairy; Capper, 2011, for beef). Research sustainabledevelopment/sustainable-devel- on the global environmental costs of animal pro- opment-goals/; accessed 11 March 2020). The duction will become increasingly important given priorities of ILRI would be organized around expected human and animal population growth; livestock’s role in meeting global goals in five do- land competition among animals, crops and trees; mains: (i) food and nutrition security; (ii) human and the projected long-term growth in the global health; (iii) climate change; (iv) livestock and carbon burden. natural resources; and (v) prosperity and growth Zoonoses are diseases that are transmissible with equity. between humans and animals through direct contact or from food, water or the environment (see Chapters 8 and 9, this volume). Perhaps 60% of all human infectious diseases and as many as Priorities for ILRI’s Research in the 75% of emerging human infectious diseases are Next 10 Years zoonotic. Zoonoses sicken several billion people each year and kill millions, mostly in low- and middle-income countries; one estimate is that Livestock and food and nutrition security emerging zoonoses cost around US$7 billion a year (World Bank, 2012). Global syntheses of Global evidence suggests that higher agricul- the impacts of zoonotic diseases, led by ILRI, tural productivity and food availability do not estimated that in least-developed countries, 20% necessarily improve food and nutrition security. of human sickness and death was due to zoo- Randolph et al. (2007) reviewed the three poten- noses or diseases that had recently jumped spe- tial pathways along which livestock can strengthen cies from animals to people (Grace et al., 2012). food and nutrition security. The risks of zoonoses in the tropics or originating One is the own production to own consump- in the tropics, including those with pandemic po- • tion availability pathway, in which house- tential, will rise as human and animal popula- hold consumption of ASF increases through tions expand. Expanded research on the animal own production, as shown empirically by aspects and the epidemiology of zoonoses will Hoddinott et al. (2015) in rural Ethiopia. become increasingly important for ILRI and its While greater own production of ASF should collaborators. The 2020 coronavirus pandemic lead to greater consumption, existing stud- has made this work even more urgent at ILRI, ies give mixed results about this pathway, which has long experience with zoonoses. and the reasons for the differences are not clear. • A second is the income/wealth access Livestock and Global Development pathway in which incremental animal pro- Goals duction results in new sales, new income and new food purchases. Research on the The research and development challenges of access pathway is the study of livestock today are more complex than when ILCA and value chains, which provide employment The Future of Research at ILRI 711 17 Partnership for the goals 1 No poverty Stateholders of the livestock sector have come together to form Many rural poor rely on livestock the Global Agenda for Sustainable Livestock and recognize the Livestock provide three major pathways out of poverty: (1) securing assets, UN SDGs in their strategy (2) improving productivity and (3) increasing market participation 16 Peace, justice and strong institutions 2 Zero hunger Numerous conflicts in areas where access to land creates tensions Food (energy and high value protein) between communities (e.g. pastoralists) Traction and fertilizer for crop production Livestock can also be a threat to biosecurity Income 15 Life on land 3 Good health and well being The major part of land is used for livestock Essential micronutrients, especially for children, women and the elderly Livestock contribute to biodiversity losses through impacts on Majority of animal diseases could cause human pandemics habitats, LUC, water, land-use change, soil pollution, grassland species, etc. Use of antimicrobial expected to rise in livestock They also contribute to preserve biodiversity and domestic animals Diseases limit livestock productivity are part of biodiversity 14 Life below water 4 Quality education Livestock use large amounts of fishmeal, which leads to A healty diet is key to learning capacities (e.g. school milk programs) overexploitation of marine resources and losses of biodiversity Livestock provides income, which supports education Eutrophication and hypoxic water conditions 5 Gender equality 13 Climate action Majority poor livestock keepers are women, especially with small Poor livestock keepers are among the most vulnerable to climate ruminants and poultry change. Women have less access to resources (land and capital) Livestock are responsible for a significant share of GHG emissions but have large mitigation potential, including through soll carbon sequestration in grasslands 6 Clean water and sanltation Livestock use large amount of water They are source of water pollution (e.g. nitrates) 12 Responsible consumption and production Water contaminated by livestock causes hygene problems Wastes and losses along livestock supply chains are high Livestock can contribute to protect water quality Rebalancing diets and the share of animal products can contribute to sustainability and health 7 Affordable and clean energy Livestock are an energy sink and source 11 Sustainable cities and communities Recycling animal manure (e.g. biogas) provides an alternative to fossil fuels or wood Hundreds of million of people in cities keep livestock Benefits for food security, nutrition, jobs creation Potential threat to health and sanitation 8 Decent work and economic growth Supports rural–urban linkages 40% of agricultural GDP is provided by livestock The sector is growing at a fast rate High rate of child labour and occupational hazards 10 Reduced inequalities Livestock are a source of income, create employment opportunities and provide market participation to poor rural 9 Industry, innovation and infrastructure households Many depend on livestock, including from jobs provided in the value chain (feed, processing, retailing) Small scale livestock keepers lack market access and inclusion Livestock products are mainly consumed locally and marketed informally Fig. F.8. Contributions of livestock to the 17 Sustainable Development Goals (SDGs). 712 Jimmy Smith et al. and income to allow the purchase of nutri- investigations of food safety, transboundary dis- tious foods. eases and zoonoses (Chapters 7–9, this volume). • A third pathway follows incremental local We can identify four evolving trends in live- sales of livestock products, which increase stock health that will influence ILRI research pri- the availability of ASF for households that orities in the coming decade (Perry et al., 2018). are not rural or that do not keep animals. The first trend is new problems. Emerging This is the dominant path in developed and other zoonotic diseases are likely to have countries and will become more important large consequences for human health and liveli- with growth and urbanization in develop- hoods. Illnesses from ASF, such as contaminated ing countries. milk, meat and eggs, can cause significant A major gap in knowledge is the extent to morbidity and mortality. The growth of anti- which these pathways are important in different microbial resistance, which is arguably a result contexts. Until we have this understanding it of misuse of antimicrobials in people and in ani- will be difficult to design interventions and pol- mals, can make some human infections hard or icies to optimize the contribution of livestock to even impossible to treat (see Chapters 7–9, this nutritional security in different settings. There- volume). ILRI has consequently expanded re- fore, ILRI will work to understand the drivers of search in these areas, as described in the section food choice through the different pathways. We on ‘Livestock and human health’ below. will investigate, for example, what drives deci- Second, technological advances have revo- sions on consumption of ASF within producer lutionized our ability to detect, diagnose, cure and and consumer households; where and under what prevent animal diseases. Some of these technolo- circumstances the different pathways operate, gies are health specific (e.g. lateral flow diagnos- including when livestock production within the tics), while others are novel applications to health household increases; and what incentives are problems (e.g. disease reporting via advanced needed to modify/increase consumption. ILRI mobile-phone applications). ILRI and its Bio- will develop new methods for estimating the in- sciences eastern and central Africa (BecA) Hub take of ASF and the minimal levels required for operate platforms for disease detection, and ILRI vulnerable groups. is studying cutting-edge technologies such as A related research domain is the mitigation metagenomics (the study of genetic material of the adverse consumption effects of ASF. Re- recovered directly from environmental samples) search to mitigate negative effects of ASF con- and phenomics (the study of phenotypes), while sumption will focus on food-borne disease, zoo- exploring other platforms to deliver animal notic diseases under the umbrella of OneHealth, health services and to control antimicrobial and equitable consumption of ASF globally. resistance in pathogens. Third, a warmer, wetter world will very likely be a sicker world. Vector- and food-borne diseases are highly climate sensitive, and their distributions may alter dramatically if climate Livestock health change causes major ecosystem disruptions. One likely outcome of climate change is the further ILRI’s lifetime programme has seen major changes extension of disease vectors and tropical dis- in the tropical animal health sciences. ILRI and eases into temperate climates, which may spur its predecessors, ILRAD and ILCA, have made greater investment in tropical animal disease major scientific achievements in animal genetics research. (Chapter 1, this volume), the control and man- Lastly, animal health constraints are in- agement of trypanosomiasis (Chapters 2 and 3, creasingly addressed through a systemic, ‘holistic’ this volume), the elucidation of problems in im- lens rather than through a ‘single disease per- munology and immunoparasitology in cattle spective’. This multi-disease, multidisciplinary (Chapter 4, this volume), the control and man- approach has necessitated stronger partnerships agement of East Coast fever (ECF; Chapters 5 among ILRI health, gender, environmental and and 6, this volume) and vector biology (Chapter economic scientists. This systemic approach 10, this volume, on ticks), while launching novel includes the ethical treatment of animals. The Future of Research at ILRI 713 We expect these trends – greater under- and the differential benefits they derive) and standing of the relationships among livestock, how these can be delivered by public and private human health and livelihoods; innovations from agents. research; interactions among climate change, crops, land use and animals; and a new systemic approach to animal health management – to Transboundary animal diseases orient ILRI priorities. In light of these trends, ILRI’s future livestock health research will focus Many animal diseases are best understood and on: (i) a ‘herd health’ approach; (ii) transbound- managed at ecosystem or regional rather than ary animal diseases; and (iii) developing vaccines local levels. Among such afflictions are trans- and diagnostics for specific diseases. boundary animal diseases (see Chapter 7, this volume), which are both transmissible and highly contagious. They can easily spread and quickly Herd health become epidemic where control, management or exclusion require costly cooperation among countries. Transboundary diseases have been The reasons for poor herd health in the tropics the main preoccupation of veterinary health include inadequate husbandry, lack of inputs, services in low- and middle-income countries, poor knowledge of disease prevention and man- and while emphasis on herd health has rightly agement, and environmental disease pressure grown, transboundary diseases remain import- created by tropical conditions. Herd health ant constraints to livestock production. ILRI’s research at ILRI will work to understand work on transboundary diseases will involve the  impacts of diseases (both infectious and global collaboration for their progressive con- non-i nfectious) in small- and medium-sized trol. ILRI will support trade that benefits poor farms and herding households, and to develop producers by developing and testing new innov- and test interventions to prevent and control ations for better control of these diseases. disease. Whereas early ILRI (ILRAD/ILCA) research focused on diagnosis and control of specific dis- eases (especially ECF and African animal tryp- Diagnostics and vaccines anosomiasis), the institute has broadened its approach to investigate co-infections in animals Vaccines and diagnostics are as important to and syndromes such as perinatal deaths rather animal health as they are to human health and than investigating individual diseases such as have been central to ILRI research since the brucellosis or toxoplasmosis. 1970s. Diagnostic tools play a critical role in ILRI’s attention is now expanding to in- understanding infection and epidemiology, moni- clude intervention delivery. Many vaccines and toring disease, discovering pathogens and devel- other animal health treatments remain on the oping control strategies. In tropical countries, shelf because insufficient attention has been there is a need for more specific, sensitive and paid to providing access to them for poor farm- field-adapted tests for livestock diseases. Some ers. For example, cheap, effective, thermostable low-technology applications have potential for vaccines exist for Newcastle disease virus, often controlling livestock diseases, such as implement- the worst killer of village poultry, yet uptake of ing biosecurity protocols, although ILRI research the vaccine in village systems has been limited has shown how difficult this is in smallholder and dependent on transient animal health pro- systems. In general, reliable diagnostics and jects. In future, we will increasingly prioritize effective vaccines represent the most sustainable animal health challenges in different systems and cost-effective solutions to preventing spe- and contexts and develop and test appropriate cific diseases, especially in under-resourced agri- packages of herd health interventions for differ- cultural systems. Although vaccines are available ent contexts. We will look at how these can be for many infectious diseases, many are unsuit- tailored for specific livestock keepers (e.g. the dif- able for use in developing countries due to high ferential access to vaccines by men and women cost or unreliable supply. In such cases, policy 714 Jimmy Smith et al. and market changes could increase the use of catarrhal fever, peste des petits ruminants, Rift livestock vaccines and other products to improve Valley fever, and tick-borne haemoparasitic herd health. In some cases, vaccines used in the pathogens and their tick vectors. developed world exhibit limited efficacy against pathogen variants present in tropical climates. Current knowledge allows veterinary compan- Animal welfare ies to rapidly respond to some livestock disease emergencies, such as Schmallenberg virus and ILRI constantly seeks to improve the welfare of bluetongue virus. Good manufacturing and vac- animals under its direct and indirect care and cine-production facilities are needed in develop- has done much to improve animal health, which, ing countries to meet such needs and to supply by necessity, improves animal welfare. However, vaccines in suitable formats for use in diverse ILRI has done little as yet on other aspects of ani- situations, such as thermostable vaccine formu- mal welfare. Animal welfare is an economic as lations for regions where use of a cold chain is well as an ethical issue because it involves other problematical. aspects of livestock development, such as animal Combining new tools in vaccinology in ‘dif- housing, processing and management. ILRI is ficult-to-make’ vaccines can ultimately benefit exploring win–win interventions that im- the most people. These tools can be used: (i) to prove animal welfare as well as productivity and monitor and map immune responses to infection efficiency. and immunization; (ii) to identify candidate vac- cine antigens; and (iii) to redesign vaccine anti- gens and immunization methods to increase Livestock genetics their efficacy. Deep sequencing and gene-expression pro- ILRI’s genetics programme has applied a dual files are leading to maps that describe networks strategy of characterizing population diversity of innate and acquired immune responses in dis- per se for the purpose of understanding the rela- ease and health. Mapping antigen epitopes has tionships and origins of different ecotypes, while given rise to algorithms that can predict anti- parallel studies have addressed individual genes genic determinants. Peptide–major histocom- or traits of interest (reviewed in Chapter 1, this patibility complex tetramer technology to track volume). The intersection to date between the and characterize antigen-specific T-lymphocytes population and individual approaches has is now available for veterinary research. Syn- been in the use of population studies to detect thetic antibody libraries and tools to engineer regions or alleles under selection in order to val- antibodies with desired properties are revealing idate candidates derived from trait-specific studies their value for both therapeutic vaccines and such as trypanotolerance (see Chapters 1 and 2, diagnostic purposes. Crystal structures of anti- this volume). gens are providing clues on how they can be ma- The population and individual approaches nipulated to serve as superior antigens. The will be joined more closely as sequencing and in- realm of adjuvant research is offering guidelines formatics technology advance. For example, the on their use to skew immune responses towards African Dairy Genetic Gains (ADGG) programme desired responses. Collectively, these approaches and its predecessors initially used single-nucleo- will find greater utility in characterizing immune tide polymorphism analysis to determine breed responses to infection or to vaccination and in composition of individual cattle and to generate defining in vitro correlates with immunity, sig- genomic relationships, where pedigree records nificantly reducing the time needed to attack did not exist. These relationships were then used neglected diseases. with phenotypic and pedigree data to derive gen- ILRI will continue to work on research to omic breeding value predictions. Valuable though develop novel vaccines or methods for the con- such predictions are, they tell us little about gen- trol of African swine fever, contagious bovine ome function. As ADGG data accumulates it will pleuropneumonia, contagious caprine pleuro- generate insights into regions of the genome pneumonia and ECF. Other diseases of interest controlling important functional genetic variation, will include foot-and-mouth disease, malignant which coupled with other studies could allow The Future of Research at ILRI 715 functional genetic variants to be identified and exploited by means of marker-assisted selection, used in future. This approach has started to rare variants or variants absent from the target approximate the livestock landscape genomics population, in which case genome editing becomes studies envisioned in the 1990s. We are now be- a possible mechanism for delivery, or genes or ginning to integrate geographic, phenotypic and variants absent from the target species, in which genetic characteristics to allow detection signals case more substantial genetic modification be- for heat tolerance, for example, and to link these comes a likely tool. There are examples of all signals to genomic make-up. These data will soon three types of variant already under study today. be complemented by targeted investigations at In all the genetic studies, it will be necessary to ILRI’s Kapiti Research Station and the Addis consider the possibility of unexpected pleiotropies Ababa poultry facility, which offer controlled (when one gene influences two or more seemingly environments for the calibration of field ob- unrelated phenotypic traits). An unambiguous servations. win–win from a particular genetic intervention This wealth of information will be trans- that nature and thousands of generations of live- formational. We look forward to a new and stock keepers have not discovered might seem un- much more integrated approach in which gen- likely and we might expect to see downsides from etic hypotheses are raised from populations, attempts at ‘improvement’. However, the simple tested in the laboratory and returned for valid- fact that current performance is primarily limited ation in the population. Such an approach allows by feed, water, disease and management suggests modelling of the possible impact of a given gen- that we have an opportunity for very significant, etic intervention, which in turn allows priori- absolute improvement in performance through tization of such interventions. This integrated improved genetics combined with a modified envir- approach is built into a livestock genetics strat- onment in the form of better husbandry. egy, which envisages three phases: (i) targeting, The final and most difficult phase is delivery. (ii) gene discovery, and (iii) delivery. This requires exploration of links with farmers, The targeting phase exploits information artificial insemination providers, commercial about the sensitivity of populations to a particu- breeders and many other value-chain agents. Our lar environmental pressure to predict the poten- approach to this has been to use mobile tech- tial values of adaptations. For example, studying nologies to collect the raw information needed, the impact of thermal stress on milk production for example to plan improved breeding systems. allows us to consider the diversity and heritability In return, we offer management information, of response in existing populations, to consider training and links to input providers and markets to the introduction of new variants, including by the farmer. There is thus a short-period cycle of in- genome editing, and to model the effects on prod- formation between farmers and researchers uctivity, interactions with the value chain, and within the delivery component, while the same environmental impact. Similarly, modelling the data form part of a long-term cycle involving the impact of trypanosomiasis allows quantification targeting and gene-discovery phases of the model. of the consequences of introducing trypanotol- ILRI’s livestock genetics programme thus erant or resistant cattle, or the prediction of climate- looks to a future that builds on wide genetics and change effects. genomics expertise and on unique field experi- The gene-discovery phase follows in a close ence while using modern informatics to create an cycle between field and laboratory data to under- integrated view of performance. It also seeks to stand the genetic basis of a targeted trait. Infor- understand the broader social and economic con- mation from other species and related traits can text for which we are ‘designing’ livestock and to be used to identify fruitful variants. This is a pro- ensure the sustainable and equitable transform- cess that relies on bioinformatics and compara- ation of animal agriculture in the 21st century. tive sequence data to explore pathways as well as on traditional mapping studies and laboratory work used in earlier studies to map regions of Livestock feed and forages the genome associated with trypanotolerance. Such studies might identify variants already Lack of quality feed has constrained livestock in the population at sufficient frequency to be productivity in the mixed and pastoral systems 716 Jimmy Smith et al. of the tropics (see Chapters 11–14, this volume). logical and socio-economic conditions across de- The feed constraint will tighten with evolving veloping countries (Sumberg, 2002). ILRI will climate change and with greater land scarcity. lead this effort by characterizing feeding systems Increased demand for cropland will further re- to enable prioritization of options based on an duce grazing areas and leave remaining grazing understanding of local and regional constraints. lands at risk of degradation. The growing scar- Feed and forage research will include the city of grazing will require an additional re- following: search effort on crop residues and on planted forages. • Work using new genetic tools on plant Crop residues will provide the main feed in productivity. mixed systems (see Chapters 11, 13, and 14, • The Feed Assessment Tool (FEAST; https:// this volume), given the constraints to grazing www.ilri.org/feast; accessed 11 March 2020), resources. Research over the past two decades which can be used to estimate feeding con- has shown that genotypic variation in fodder straints and options to relieve them (see traits can be captured to improve feed quality Chapter 13, this volume). without loss of grain yield and quality (see • Defining feed options based on expert know- Chapter 14, this volume). With perhaps 50 bil- ledge, national survey and census data, and lion t of cellulosic biomass annually produced global comparators. globally, the opportunities for leveraging in- • Mainstreaming near-infrared spectroscopy vestment in second-generation biofuel tech- analysis as a default platform for feed ana- nologies could be a turning point in converting lysis as an extension of historical work on lower-feed-quality biomass into higher-value the feed value of multi-purpose crops (see feeds. ILRI is exploring with partners the busi- Chapter 14, this volume). ness models that would allow these technolo- • Optimized feed strategies and sourcing. gies, which break the lignocellulose bonds in • Reduced cost of feed and improved feed plant cell walls to release their sugars, to be value-chain function. exploited. Planted forages are an important source of Plant improvement of feed and forages livestock feed after crop residues, despite histor- ical disappointments with planted forages in the New tools in molecular genetics and genomic tropics (see Chapters 12 and 13, this volume, technologies, together with high- throughput and the meta-analysis of White et al., 2013). phenotyping capabilities, offer rapid, accurate The challenge will be to meet new feed demands and cost-effective methods to quantify traits, al- on a narrower land base without damaging the lowing the capture of genetic diversity in plants land resource or destroying biodiversity, and that can be used for feed. These advances have while seeking to overcome past failures with been widely applied to crop improvement and planted forages in tropical and subtropical envir- offer the opportunity for new approaches to onments. A focus of new research will be to enhance the quality and performance traits of understand the biotic and abiotic reasons for concentrate feeds, planted forages and crop slow adoption of planted forages on small farms residues. in the semi-arid and subhumid tropics. Opportunities to capture the genetic diver- There will be two broad areas of feed work: sity of crops have been revolutionized by advances (i) improvement of feed systems and (ii) improve- in genomics and automated phenotyping. ILRI ments of feed materials. research will use these and other advances to en- Improvement of feed systems includes hance relevant traits in planted forages and crop enhanced targeting of simple technologies, com- residues. These advances offer new approaches to mercialization of forage seeds, better posthar- enhance the performance traits of both feeds vest processing, and application of new feeding and forages at a relatively low cost, including methods under local conditions. Considerable targeting improved agronomic performance by effort has been invested in promoting livestock tapping the genetic variation in ILRI’s forage feed interventions using blanket approaches, gene bank and partner collections (see Chapter which failed to recognize the various agroeco- 12, this volume). The Future of Research at ILRI 717 There is inadequate information on genetic efficacy in feeding strategies. A further strand diversity, genetic maps and genome sequences for of research will assess wide-scale feed supply feed crops. ILRI will therefore seek to develop a plat- and demand to investigate options for improved form of genomic and other biological technologies feed sourcing and to support national decision integrated into a forage improvement programme, making for improved feed supply. One such tool including, as set out for animal genetics above, is FEEDBASE, which constructs feed supply- genomic selection and genome editing. The work and-d emand scenarios from secondary data will generate superior cultivars of planted forages sets of cropping patterns, land-use patterns, to be disseminated by public and private partners. and livestock census. Initially established in Enhanced multidimensional crops will be made India, FEEDBASE has now been developed for available with the same wide collaboration. Ethiopia, based on more than 60 districts in the country, and there are plans for tests in Malawi, Reduced cost of feed and more efficient Mali, and Vietnam. feed value chains Optimized feed strategies and sourcing Feed is the largest single cost in most livestock enterprises. Work to reduce costs will include Feed demand-and-supply scenarios provide an the development of approaches for rapid eco- important tool to support rational decisions nomic appraisal of feeds. We will also develop a around investments in livestock and feed and fod- data collection and analysis framework to calcu- der value chains at local levels as well as policy late the cost:benefit ratios for individual feed decisions at higher levels. These scenarios can: (i) technologies already identified by the FEAST identify hotspots of feed surplus and deficit areas; tool, and we will pilot-test candidate technolo- (ii) identify priorities for investment in feed re- gies to ground-truth cost–benefit analyses (see sources, considering projected national targets; Chapter 13, this volume). and (iii) quantify feed-dependent yield gaps in Lack of good-quality seed constrains forage livestock productivity. Work will involve locating development. ILRI will address this by building and matching suitable secondary data sets to es- on recent research in Ethiopia that piloted devel- timate feed availability from data sets on crop- opment approaches to seed supply. This will in- ping (crop residues, by-products and grains), volve developing local capacity in seed and forage land use (pasture, common property, forests, and production, as well as testing new approaches planted forages) and location of agro-industries for the promotion, sales, and marketing of forage to calculate feed supply. Feed demand will be esti- seeds (especially in small quantities for small- mated using data sets on livestock populations holders) and clarifying the roles of the public and distributions combined with calculations of and private sectors. ILRI will also support seed animal maintenance, production, and reproduc- certification and seed-quality programmes for tion. Work will also involve ground-truthing the forage sector in target countries. scenarios using project findings and direct sur- A further area of work will focus on feed veys and inventories of feed resources and de- quality analysis/certification. Farmers are often mand. We will develop interactive web, tablet, reluctant to buy feed supplements and complete and phone-based feed demand-and-supply tools feeds for ruminants because of uncertain qual- for use by different stakeholders. We also envi- ity and distorted feed quality-price relationships. sion livestock line departments, non-governmen- This work will be used to promote livestock feed tal organizations, and private companies using options based on their costs and benefits. Finally, ILRI’s feed supply-and-demand scenarios to in- we envision feed producers branding products form national and subnational decision making with quality marks based on certification sup- around livestock feed development. port originating in ILRI. The prioritization of feeding options will build on developments of FEAST and the Tropical Livestock and human health Forages Tool (TFT). FEAST and TFT are already in widespread use (see Chapter 13, this volume), Work on livestock and human health fits so the future emphasis will be on assessing their within a framework of OneHealth, defined as 718 Jimmy Smith et al. multidisciplinary collaboration to attain optimal and 2017, approximately 80% of all food-safety health for people, animals and the environment. projects were focused on three areas – national Under the umbrella of OneHealth, ILRI’s research food-safety control systems, aflatoxins and pesti- on the links between livestock and human cide residues – yet the public-health benefits of health will proceed under four rubrics: (i) food- these investments were hard to determine (Alon- borne disease; (ii) emerging zoonoses; (iii) neg- so, 2019). lected zoonoses; and (iv) antimicrobial resistance ILRI has been a leader (see Chapter 9, this associated with farming. volume) in generating evidence for new food- safety priorities. We believe the following will be- Food-borne diseases come more important: Food-borne disease was a minor topic in ILRI’s • Understanding the presence, prevalence, early years but in recent years has risen rapidly and risk factors associated with food-borne in the global health agenda (see Chapter 9, this disease. volume). A recent global assessment of the health • Identifying food-safety risks in informal burden of food-borne disease found that it was markets, where most of the poor trade and comparable to that of the ‘big three diseases’ – where the heaviest burden of food-borne human immunodeficiency virus/acquired im- disease is felt. mune deficiency syndrome (HIV/AIDS), mal- • Quantifying the health and economic aria, and tuberculosis (Havelaar et al., 2015). burdens of food-borne disease. Food consumers are increasingly concerned • Exploring relationships among food safety, about food-borne disease: in several countries it gender, and nutrition. ranks as their single highest concern (Grace, • Discovering, developing, and testing tech- 2015). Current evidence indicates that ASF is nologies to mitigate hazards, or improving responsible for most food-borne disease (Hoff- their detection and removal. mann et al., 2015) and that the economic bur- • Testing the ‘triple approach’ to food safety den of food-borne disease is high, but national – regulation, capacity, and incentives for data are lacking. While most communicable behaviour change. diseases are declining, it appears that food- borne diseases may worsen as a result of agri- Emerging zoonoses food system transformation in low- and m iddle-income countries, making food-borne Diseases that jump from other species to humans, diseases increasingly important and relevant to known as zoonoses, are a relatively new area for development. ILRI (Grace et al., 2012). Some 75% of new A first step is to generate data on human infectious illnesses are classed as zoo- food-safety priorities, based on the human noses. Domestic animals are involved in some health risks of food-borne diseases and their zoonoses with the highest pandemic potential. economic burdens. ILRI research has helped As the spread of COVID-19 is showing, human to disclose an underinvestment in food-borne pandemics are one of the highest ranked global disease research. In Africa alone, around catastrophes in terms of probability and impact US$2.4 billion is invested in the ‘big three’ and yet are among the most predictable and pre- (malaria, tuberculosis, and HIV/AIDS) each ventable catastrophes. Livestock pandemics can year compared with around US$40 million in also result in widespread human misery and food safety, an imbalance that is not likely to animal suffering, especially in communities change soon. dependent on livestock. Risk management follows risk assessment. We will continue to work on high-priority A recent assessment carried out by the Global zoonotic diseases with a livestock interface and Food Safety Partnership (GFSP) in collaboration pandemic potential, including highly pathogenic with ILRI concluded that many previous avian influenza, Rift Valley fever, Middle East re- food-safety investments were poorly aligned spiratory syndrome, and animal aspects of Ebola. with public-health priorities. Their work esti- We will further develop tools for mapping and mated that in sub-Saharan Africa between 2010 targeting the risk of zoonotic disease, modelling The Future of Research at ILRI 719 zoonotic pandemics, forecasting short-term dis- at ILRI in Nairobi in 2019, to influence the ease risks, developing and testing surveillance development agenda on antimicrobial resist- systems, crafting decision-support methods, and ance issues. advising on vaccination strategies. We will also assess the impacts and costs of pandemics and the benefits of their prevention. Livestock and climate change Neglected zoonoses For nearly 20 years, ILRI has contributed to the While emerging diseases often receive the most scientific understanding of livestock and climate attention, neglected zoonoses continue to exact change (see Chapter 16, this volume) in three a high toll on poor consumers and poor live- areas – livestock-based climate-change mitiga- stock-keeping communities. Among the ‘unlucky tion, adaptation, and policy. The main findings 13’ zoonoses responsible for millions of deaths of this work were as follows: and billions of illnesses each year are brucellosis, • Climate-change mitigation requires greater cysticercosis, and tuberculosis (Grace et al., 2012). focus on results in tropical conditions and Most neglected zoonoses are highly amenable to must rely less on direct transfer of models progressive control: we plan to support this from Organisation for Economic Co-operation through the development of tools for assessing, and Development (OECD) situations, imply- managing, and evaluating control of neglected ing more empirical study of environment– zoonoses. Cysticercosis, caused by the pig tape- animal–climate interactions. worm, Taenia solium, has been identified as a • Adaptation of livestock production to cli- priority neglected zoonosis because of its high mate change requires a broader view of its burden and high potential for control and will be impacts. This demands a greater effort on an immediate focus for ILRI, which has already the tropical problems of heat stress and ani- contributed to a roadmap for its progressive eradi- mal performance, livestock reproduction cation, led by the World Health Organization. under heat stress, and livestock and forage pests and diseases that do not exist in tem- Antimicrobial resistance perate climates or which will expand more aggressively in the tropics. Antimicrobial resistance is a special case of an • Climate-change policy recommendations emerging disease (Grace, 2015). Recent years from agricultural research require a have seen rapidly growing numbers of failures much closer integration of supply-and- of previously effective drugs to treat human in- demand modelling and a much more real- fectious diseases. The use of antimicrobials in istic use of feasible policies and their agriculture may have contributed significantly likely outcomes. to these failures of medical treatment, suggest- ing that a OneHealth approach to managing the Mitigation rise in antimicrobial resistance in pathogens is appropriate. ILRI is developing tools for map- Global research has established the climate ping, measuring and mitigating antimicrobial change mitigation potential of technical changes use globally. This includes work to develop vac- in animal production systems. The best under- cines and other alternative means of controlling stood systems are dairying and beef, which ac- livestock disease. The goal of ILRI’s antimicro- count for nearly 70% of GHG emissions (Gerber bial resistance research is to develop, test, and et al., 2013, p. 18) from livestock supply chains. evaluate interventions to mitigate antimicrobial Gerber et al. (2011) studied intensive dairying in resistance risks in livestock value chains. ILRI a temperate climate and estimated mitigation will increasingly support judicious policies on gains from technical changes in feed, genetics, antimicrobial resistance and national action and health. This work found that the potential to plans and capacity-development programmes. reduce GHG emission intensity, defined as GHG This work will be operationalized through the emissions per unit of product, in tropical sys- CGIAR Antimicrobial Resistance Hub, launched tems is large due to the current low productivity 720 Jimmy Smith et al. per animal, but the scientific potential to reduce Policy modelling emissions from beef cattle in the tropics is not well established (see Chapter 16, this volume). While there is extensive process and agent mod- The future of climate-change mitigation re- elling of climate problems in tropical livestock search at ILRI is: (i) to estimate potential mitiga- systems, the literature comparing supply-and- tion gains in less-well-studied systems, such as demand policies for climate-change mitigation extensive beef raised on tropical pastures and is limited. Future modelling must involve closer intensive meat production on tropical mixed integration of supply-and-demand models, as crop-livestock systems so that we are not reliant was done by Weindl et al. (2017). The depend- on estimates from temperate regions; (ii) to iden- ence of arid rangelands on livestock demands an tify the sources of these gains through improved extended data collection, research, and policy feed, genetics, health, and management; (iii) to modelling effort that recognizes that technical refine estimates of the probability of mitigating options are limited (Herrero et al., 2016). Pol- GHG emissions by, for example, improving trop- icies to enable responses to climate change need ical feed-use efficiency; (iv) to identify profitabil- to consider future uncertainties and system ity and social constraints, including policies and transformations that could affect development gender norms, to potential application of tech- goals, such as economic growth, food security, nical gains; and (v) to extend field tests under and livelihood maintenance. tropical conditions of actual emission levels and While there has been extensive modelling possible reductions. of the interactions among livestock systems and climate, the scope for mitigating livestock’s con- tributions to climate change remains unclear in Adaptation terms of productivity and income distribution. We will model and quantify climate-change im- Climate-change adaptation requires a broader pacts such as on pre- and post-production losses, view beyond technical change, including changes infectious/vector-borne livestock diseases and in behaviour, institutions, and cultures. Priority animal health costs, costs of loss of productivity areas for adaptation are the following: (with potential gains in productivity in some • Understanding the genetics of climate areas), the economic disease burden, and impli- adaptation in domestic livestock. cations for public policy and investment. One • Improving technical and institutional modelling focus will be on the economic costs of support for rangelands to promote their climate change and another will be on policies sustainable management. and investments for mitigation. • Improving capacity for surveillance of cli- mate-sensitive diseases, coupled with new diagnostics for testing for these diseases. Livestock and natural resources • Testing and disseminating perennial forage species adapted to hotter and more variable Tropical livestock research recognizes the climates. interactions among agriculture and natural • Developing decision-support tools to target resources in the main production systems. ILRI agricultural adaptations and to monitor historically focused on extensive dryland sys- their impacts on production, adaptive cap- tems (see Chapters 11 and 15, this volume) and acity and the environment. intensive mixed systems (see Chapter 15, this • Developing measures of adaptation suc- volume). cess. Adaptation cannot be observed dir- Rangelands research has been reinvigor- ectly because it is a synthetic indicator ated by renewed global interest in drylands and constructed from sets of variables. An pastoral livelihoods. The economic potential of approach is needed that combines house- pastoral (and agro-pastoral) production systems hold enterprise, assets, and poverty depends on research that stimulates rangeland measures as functions of governance and productivity. The key challenges facing productive policy factors at community and national rangelands remain competition over resources, levels. heightened by increasing conflicts, overuse of The Future of Research at ILRI 721 rangeland resources as land fragmentation rainfed systems that dominate much of Africa continues, and the failure of past governance and Asia, the water used to grow fodder crops approaches. At the moment, the outlook for and for animals to drink is ‘green water’, which rangelands is worrying – as their productivity is not diverted from other uses. declines, so does their capacity to sustain pas- ILRI’s future mixed crop-livestock system toral livelihoods. agenda includes the following: Various community-based approaches have been tested in rangelands. Evidence of what • Measuring livestock’s impact on soil, water, makes participatory and community-based range- and GHG emissions in intensifying mixed land management successful is accumulating, systems, which will be the systems that but researchers have not yet systematically con- expand most rapidly. solidated this evidence (see Chapter 11, this • Testing how livestock can contribute to sus- volume). Evidence is emerging that simple tainable water use across landscapes. interventions by communities to establish (or • Analysing the contributions of livestock to re-establish) seasonal grazing patterns can have sustainable agricultural intensification at a quick and significant effect on rangeland con- the landscape level, using a range of data dition. To assess the feasibility of rangeland sets, indicators from several different frame- management, however, we need more trials of works, and simple trade-off analysis tools. interventions such as planned seasonal grazing, • Using ex ante evaluations of sustainable rangeland rehabilitation, and management of intensification interventions to pilot inter- bush encroachment. Cost–benefit analysis of ventions, addressing barriers to adoption rangeland management must be joined to im- such as lack of input and output market pact assessments of their economic, social, and linkages and lack of incentives to manage biophysical impacts. soil and water sustainably. Scaling local successes beyond pilot sites often proves challenging, given the costs of Livestock and prosperity: growth with equity resource access, the site specificity of many areas, and the need for flexibility. Validation and Markets and consumption patterns are chan- dissemination of evidence for appropriate models ging dramatically with the growth in demand of governance and technology in diverse com- for livestock products in developing countries. munities is thus another priority. There has been a rise in the share of purchased Documentation of the ecosystem services food in rural consumption – up to 45% in East provided by rangelands, including but not and southern Africa and up to 70% in Vietnam. limited to carbon sequestration, is sparse. More Urban centres are no longer the sole food-pro- work to quantify and map these services would cessing areas, and now more than 50% of food allow more careful assessment of investment po- purchased is processed in rural areas in Africa tential in ecosystem services (IPBES, 2018). and Asia. As markets transform to meet demand, Much of ILRI’s previous work on ecosystem ser- the ‘midstream’ of the food-supply chains (e.g. vices focused on wildlife biodiversity. However, processors, wholesalers, truckers) comprises 40% carbon sequestration has recently attracted of the value chain. As incomes increase and greater interest as an option to offset other GHG tastes change, the types of food demanded are emissions. While rangelands offer theoretical also moving away from traditional staples. In potential for carbon sequestration, there is inad- Africa, more than 50% of purchased food is now equate quantitative evidence about feasible se- non-grain, including vegetables and livestock questration in arid and semi-arid environments. products (Reardon et al., 2019). For mixed crop-livestock systems, ILRI’s en- This ‘quiet revolution’ is occurring largely vironmental agenda will address the impacts of in informal markets, and in the case of livestock livestock production on soil and water resources. is managed by small- to medium-scale agents, While livestock are often criticized for being many of whom handle live animals, raw prod- heavy water users, eroding soils, and degrading ucts, or locally processed products. Some agents lands, this narrative is based on insufficient are adding modern, small-scale processing of data from tropical systems. For example, in the products such as yogurt, or hygienic packaging 722 Jimmy Smith et al. of traditional products, at the same time that in http://whylivestockmatter.org/ (accessed 11 larger-scale modern processing, marketing and March 2020). The evidence generated will in- distribution grow. form policy making and public investment. The These structural changes have important factors behind smallholder success will be used implications for understanding how small and in models of livestock systems to guide public in- medium producers can participate in the grow- vestment plans that decide which livestock sys- ing livestock markets. Will they be marginalized tems are prioritized to assist in meeting national due to their inability to meet market demands for and global Sustainable Development Goals. higher quality or for reliable delivery? What are the market mechanisms and organizations that Market links will allow small to medium agents to participate while meeting increasing demand? What are the Business models for linking smallholders to mar- policy approaches to allow both women and kets and services have had a mixed record. For men to benefit as livestock keepers and as market example, contract farming has usually failed in agents? ruminant systems such as dairy and small-scale cattle or small-ruminant production due to the The future of smallholders ability of producers to rely on spot markets. Con- tract farming has, however, proven successful in Despite the fact that smallholders play a large broiler production, where the risk of having a role in livestock supply in many countries, many large batch of broilers at market-ready weight observers assume that larger-scale production with no reliable buyer provides incentive for more will replace smallholder producers. This assump- formal producer–buyer arrangements. Other ap- tion is based on the trend towards larger-scale proaches use a ‘hub’ model, particularly in production in developed nations and on the fact small-scale dairy systems, where a milk-collection that economies of scale are inevitable. The evi- centre provides a location for business activities dence to support these assumptions is mixed. to be consolidated, including provision of veter- While economies of scale for monogastrics (pigs inary services, genetics, and feeds, potentially and poultry) are observable, where low-margin with integrated financial services to facilitate production relies on efficient feeding and on savings and credit for producers to access. biosecurity best managed by professionals, in New tools in information and communica- contrast, economies of scale are not easily achieved tions technology are being explored to provide for ruminants, particularly in dairy but also in market information, production, and process small ruminants (Delgado et al., 2008). A large monitoring and feedback, or extension informa- part of this reality is due to the fact that for tion, which can be linked to these organizational ruminants, smallholders can make use of own- models. Such models suffer in the same way that grown fodder and crop residues, communal fod- other cooperative approaches often do, from poor der sources and family labour. These factors are management and inadequate or excessive regula- only likely to change when wages rise due to eco- tion, and many have failed despite significant nomic growth and so the opportunity costs of public or private investment. In addition, such labour require its substitution by capital invest- models are most likely to succeed in intensive pro- ment, which in part depends on scale to be effi- duction systems where there is a lot of economic cient. The question is thus when and under what activity to drive demand for such services. Link- circumstances should decision makers and in- ing informal to modern markets is often a chal- vestors prioritize larger production systems over lenge, for example, due to the need to address food the smallholders who dominate production in safety and quality. Much less certain are success- much of the tropics. ful models to link producers to output markets ILRI will pursue analysis of the conditions and particularly to services in remote and eco- in which smallholders remain viable, across dif- nomically less dense settings, such as in dryland ferent settings and species, particularly in sub- and pastoral systems where many livestock are Saharan Africa and South Asia. This work will situated. Mobile service providers face high trans- generate more rigorous findings than the mostly port costs there, and producers cannot easily as- case-study evidence available to date, as compiled semble to aggregate demand for the same reason. The Future of Research at ILRI 723 ILRI’s future research on markets will in- and geographically spreading potentially dev- volve the following: astating animal diseases has led to the creation of sanitary and phytosanitary (SPS) standards • Analysis of organizational models that can imposing strict limitations on places from create market and service links for livestock which livestock products can be sourced and on producers, including: (i) the policy, regula- their manner of quality control, processing, pack- tory, and private-sector settings required aging, and so on. Some countries, such as those for success; (ii) the business and capacity- where foot-and-mouth disease is endemic, are development support required in typical limited from exporting some bovine or pig prod- cases; (iii) the manner in which informa- ucts except to similarly infected zones. tion and communications technologies can In spite of such limitations to international facilitate the better functioning of such livestock trade, many livestock trade opportun- organizational models; and (iv) how organ- ities for low- and middle-income countries are izational models can benefit value-chain act- underexploited, particularly where SPS constraints ors, women, youth, and other marginalized may be less binding among countries with simi- groups. Particular attention will be paid to lar disease prevalence. ILRI’s future research remote and economically marginal settings will consider the following: and to exploring how small-scale mobile services can be linked to information and • The potential place and role for approaches communication technologies and GIS tools to overcome SPS barriers such as commodity- to allow cost-effective flows of products and based trade and disease-free zones. Com- services. modity-based trade aims to establish nar- • Analysis of gender-equitable approaches to rowly controlled supply chains that allow a ensure that the needs and capabilities of specific product to meet SPS regulations both women and men are taken into account across borders. Disease-free zones aim to set when designing and evaluating organiza- up broader systems to control disease. In tional models, as well as assessing the im- both cases, the economics of the required pact on women’s empowerment and social investment versus the potential market re- equity more generally. turns are uncertain and often suspect. ILRI • Regarding the specific case of local and in- will identify in which cases such invest- digenous species and products and given ments potentially make sense and where consumers’ persistent preferences for many they do not, and inform policy makers and of these products in sub-Saharan Africa private investors of this. and much of Asia, an examination will be • National statistics often fail to provide good made of the scale of speciality livestock mar- information on the consumption of specific kets and an assessment made of smallholder meat cuts, including beef and poultry cuts opportunities to meet this demand. This and offal, flowing into developing countries. work will include the development of busi- ILRI will examine these nuances, cataloguing ness models with backward links to livestock approaches with the potential to enhance producers and suppliers of genetics and for- livelihoods, food security and nutrition via ward links to specialized restaurants and local livestock value chains faced with retailers. lower priced but imperfectly substitutable livestock food product imports. Trade in livestock products • The national governments of poorer coun- tries are often attracted to the goal of export- Only about 10% of global livestock production ing livestock and livestock products to rich is traded internationally. This is partly due to countries, particularly with the aim of gen- perishability because many livestock products erating hard currency. Evidence suggests, must be transformed to be safely traded and however, that larger and more attainable such transformation – requiring drying, can- markets are often closer to hand – either ning, freezing, or other processing – adds to domestically or among regional neighbours. costs. In addition, the threat of transmitting These markets are, however, often constrained 724 Jimmy Smith et al. by poor infrastructure and inconsistent can only partially capture the direct effects of standards and rules at their borders. ILRI climate change on livestock systems, such as will examine the levels of comparative ad- impacts on animal performance at higher tem- vantage for livestock exports across key de- peratures, or not at all, and even when they do, veloping countries. We will also examine the results are not analysed in an economic- the main constraints to greater regional optimization context. trade, particularly in Africa, and assess ILRI has collaborated with governments’ their impacts and the costs and returns of efforts to develop livestock master plans in Ethi- investments required to alleviate these con- opia, Rwanda, and Tanzania (Shapiro et al., straints. 2015). In Ethiopia, this work has stimulated dir- • While some work has been done to assess ectly and more recently in the State of Bihar in postharvest losses in livestock value chains, India or indirectly several hundred million US$ which found relatively low levels of loss, in new investments in the livestock sector, from almost no work has addressed the much both the public and private sectors. As new re- larger (by volume) markets in live animals quests for livestock master plans emerge, it has in low- and middle-income countries. These become apparent that a wider range of metrics is pose greater problems methodologically desirable from the plan’s outputs, which cur- and practically. ILRI has begun to develop a rently focus only on the supply and demand of suitable methodological approach and will livestock products and their economic impacts apply it in several countries, particularly where under alternative scenarios. In addition, the cur- there are long and complex supply chains rent tools do not address substitutability among for live animals, especially ruminants. livestock products and market dynamics. Devel- oping broader system indicators will allow pub- Modelling livestock systems lic and private agents to make better decisions on and economies livestock investments. Demand for livestock products in low- and mid- data quality. A joint activity is being initiated dle-income countries will continue to grow sig- to improve and harmonize data, tools and nificantly in the coming decades. Managing the methods of global and regional livestock systems external costs of this demand on health and the modelling at ILRI, FAO, IFPRI, the Common- environment will require better models. Current wealth Scientific and Industrial Research tools for livestock-sector modelling are disjointed Organisation (CSIRO), the Centre de coopération and poorly coordinated, and so fall short of their international en recherche agronomies pour le potential. Both FAO and the International Food développement (CIRAD) and other partner insti- Policy Research Institute (IFPRI) have developed tutes. This is required to exploit the common- global models that address agriculture and the alities and complementarities of the multiple ef- environment, but their suitability for addressing forts to develop tools and databases, which are the livestock sector is still limited. IFPRI’s IM- currently not well harmonized. A key objective PACT model and FAO’s model, the Global Agri- is not only to improve the quality of projections culture Perspectives System (FAO, 2016), both fail by much better incorporations of feed and to capture the contribution of grazing land, crop-related interactions but also to improve the planted forages, crop residues and other feeds to metrics for environment, climate change, and livestock production and do not yet adequately other socio-economic and livelihood indicators. model the herd growth dynamics that are im- Part of this work would look at investment in portant in livestock production (Msangi et al., different types of livestock technologies, includ- 2014). Current analytical tools, while able to ing those for disease control, assessing short- conduct supply-and-demand projections, may versus long-term gains and benefits. There is be inadequate to answer important questions also the potential to begin to incorporate the regarding the potential implications for socio- long-term implications of the growth in ‘cultured economic, public health and environmental im- (laboratory-grown) meat’, once supply-and- pacts and for returns on investment. These models demand parameters are better understood. The Future of Research at ILRI 725 public expenditure. Data from the OECD indi- The history of gender research at ILRI is cate that some 4% of official development as- relatively brief but productive. We have evidence sistance goes to agricultural development, of that livestock can be a cornerstone of economic which only a small part is invested in livestock empowerment for women (see Chapter 18, this development and the majority is invested in volume) for several reasons. Livestock are easier crops. This amount is a small fraction of the for poor women to own or to manage than land. share of livestock in global agricultural GDP. The Livestock propagate themselves and can provide reasons for this apparent bias against livestock regular income through the supply of milk and include the dominant role of crop scientists in eggs. A woman can take livestock assets with her agricultural investment decisions, the percep- in case of divorce or conflict. The development of tion that crops support food security more than the Women’s Empowerment in Livestock Index livestock do, and public pressure on donors to re- (Galiè et al., 2018) was an important step in cre- duce support to livestock for environmental and ating a metric of progress. health reasons. Research on gender equity will be under- ILRI will investigate trends in public invest- taken under the following specific themes: ment in livestock research and development work and the reasons for these trends where pri- • Quantifying the links between gender and in- vate research has lagged. More work is needed to creased agricultural production. We need to determine returns to investment (and the broader know whether increasing women’s access welfare impacts) in livestock research and devel- to resources and innovations leads to greater opment, including identifying the highest re- livestock production and higher efficiency. turns to guide resource investments. Part of this We need to know if and when this greater work will be to understand the right balance of access leads to greater female empower- investments in smallholders and commercial ment and if and when it leads to losses of agents. A total economic valuation approach power to men. will also be used given the multiple benefits of • Identifying gender-specific interventions in re- livestock. search for development. More study is needed One area for model improvement in the to understand gender roles in different pro- analysis of public expenditure will be in stake- duction systems. One example is the urgent holder engagement. ILRI will develop model ex- need to conduct time and labour studies to tensions: (i) to apply parameters defined by understand task, ownership, and decision- users, such as national policy makers; (ii) to gen- making roles within animal production erate results that are easy to communicate; and systems. These studies will inform a much- (iii) to represent the objectives of decision makers needed empirical underpinning of how while creating a feedback loop between users households work – individually and collect- and developers that enhances continuous model ively – and will allow a better understand- development. ing of gendered labour dynamics. With this more detailed knowledge of labour dynam- ics, we will be able to better ascertain the Livestock and gender equity implications for given interventions and ap- proaches for women as well as for men. Ensuring that the benefits of livestock produc- • Framing gender policy problems specific to tion are broadly shared is the present and future agriculture. One class of gender policy ques- goal of equity research at ILRI. Improving gen- tion can be posed as: which animal species der equity is important for livestock because and therefore which livestock diseases are research shows that women often do not receive most important to women? Regarding any benefits commensurate with the labour and cap- given species, what role do women play in ital they invest in animals (see Chapter 18, this the care and management of that species – volume). Some of ILRI’s portfolio over the next and how can that be tapped for better ani- 5 years will concentrate on gender equity, which mal health care? In our forage research, provides opportunities for women and men alike how often and when are women respon- to benefit from ILRI’s interventions. sible for feeding the family livestock? How 726 Jimmy Smith et al. do we ensure that we are selecting and pro- are wide paths for youth employment, the evi- moting animal genetic traits that may be dence on youth employment in agriculture and preferred by women? In our genetics re- livestock remains thin. ILRI and partners will search, what are the genetic traits of a given target youth employment creation within the breed of animal that are preferred by agricultural sectors of mandate countries as a women? Why are these traits preferred? response to these demographic opportunities. Can we give women greater access to the Three lines of inquiry will be pursued: (i) devel- traits they prefer? An important and un- opment and adoption of new livestock practices, der-researched area is ensuring that our including digitization and mechanization, and interventions do not increase the labour institutions on farms and in value chains; (ii) burden within households, particularly for promotion of youth employment and entrepre- women, whose labour is already often unre- neurship in the small-scale livestock sector; and warded in monetary terms. For example, (iii) strategic research on youth and livestock. making improvements to animal feeding The results of work by ILRI and partners fo- may entail more work for women, who may cusing on gender equity and youth employment need to spend more time gathering fodder should lead to a series of inclusive business and thus have less time to care for their chil- models that allow women to get competitive dren. Routinely including gender aspects in returns on their capital and labour investments our work creates potential for higher agri- and to benefit fully from these returns. A recent cultural productivity. example of such work is the integration of gen- • Extending livestock as a business opportun- der in the development of livestock master ity. We need to understand what women plans of nations and states (see Chapter 18, this need to start and be successful at commer- volume). cially oriented livestock enterprises. Mov- ing from keeping a few backyard chickens to raising hundreds of birds in a poultry business requires new skills and inputs: ILRI Operations in a Competitive which are most important and how do we Global Environment ensure that women have ready access to these? We also need to better understand Changed funding environment which women would be interested in mak- ing such a transition. We need to know The work summarized in this volume has con- under what circumstances a commercial firmed the major scientific impacts of research activity like this gets taken over by the by ILRI and its partners. We have seen that household men. greater investment in livestock and livestock sys- • Expanding the nutritional benefits of women’s tems research can lead to significant benefits in empowerment through animal production. food and nutrition security, in economic devel- Regarding ASF and nutrition, it is ILRI’s opment, and in natural resource management. aim not only to help women make the best Despite good evidence for these benefits, the glo- nutritional choices for their families but also bal funding environment relative to the scale of to help ensure that females have adequate the research problems has become more chal- access themselves to milk, meat, and eggs, lenging. especially young girls, pregnant and lactat- The CGIAR funding environment was very ing women, older women, and those who different when ILRI’s predecessors – ILRAD and are ill. ILCA – were founded in the early 1970s (Özgediz, 2012). Unrestricted (‘core’) funding was pro- An aspect of equity is youth employment. Popu- vided to centres to do public-goods research on lation growth in Africa is expanding the num- agricultural productivity to improve food security. bers of young people seeking jobs. While labour As more centres joined CGIAR in the force growth presents opportunities, it also rais- 1990s, the historical growth in real system es fears of un(der)employment, migration, and funding began to fall after 19951. Associated violence. While the food and agricultural sectors with a decline in the rate of the growth of real The Future of Research at ILRI 727 funding was a collapse in the share of core fund- building on ILRI’s successful partnership with ing, which fell from almost 100% in the 1980s the Chinese Academy of Agricultural Sciences. to some 40% in 2002. In response to the decline ILRI will generally continue to collaborate with in total funding emerged new multi-centre fund- national systems as a means of doing research ing models, including ‘challenge programmes’; and of developing human and institutional these multi-centre programmes managed to capacity. attract limited additional funding but at the cost of diverting core funding to external partners in the challenge programmes. A new financing mechanism began in Impact at scale 2012 in response to the spending decline that began in the 1990s. Instead of core funding going The founding model of the international agri- directly to centres, supplemented by project cultural research centres was that their research funding, the core funding was converted to pro- would generate technologies and interventions gramme funding (Windows 1 and 2) through that were applicable by hundreds of millions of multi-institutional CGIAR research programmes, poor farmers in developing countries. These known as CRPs. Restricted projects (under the technologies would first be adapted for local use so-called Window 3) were aligned to the CGIAR by national partners and ultimately extended to research programmes. After an initial increase, farmers. While this model did achieve much, as the ‘programme funding’ declined from about testified by the many achievements detailed in 40% in 2012 to 20% in 2018. Despite unreal- this volume, its linear character – from research to istic statements made by some officials of what adaptation to development – no longer serves the was then the ‘CGIAR Consortium’, there has world well, because the challenges of agricul- been no widening of the donor base; the Bill & ture in the tropics have become more complex. Melinda Gates Foundation is the only new donor ILRI’s core expertise will continue to be live- to provide significant funding to livestock. stock research for development, but there will be ILRI and collaborators will continue to two ways in which ILRI departs from its past. argue for more investment in tropical livestock, The first way is that research programmes and focusing on low- and middle-income countries projects are increasingly being conceptualized and on excluded producer groups, but these by ‘starting with the end in mind’. This means efforts are unlikely to provide major resources we start by identifying the development chal- without substantial commitments from new lenge(s) we want to meet – the impact we want donors. ILRI’s funding from 2014 to 2018 aver- to make, where and on whom. Then we work aged about US$75 million in 2015 US$, or back to what outcomes are required to achieve perhaps 7–8% of CGIAR system funding from this, what research outputs are needed to produce Windows 1, 2, and 3, and it is unlikely that the those outcomes, and what research activities ILRI total or the ILRI share of the global amount and resources are needed to produce these out- will grow significantly in real terms. puts. This approach requires ILRI to be more purposeful in identifying which partners can increase impact. Given the increasing role that Collaboration with national agricultural the private sector will play in development, the research systems (NARS) new approach includes greater engagement with private agents at all scales. Collaboration with the NARS of the developing The second way in which ILRI’s business world has been a foundation of ILRI’s work since will evolve is to develop mechanisms to engage the institute’s founding. While some of these na- more effectively with development agencies. tional systems have strengthened considerably In 2017, ILRI established an ‘Impact at Scale’ since the 1970s, many still have limited scien- programme, which is designed to manage the tific capacity relative to their needs. ILRI will in- interaction between research and development. creasingly work with the world’s stronger NARS It identifies products and innovations from ILRI to leverage its expertise. This will be achieved and other research organizations that have been through joint programmes and laboratories, piloted and are ready to be deployed at scale, 728 Jimmy Smith et al. packages them (with adaptations as required) committed to bringing to bear the best science in and then works with development agencies, in- tackling these challenges and in partnering with cluding classic public extension, NGOs, and pri- research institutions in both developed and vate companies. While not engaging directly in developing countries in pursuit of this. field development activities, ILRI will continue to We are also committed to ensuring that develop the expertise needed to engage effect- ILRI’s research is used effectively to make a real ively with partners that do field development. difference to the hundreds of millions of people This more complex agenda requires a differ- in developing and emerging economies that ent way of working, including more diverse depend on livestock – and that our research partnerships and applying different skills and manages this in ways that are socially equitable disciplines across the biophysical and social sci- as well as environmentally sustainable. We will ences. The ‘omics revolution’ has created re- continue to work closely with our funding and search opportunities that were unimaginable development partners to ensure that our research 40 years ago, allowing the creation of new gen- is supremely relevant and reaches all those who etic, biophysical, and economic models. ILRI is need it most. Note 1 Real CGIAR system spending, in 2015 US$, averaged US$271 million during the period 1971–1993, during which it rose at a real annual rate of 7.4%; real spending from 1994 to 2011 averaged US$510 mil- lion and rose at an annual rate of 2.9% from 1994 to 2010; real spending averaged US$876 million during the period 2011–2018 and rose at a real annual rate of 1.3%. It should be noted that real spending took a great leap forward from 2010 to 2011, when it rose by slightly more than 20%. References Adu-Afarwuah, S., Lartey, A. and Dewey, K.G. (2017) Meeting nutritional needs in the first 1000 days: a place for small-quantity lipid-based nutrient supplements. Annals of the New York Academy of Sci- ences 1392, 18–29. Alexandratos, N. and Bruinsma, J. (2012) World agriculture towards 2030/2050: the 2012 revision. ESA Working Paper No. 12-03. FAO, Rome. Alonso, S. (2019) Food safety in Africa. Presented at the Launch of Global Food Safety Partnership (GFSP) Report – Food Safety in Africa: Past Endeavors and Future Directions, 11 February. Addis Ababa. ILRI, Nairobi. Capper, J.L. (2011) The environmental impact of beef production in the United States: 1977 compared with 2007. Journal of Animal Science 89, 4249–4261. Capper, J.L., Cady, R.A. and Bauman, D.E. (2009) The environmental impact of dairy production: 1944 compared with 2007. Journal of Animal Science 87, 2160–2167. Delgado, C., Rosegrant, M., Steinfeld, H., Ehui, S. and Courbois, C. (1999) Livestock to 2020. IFPRI, Washington, D.C. Delgado, C.L., Narrod, C.A., Tiongco, M.M. and de Camargo Barros, G.S.A. (2008) Determinants and implications of the growing scale of livestock farms in four fast-growing developing countries. IFPRI Research Report No. 157. IFPRI, Washington, D.C. FAO (2011) Mapping supply and demand for animal-source foods to 2030. Animal Production and Health Working Paper No. 2. FAO, Rome. FAO (2016) The Global Agriculture Perspectives System (GAPS) Version 1.0. ESA Working Paper No. 16-06. FAO, Rome. Available at: www.fao.org/3/a-i6112e.pdf (accessed 11 March 2020). FAO/IFAD/WFP (2018) The state of food security and nutrition in the world. FAO, Rome. Galiè, A., Teufel, N., Korir, L., Yount, K., Webb, A., et al. (2018) The Women’s Empowerment in Livestock Index. Social Indicators Research 142, 799–825. Gerber, P., Vellinga, T., Opio, C. and Steinfeld, H. (2011) Productivity gains and greenhouse gas emissions intensity in dairy systems. Livestock Science 139, 100–108. Gerber, P., Steinfeld, H., Henderson, B., Mottet, A., Opio, C., et al. (2013) Tackling climate change through livestock: a global assessment of emissions and mitigation opportunities. FAO, Rome. Grace, D. (2015) Review of evidence on antimicrobial resistance and animal agriculture in developing coun- tries. DFID, London. The Future of Research at ILRI 729 Grace, D., Gilbert, J., Randolph, T. and Kang’ethe, E. (2012) The multiple burdens of zoonotic disease and an ecohealth approach to their assessment. Tropical Animal Health and Production 44 (Suppl. 1), S67–S73. Grace, D., Dominguez-Salas, P., Alonso, S., Lannerstad, M., Muunda, E., et al. (2018) The influence of live- stock-derived foods on nutrition during the first 1,000 days of life. ILRI Research Report No. 44. ILRI, Nairobi. Havelaar, A.H., Kirk, M.D., Torgerson, P.R., Gibb, H.J., Hald, T., et al. (2015) World Health Organization global estimates and regional comparisons of the burden of foodborne disease in 2010. PLoS Medicine 12: e1001923. Herrero, M., Havlík, P., Valin, H., Notenbaert, A., Rufino, M.C., et al. (2013) Biomass use, production, feed efficiencies, and greenhouse gas emissions from global livestock systems. Proceedings of the National Academy of Sciences USA 110, 20888–20893. Herrero, M., Henderson, B., Havlík, P., Thornton, P.K., Conant, R.T., et al. (2016) Greenhouse gas mitigation potentials in the livestock sector. Nature Climate Change 6, 452–461. HLPE (2016) Sustainable agricultural development for food security and nutrition: what roles for livestock? A report by the High Level Panel of Experts on Food Security and Nutrition of the Committee on World Food Security, Rome. Hoddinott, J., Headey, D. and Dereje, M. (2015) Cows, missing milk markets, and nutrition in rural Ethiopia. Journal of Development Studies 51, 958–975. Hoffmann, S., Maculloch, B. and Batz, M. (2015) Economic burden of major foodborne illnesses acquired in the United States. Economic Information Bulletin No. EIB-140. USDA, Economic Research Service, Washington, D.C. IPBES (2018) Assessment report on land degradation and restoration. Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) Secretariat, Bonn, Germany. Masters, W.A., Djurfeldt, A.A., de Haan, C., Hazell, P., Jayne, T., et al. (2013) Urbanization and farm size in Asia and Africa: implications for food security and agricultural research. Global Food Security 2, 156–165. Milne, E., Aynekulu, E., Bationo, A., Batjes, N.H. and Boone, R. et al. (2016) Grazing lands in Sub-Saharan Africa and their potential role in climate change mitigation: What we do and don’t know. Environmental Development 19, 70–74. https://doi.org/10.1016/j.envdev.2016.06.001. Msangi, S., Enahoro, D., Herrero, M., Magnan, N., Havlik, P., et al. (2014) Integrating livestock feeds and produc- tion systems into agricultural multi-market models: the example of IMPACT. Food Policy 49, 365–377. Özgediz, S. (2012) The CGIAR at 40: institutional evolution of the world’s premier agricultural research network. CGIAR Fund Office, Washington D.C. Perry, B., Robinson, T. and Grace, D. (2018) Review: animal health and sustainable global livestock sys- tems. Animal 12, 1699–1708. Randolph, T.F., Schelling, E., Grace, D., Nicholson, C.F., Leroy, J.L., et al. (2007) The role of livestock in human health and nutrition for poverty reduction in developing countries. Journal of Animal Science 85, 2788–2800. Reardon, T., Echeverria, R., Berdegué, J., Minten, B., Liverpool-Tasie, S., et al. (2019) Rapid transformation of food systems in developing regions: highlighting the role of agricultural research and innovations. Agricultural Systems 172, 47–59. Rosegrant, M.W., Fernandez, M., Sinha, A., Alder, J., Ahammad, H., et al. (2009) Looking into the future for agriculture and AKST. I: McIntyre, B.D., Herren, H.R., Wakhungu, J. and Watson, R.T. (eds) International Assessment of Agricultural Knowledge, Science and Technology for Development (IAASTD): Agriculture at a Crossroads, Global Report. Island Press, Washington, DC, pp. 307–376. Shapiro, B.I., Gebru, G., Desta, S., Negassa, A., Negussie, K., et al. (2015) Ethiopia livestock master plan: roadmaps for growth and transformation. ILRI, Nairobi. Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M. and de Haan, C. (2006) Livestock’s long shadow: environmental issues and options. FAO, Rome. Sumberg, J. (2002) The logic of fodder legumes in Africa. Food Policy 27, 285–300. United Nations Economic Commission for Africa (2016) The Demographic Profile of African Countries. Economic Commission for Africa, Addis Ababa. Weindl, I., Bodirsky, B.L., Rolinski, S., Biewald, A., Lotze-Campen, H., et al. (2017) Livestock production and the water challenge of future food supply: implications of agricultural management and dietary choices. Global Environmental Change 47, 121–132. White, D.S., Peters, M. and Horne, P. (2013) Global impacts from improved tropical forages: a meta- analysis revealing overlooked benefits and costs, evolving values and new priorities. Tropical Grass- lands-Forrajes Tropicales 1, 12–24. World Bank (2012) Zoonotic disease prevention and control, one health and the role of the World Bank. World Bank. Washington, D.C. World Bank (2019) The World Bank and nutrition. Available at: www.worldbank.org/en/topic/nutrition/ overview (accessed 1 March 2020). APPENDIX 1: CIAT and ICARDA Livestock Research CIAT Livestock Research, 1967–20181 Pastures Programme in 1979, later becoming the Tropical Forages Programme (CIAT/TFP) A thorough history of the Centro Internacional and shedding along the way the work on animal de Agricultura Tropical (CIAT) stated that one health, breeding and reproduction. purpose of ‘…CIAT’s research over the past The CIAT Tropical Forages Programme had 50  years has been to establish fundamental three broad themes after 1980: (i) forage knowledge on tropical forages’ (Lynam and germplasm, in response to the identified need to Byerlee, 2017, p. 81). The search for this know- increase the available species and genotype pools of ledge began with CIAT’s Beef Program in 19692. forages, both grasses and legumes, for screening The programme’s objective was to ‘… increase for adaptation to limiting abiotic (mainly soil) and cattle productivity in the lowland tropics of biotic (pests and diseases) constraints; (ii) on-farm Latin America…at elevations below 1000  m’ forage development using materials that could (CIAT, 1973, p. 4). The farm type targeted by thrive on acid soils and resist insects and diseases, CIAT animal research was ranches of varying notably anthracnose on Stylosanthes spp.; and (iii) size, typically with sown pastures, and some natural resource management, involving plant, small- and medium-scale dairying in Central animal and soil components. America3. Research themes were ambitious, seeking to: (i) improve the quantity and quality of feed; (ii) control diseases and parasites; and Forage germplasm (iii) ‘devise production systems that produce good quality beef efficiently and cheaply’ (CIAT, Two phases can be distinguished in the develop- 1973, p. 5). Feed work involved collecting and ment of the CIAT forage gene bank: (i) a first evaluating legumes and grass species, initially in phase with the main focus on assembling and Latin America and later in sub-Saharan Africa using the forage germplasm collection (1972– and in Asia, with related soil analyses. Work on 1993); and (ii) a second phase that comprised production systems included field surveys, mod- continuing use of germplasm, diversity studies elling and economic analysis. Animal breeding and routine germplasm management and its and health were initially seen as major themes in optimization (1993–2017)4. the Beef Production Systems Programme but In the early 1970s, CIAT began systematic were later left to the national programmes and the missions throughout tropical America to collect private sector in Latin America (CGIAR/TAC, germplasm of wild species with forage potential. 1977, p. 47). The Beef Production Systems The objective was to create a diverse germplasm Programme ultimately evolved into a Tropical pool for cultivar development via selection or, if © International Livestock Research Institute 2020. The Impact of the International Livestock Research Institute (eds J. McIntire and D. Grace) 731 732 Appendix 1 natural variability failed to provide the desired Forage adoption traits, via breeding (Lynam and Byerlee, 2017, pp. 81–82). Collecting missions ranged from ex- The greatest achievement in tropical forages has cursions of short duration, particularly in Co- been the wide-scale adoption of such materials lombia, to several weeks-long expeditions. An- in Brazil (Schultze-Kraft et al., 2020), for which other germplasm source was opportunistic most of the credit goes to the national program collections made by CIAT staff and collaborators of Brazil, with some more recent input from during field visits. There was a combined target CIAT. It is estimated that about 120 million ha focus on acid-soil regions and plant genera of are planted to forages, of which nearly 100 mil- particular interest. The emphasis was on leg- lion are Urochloa spp. and approximately 17 mil- umes, in many cases including associated rhi- lion ha are Megathyrsus maximus (Jank et al., zobia, as the Neotropics are the main centre of 2014). Arguably about 50 million ha are plant- diversification of the Fabaceae (Leguminosae) ed to one cultivar, Urochloa brizantha cv. Maran- family. From 1979 onwards, collection mis- du, in Brazil alone (Lynam and Byerlee, 2017, p. sions expanded to South-east Asia, an import- 87). Particularly impressive is the rapid adoption ant centre of legume diversification (e.g. Puer- from an estimated 16 million ha in the mid- aria and Desmodium spp.). Collections extended 1980s to nearly 120 million ha by the early in the 1980s to Africa, with a focus on grasses, 2010s, indicating the transformative potential as sub-Saharan Africa is the main centre of Po- of improved forages when respective support aceae diversity. Maintenance of the forage col- structures, including involvement of the seed lection passed on to the CIAT Genetic Re- industry, are in place. Together with parallel sources Unit (GRU) after its foundation in improvements in animal breeds and animal 1977. The GRU manages seed testing, seed in- health, more productive forages contributed to a crease, germplasm preservation, safety back- fourfold increase in productivity per area and ups of the collection, maintenance of living per animal; this is recognized as one of the major collections, seed distribution, etc., all of which successes of global agriculture in the past are routine and costly germplasm conserva- 30 years (Lynam and Byerlee, 2017, p. 87). tion measures. In 2002, a Urochloa (syn. Brachiaria) de- Achievements in forage germplasm cumbens × brizantha × ruziziensis cultivar (paraphrased from Lynam and Byerlee, 2017, from CIAT’s breeding programme was released pp. 85–88) were: (i) the collections conserved as the first bred Urochloa sp. cultivar worldwide. in the gene bank with provision of germplasm Uptake based on documented seed sales until for selection and breeding programmes; and the end of 2016 is estimated at 750,000  ha (ii) forage adoption particularly in tropical mostly sown in the past decade. Adoption in- America. cludes more than 40 countries in Latin Amer- ica and the Caribbean, tropical Africa, tropical The forage gene bank Asia, Australia and Oceania, tropical/subtrop- ical North America and southern Europe; most With a total of approximately 23,000 acces- adopters appear to be small- and medium-sized sions (about 21,500 legumes and 1500 livestock producers6. grasses) from some 75 origin countries, the CIAT has released materials from its CIAT collection is the largest tropical forages germplasm base, including 11 grasses and germplasm collection worldwide. Its particu- 16 legumes in Mexico and Central America. The lar value lies in its focus on: (i) plants from, CGIAR Standing Panel on Impact Assessment and subsequently adapted to, acid, low-fertili- (SPIA) found limited published work that evalu- ty soils; and (ii) legumes. However, as far as ated the adoption and impact of these materials countries and regions on which germplasm in Mexico and Central America (Jutzi and Rich, collecting missions concentrated in the past 2016, pp. 46–55). are concerned, there are still important gaps: Better documented is the uptake of Urochloa the collection is probably far from being repre- grasses in Mexico and Central America where sentative of the geographic diversity of trop- plantings of over 3 million ha were reported up ical Poaceae and Leguminosae. to the early 2000s (Holmann et al., 2004). Appendix 1 733 Surveys in Colombia’s Eastern Plains carried out Conserving and using forage germplasm in 2017 suggested that about one-third (about 3 million ha) of improved pastures is sown using ICARDA plays a crucial role in conservation and Urochloa cultivars selected by Corporación use of global forage genetic resources. It holds Colombiana de Investigación Agropecuaria more than 155,000 accessions in trust, includ- (CORPOICA)7 and CIAT, or bred by CIAT (Labar- ing 38,955 accessions of temperate forage and ta et al., 2017). Empresa Brasileira de Pesquisa range species. ICARDA gene banks hold a highly Agrícola (EMBRAPA), with CIAT contribution, diversified collection of temperate/Mediterra- achieved another success with about 1.5 million nean forages including globally important and ha sown to Andropogon gayanus up to 2000 (Rivas unique collections of members of the genera Ríos, 2002). The national system of Nicaragua, Lathyrus, Medicago, Pisum, Trifolium and Vicia supported by CIAT, promoted the adoption of representing 16.6% of holdings reported in improved forages such as Urochloa spp. and Genesys. The bulk of the collection is still con- Canavalia brasiliensis, benefiting about 2000 served in the gene bank in Syria, although after smallholders and giving a 28% increase in daily 2014, gene bank core activities were relocated milk yield (Pinillos et al., 2018). to Lebanon and Morocco where new facilities Notable production gains were achieved in were established and efforts to regenerate and the Eastern plains of Colombia as part of the col- characterize the active and base collections were laboration between CORPOICA and CIAT. These undertaken. gains were due to the inclusion of Urochloa spp. The ICARDA forage collection is unique in in crop–pasture rotations leading to a twofold its geographical coverage (originating from 112 gain in carrying capacity over degraded pasture countries) and its species coverage (631 taxa in- and a tenfold gain over native savannah (Rincón cluding many neglected species). The base col- and Ligaretto, 2008). lection has a total of 30,008 accessions repre- senting 77% of the active collection. Only 62.5% of the collection is safety duplicated in four gene ICARDA Livestock Research, banks representing mostly the accessions col- 1977–20188 lected by ICARDA. ICARDA also conserves an important Rhizobium spp. collection totalling Research at the International Centre for Agri- 1483 strains belonging to 73 taxa, most of cultural Research in the Dry Areas (ICARDA), which are related to forage legumes. In 2008, established in 1977, has covered improvement ICARDA started sending the accessions in its of barley, chickpea, faba bean, grass pea, lentil active collection to the Svalbard Global Seed and wheat, as well as development of water and Vault in Norway, and approximately 23,360 land management and crop–range–livestock inte- accessions have been sent there for long-term gration. ICARDA livestock research has focused conservation. on three areas: (i) introduction of Medicago sati- The ICARDA forage gene bank has distrib- va into North African and West Asian farming uted over 30 years approximately 205,000 sam- systems; (ii) conserving and using forage ger- ples for use by ICARDA scientists and external mplasm; and (iii) understanding soil–water– partners. Most have been used to select ecotypes plant–animal livestock interactions in the mixed for pasture improvement or as sources of traits grazing systems of North Africa and West Asia. for plant-breeding programmes. Notes 1 Material on CIAT is derived from Lynam and Byerlee (2017) and from Schultze-Kraft et al. (2020, forth- coming). 2 Lynam and Byerlee (2017) present a history of CIAT from 1967 to 2017. 3 CIAT’s early swine programme closed in 1978 (Lynam and Byerlee, 2017, p. 29). 4 CIAT germplasm collections and its accessions database are described in Schultze-Kraft et al. (2020, forthcoming). 734 Appendix 1 5 See Chapters 12 and 13 (this volume) for forage gene bank results from ILRI, CIAT and ICARDA. 6 This is probably not equivalent to ‘smallholders’ as used in the African context. 7 CORPOICA is now known as Agrosavia. 8 The following section on ICARDA is derived from material found at www.ilri.org/dataportal/impact/forage. References CIAT (1973) External review team: beef production systems program. CIAT, Cali, Columbia. CGIAR/TAC (1977) Report of the TAC quinquennial review mission to the International Center for Tropical Agriculture (CIAT). CGIAR and TAC, Washington, DC. Holmann, F., Rivas, L., Argel, P.J. and Pérez, E. (2004) Impact of the adoption of Brachiaria grasses: Central America and Mexico. Livestock Research for Rural Development 16, 98. Jank, L., Barrios, S.C. and do Valle CB, Simeão, R.M. and Alves, G.F. (2014) The value of improved pastures to Brazilian beef production. Crop and Pasture Science 65, 1132–1137. Jutzi, S.C. and Rich, K.M. (2016) An evaluation of CGIAR Centers’ impact assessment work on livestock-re- lated research (1990–2014). Standing Panel on Impact Assessment (SPIA), CGIAR Independent Sci- ence and Partnership Council (ISPC), Rome. Labarta, R., Andrade, R., Marin Salazar, D., Rivera, T., Orrego, M. and Pinillos, J. (2017) The Impacts of CIAT’s Collaborative Research. CIAT, London. Available at: https://cgspace.cgiar.org/bitstream/han- dle/10568/89160/CIAT50_The_Impacts_of_CIAT%27s_Collaborative_Research.pdf?se- quence=1&isAllowed=y (). Lynam, J. and Byerlee, D. (2017) Forever Pioneers – CIAT: 50 Years Contributing to a Sustainable Food Future… and Counting. CIAT, Cali, Colombia. Pinillos, J., Labarta, R., Rivera, T., Vasquez, A., Martinez, A. and Mojica, E. (2018) Study of Pasture Adoption in the Colombian Caribbean. CIAT, Cali, Colombia. Rincón, Á. and Ligarreto, G. (2008) Productividad de la asociación de maíz-pastos en suelos ácidos del Piedemonte Llanero colombiano. CORPOICA, Villavicencio, Colombia. Rivas Ríos, L. (2002) Impacto económico de la adopción de pastos mejorados en América Latina tropical. CIAT, Cali, Colombia. Schultze-Kraft, R., Peters, M. and Wenzl, P. (2020) A historical appraisal of the tropical-forage collection conserved at CIAT. Manuscript submitted to Genetic Resources. Index Note: The locators in bold and italics represents the tables and figures respectively. AAT immunoprophylaxis 136, 188 African agricultural systems AAT vaccine antigens changes in species mix 527, 530, 531 Big Pharma 117 climate to 2000 536 conserved plasma membrane proteins 115 human populations 525–526, 526 cysteine protease (CP) 116 inequality, livestock holdings 536 drug-screening 117 land use, fertilizer use and cereal yields 531, ILRAD/ILRI 117, 118 532–534 macromolecular nutrient receptors 115–116 livestock production 527, 528 metacyclic and bloodstream-stage trypomastigote poverty 536–537 antigenic variation 113 African animal genetics 61 bloodstream-stage parasites 113 African animal trypanosomiasis (AAT) cross-reacting determinant (CRD) 114 anaemia 109–110 direct analysis of VSG 113 antigens 108–109 GPI-PLC 115 causative agents 107–108 immune sera and mAbs 113 community-based tsetse control 151 infection-and-treatment method 113 definition 104 interference phenomenon 114 economic losses 105–106 interferon-γ (IFN-γ) 114 Glossina (tsetse) 110 skin defence system 114 ILCA/ILRI research 149, 150 research themes 112 ILRAD, ILCA and ILRI research, 1979-2017 tomato lectin binding (TL) antigens 117, 117 (see AAT vaccine antigens) trypanosome endocytic vesicles 116 pathogenesis 109 acaricides 370 research collaborations 106 adaptation scientific contribution 104–105 CGIAR research 617–618 trypanocide resistance 151–152 costs 617 trypanotolerance 149–151 knowledge gaps 618–620 tsetse infested areas 108 mixed crop–livestock systems 619 African livestock genetics 61 pastoral systems 619–620 African livestock policy analysis network tropical livestock systems 630 (ALPAN) 25, 641 adoption African livestock system, mid-1970s Asia 462 affected animals 8 Latin America and Caribbean 462–463 agricultural GDP 6 sub-Saharan Africa 457–462 economic benefits 9–10 735 736 Index African livestock system, mid (continued ) Animal Health, Food Safety and Zoonosis 306 growth potential 3–6 animal health services/productivity 653–655 infested areas 7, 7–8 animal-source foods (ASFs) land use 3 nutritional benefits 701 productivity effects 8–9 overconsumption 701 rural population 3 antimicrobial drugs 319 sub-Saharan Africa 4 anti-parasitic drugs 319–320 tropical livestock units 5 CGIAR Antimicrobial Resistance Hub 320–321 trypanosomiasis 6–7 resistance/transmission 320 African rangelands research environments use of antibiotics 320 access to grazing resources 405 antimicrobial resistance 719 East Africa 401–402 ASFs see animal-source foods grazing regimes 404 ATLN see African Trypanotolerant Livestock network; people, livestock and rangelands interactions 405 African Trypanotolerant Livestock Network predecessors and partners 403–404 (ATLN) quantitative assessment and monitoring 404 avian influenza 229–230 West Africa 402–403 African range systems bibliometrics 387–388 arid and semi-arid pastoral systems 397 bioeconomic models 578, 642, 644–647, 664 climate 398 Bovine HapMap Consortium 32, 37, 61 human populations 397–398 bovine immune system livestock populations 398 cell cloning 177 mobility and grazing access 399 cluster-defined (CD) antigens 181 vegetation 398–399 FACS 178 African swine fever (ASF) 275 international peer-reviewed journals of field epidemiology and control immunological scientific societies 183 control 280 milestones 180–181 risk factors 279 T- and B-cell 178 socio-economic studies 280 T-cell antigen-specific receptor (TCR) 178, 181 transmission dynamics 279 T-cell immunology developments 179 molecular epidemiology 278–279 γδ T-cells 181, 182 research 277–278 tumour necrosis factor (TNF)-α 183 technologies bovine immunology diagnostics 278 BoLA typing 170–171 vaccines 278 FACS 170 African Trypanosomiases 106, 109 mAbs 170 African Trypanotolerant Livestock Network (ATLN) T-cell clones 171 26, 72, 118, 128, 132, 151, 154, 547, 641 bovine lymphocyte antigens (BoLA) 170, 171, 195 Ag-ELISA 120, 121 bovine tuberculosis (bovine TB) 303, 304, 306, chemoprophylaxis/chemotherapy, absence 122 321–322, 327 data analysis programmes 119 brucellosis 303, 304, 305, 306, 307, 310, 312, 321, field operations 119 322, 325–327, 561, 719 livestock production, improvement 119 microhaematocrit and parasites estimation 119, 119 natural tsetse/ trypanosome challenge 122 CAAS see Chinese Academy of Agricultural N’Dama cattle 118 Sciences (CAAS) orthogonal-field-alternation gel electrophoresis Centro Internacional de Agricultura Tropical (OFAGE) 120 (CIAT) 2, 451, 519, 603, 686, 731 phase-contrast diagnostic technique 119 Centro Internacional de Mejoramiento de Maíz y polymerase chain reaction (PCR) 119, 120 Trigo (CIMMYT) 11, 482, 483, 490, 567 treatments, PCV value 123 CGIAR trypanotolerant status, N’Dama 123 development impacts 425–426, 443 Agroecology 520, 526, 602, 627 forage gene banks 426, 433–443, 434, 446 altmetrics 61, 507–509, 508 germplasm (see forage germplasm) animal genetic resources (AnGR) 60 International forage collections 427–431 characterization 66 knowledge bank 445 animal genetics 54–57 scientific impacts 424–425, 443–444 Index 737 CGIAR Research Programme (CRP) 275 studies 567, 568–570 Chinese Academy of Agricultural Sciences sub-saharan Africa 572–573 (CAAS) 70, 71, 727 technical changes 567, 571 CIAT livestock research 731 crop residues (CRs) classical swine fever (CSF) 275, 290, 641 digestibility, grain and yield 496–499, 498, 499 climate change farm mechanization 581 adaptation 602, 617–620, 720 feed supply and demand scenarios 481, 483 capacity development and partnerships 604 fodder markets 484–487, 486, 486 components 630 fodder quality (see CR fodder quality) development impact 603–604 groundnut and faba bean cultivars 488, 488 ILRI research 602–603 lignocellulosic biomass 481, 483 knowledge gaps and research hypotheses 611, livestock feed and forages 716 614–615 livestock productivity 487–488 mitigation 719–720 milk potential 487, 488 policy modelling 720 multidimensional crop improvement 484, 499–500 research spending and bibliometrics 603 N content, grain and yield 496, 496, 497 scientific impact 603 postharvest treatments 481, 484 weather data 610–611 primary and secondary traits 495–499 See also ruminant livestock scientific impacts 481–482 commercial dairying 522 small- and medium-sized seed enterprises 500 Committee on World Food Security (CFS) 350 trait identification and tools 488–490 Commonwealth Scientific, Industrial and Research CRs see crop residues Organisation (CSIRO) 373 cultivar-dependent variation community-based sheep 86 food–feed crop cultivars 492 competitive global environment GWAS and GS 494–495 changed funding environment 726–727 hybrid breeding, dual-purpose maize 493, impact at scale 727–728 493–494 NARS 727 phenotype pipeline and releases 490–492 Comprehensive Africa Agriculture Development QTL identification and backcrossing 494 Programme (CAADP) 351 recurrent selection 492 Consultative Group on International Agricultural stover N vs. grain yield 490, 491 Research (CGIAR) 106–107 cysticercosis 322–324 contagious bovine pleuropneumonia (CBPP) 30, 196, 275, 561, 689, 714 contagious caprine pleuropneumonia (CCPP) DAGRIS see Domestic Animal Genetic Resources 30, 275, 714 Information System (DAGRIS) continual engagement 406 Dairy Genetics East Africa (DGEA) 62, 84, 85, 92 COVID-19 306, 313, 318, 328, 342, 718 dairy training 352 CR fodder quality deliver adapted chickens 87 crop-improvement approaches 482 densified total mixed ration (DTMR) 487, 488 existing cultivar-dependent variation 490–495 development impacts livestock productivity 487–488 capacity building 453 NIRS tools, calibration and validation 489–490 capacity building and partnerships 426 validation 488–489 climate change 603–604 crop-improvement 424, 427 CRs 482 multidimensional (see multidimensional crop defence of pastoralists’ interests 517–518 improvement) development gaps 426 sensory phenotyping 489 economics/policy research 641 timespan 490 FBD 339–340 whole-plant optimization 482 gender research 681–682 crop–livestock interactions germplasm distribution as proxy 426 elements 567 grazing areas 580 goals 567 information provision 482 mechanization 571–572 LSR 517 predictions 571 multidimensional crop improvement 483 semi-arid tropics, India 573 option value preservation 425–426 soil fertility management 573–574 partnerships 453 738 Index development impacts (continued ) DNA-based strain identification 249–250 pastoral areas 578–579 irradiation of sporozoites 253–254 planted forages 452 molecular tools 247–248, 253 potential option values 443 monoclonal antibodies (mAbs) 248–249 rangeland ecology 397 Muguga cocktail 246–247 TAD 276 performance monitoring 253 technical change 518 vaccine production 253 uptake summary 425 life cycle of T. parva 243 valuing land rights 518 model gaps 267 development programme implementers 351 production and net imports 264 DGEA see Dairy Genetics East Africa (DGEA) subunit vaccine 250 Domestic Animal Genetic Resources Information exposed animals 250 System (DAGRIS) 37, 62, 78–79 T. parva parasite 243–244 Dorper sheep 63, 75–76, 77, 85, 87 T. parva-infected lymphocytes 254–255 Dryland Agricultural Research Institute (DARI) 440 vaccine adoption 261, 263 DTMR see densified total mixed ration Ebola 318–319 risk assessment 319 economics/policy research East Africa animal health services and productivity 653–655 land-cover/land-use changes 405, 406 bioeconomic models 642, 644–647 Mara ecosystem 406 class and problem 642, 643 modelling 407–409 development impacts 641 Napier grass 459, 460, 461, 461 goals 640, 670 research environments 401–402 livestock revolution (see livestock revolution) wildlife and livestock conflict 406–407 policy problems 642, 648–655 East African Veterinary Research Organization research spending 640 (EAVRO) 14, 240 scientific impacts 640–641 East Coast fever (ECF) 14, 49, 60, 107, 111, 149, empowerment 692 152, 165, 209, 215, 239, 242–243, 292, equilibrium vs. non-equilibrium models 400–401, 401 306, 367, 610, 689 Erysipelothrix rhusiopathiae 312 agricultural production 265 exoantigen anti-schizont vaccine bloodstream-stage trypanosomes 110 antigenic diversity 251–252 definition 108 CTL 250–251 external programme review (EPMR) 18 lack of protection with serum 250 anti-sporozoite vaccine 252 benefit:cost ratios 266 farming system 507, 508 control methods 244 farming systems economic impacts animal health and food safety 688–689 grazing type 259–262 environment 690 ITM price 259 fodder and forages 687–688 methods 257, 259 genetics 689–690 modelling 256 institutions 687 parameters 258 markets and value chains 691–692 global context 240–241 nutrition and food security 690–691 ILRI research farm mechanization capacity building 242 assumption 580 development impacts 241–242 crop residues 581 economic impacts 242 inadequate crop residue research 581 partnerships 242 nutrient cycling 582 policy impacts 242 sustained growth 580–581 scientific impacts 241 fat-tailed sheep 69 impacts 244–245 feed assessment tool (FEAST) 452, 463, 464, 686, investment 262–263 716, 717 infection-and-treatment method (ITM) 245–246 feed systems buffalo problem 252–253 animal fattening on farm 575 cold chain 253 animal health 561 development 246 bush encroachment 558 Index 739 crop residue management 574 number of associates 441, 442 experiments 556–557 scientific and field benefits 438 field characterization 556 usage, forage germplasm 438–440 fodder trees 574–575 forage germplasm impact 558, 559–560, 561 achievements 732 pasture improvement 557–558 adoption 732–733 sown forages and ley farming 574 CIAT gene bank 428–429, 429 fluorescence-activated cell sorting (FACS) 170 conservation and usage 733 food-borne disease (FBD) 339, 340–341, 718 distribution 433–435, 434, 435 approaches 346, 347 gaps in collections 430–431 capacity building 340, 347–348, 348 gene bank 732 case studies 350 ICARDA gene bank 429, 429–430 dairy training 352 ILRI gene bank 428, 428 development impacts 339–340 phases 731 development programme implementers 351 traits 431–433 economic impact 223 value 433 empirical evidence 344 forage production 404, 434, 438, 446, 456, 457, enablers 351 462, 464, 608, 617, 717 global context 339 history 341–343 ILRI 343 GCM see general circulation model ILRI initiative 349 gender equity 725–726 ILRI interventions 355–356 gender research ILRI research 349 capacity development and partnerships 692 impact 351 CGIAR 698 informal sector agents 351–354 CRP (2010 onward) 684–685 international/regional/national data and statistics improvement 693 policies 348, 350, 351 development impact 681–682 livestock keeping/human nutrition 358 discrimination 681 low- and middle-income countries 224–225 empowerment 692 multi-country evaluation 348 farming systems and technologies 687–692 pathways to impact 343–344 institutionalization (2005-2010) 683–684 Peru/Colombia/India 224 landscape level 686–687 policy impacts 340 merger and clash of cultures (1990s- product line 347 early 2000s) 682–683 research evidence 344–345 policy impacts 693 research methods/tools 345–346 research spending and altmetrics 681 researchers 351 scientific impact 681 scientific impacts 339 scientific perspectives, methods and level of southern Africa 224 analysis 685–686 surveillance 346–347, 347 1970s–1980s, evolution 682 technologies 346, 347 general circulation model (GCM) 469, 470 value chain agents 351 genetic diversity 431–432, 444–445 Foodborne Disease Burden Epidemiology Reference genetics of disease resistance Group (FERG) 350 African cattle 74 food–feed crop cultivars 492 biorepositories/biobank facilities 79 food production index 531, 535 breeding programme design 84 food safety 54–57, 338–339 breeding programmes in Africa 83–84 foot-and-mouth disease (FMD) 210, 223–225, 275, conservation decisions, tools 80–81 289–290, 714, 723 economic valuation 79–80 forage diversity 38, 387, 390–391, 423–445, 451 gastrointestinal parasites 74–75 forage gene banks 23, 27, 33, 38, 390, 432 Menz vs. Horro sheep, Ethiopia 76 capacity building 441–443 national experiments 76–77 costing 435–436, 437, 437, 437–438 Red Maasai sheep 75–76 estimating economic benefits 435 resistance and productivity 77 germplasm distribution 433–435, 434, 435 Small East African vs. Galla goats 76 management 436–438 small ruminants 77–78 Napier grass germplasm 440–441 genetic improvement of livestock 81–82 740 Index genetics of disease resistance (continued ) ICARDA livestock research 733 germplasm for smallholder dairy farmers ILCA’s programme adapted and productive chickens 85–87 Borana, Ethiopia 548–549 on-farm selective breeding 84–85 Fakara, Niger 551 optimum breed combinations 84 generations 547 potential for within-breed selection 85 grazing systems 546–547 sustainable germplasm supply 84 Kaduna, Nigeria 550–551 long-term breeding programmes 82–83 livestock systems 547 milk production 83 Maasailand, Kenya 549–550 NARS capacity 88 milestones 547 development 89 Niono, Mali and Malian Delta 550 future 91–92 ILRI climate research 616 group training 89 immunology 54 ILRI-SLU project 89–91, 90, 90 immunology and immunoparasitology research reproductive technologies 88 AAT 165 T. congolense infection 74 assessment questions 197, 198 trypanotolerance 72–73 bovine immunology technology, ILRAD genome-wide association studies (GWAS) 482, 494–495 (see bovine immunology) genomic selection (GS) 482, 494–495 fundamental immunological questions 169 global warming 604 goals and achievements, ILRI 165, 166 Global Foot-and-Mouth Disease Research Alliance ILRAD 167 (GFRA) 225 ILRAD/ILRI benefiting diseases 195, 196 Glossina palpalis 151 ILRAD/ILRI, diseases benefiting research 196 grade livestock 63 ILRI goals and achievements 165–166 G-range 408, 414 immune systems, ILRAD’s founding 169, 170 Grass pea (Lathyrus sativus) 435, 440 quality of research 197 grazing 507, 508, 523 scientific impact measurements grazing organization 553–554 citation data 171–172, 172 greenhouse gas emissions cost comparisons 175–177 anthropogenic emissions and livestock’s disease vaccine comparison 177 share 620, 621 highly cited manuscripts 172, 173–175, 176 AR5 review 620–621 h-index of scientists 172, 175, 176 demand-side options 628, 629, 630 manuscripts quantity 171 estimation 620 scientific impacts 166 global livestock 620, 621 change in behaviour concept 168 mitigation research 623–627, 625–627 outputs and outcomes 167 supply-side options 621–623, 624 practical outcomes 168 Green Revolution 11, 537, 661 scientific truths 168 GS see genomic selection traditional bibliographic measurements 168 GWAS see genome-wide association studies survey responses graph 194 T. parva 166 herd health 713 work environment questions 194–195 High-Level Panel of Experts (HLPE) 350 immunoparasitology highly pathogenic avian influenza (HPAI) theileriosis 210, 229–230, 287, 303, 718 anti-pathology vaccine 187 evaluation of control, Indonesia/Nigeria 317 CD8+ T-cells 185 response models, Egypt/Indonesia 316–317 humoral immunity, findings 187 supporting surveillance, Africa 316 live vaccine approach 184 high-potential Napier adoption sites 469, 471 Muguga cocktail 183 human African trypanosomiasis (HAT) 16, 106, R. appendiculatus 184 120, 149, 154, 192, 220, 303, 324–325 T. parva 183, 184 human nutrition research 354, 357, 359, 360 T. parva-infected lymphocytes immunity 185, 186 T. parva sporozoites 187 IBLI project trypanosomiasis household surveys 668 AAT immunology, ruminants 190, 190–191 impact 669 African trypanosomes 187 policy 669 Cape buffalo 191, 192 Index 741 CD4+ T-cells 194 international livestock research centres natural killer (NK) cells 193 Beck report 12 N’Dama cattle 192 Green Revolution 11 non-variant immunoprotection 189 Rockefeller Foundation 11 quantitative trait loci (QTLs) 193 Tribe report 12–13 T-cell-specific antibodies 192 International Livestock Research Institute (ILRI) Trypanosoma brucei 188 2, 104, 105, 106, 149 trypanotolerance 189 achievements 32–35 trypanotolerant Boran cattle 194 achievements (1975–2018) 36 2B4 molecule 193 animal genetics and health 37–38 VSG-coated trypanosomes 188 citation scores 40 xanthine oxidase 191 development impacts indigenous livestock 63 capacity 62 infection-and- treatment method (ITM) 16, 113, partnerships 62–63 213, 240, 245–246, 376 economics and policy 39 infectious bursal disease 290 evolution 65 innovation systems theory 467 genetic improvement 63–66 international agricultural research centres global environmental 38–39 (IARC) 11 ILRAD/ILCA merger 27–29 International Center for Agriculture Research in the CGIAR allocations 29 Dry Areas (ICARDA) 2 funds and period 28 International Centre for Rangelands 12 new priorities 29 International Laboratory for Research on Animal impact analysis 39–40 Diseases (ILRAD) 2, 104, 105, 149 livestock systems 38 achievements 19–20 phenotypic characterization 66 capacity development 20 African chickens 70 challenge 60 African dromedary 70 ECF control 16–17 African livestock populations establishment of 13 67–68 evidence-based research 111 African sheep 69 institutional evolution Beijing/Asia 70–71 EPMR 18–19 breed surveys, tools/methods QQR 18 66–67 1975–1979 international interest 71–72 vaccine development and control, linking livestock, human history 69 experimental questions molecular genetic characterization 111–112, 112 68–69 priorities 14–15 primary production 38 research domain 17, 19 priorities International Livestock Centre for Africa (ILCA) 2 scientific 29–30 achievements species mandate 29 animal genetics and health 24–25 rangeland ecology 396 livestock systems 25 reform of CGIAR 31 primary production 25 reorientation after 2000 30–31 scientific accomplishments 26 resources 30 scientific impact 25 scientific impacts systems characterization 26 breeding technologies 62 systems evolution 25–26 genetic characterization 61–62 trypanotolerance 26–27 history 61–62 establishment of 13 livestock genetics and breeding 61 feed quality 27 search expressions forage gene bank 27 cost–benefit analyses 40–41 institutional evolution 22–24 literature reviews 40 nutrition research 27 meta-analyses 40 priorities 21, 23 project evaluations 41 research domain 22, 24 theileriosis 36–37 resources 21–22 trypanosomiasis 36 trypanosomiasis 15–16 trypanotolerance 36 742 Index International Range Land Institute 12 land 606, 606–607 International Treaty on Plant Genetic Resources for resource usage 605–607 Food and Agriculture (ITPGRFA) 427–428 water quality 607 International Trypanotolerance Centre (ITC) water quantity 607 50, 63, 106, 150 livestock productivity 487–488 livestock revolution 655–656 dairying in South Asia vs. East Africa 660–661 Kenya Agricultural Research Institute (KARI) food security and nutrition 667–668 20, 170, 212, 352, 371, 658, 687 IBLI project, Kenya and Ethiopia 668–669 Kenya Livestock Insurance Project (KLIP) 669 land rights 661–666 Kenya Veterinary Vaccines Production Institute policy interventions 656 (KEVEVAPI) 282–283, 286 poverty 666–667 knowledge products sector analyses and master plans 669–670 FEAST 452, 463, 464, 686, 716, 717 smallholder dairy development 656–660 tropical forages tool 463–464 smallholder interests 655 Tropical Grasslands-Forrajes Tropicales 464 livestock systems research (LSR) African agricultural systems 525–537 characteristics 524 land rights climate and growing period 522 land tenure, resource allocation development impact 517–518, 578–580 and productivity 663–664 ILCA’s programme 546–551 pastoral systems 662–663 impacts 537 resource management 664–666 mixed-systems 564–575, 580 tenure and fodder trees 666 modes of production 522–525 themes 661 objectives 519–520 land use change, impacts and dynamics pastoralism 537, 542–543, 546 (LUCID) 405, 406 pastoral systems 551–564 LGA see livestock/grazing/arid potential impacts 521–522 LGH see livestock/grazing/humid research spending 516–517 LGT see livestock/grazing/tropical highlands scientific impact 517, 575–578 lignocellulosic biomass 481, 483 scientific into development impact, translating live-bait technique 151 problem 518–519, 582–584 livestock stages 520–521 animal welfare 714 water 554–564 climate change 702, 719–720 Lucerne (Medicago sativa) 439, 455, 459, 461, 558, 733 demand, products 702, 703 LUCID see land use change, impacts and dynamics diagnostics and vaccines 713–714 lumpy skin disease (LSD) 290 farming 507, 508, 509–510 feed and forages 715–717 food and nutrition security 700–701, 710, 712 major histocompatibility complex (MHC) 36, 61, 112, gender equity 725–726 250, 278, 714 genetics 714–715 manufacturers’ unit value (MUV) 49, 387, 507 global development goals 710 MarkSim 510, 603, 610, 611, 612–613, 623, 630 growth with equity 721–726 Medicago littoralis 439 health 712–713 Medicago polymorpha 435, 439 herd health 713 Medicago truncatula 439 human health 717–719 MIA see mixed/irrigated/arid and semi-arid natural resources 701–702, 720–721 Middle Eastern respiratory syndrome-coronavirus prosperity 701 (MERS-CoV) 303, 306 resource usage 707–710 national response 318 supply, products 702–707 reservoirs and transmission 318 transboundary animal diseases 713 MIH see mixed/irrigated/humid and subhumid livestock/grazing/arid (LGA) 523–524 MIT see mixed/irrigated/temperate and tropical livestock/grazing/humid (LGH) 523 mixed farming by smallholders 524–525 livestock/grazing/tropical highlands (LGT) 523 mixed/irrigated/arid and semi-arid (MIA) 525, 526, livestock production systems 536, 537, 605 air 607 mixed/irrigated/humid and subhumid (MIH) 525, 526, ecosystem services 607 537, 605 Index 743 mixed/irrigated/temperate and tropical (MIT) 525, 526, neglected zoonotic diseases 305, 306, 310, 313, 537, 565, 605 321–322, 327–328, 719 mixed/rain-fed/arid and semi-arid (MRA) 525, 526, Newcastle disease (ND) 196, 275, 287, 713 536, 537, 565, 605, 627 field epidemiology and control 287–288 mixed/rain-fed/humid and subhumid (MRH) 525, 526, research 287 537, 547, 565, 605, 627 socio-economic studies mixed/rain-fed/temperate and tropical (MRT) 525, 526, vaccine adoption 288 537, 547, 550, 565, 605, 627 technologies 287 mixed-systems 507, 508 nomadic/semi-nomadic grazing 522–524 classification 525, 564–565 normalized difference vegetation index (NDVI) 410, 411 crop–livestock interactions 567, 568–570, nutritional traits 432 571–574, 580 densely cultivated areas 585 feed systems 574–575 Open Nucleus Breeding System 82 fertilizer use 567 Overseas Development Institute (ODI) 25 improved plant cultivars 566–567 levels and rates, farm sizes 566 pastoralism 565 Pastoral Household and Economic Welfare Simulator smallholder 565 (PHEWS) 407, 408 Monoclonal antibodies (mAbs) 36, 110, 165, 170, pastoralism 507, 508 195, 248–249 animal mobility 546 MRA see mixed/rain-fed/arid and semi-arid animal productivity 542 MRH see mixed/rain-fed/humid and subhumid calf nutrition 563 MRT see mixed/rain-fed/temperate and tropical herd demography and stocking rate 562 multidimensional crop improvement livestock pessimism 542 capacity building and partnerships 483 new grazing model 563–564 impacts 499–500 primary productivity 562 multidimensional crops 38, 387, 391, ranching vs. 543, 544–545 392–393, 451, 484, 521, 558, 574, scientific impact 575–576 582, 717 semi-nomadic systems 543 multi-purpose trees 452, 458–459 subsistence and commercial grazing mycoplasma disease 280–281 systems 543, 546 field epidemiology and control vector control 563 epidemiological surveys 284 pastoral systems modelling control strategies feed systems 556–561 283–284 governance research 585 vaccine trial 284 grazing organization 553–554 host-mycoplasma interactions 283 land rights 662–663 molecular epidemiology 283 modelling pastoralism 561–564 research 281 sedentarization 551–553 socio-economics studies 537, 538–541 demand for vaccines 284–285 water 554–556 policy analysis 285 peste des petits ruminants (PPR) 275, 276, 285, 294 vaccine delivery systems 285 field epidemiology and control technologies global eradication 287 diagnostics 281 vaccination strategies 286–287 therapeutics 283 research 285 vaccines 281–282 technologies 286 PHEWS see Pastoral Household and Economic Welfare Simulator Napier grass 459, 460, 461, 461, 470 pigs 650–653 national agricultural research systems (NARS) planted forages 390, 390–391 11, 29, 60, 452, 727 adoption 457–463, 472 natural resources 217, 407, 426, 435, 537, 613, environmental benefits 466–467 701–702, 720–721 global impacts 468–469 N’Dama cattle 63, 72, 105, 107, 118, 122–127, impacts 452–453 133, 192 knowledge products 463–464 NDVI see normalized difference vegetation index livestock 716 744 Index planted forages (continued ) Red Maasai 63, 75–76, 77, 78, 85, 87 livestock production 453–454 research spending 49 new adoption potential 469–472 resource usage potential 451 environmental costs 710 research spending, ILRI 451 labour 707–708 seed technical support and distribution land 708–709 464–466, 465 RHoMIS see Rural Household Multi-Indicator Survey tropical Africa and Asia 467–468 Rift Valley fever (RVF) 37, 54, 210, 225–227, 303, tropical farming systems 454–463 313–316, 609, 654, 714, 718 point-centred quadrant 410 economic impact 226 policy problems infection and disease dynamics 226–227 markets, institutions and competitiveness 648–653 risk maps for eastern Africa 226 pigs and poultry 650–653 supporting response, East Africa 315–316 research 642 transmission and burden 315 ruminants 648–650 vaccines 315 supply response 642, 648 Rinderpest 288–289 TAD 276 Rockefeller Foundation 11, 12, 13, 14, 212, 309 porcine reproductive and respiratory syndrome ruminant livestock (PRRS) 290, 291 AR5 608–609, 609 poultry 650–653 CGIAR research 610–611 poverty 666–667 food security and production systems 608, 608 primary production greenhouse gas emissions 620–630 bibliometrics 387–388 impacts 611, 613 forage diversity 390–391 knowledge gaps 613, 617 high citation papers 393 mitigation and supply- and demand-side efforts 604 multidimensional crops 391, 392–393 productivity 604 planted forages 390, 390–391 See also climate change rangeland ecology 388–390, 389 ruminants 648–650 research spending 387, 388, 392 Rural Household Multi-Indicator Survey ‘Primary Production in the Sahel’ 403 (RHoMIS) 693, 694 public policy 642 push–pull technology 461–462 Sahelian Transpiration, Evaporation and Productivity (STEP) 409 quantitative trait loci (QTLs) 32, 61, 193, 482, 494 SAVANNA model 407, 408 quinquennial review (QQR) 18, 22, 23, 27 scientific impacts African livestock systems 577 basic and applied research 444 ranching 522 bioeconomic models 578 rangeland ecology 388–390, 389 characterization 424–425 adapting to climate change 414 CRs 481–482 African rangelands research data sets 603 environments 401–405 distribution 425 African range systems 397–399 economics/policy research 640–641 East Africa 405–409 environmental benefits 443 global context, ILRI role 396 FBD 339 grazing regimes, animal and rangeland gender research 681 performance 411–412 germplasm collection 424 impacts, ILRI’s research 396–397 LSR 517 mitigating climate change 413–414 mapping systems 577 mobility and access to grazing 412–413 models 603 monitoring rangeland conditions 413 network research model 578 productivity 414 pastoralism 575–576 range governance 413 planted forages 452 sub-Saharan Africa 396 productivity parameters 577–578 tropical range 399–401 rangeland ecology 396–397 West Africa 409–411 results 603 Index 745 scientific gaps 425 scientific findings 291–293 TAD 276 space-time cluster analysis 291 Scopus database 40, 61, 507 study area 292 SDGs see Sustainable Development Goals transboundary diseases 30, 37, 49, 54–57, 56, 279, sedentarization 551–553, 306 288–290, 713 severe acute respiratory syndrome (SARS) 303, 306 triple-path model 354 single-nucleotide polymorphisms (SNP) 61, 37, 62, tropical farming systems 65, 74, 84, 92, 136 benefits, forages 454 smallholder dairy development 353 crop–livestock systems 454 Ethiopia 656–658 extent 456–457 Kenya 658–660 forage legumes 454, 456 SNP see single-nucleotide polymorphisms forage species 454, 455, 456 space-time cluster analysis 291 tropical forages tool 463–464 STEP see Sahelian Transpiration, Evaporation and Tropical Grasslands-Forrajes Tropicales 464 Productivity tropical livestock systems Streptococcus suis 312 altmetrics 507–509, 508 Stylosanthes 26, 424, 432, 435, 452, 455, 456, 457, climate change 510, 512 462, 463, 468, 521, 551, 574, 645, 731 economics and policy 510–511, 511 Sub-Regional Representation for South-East Asia high citation papers 511, 513 (SRR-SEA) 232 research and publications 509, 509–510 supply, livestock products research spending 507, 508 estimation 702–704, 704–708 tropical livestock unit (TLU) 3, 5, 527, 529 imports 704–705 tropical range ecology industrial livestock production 705–706 climate, short- and long-term effects 400 small- and medium-sized producers 706–707 equilibrium and non-equilibrium Sustainable Development Goals (SDGs) 359, 444, models 400–401, 401 700, 710, 711, 722 rangeland activities, effects of 400 technical and economic feasibility 399–400 trypanosome antigen enzyme-linked immunosorbent taurine cattle 68, 69, 74, 107, 118, 214 assay (Ag-ELISA) 120, 121, 323 theileriosis 50, 52–54 trypanosomiasis 6–7, 49–52, 50–52 thin-tailed sheep 69 control in Uganda 220 ticks economic burden 10, 218–219 control epidemiology 219 acaricide treatment 370–371 evaluating control options 219–220 cattle 369–370 ILRAD’s founding 15–16 Lowveld habitats 372 resistance in cotton zone 220–221 Pesticides 371 tsetse species 220 development impacts 368 vaccine 10 developments 380–381 trypanotolerance, cattle 118 future 381–382 AAT-induced anaemia global context 367 anti-CD4 treatment 127 pathogens 366–367 erythrophagocytosis 125 research capacity, application haematopoietic stem cells 126 genetic complexity, R. appendiculatus 377 haematopoietic-system intrinsic vectors 375–377 mechanism 127 research capacity, development N’Damas and Borans 126 artificial feeding systems 373–374 pro-inflammatory cytokine TNF-α 127 genomics and molecular biology 374–375 reticulocyte response 127 ILRI Tick Unit 373 T-cell marker, CD5 126 scientific impacts 367–368 T. congolense 126 vaccines 378–380 AAT-induced lymphopenia 128 TLU see tropical livestock unit AAT-induced T-cell transboundary animal diseases (TADs) 275, 287, B-cells 130, 131 290, 294, 713 congopain 132 capacity building 276–277 natural tsetse/trypanosome global context 275 challenge 131 746 Index trypanotolerance, cattle (continued ) value chain agents 340, 342, 346, 347–348, 715 351 TD antigens 130, 131 veterinary epidemiology (1987-2018) TI-2 antigens 130, 131 animal health research for poverty antibody responses, N’Dama and Boran reduction 227–228 antigenic epitope 129 avian influenza 229–230 IgG1 antibodies 130 ECF risks by region 214 IgM:IgG ratios 129 economic impact 228–229 ILNat 3.1 VSG 129 economics 212 VSG analysis 128 Field studies, Kenya 212–214 N’Damas and Borans FMD (see food-borne disease (FBD)) heterologous infections 124 gastrointestinal parasites 227 homologous reinfection 124, 125 GFRA 225 parasitaemic levels and kinetics 124 heartwater studies, Zimbabwe 215–216 T. congolense infections 125 ILRAD and ILRI, epidemiology 230 T. congolense serodeme 124, 125 ILRAD and ILRI, impacts trypanotolerance, genetic basis capacity development 231 Bovine Trypanotolerance Mapping developing countries 231 Programme 132 national animal health 231 ILRAD/ILRI Trypanotolerance Gene Mapping strategy 231 Programme 132 ILRI research linkage mapping 133 animal health 211 QTLs and bovine 134–135 capacity development 211 QTLs and murine 132–134 developmental impacts 210 trypanotolerant cattle 8, 61, 67, 72, 105, 107, 123, economic impact assessment 210 149, 151, 154, 155–156, 159, 160 human resource capacity 211 Tsetse and Trypanosomiasis Control, ILRI impacts partnerships 211 anthropogenic 160 scientific impacts 210 contingent valuation surveys 159 strategy 211 economic analysis 158–159 Inter-Agency Donor Group (IADG) 227, 228 epidemiological modelling 160 ISVEE experience 230 Ethiopia rabies research 221–222 dichlorodiphenyltrichloroethane rinderpest 222 (DDT) 153 RVF (see Rift Valley fever (RVF)) land-usage 154 tick-borne disease 214–215 mobile targets 153 tick-borne diseases sticky traps 153 distribution modelling 217–218 Kenya 152 economic impact 216–217 trypanocide prophylaxis 159 trypanosomiasis (see trypanosomiasis) Uganda 154 vector-borne diseases 218 West Africa Wellcome Trust 228 ATLN 154 best-bet strategies 155 community-based trypanosomiasis water development control 155, 156 adverse effects 555 dummy customer’ or ‘secret shopper’ experiments 554–555 methods 158 footprint and productivity 555–556 epidemiological model 155 literature reviews 554 harm reduction 156 system studies and field investigations 555 isometamidium 157 WELI see Women’s Empowerment in Livestock Index rational drug use 157 West Africa trypanosomes resistance 155 assessment and monitoring, rangelands 409–411 Tumaini 61, 88 fodder bank 457–458 research environments 402–403 White clover (Trifolium repens) 439–440 United Nations Development Programme wildlife–livestock interactions 397 (UNDP) 14 Women’s Empowerment in Livestock Index Urochloa 432, 446, 472, 732, 733 (WELI) 681, 685, 686, 725 Index 747 Zebu 8, 16, 25, 61, 67, 68, 69, 72, 73, 74, 83, 109, international initiatives 312–313 112, 121, 132, 152, 153, 155, 160, 165, 213, intervention to control 311 214, 240, 242, 244, 293, 321, 369, 554 neglected 719 zoonoses 54–57, 300 participatory disease surveillance 308–309, 309 capacity development 305 poor livestock keepers 307 drivers of disease emergence 308 prioritization 306–307 emerging infectious diseases 313, 718–719 smallholder pig systems 311–312 global context 303–304 systems 309–311 ILRAD 306 zoonosis 317, 327, 719 ILRI research 304–305 Zoonosis Technical Working Group 307