The Internationa l Potato Center (CIP) is a not-for-profit, autonomous scientific institution established in 1971 by agreement with the Gov- ernment of Peru. CIP is one of 16 international resea rch and training centers supported by the Consultative Group on International Agri- cultural Research (CGIAR). The CGIAR is sponsored by the Food and Agriculture Organization (FAD) of the United Nations, the United Nations Development Programme (UNDP), the United Nations Envi- ronment Programme (UNEP), and the World Bank, and comprises more than 45 countries, international organizations, and private foun- dations. The information and conclusions reported in this publication reflect the views of Cl P. Cover art based on a 9xl 2-inch mixed pastel titl ed Partnership by Josephine Dell ' Oreo Prain PROGRAM REPORT 1995- INTERNATIONAL POTATO CENTER International Potato Center Apartado 1558 Lima 12, Peru cip@cgnet.com http://www.cipotato.org International Potato Center Program Report 1995-1996 ISSN 0256-6311 September, 1997, 2 M Copyright © International Potato Center 1997 The International Potato Center. 199 7. Program Report 1995-1996. Limo, Peru . 323 p. PR OGRA M REP ORT Providing Better Access to CIP Research Hubert Zandstra Alleviating Poverty Through Research Partnership Regions Overview of CIP Work in Latin America and the Caribbean Fern ando Ezeta Overview of CIP Work in the Middle East and North Africa Aziz Lagnaoui Overview of CIP Work in Sub-Saharan Africa Peter T. Ewell Overview of CIP Work in South and West Asia Sarath lfangantileke Overview of CIP Work in East and Southeast Asia and the Pac ifi c Peter Schmiediche Program 1: Production Systems Thomas S. Walker Prospects for Sustaining Potato and Sweetpotato Cropping Systems in Southwest Uganda }an W Low Page i x x i 5 8 12 15 19 22 iii Risk Analysis of Potato Production in the Altiplano: Quantifying Farmers' Beliefs 29 R. Valdivia Alatrista, R. Quiroz, R. Va ldivia Fernandez, and V. Choquehuanca Analyzing Potato Productiv ity in Farmers ' Fi elds in Bolivia 34 F. Terrazas, V. Suarez, G. Wa tson, G. Thiele, T. Walker, and A. Devaux Preferences of Urban Consumers for Andean Roots and Tubers in Ecuador 40 P Espinoza A. and C.C. Crissman Selection of New Sweetpotato Vari eties for High Dry Matter Content in Indonesia 45 JI-Gin Mok, Lisna Ningsih, and Tjintokohadi Co llaborati ve Sweetpotato Breed ing in Eastern, Central, and Southern Africa 49 E.E. Carey 5. T. Gichuki, PJ.Ndolo, G. Turyam ureeba, R. Kapinga, N.B. Lutaladio, and J.M . Teri Tradeoffs in Agriculture, the Environment, and Farmer Hea lth 58 C.C. Crissman, J.M. Antle, and S.M. Capalbo Program 2: Germplasm Management and Enhancement 67 Ali Golmirza ie Advances in Potato Cryopreservation by Vitrification 71 A.M. Golmirzaie and A. Panta Genetic Diversity Analys is in a Cultivated Andean Potato, 5. phureja Juz. et Buk. 77 M. Ghislain, D. Zhang, D. Fajardo, Z. Huam an, and R. Hijmans DNA Markers for the lntrogress ion of Late Blight Resistance in Potato 84 M. Ghislain, B. Trognitz, C. Herrera, A. Hurtado, and L. Portal Identify ing Duplicates in Sweetpotato Germplasm Using RAPD 90 D.P Zhang, M. Ghislain, Z. Huam an, F. Rodriguez , and J.C. Cervantes RAPD Variation in Sweetpotato Cultivars from South America and Papua New Guinea D.P Zhang M. Ghislain, Z. Huaman, A. Golmirzaie, and R. Hijmans Sweetpotato Breeding Strategy and Germplasm Testing in Southeast Asia II-Gin Mok, Tjintokohadi, Lisna Ningsih , and Tran Due Hoang Advances in the Morphological Characterization of Oca, Ulluco, M ashua, and Arracacha Collections C. Arbizu, R. Blas, M. Holle, F. Vivanco, and M. Ghislain I V 97 104 110 Program 3: Disease Management 11 9 Luis F. Sa laza r Developing Hori zontal Resistance to Late Blight in Potato 122 J.A. Landeo, M. Gastelo, G. Forbes, J.L Zapata, and F.J. Flores Res istance to Late Blight from Diverse Wild Sources 127 B. Trognitz, M. Eslava, L. Portal, and P. Ramon Evaluation of Late Blight Res istance in Populations of Dip lo id Potato Hybrids fo r Genetic M apping 132 B. Trognitz, M. Ghislain, G. Forbes, P. Oyarzun, M. Eslava, R. Herrera, L. Portal, P. Ramon, G. Chacon Host Spec ifi c ity of Late Blight Pathogen on Potato and Tomato in Ecuador 138 G.A. Forbes, P.J. Oyarz un, A. Pozo, and ME. Ordonez Use of Natural Resistance Genes for Transgenic Resi stance to Potato Viruses 144 M. Querci, G. Brigneti, }. Garc fa-Mas, and D.C. Baulcombe Encapsidation of PSTVd in PLRV Particl es and Its Transmi ss ion by Aphids 150 M. Querci, R.A. Owens, and L.F. Sa lazar Progress in Identifying Viruses Infecting Andean Root and Tuber Crops 156 C. Lizarraga, M. Santa Cruz, and L.F. Salazar Program 4: Integrated Pest Management 159 Fausto Cisneros Integrated M anagement for the Potato Tuber Moth in Pil ot Units in the Andean Region and the Dominican Republi c 162 M. Palacios and F. Cisneros Integrated M anagement fo r Andean Potato Weevil s in Pil ot Units 169 }. Alcazar and F. Cisneros Developing IPM Components for Leafminer Fly in the Canete Valley of Peru 177 N. Mujica and F. Cisneros Large-Sca le Implementation of IPM for Sweetpotato Weevil in Cuba: A Collaborative Effort 185 }. Alcaza r, F. Cisneros, and A. Morales Integrated Pest Management for Sweetpotato in East Afri ca 191 N. Smit and B. Odongo Farmer Field Schools for Sweetpotato in Indonesia 198 A.R. Braun, E. Priatna, R. Asmunati, Wiyanto, Y Widodo, and E. van de Fliert v Developing Weevil Resistance in Sweetpotato with Genetic Transformation 205 O.P. Zhang, A. Golmirzaie, G. Cipriani, A. Panta, M. Ghislain, N. Smith, I. Rety, and 0 . Michaud Program 5: Propagation and Crop Management 211 Mahesh 0. Upadhya Simultaneous Selection for Yield and Stability in Hybrid True Potato Seed Families 214 M. Upadhya and R. Cabello Temperature and Moisture Affect Dormancy and Deterioration of True Potato Seed During Storage 221 N. Pallais and R. Falcon Frost-Tolerant Potato Varieties for the Andean Highlands 227 E. Carrasco, A. Devaux, W Garcia, R. Esprella Sweetpotato Seed Units for the Dissemination of Planting Materials of Improved Varieties 233 E.E. Carey, S. T. Gichuki, 0. Hidalgo, and D.P. Zhang Program 6: Postharvest Management, Marketing 239 Gregory }. Scott Making Sense of Agricultural Marketing in Asia, Afri ca, and Latin America 242 C.}. Scott Perceptions Versus Projections for Potatoes: New Estimates Point to a Changing Global Research Agenda 244 G.}. Scott and A. Coccia Sustainability of Potato Consumption in Developing Countries: The Case of Bangladesh 249 G.}. Scott and H.E. Bouis Improved Rustic Storage in South Asia 256 S.G. llangantileke, VS. Khatana, J.P.Singh, and D.Kumar Recent Advances in CIP's Strategy for Collaborative Postharvest Research on Sweetpotato 264 G.}. Scott and C. Wheatley Enhancing the Role of Small-Scale Sweetpotato Starch Enterprises in Sichuan, China C. Wheatley, Lin Liping, and Song Bofu Feasibility, Acceptability, and Production Costs of Sweetpotato-Based Products in Uganda V Hagenimana and C. Owori vi 270 276 The Potenti al of Root Crop Process ing for Rural Development in Vietnam C. Prain, C. Wheatley, and Nguyen Doy Due Perspectives on Sweetpotato: Dual-Purpose Variet ies C. Leon-Velarde, }. Roca, }. Arteaga, L. Quispe, and A. Parraga Crop Growth and Starch Productivity of Edibl e Canna M. Hermann, R. Uptmoor, I. Freire, and J.L. Monta lvo Cross-Program Training Courses 1995-1996 Scientific Publications 1995-1996 CIP Board Members 1995-1996 CIP Staff in 1995-1996 Acronyms and Abbreviations 282 291 295 303 305 313 314 320 vii Providing Better Access to CIP CIP's Program Report, published every two years, takes on a new look for 1995 -96, and is more comprehensi ve th an ever befo re. Previous reports covered scientifi c projects so that they cou ld be reviewed quickly, but required many readers to request fo ll ow-up informat ion. In presenting this new report, our intention is to provide resea rch partners w ith a fu ll er understanding of CIP science, so that our work can be more eas ily rep li- cated. The new format is intended to be more usefu I to a broader spectrum of sc ienti sts, including our traditional partners in nat ional agricultural research systems, and research- ers in advanced laboratories and the private sector. Together with our sc ience reports, we have inc luded reg ional overvi ews of CIP work worldwide, which provide some of the needs assessment and ration ale for our re- sea rch strategy. Presented here in pub li shed form, th e entire report is also access ible throu gh CIP 's hom e page (http:// www.c ipotato.org) on the World Wide Web. Techni ca ll y, our Program Report cove rs CIP research conducted in 1995-96 under the able leadership of Dr. Pete r Gregory, fo rm er Deputy Di rector Genera I for Re- search. Readers should keep in mind, how- ever, that CIP resea rch is in a per iod of tran- siti on fo ll ow ing our 1995 extern al rev iew and because preparations for the ini t iat ion of th e Center's 1998-2000 Medium-Term Plan we re advanced by one yea r. Th ese events, combined with more comprehensive impact assessment and priority sett ing, have led us to make some strategic adj ustments in our resea rch programs. For example, readers may note th at this report contains more articl es in the area of biotechnology, particul arly in our Disease Management and Germplasm Management and Enhancement Programs. This represents a sh ift in CIP research emphas is brought about by the recommendations of an exter- nal review and by a recognition among CIP sc ient ists of the urgent need to avert a late bl ight emergency in deve lop ing countri es. In add it ion, readers c an exa min e th e progress being made in virology, as CIP sc i- enti sts are employing genet ic engineering in their work to incorporate naturally occu r- ring virus res istance in to advanced breed- ing lines . The 1995-96 report also points to how CIP w ill condu ct its resea rch under its next IX M edium-Term Pl an. As w e gea r up for a stream Ii ned project portfol io th at w i 11 be- gin in January 1998, the Center is adjusting its work programs, phys ical pl ant, and bud- get and management systems to accommo- date new prioriti es and initiati ves. The Me- dium-Term Pl an-w hich rece ived hi gh marks from the CG IAR Techni ca l Adv isory Co mmittee-g ives hi gh pri o r ity to la te bli ght, the deve lopment of sweetpotato cul- ti va rs w ith higher dry matter co ntent, and viru s di seases . In keeping w ith th e Center' s mi ss ion , these pri oriti es are heavil y weighted to those regions of the deve loping world w ith large numbers of poor peopl e. Without exception, x th e research cove red in the report fa vors low-in co me- espec iall y women-farm ers, labor-intensive techno logy th at encourages rural empl oyment, and urba n co nsumers w ith limited funds fo r purchas ing enough nutrient-ri ch food s. CIP resea rch emphasi zes work in and fo r those reg ions w here agri - cul tural producti v ity is threatened by envi- ro nm enta l degradati o n and, conve rse ly, w here in approp ri ate fa rmin g p rac ti ces threaten the natural resource base on which agri culture depend s. I beli eve that our new repo rting form at w ill assist our many re- sea rch partn ers, and w ill help br ing into greater focus the need to meet these chal- lenges. Hubert Zandstra Director General Introduction: Alleviating Poverty Through ""-~~~,,_,,,,,,,,,.,.~_,,.. Research The 54 papers presented in thi s program re- port document a large segment of the ma- jor research initi atives conducted at CIP in 1995-96 . Center sc ientists from each of our six program areas-together with represen- tatives from the consort ia and advanced laboratories with whom we conduct our re- search-share authorship for these reports. The papers present various aspects of th e Center's resea rch agenda, parti cul arl y those elements of our program focused on envi- ronmentally fr iendly production systems and the many yie ld-reducing factors that affect root and tuber crops. CIP's resea rch also em- phas izes deve loping techno logies that pro- vide farmers w ith greater access to more diversifi ed markets and, at the same time, protect endangered upper watershed eco lo- gies . Potato and Sweetpotato In pursuit of these goa ls, CIP began to shift a higher percentage of its resources into upstream research in 1995. The b lending of conventional and molecular approaches in CIP's breeding program, for example, illus- trates the important changes takin g place throughout much of the Center's research portfoli o. In the case of late blight di sease, our number-one pr iority, this includes a pro- gram of rec urrent se lec ti on and multi- locat ional testing of host-pl ant res istance, co mpl emented by dip loid prebreed in g, marker- ass isted se lect ion, and direct gene transfer. The feas ibility of this approach was il- lustrated over the past two years by Center v iro log ists work in g w ith partners in ad- va nced research laboratori es. As a result of their co llaborat ion, we now have ava il able a genet ic map that provides the prec ise lo- cation of res istance genes for several key potato viruses. Projections indi cate that the c loned genes ca n be successfull y int ro- duced into advanced breeding lines and varieties and made ava il abl e for test in g w ithin three years. In th e case of sweetpotato-now th e world 's seventh-ranking food crop-CIP re- sea rch was direc ted towa rd a signi f icant, but more limi ted, number of object ives. CIP soc ial sc ientists, work ing closely w ith co l- leagues at th e Intern at ional Food Po li cy Research Inst itute and in China, estimated XI that sweetpotato produ ction in developing co untri es wil l increase at an average an- nual rate of 1 percent fo r the period 1990- 2020. Thi s calculation is co nsistent w ith long-term trend s and emerging ev idence from severa l countries, whi ch indi ca te a bottoming-out in the decline in area planted or, as in the case of sub-Saharan Africa, a sharp increase in producti on. After more th an fi ve years of diagnostic work in genet ics, production and food sys- tems, and socioeconomics, CIP researchers formul ated a renewed sweetpotato resea rch agenda. The new strategy is based on the hypothesis that production and use in sub- sistence fa rming systems are constra ined by var ieti es that are poorl y adapted and low in dry matter. In response, CIP sc ientists be- gan to use the high-dry-matter accessions found in Center-he ld co ll ections for varietal improvement targeted at poor countries in sub-Saharan Afri ca and Southeast Asia , and the poorer prov inces of Chin a. Thi s work shou ld produce high-qu ali ty sweetpotato cultivars suitab le for direct consumption and for use as animal feed, starch, and flour prod- ucts . In add iti on, CIP sc ientists began to sc reen culti va rs suitab le for production in areas subject to drought. A major objecti ve of thi s work is to make sure that farmers, particu- larly in the harsher env ironments of sub-Sa- haran Africa, have enough seed cuttings for planting once the rainy season begins. The Center' s sweetpotato breeding program in Africa also reflects th e crop 's potenti al to offset th e im pact of cassava mosaic v irus and bl ac k Si ga toka , diseases th at now threaten food produ ction in much of east- ern and central Africa's most popul ated ar- eas. Genetic Conservation The work reported here on genetic d ivers ity analys is is the first use of molecular mark- ers at CIP to f ingerprint an entire cu lti vated potato co ll ecti on (Solanum phureja Juz. et Buk.). CIP scienti sts also deve loped a cost- effective protoco l for ge notypin g large X II ge rmplasm co ll ect io ns using RAPD tech- niques, or randomly amp li f ied po lymorphi c DNA. Integrating these and other molecu- lar methods of germp lasm management- together with more conventional conserva- tion meth ods- cou Id lowe r ge rmpl as m mai ntenance costs 30-40 percent, and al- low greater in vestment in eva lu ation and utili zation. A long simil ar Ii nes, cryopreservation re- search begun in 1995 should prov ide long- term so lutions to the probl em of maintain- ing vast numbers of trad it iona l root and tu- ber va ri eti es in conventi ona l gene banks. CIP continued its work through national sys- tems to safeguard nine exot ic Andean crop spec ies for fa rmers working in marginal ar- eas, especia ll y in the highlands. By the year 2000, Center sc ienti sts believe it w ill be possible to reduce investments in conserva- tion and shi ft emphas is to resea rch on uti Ii- zat ion and marketin g. CIP has also begun efforts to support com- munity groups th at w ish to conserve thei r germplasm in situ, and wo rks to improve local va ri eti es throu gh part ic ipatory p lant breed ing programs. We expect these initi a- ti ves to red uce pressure on conventi ona l gene banks, encourage natural evo lut ion, and co ntribute to geneti c d iversity. Natural Resources Management Th ough less advanced th an our work on potato and sweetpotato, conse rvation of lesser-known Andea n genet ic resources is con ducted throu gh CONDESAN, the Con - sortium for the Sustainabl e Development of th e Andean Eco re gion . In 1995-96, CO NDESAN invol ved work by more th an 100 publi c-sector organi za tions, uni ve rsi- ties, and NGOs from seven Andean coun- tri es . One particul arly important study com- p leted in 1995 presents startlin g informa- t ion on the hea lth risks assoc iated with in- sect ic ide use and prov ides poli cymakers with new methods for assessi ng the tradeoffs between agri cultural product ivity, environ- ment, and hea lth . CONDESAN, working in the established traditions of other CIP-coordinated research networks and consortia, is also a key ele- ment of the CGIAR's Global Mountain Pro- gram. The Program's goal is to compare land use patterns and provide technological op- tions that can be used across the world's great mountain systems. As with all CIP re- search, the overriding objective is to seek technological solutions that help alleviate poverty through increases in productivity and the enhancement of natural resources. We ask readers to recognize that CIP sci- entists work from the premise that the poor should benefit proportionally more from Center research and training activities than do others. Implicitly, gender and poverty concerns strongly overlap with this goal. Although the reports presented here may not always explicitly express these concerns, the research behind them is geared exclu- sively toward improving the lives of the ru- ral and urban poor in ways that are effec- tive, equitable, and ultimately sustainable. XIII PROGRAM REPORT Regi Overview of CIP Work in Latin America and the Caribbean Fernando Ezeta1 Potato production in Latin America has grown at an average annual rate of 2% over the past 30 years. This expansion of the total output resulted mainly from improved productivity since the area in potato remained stable at nearly 1 million hectares for the region. But the aggregate regional figures do not reflect important differences among subregions and countries. Productivity gains have been greater in the Southern Cone countries and Central America than in the Andean region. Expan- sion of area was observed in Colombia, Cuba, Ecuador, and Mexico only, whereas in most other countries of the region potato area de- creased or remained the same. · Among the technological interventions that have led to remarkable productivity 1 Regional Representative, Lima, Peru. gains are better varieties, access to quality seed, and improved agronomic practices, including dose and timing of fertilizer ap- plication. In spite of the obvious progress in potato agriculture observed over the past few years, average yields per unit area re- main well below productivity levels in in- dustrialized nations. Several constraints could be removed to increase overall po- tato output, basically via increased produc- tivity. Over the next few years, CIP will concentrate efforts in developing technol- ogy and assisting Latin American and Car- ibbean NARS in improving potato pest and disease management, improving seed qual- ity, and introducing processing attributes to new varieties . (IP Prog ra m Report 1995-96 l Disease Management Late blight (LB) is the main constraint to potato production in most of Latin America and the Caribbean (LAC). CIP's breeding program for resistance to LB with testing in Peru , Mexico, and Colombia (Rionegro), which led to population A, has produced several LB-resistant va rieti es for the region. In Costa Rica, three clones of population A, introduced in 1991 , were released as vari- eties in 1996. About 50% of the potato area in Costa Rica is now planted to these new varieties that are rapidly replacing cu lti va rs Granola and Atzimba, which are extremely susceptible to LB. In Panama, two varieties selected from population A were released by IDIAP1 • One of them, named IDIAFRIT, has excel lent attributes for the french fries industry. In the Andean countries, CIP's LB- resistant clones have been released in Bo- livia, Colombia, Ecuador, Peru, and Venezu- ela. In the Southern Cone, clones of popu- lation A have been used as progenitors in breeding programs. Direct adaptation of clonal selections from Andean env iron- ments to the growing conditions of the South- ern Cone has been limited by the demand there for earliness and good tuber appear- ance. The spread of a new, more aggressive popu- lation of the LB causal agent, Phytophthora infestans, together with growing environmen- tal concerns have triggered regional efforts to support the Global Initiative on Late Blight, sponsored by CIP. The principles of searching for durable resistance without R genes and integrated disease management (IDM) are widely accepted all over the region. Argen- tina is actively screening R-gene-free popu- lations adapted to long daylength. Other Southern Cone countries are equally interested in early screening of R-gene-free populations in the search for well-adapted, early LB-resis- tant materials. It is expected that PICTIPAPA and the potato research networks of the re- gion will support research for durable resis- tance in their subregions. 1 Acronyms ci ted in th is section can be found w ritten out in the section Acronyms, p. 320. 2 lAC Region The introduction of virus resistance is also of high priority in the region. Several breed- ing programs are making intensive use of CIP's PVX- and PVY-res istant populations to introduce res istance into their populations through conventional breeding. In Uruguay, variety IPORA, generated by INIA, incorpo- rates PVY resistance from CIP's PVY-resistant population. The in~roduction of vi rus res is- tance is likely to have a remarkable impact on seed renovation rates, especially among low-income potato producers. Bacteri al wilt continues to be an impor- tant constraint to potato production through- out the region. Movement of large quantities of seed from infected areas seems to be the main dissemination mechanism. Over the past three years, CIP has organized workshops to alert research institutions and government plant health services to the problem. These workshops provided the opportunity to rec- ommend actions to reduce, or possibly elimi- nate, the problem through IDM, quarantine, and the production of healthy seed. Pest Management Progress on integrated pest management (IPM) in LAC has been outstanding. In the Andes, pi lot projects for control of the potato tuber moth (PTM) and the Andean potato weevil have proven the viab ility of the inte- grated approach to keep these pests below economic threshold leve ls. Technology com- ponents developed at CIP have been tested in farmers ' fields to evaluate not only the tech- nology itself but also training and communi- cation materia ls. The program had a strong training component that reached technicians as well as potato farmers at the community level. In Peru, incorporating NGOs into the program increased the dissemination of IPM practices through extension services. The successful application of IPM tech- nology to the sweetpotato weevil (Cy/as formicarius) in Cuba and the Dominican Re- public must be highlighted. The sweetpotato weevil had gotten out of control in Cuba un- til the IPM program began. Now, the program is expanding rapidly, thanks to a well-orga- nized extension service and well-defined in- dividual control components. Among the components are effective sex pheromones that attract males into traps where they are eliminated by physical methods or by local- ized pesticide applications. Significant progress has been made in con- trolling the leafminer fly (LMF) through IPM. In the central coast of Peru, where potatoes are grown in the winter, farmers have been able to reduce multiple insecticide applica- tions to only two by adopting IPM practices. The use of yellow traps to catch adult flies dramatically reduces LMF populations. At the same time, the reduction in pesticide appli- cation favors biological control, which has proven to be extremely efficient. Planting Materials Although seed production technology is quite well disseminated in the region, the progress of seed production systems is below expec- tations. Pub I ic efforts to create a certified seed system have been unsuccessful in most cases because of the strong informal seed systems that supply seed to commercial growers. Nev- ertheless, in recent years the seed business has attracted the interest of the private sec- tor. Private seed growers are adopting mod- ern seed technology for rapidly multiplying healthy seed stocks. The progress made in pathogen detection of viruses and viroids is also a major contribution to improved seed quality standards. In summary, thanks to pri- vate efforts, local capacity for seed produc- tion in LAC has expanded in recent years. True potato seed (TPS) as an alternative source of healthy planting material has been successfully used in Nicaragua and Peru. Other countries in Central America are ex- ploring its potential. In Nicaragua, potato pro- duction depends on imported seed, which is expensive and physiologically immature. The experience with TPS has been rapidly assimi- lated into the production system of Nicara- gua with good market acceptance. In the Pe- ruvian highlands, the experience with Chacasina, a hybrid of a Peruvian variety and CIP clone 104.12 LB, has been so successful that farmers now have 200 kg of TPS as a source of seed for the coming years. Chacasina is a high-yielding TPS progeny with good acceptance among the consum- ers of this remote highland region, where good-quality clonal seed was extremely dif- ficu It to get before the project started. Changing Markets All over Latin America there is a growing in- terest in improving the processing quality of potatoes for chips and fries. The rapid expan- sion of the fast-food business in the region has created a demand for precooked frozen fries that, with few exceptions, is being met with Canadian imports. Most countries, how- ever, would like to substitute locally produced potatoes for imports. Dry matter content and frying quality of advanced clones have been considered as important selection attributes in most breeding programs of the region. With the present rate of development of the potato processing industry in LAC, varietal accep- tance is likely to be closely linked to process- ing quality in addition to agronomic traits and pest resistance. Globalization and free-market economics are introducing important changes in potato production systems and markets of LAC. The progressive removal of trading barriers be- tween countries within subregional common markets is increasing trade and competition among them. This more competitive environ- ment poses a threat to small farming units and favors large-scale, high-input production units. On a more global scale, LAC potato agri- culture may face severe competition in the future from producers in industrialized coun- tries. Production costs in most LAC countries are usually higher than in North America, an issue of major concern as LAC economies open up and reduce trading tariffs. Collaboration Collaborative potato research networks in LAC (PRECODEPA, PRACIPA, and PROCIPA) have confirmed their validity as mechanisms CIP Program Report 1995-96 3 fo r the hori zontal transfer of technology among members. A lthough at present only PRECODEPA counts on external funding, the links established between researchers in PRO CIPA and PRACIPA continue to facili- tate the exchange of research resu Its. Electronic communication is essential in today' s research and development efforts . 4 LAC Region CIP is therefore fostering the consolidation of an e lec t ro ni c informat io n netwo rk , INFO PAPA, th at aims to link potato re- sea rchers of all research networks w ithin LAC. We expect INFOPAPA to accelerate the exchange of knowledge and in fo rma- tion among members effi c iently and inex- pensive ly. Overview of CIP Work in the Middle East and North Africa Aziz Lagnaoui1 The close proximity to Europe, the relative wealth of the Middle East and North Africa (MENA) region, and the strong commercial appeal of potatoes make this commodity more of a cash and export crop. Up to 75% of Dutch seed exports, out of Europe, are destined to countries in the MENA region. These same countries are in turn potato exporters during the winter months, mainly to the United King- dom, France, and Germany. The special trade arrangements between the European Commu- nity and several MENA countries will certainly continue to influence this trend. CIP's com- parative advantage is the potential to increase yields by concentrating on crop and pest man- agement research. Our Regional Action Plan and priority-set- ting for 1995-96 focused on reducing environ- mental hazards of potato production systems by developing and disseminating biological control methods for major pests. CIP has a clear comparative advantage in these activi- ties and research results can be effectively used immediately. CIP showed impact in activities and places where a staff scientist was posted to work closely with counterparts in the national program. In response to our challenge to "focus on fewer things in fewer places," areas that held uncertain promises of success were avoided and promising re- search was continued to avoid a dilution of effort. The research team concentrated on in- tegrated pest management (IPM), true potato seed (TPS), and healthy sweetpotato planting material. With the strong belief that IPM is at the foundation of sustainable agricultural devel- opment, we continued to implement CIP's IPM 1 Regional Representati ve, Tunis, Tunisia. strategy through participatory research and training, promoting public awareness of the benefits of IPM, and involving policymakers. Integrated Pest Management of Potato Tuber Moth Our IPM strategy was to implement a process whereby potato growers in the MENA region could begin to substantially reduce their de- pendence on toxic pesticides for control of the potato tuber moth (PTM). Over the past few years, CIP and its partners in the Tunisian national program have developed effective practices for controlling PTM, the insect pest most responsible for significant losses in the region. Implementing the PTM management strategy was basically reduced to an educa- tional process. Recognizing the considerable opportunity to alter present PTM control practices in both Tunisia and Egypt, we invested our efforts in • demonstrating advanced IPM technologies on-farm; • promoting national policies that support the use of IPM; and • training national scientists and extension specialists to implement IPM on a large scale. By capitalizing on the soundness of the IPM strategy in Tunisia, we achieved a gradual shift in priorities as the Tunisian national pro- gram was encouraged to begin the large-scale implementation of IPM. We worked to deter- mine the best way to extend IPM practices on a larger scale and to accelerate adoption. In Tunisia, adoption of improved control practices has all but eliminated unnecessary CIP Program Report 1995-96 5 insecticide sprays for PTM control in fields. The adoption of Bacillus thuringiensis (Bt) and granulosis virus (GV)-based insecticides is steadily increasing. Complementary research focused on en- hanc ing GV activity by formulating it with various compounds to put an additional stress on the insect. That should reduce the time required for the virus to kill PTM and lower the pathogen dose needed. Improving traditional storage structures was also deemed necessary, so we ca lled on CIP storage specialists to design an improved tra- ditional storage facility. The Tunisian Minis- try of Agriculture is funding the construction of pilot storage units in the major potato-pro- ducing areas. The involvement of policymakers, includ- ing the President of Tunisia, generated unprec- edented excitement for IPM in the research and extension community. That is likely to accelerate the adoption of IPM. As antici- pated, the success of the Tunisian project is providing the foundation for the adoption of IPM in the other potato-producing countries in the MENA region. In Egypt, a ban on the most widely used pesticide for PTM control in storage precipi- tated interest in the use of 8. thuringiensis and granulosis virus. As a result, mass production of GV and Bt is increasing and quality is im- proving. In addition, 400,000 capsules of PTM sex pheromone for use in pheromone traps have been produced. Off-farm implementa- tion is a trend in newly reclaimed areas. Grow- ers contract with agencies or companies to control PTM using Bt and GV. A similar sys- tem was used in the past in the cotton IPM program, which is mainly designed to protect farmers from crop losses. Integrated Disease Management Causal agents of soft rot and dry rot A survey of potato diseases in fields and in summer rustic stores was conducted in Tuni- sia to identify the main pathogens responsible for storage losses. 6 MENA Region No incidence of soft rot or blackleg was found, although Erwinia carotovora and f. chrysanthemi were isolated from imported seed lots. In traditional stores, watery wound rot (WWR), also called potato leak, caused by Pythium aphanidermatum and P. ultimum, was prevalent. Tubers exhibited a discolora- tion similar to the pink rot symptom. Although Fusarium roseum var. sambucinum was iso- lated from imported seeds, it was dry rot caused by F. solani that showed a high inci- dence in stores. Losses in summer rusti c stores were correlated to the seed health status of the previous spring crop. In the field, wilt was caused by soil patho- gens such as Verticillium dah/iae and F. solani var. eumartii. P. aphanidermatum was also isolated from wilted plants; losses at harvest were mainly due to WWR. Wilt inc idence was not correlated to seed health status but to the cropping system (successive planting of solanaceous plants), thus showing the impor- tance of soil infestation in the transmission of disease. Watery wound rot management Pythium aphanidermatum and P. ultimum were found to be seed-transmitted as well as persistent in Tunisian soils. Because of the high incidence of these pathogens on-farm and in large-sca le stores of the Tunisian seed program, we investigated various components of inte- grated management of WWR. Comparison of varietal susceptibility. We developed a method to assess susceptibility to WWR and tested several European variet- ies from the national list. Among these, only varieti es Korrigane, Superstar, Safrane, and Yesmina were found to be moderately suscep- tible. Biological and chemical control. Seed treatm ents with the antagonistic soil saprophyte Trichoderma harzianum or with fungicide solutions of Ridomil (metalaxyl 10%, mancozeb 48%), mancozeb, maneb, or hymexazol efficiently controlled WWR de- velopment in the laboratory. Disinfecting soil by solarization. Popu- lations of Fusarium solani and of Pythium sp. w ere significantly reduced by soil mulching. The benefit of the method on potato growth and weed control was demonstrated in ex- periment station and on-farm trials. Baiting technique. A simple baiting tech- nique using oat seeds to detect and quantify soil populations of P. aphanidermatum was established. A detection threshold of 1-10 oospores/g of soil will allow us to identify noninfested soils for local seed multiplication. Research begun on the control of WWR will help to formulate effective integrated control recommendations for local seed mul- tiplication. Al I surveys and research efforts have been performed in collaboration with the staff of the Tunisian seed program and the Ministry of Agriculture. Extension specialists have been trained to improve their ability to diagnose storage disease. The identification of the main potato pathogens in Tunisia also permitted us to establish tolerance thresholds for certifica- tion schemes and importation agreements more appropriate to local conditions. Furthermore, these findings will greatly assist scientists in formulating integrated dis- ease management (IDM) approaches for po- tato soft and dry rots. Sweetpotato in Egypt Newly released varieties Kafr El-Zayat I and Kafr El-Zayat 11, varieties Jewel and Beauregard, and the selected advanced clone A 193 were produced in large quantities in the newly constructed screenhouses at CIP's Kafr El-Zayat station. So far, 300,000 healthy plant- lets were produced at the CIP station and dis- tributed to farmers by the national program for on-farm evaluation; over 2 million plant- lets were produced by the private sector. A similar operation is being set up in Syria (spinoff effect), where local planting mate- rial is degenerating. We will continue to em- phasize the quality of planting material and promoting wider use of the crop. True Potato Seed The performance of TPS hybrids and recent advances in TPS use led to greater transfer of this technology to Egyptian farmers. A num- ber of entrepreneurial farmers started to pro- duce seedling tubers from TPS at 50% the cost of locally produced seed tubers. Our liaison office in Egypt is catalyzing interactions with the private sector and government organiza- tions. Already a spinoff effect is taking place in neighboring countries as Syria has begun experimenting with promising progenies. The Future Over the years, CIP has ensured greater im- pact on-farm by strengthening the research capabi I ities of our national partners, by en- couraging farmer participation, and by en- hancing communication between the research and implementation institutions. In the years ahead, we will focus on the transfer and imple- mentation of acquired knowledge on IPM and IDM to ensure greater impact on-farm and contribute to the sustainability of production systems in key countries in the region (Egypt, Tunisia, Morocco). We will also seek to pro- mote linkages with institutions in countries of the Middle East Uordan, Yemen, Syria). Persistent cha I lenges in the region are the ever-increasing threats posed by potato tuber moths, whiteflies, and leafminer flies as a re- sult of intensified potato cropping in several MENA countries. Furthermore, special empha- sis should be placed on the evaluation of the potential contribution of biotechnology to potato and sweetpotato production in the re- gion. In accordance with IPM principles and biosafety guidelines, transgenic material will be carefully evaluated for its effectiveness and effects on nontarget species before deploy- ment. The most important challenge is to en- sure long-term continuity of CIP's work in the region and to respond to new requests while maintaining a minimal physical pres- ence. One way of meeting that challenge is to develop special country projects as needs are identified and defined. CIP Program Report 1995-96 7 Overview of CIP Work in Sub-Saharan Africa Peter T. Ewell1 Sub-Saharan Africa is a large and diverse region. Of its 47 countries, 36 report potato and/or sweetpotato production to the FAO. In the aggregate, food production has fa iled to keep pace with population growth, and per capita production has declined. Chronic deficits are made up with imports and food aid . According to the World Bank, 21 of the 30 poorest countries in the world-with GNP per cap ita per year below $430-are in sub- Saharan Africa. Natural environments, cul- tures, and economic conditions vary mark- ed ly among countri es. Vast regions are sparsely populated deserts and fo rests. Semi- arid lands, subject to periodic droughts, ac- count fo r large areas. Both potato and sweetpotato are grown primarily in densely populated, intensively cultivated highland and mid-elevation zones. They p lay impor- tant, contrasting roles in the food systems of the region. Potato and Sweetpotato in Africa Potato is a short-season, high-va lue crop, grown as a cash crop and for household con- sumption. According to the FAOSTAT data- base on the Internet, about 400,000 hectares are harvested in sub-Saharan Africa. Potato production reported in Africa as a whole has nearl y tripled over the past 35 years, from 1.3 million tons in the earl y 1960s to 3.7 million tons in 1996. This growth has been consis- tently and significantly higher than that of the population. This refl ects the crop's growi ng importance as a food in rapidly growing ur- ban areas. Broad ly speaking, there are three major potato systems: 1 Reg ional Representati ve, Nairobi , Kenya. 8 SSA Region 1. In the densely populated, high-poten- ti al highland areas (1,800-2,750 m) of east- ern and central Africa, potatoes are grown by sma ll farme rs (0.5-2 ha) both for the market and fo r home consumption. Yields vary between 5 and 20 t/ha, w ith a mean of 8-10 t/ha. Rates for chemica l ferti lizer and fungicide use vary widely among countries and production areas, but are generally low because farmers cannot affo rd chemica ls and supplies are unreliab le. Seed is obta ined primarily from informal, local sou rces and average seed quality is low. Late blight and bacterial w ilt, as we ll as v iru ses, cause chronic, signifi cant y ield losses. 2. In southern Africa, particularly in South Africa, potatoes are grown on a relatively large scale in the modern fa rming sector. Irri- gation is becoming more important, good- quality seed is ava il able, inputs are used in- tensively, and average yields are high-from 15 to over 25 t/ha in South Africa. 3. In Cameroon and Nigeria, potatoes are an important smallholder crop in higher ar- eas. El sewhere in West Africa, the crop is grown on a very small sca le as a high-va lue vegetable, usually under irrigation. Sweetpotato is an important food security crop grown in almost every country on the continent. Approx imately 1.5 million hectares are planted, primarily in rural areas fo r home consumption. Although the overa ll growth rate in reported production has fa ll en behind tota l population growth, sweetpotato has sig- nifica nt potential for increased use. Sweetpotato is a low-input crop, eas ily propagated from vine cuttings, that is grown under a wide variety of conditions, from in- tensive irrigation to commercial rainfed fields and to milli ons of sma ll plots in and around fields of other crops, along roadsides, in backyard gardens, and in urban plots. The crop is most important in eastern and cen- tral Africa, including Cameroon, in densely populated, intensively cultivated mid-eleva- tion (1,200-1,800 m) areas, slightl y lower than where potatoes are grown in most of the same cou ntries. Elsewhere in eastern and southern Africa, sweetpotato is an important secondary food in diets featuring maize and other cerea ls. The crop is important in cer- tain regions and periods of the year, such as in the " hungry months" w hen stores are ex- hausted and the next grain crop awa its har- vest. Diets in the lowlands of West Africa are dominated by cassava, yams, and other staples, and sweetpotato plays a minor ro le. Research in Partnership with National Institutions CIP set up its regional program in sub-Saharan Africa in the mid- l 970s. The goa l has been to work with national potato and sweetpotato research programs and other partner institu- t ions on key issues fac ing increased produc- tion and use. In 1996, nine CIP international and regional sc ienti sts were worki ng in the region: five in Kenya, three in Uganda, and one dividing time between Cameroon and Nigeria . These scienti sts were directly in- volved in core-funded, co llaborative research with national scientists in Kenya, Uganda, Cameroon, Ethiopia, Tanzania, and Nigeri a. Strong links were maintained with other potato and sweetpotato programs through the PRAPACE 1 network for centra l and east- ern Africa and the SARRNET network for the SADC countries of southern Africa. These networks provided the context for CIP's par- ticipation in seed relief and program reha- bilitation projects in war-torn Rwanda and Angola. M ore limited contacts are maintained with countries that are not network members, fo r the distribution of in fo rm at io n and germplasm. 1 Acronyms cited in this section can be found w ritten out in the section Acronyms, p. 320. Co llaborative research is concentrated in six major project areas: 1. Potato late blight. Late blight is the single most important potato disease in the region, particularly in the tropica l highland env ironment of central and eastern Africa. Temperatures in the major production zones (1,800-2,750 m) are relatively even. The rainfa ll pattern is bimoda l, and inoculum is almost continuously present. Farmers do not have the income to spray fungicides regularly, and the disease causes serious yield reduc- tions in most years. The disease is less signifi- cant in the drier areas of southern Africa. Over the past 20 years, blight res istance has been the principal trait of new varieties adopted by fa rmers in countries throughout the region. Several advanced clones are near release and are expected to do well , although R genes are present and their resistance is likely to break down eventually. Sets of geno- types from population B, with better horizon- tal resistance, are in the earl y stages of se lec- tion in Kenya, Uganda, and Ethiopia. As the Globa l Initiative on Late Blight develops, CIP and its national partners will increase their investments in research, parti cularl y to pro- mote integrated disease management on- fa rm. 2. Farmer-based potato seed systems in Kenya, Uganda, Ethi opi a, and Cameroon. Farmers need healthy seed, which is avail- able on their farms at planting time, in the right physiological state, at a price they can afford . Centralized systems for seed multi- plication and certifi cation modeled on those in Europe and North America have proven difficult to implement and maintain in pub- lic-sector institutions in Afri ca. An alterna- tive is to provide selected farmer-multipli- ers w ith clean starter seed stock, and sup- port th em to become spec iali zed in the multiplication of seed for sa le to other farm- ers in their area. This also provides a mecha- ni sm to get new varieties to fa rmers. Th e fac i I ities that CIP developed over many years in co llaboration with the Plant Quarantine Station in Kenya fo r the intro- CIP Program Report 1995-96 9 duction and testi ng of germ p lasm for re- gional distribution have been converted into a seed unit. The best avai lable varieties are multipli ed in v itro as "sta rter stocks" for seed programs in target countries. Linkages have been deve loped with NGOs with contacts in pilot potato-produc ing communities to fac ilitate selection of suitab le fa rmers, train- ing in seed techniques, and support for con- structing diffused-light stores. 3. Integrated disease management for potato. IDM is the onl y practi ca l approach to bacterial wi lt, as the disease organism pers ists in the soil, is carr ied by infected seed tubers, and cannot be controlled with chemica ls. In Uganda, CIP is co llaborating with NARO, the national research institute, in th e African Hi ghl ands Initi at ive, an ecoregional program linked w ith the Global Mountain Initiative. Its purpose is to take a multi-institutional, multicommodity approach to improving resource management in the in- tensively cultivated high-potential areas of the highlands. One project is the integrated management of diseases associated with increasing land use intensity and decreasing soil fertility. Bac- terial w ilt is one of four diseases being inves- t igated. A full-time regional scientist attached to CIP to work on the project has shown that a package of improved rotation, clean seed, tolerant varieties, and improved on-farm sani- tation reduces losses, even o n very smal I fa rm s. Strategies for improved management provide an entry point, to encourage farmers to improve their land use management, with immediate payoff in increased potato y ields and longer-term payoffs in stab le soil fertility. 4. Sweetpotato improvement. From the late 1980s, CIP's regional program has im- ported advanced sweetpotato cu lt iva rs from throughout the world fo r distribution and eva luation. At the same time, true seeds pro- duced from crosses made by Uganda's na- tional program and from elsewhere are be- ing tested at a few locations. Farmers grow diverse combinations of lo- ca lly adapted varieties, w hich tend to be late- l 0 SSA Region maturing and low-yielding. Pl anting mate- ri al is exchanged informall y among rela- ti ves and neighbors, and on ly a few intro- duced va ri eties have become establish ed in sub-Saharan Africa over the past 25 -30 years. African farmers are interested in new va ri eties if they meet certain cr iteri a: high y ield, ear liness, and persistence of plants in fields so that planting materi al is avai l- able for the next season. Drought-tolerant va ri eti es would permit the expa nsio n of sweetpotato cultivation in drier areas and drier peri ods in the year. Good taste is very important both for home consumption and for sa le in the market. Consumers like roots with high dry matter (usually at least 2 7- 28%) and moderately sweet taste. Res is- tance to or tolerance of v iruses, w eevils, and other pests and diseases could help ex- pand the range of the crop and increase y ield s. New varieti es for new or expanded uses also have significant potential. Most Afri can consumers prefer roots w ith w hite or cream- colored flesh. Orange-fleshed varieties can provide significant quantities of v itamin A, w hich is criti ca lly short in the diets of many rural people, particularly children. Prelimi- nary evidence shows that mothers w i II grow, prepare, and serve new varieties, if they un- derstand the implications fo r the hea lth of their families. Sweetpotato v ines are high in protein and are an excellent forage for ani- mals, parti cularly for weaning ca lves and kids. They are currentl y fed as a by-product, but dairy fa rmers will plant new varieti es with high forage yields specifically for th is purpose. Varieti es fo r processing will be pl anted if se- cure markets become available. 5. Integrated sweetpotato crop manage- ment. The most w idespread and damaging pest in eastern and southern Africa, parti cu- 1 arl y in dry areas and in d ry years, is sweetpotato weev il (Cy/as spp.), a focus of CIP's co ll aborative research. A number of other insects, viruses, and fungal diseases are important in certa in areas. Co ll abora- tive research in Uganda and Tanzania, with ass istance from the Natural Resources Insti- tute, has isolated sex pheromones of all three spec ies found in Africa, and is testing their use fo r monitoring and possibly mass-trap- ping. Several cultural practices, such as main- taining an adequate sepa rati on between sweetpotato fields, careful hilling, and re- moval of all res idues from plots after harvest, have been shown to reduce damage. Pilot projects are being estab li shed in Uganda to work w ith farmers to see how these components ca n be adjusted and combined to provide effective protection aga inst wee- vi I damage. M ost successful IPM programs have been adopted by fa rmers because they help to reduce pestic ide applications, thereby reducing costs. As pesticides are not used on sweetpotato except by a small minority of farmers, the challenge is to develop packages that increase yields suffi ciently to justify the increased labor, cash costs, and management attention. Broadening the approach to include additional key constraints in an integrated crop management package is a key to suc- cess. 6. Expanded postharvest use of sweetpo- tato. In Afri ca, sweetpotato roots are con- sumed almost exc lu sive ly in fresh form; most are usually just boiled. In other parts of the world , particularly in Asia, use of sweetpotato as a raw materi al in the pro- duction of processed foods, feed, and indus- trial products has inc reased signifi ca ntl y over the past 40 years . The adaptat ion of known products and processes to Afri can conditions could open up new markets for farmers. The use of a loca lly grown, low- input crop in processed products could also red uce the countries' need for food imports to meet the needs of rapidly growing c ities. Co ll aborative resea rch in Ke nya and Uganda has demonstrated that a ready mar- ket ex ists for homemade products such as flat bread (c hapa tis) and dou ghnuts (ma ndazis), which substitute cooked and mashed sweetpotato for a certain proportion of the usual wheat flour. Farmers in some areas cut roots into chips and dry them in the sun to preserve them. These processes are being improved so that the dried chips can be ground into a high-quality flour, w hich can be eas ily stored and transported, and used in mixtures with w heat and other flours in baked goods. Other Activities CIP's reg ional office provides a link be- tween sc ientists working in national institu- ti ons and research throughout the world . Courses, short-term attachments, and study tours provide training in research and ana- lytica l techniques. CIP sc ientists act as ad- visors for many student thesis projects. Sup- port for networks and a number of bilateral , reg ional, and multi-inst ituti onal projects help ensure that research responds to the needs of potato and sweetpotato farmers in Africa. CIP Program Report 1995·96 l l Overview of CIP Work in South and West Asia Sarath Ilangantileke1 The largest potato producer in the South and West As ia (SWA) region is Ind ia, fo llowed by Bangladesh, Nepal, Pakistan, Sri Lanka, and Bhutan. Sweetpotato prod uction is mainly in India, Bangladesh, and Sri Lanka; production in other countri es is localized and of littl e si gnifi cance. Potato The SWA region produces about 21 million tons of potatoes on 1.3 million hectares. In the past decade, total potato production in India has increased rapidly, with compara- tively slowe r in creases in Bangladesh , Bhutan, Nepal, Sri Lanka, and Pakistan. In- creases in total production in cou ntries other than India and Bhutan are attributed more to a gradual expansion in area than to sig- nifi cant y ield increases. Among the six re- gio nal countries, Indi a current ly has the highest nation al yield average (16 t/ ha), w hereas Nepal has the lowest (8.3 t/ha). Potato is used mainly fo r the fresh market. However, local and multinational snack and fast-food industri es are develop ing rapidl y in the region, and India is stri v ing to prov ide potato for the domestic processing industry. Estimates indicate that about 9,000 tons of the crop are now processed annually in the unorgani zed sector alone in India. Trade w ithin the region in seed and ware potato is expected to increase in the future w ith new trade agreements between the SWA coun- tries. Demand for selected processing and high-yie lding varieties will therefore increase stead ily. Problems of cultivation in the region, al- though similar, are location and ecoregion 1 Regional Representative, New Delhi , India. l 2 SWA Region specifi c. A major problem is the ava il abil- ity of high-quality loca l seed for ware and process ing. Countries outside India depend on either imported seed or low-q uality lo- ca l seed. But late bli ght and bacteri al w ilt significantl y affect yields in all SWA coun- tries. Postharvest problems of storage and marketing already exist and may become more pronounced w ith increased production and future trade in the region. The intensi- ti es of problems differ significantly. As a regular m ajor acti v ity, potato germpl asm was distributed from CIP-Lima and the regional office to NARS, networks, NGOs, and private-sector organizations in the SWA region. Thi s catered to the increased demand for germplasm for specific breeding requirements such as processing and disease resistance. Development of healthy seed material is an important acti v ity. Inc reased interest among farmers in TPS technology has resulted in a remarkable increase in crop production area in d iffe rent agroecologica l reg ions of India. In addition to domestic use, large quan- titi es of TPS produced in India have been ex- ported to Vietnam, Egypt, Indonesia, and the Philippines outs ide the SWA region, and to Bangladesh, Nepal , and Sri Lanka within the region. Regular fa rmer training programs and ex- tension activ ities by NARS in co llaboration with CIP have helped increase TPS produc- tion and use in Bangladesh. TPS demonstra- ti on projects and farmer t raining on loca- tion-specific agronomic practices in Nepa l have resulted in increased use of TPS in the mid-hill s and low lands (Terai region). Al- though production of true seed in Nepal is limited, use of TPS from India has shown an upward trend. The Asian Development Bank TPS project in Sri Lanka has provided assis- tance in evaluation and farmer demonstra- tion trials. Research on TPS conti nued for developing earliness and disease res istance. Training in the region has been dominated by TPS-related training, a major reason for the rapid increase in TPS use. An interre- gional workshop on TPS production and use held in India had participants from China, Vietnam, Sri Lanka, Indonesia, the Philip- pines, Nepal , Bangladesh, and India, as well as sc ientists from CIP-SWA and the ESEAP region. The private sector, progressive farm- ers, and NGOs were also represented. CPRI and CIP-SWA jointly gave field training on TPS production and use to sc ientists from Sri Lanka and Bangladesh. Individual and group training activiti es on TPS in the re- gion included training on hybr id TPS pro- duction and use fo r research workers, farm- ers, agricultural officers, field ass istants, and NGOs in India, Bangladesh, and Nepa l. The search for new areas for potato pro- duction in India has resulted in the use of ri- verbeds to cultivate potato in Orissa, w ith ass istance from CIP. We have already ob- served an increase in production area. This technology targeted spec ifica lly to resource- poor farmers and to the use of underused land resources is ga ining popularity and has pro- vided significant increases in potato yields. Seed and ware potato storage performs an important function in the production and marketing of the crop in the region . In In- dia, large quantities of potato are stored in co ld stores immed iately after harvest. Farm- ers unable to avai l themselves of co ld-store fac iliti es store the produce on-farm until prices begin to increase two to three months after harvest. Storage in heaps and in tradi- tional structures results in losses. Research on an alternate method of storage using rus- tic evaporative coo l stores was continued in three large potato-producing districts. A nat ional workshop on problems of potato storage in India was held for sc ienti sts, farm- ers, policymakers, and co ld-store owners as a way to identify future strategies for stor- age research. In the future, CIP needs to address sev- eral potato act iv ities in the SWA region: varietal improvement for end use, disease resistance, and productivity; ava i la bi lity of high-quality seed and effic ient seed sys- tems; postharvest and marketing; and de- velopment of human resources. Sweetpotato The SWA region produces about 1.6 million tons of sweetpotato on 0.2 million hectares. India accounts for 68% of the total produc- tion, fo llowed by 27% in Bangladesh and about 5% in Sri Lanka. Sweetpotato is an important crop for Sri Lanka and occupies 1.1 % of the total cropped area, followed by 0.5% in Bangladesh and less than 0.1 % in India. Although production area in Pakistan is insignificant, yields of more than 10 t/ha are reported, followed by 9.5 t/ha in Bangladesh, 8.3 t/ha in India, and 6.5 t/ha in Sri Lanka. In India, y ields have improved since the mid- 1980s, whereas yields in other countries have declined gradually after the respective peak yield years. Use is limited to consumption as a veg- etab le in all SWAcountries, with an insignifi - cant quantity of processed snack food, lim- ited to India, followed by Bangladesh. Indus- trial use is nonexistent. A major problem is the avai labi lity of idea l varieties for different agroeco logies and different cropping systems in the region, where crop duration influences yield poten- tials. High inc idences of sweetpotato wee- v i I and other pests and diseases limit achievement of maximum yie lds, w hile a lack of postharvest utilization methodology and marketing chains limits enthusiasm to elevate the crop beyond its present status as a "poor man's" crop. The production and use of sweetpotato in the diets of people of South and West CIP P1og1om Repo11 l 995-96 l 3 Asia are expanding. The striving to popular- ize the use of sweetpotato in diets has re- sulted in the exchange of germp lasm from neig hbo rin g countri es. Sweetpotato germplasm distribution was done in v itro and as stem cuttings, as requested by NARS of Bangladesh, Nepa l, and Sri Lanka; networks of different organizations; and NGOs and private-sector o rgani zation s in the region. Varieti es having high carotene content were supplied to a USAID nutriti o n project in Nepal. CIP-SWA supported travel and atten- dance at international seminars and confer- ences. Sweetpotato breeding in co ll abo ration with NARS of India has resulted in a transfer of adva nced c lo nes for eva lu atio n to Bangladesh and Sri Lanka, and to research organizations in the sweetpotato-growing ar- eas of India. A sweetpotato production and weevil control course in India for fa rmers and agricultural officers of Goa was held to en- courage increased production. Such activi- ties wil l be increased to help popularize the crop. Like for potato, var ietal improvement for end use, increased ava ilab lity of high-qual ity planting material, and postharvest/marketing and development of human resources are important acti v ities for CIP-SWA in develop- ing sweetpotato production in the region. Collaborating Institutions Research was done in co llaborat ion w ith NARS. In Ind ia, with the Indian Counci l for 1 4 SWA Region Agricu ltural Resea rch and its institutes-the Central Potato Research Institute and the Central Tuber Crops Research Institute (for sweetpotato), the Department of Horticulture, state univers ities, NGOs, farmers, and private- sector organizations. In Bangladesh, co llabo- rat ion was w ith the Bangladesh Agricu ltural Research lnstitute 's Tuber Crops Research Center. A notable feature in Bangladesh was NGO interactions to help di ffuse TPS. Regul ar trainin g programs and travel grants were prov ided to NARS scientists. CIP p rov ided MSc and PhD programs for re- searchers in Bangladesh and India on TPS techno logy and on sweetpotato varietal se- lect ion and eva luation, respecti ve ly. Out- side funding sources were so lic ited through the GBF (Soc iety for Biotechnologica l Re- sea rch) for NARS sc ientists for a three-month train ing in biotechnology in Germany. Expected Accomplishments Potato and sweetpotato are important crops in the diet of peop le in the SWA reg ion. Act ivities conducted by CIP are expected to lead to sustainable seed and planting ma- teria l systems in countries wi th needs. CIP wi 11 a lso wor k toward marketing and postharvest strategies that w ill result in seed se lf-suffic iency and a production environ- ment with economic benefits for potato and sweetpotato fa rmers. Overview of Cl P Work in East and Southeast Asia and the Pacific Peter Schmiediche1 Potatoes account for roughly ha If of the world 's annual output of all roots and tubers and, since the ea rl y 1960s, the increase in area planted in developing countries has been higher than fo r any other major food crop . Annual world production currently totals 274 million tons on 18 million hectares, and China and India alone produce 22% of this total. The potato sector worldwide is in transi- tion . Europe and the fo rmer Soviet Union account fo r the bulk of production, but the situat ion is chang ing rap idl y. In the early 1990s, about 30% of the globa l potato out- put was produced in developing countries, up from 11 % in the early 1960s. If this trend continues, in less than a generation, most of the world's potatoes will be harvested in Asia, Africa, or Latin America. As a result, potato is becomi ng an in c reas ing ly importa nt source of food, rural employment, and in- come for the growing populations in these regions. Over the past three decades, Asia has ex- perienced the world's highest annual growth rate in potato production. With its short crop- ping cycle, potato fits well into Asian food production ca lendars, and is particularly com- patible with fast-growing hybrid cerea ls. The expansion of co ld storage, the rapid emer- gence of processing fac i I ities for the fast- food industry, and the indirect influence of improved ri ce and wheat irrigation systems have also contributed to the expansion of potato production in Asia. Low prices rela- tive to cerea ls combined with the economic impact of improved storage systems have st imulated greater demand. Thi s trend is continuing strongly in all Southeast Asian 1 Regional Representative, Bogor, Indonesia. potato-growing areas, particularly in those that feed the rapidly expanding urban con- centrations and megametropolises of this region. Demand is strongest fo r processed potatoes that supply the fast-food industry in these urban centers where processed po- tato products fetch top prices, thus convert- ing this tuber into a highly profitable cash crop . Thi s increas ing demand puts strong pressure on national and international breed- ing institutions to produce high-quality pro- cess ing va ri et ies that are well adapted to the w ide array of agroecologica l conditions in Southeast As ia. Even if more and more varieties with the required processing and culinary qualities become ava ilable, the lack of adequate seed systems is the si ngle most important imped i- ment to the development of the crop. A prof- itable crop starts with high-quality seed, but unfavorab le agroeco logical conditions in- terfere with the production of such seed in most Southeast Asian potato-growing areas. One solution to the seed problem is the use of botanical or true seed instead of con- ventional propagation from tubers. True po- tato seed (TPS) has promise in seve ral agroecological niches of Southeast Asia. One such area is the Red River Delta (RRD) around Hanoi, Vietnam, where several thousa nd hecta res of potatoes will be grown from TPS during 1997-98 to avoid the disease prob- lems assoc iated with traditional tuber plantings. The hybrid seed grown in Viet- nam was developed in CIP-sponsored projects in India, where seed production is now in the hands of the private sector. The TPS project in Vietnam is sponsored by the Asian Development Bank, and this project CIP Progrom Report 1995-96 15 is in the process of revo luti oniz ing potato culture in the RRD, w here more than 90% of Vietnam's potatoes are grown. It is also introducing hybrid TPS product ion in the RRD and in the highland area around Dalat in south Vietnam. In additi on, Vietnam is using numerous introductions of CIP potato germpl asm to great effect. Several variet- ies or potential varieti es have been selected for those areas w here the TPS technology is not yet in use. According to Chinese sources, CIP potato germpl asm has boosted China's successful breeding program since the ear ly 1980s w hen it had sta lled because of the lack of genetic vari ability. Potato variety CIP-24 now occu- pies more than 250,000 hectares in China. Results of CIP's basic phytopathology re- search have been successfully used in China's seed potato program and in the integrated control of late blight and bacterial w ilt. Apart from transferring technology, CIP has assisted w ith numerous training events for groups and individuals inside and outs ide China. Because potato is affected by more pests and diseases than any other food crop, it has become a heavy user of pesticides, with the assoc iated negati ve effect on the environ- ment. To reduce pesticide use, programs of integrated pest management (IPM) have been deve loped for this region. CIP's ESEAP re- searchers collaborate w ith governmental and nongovernmental organizations to promote integrated pest and disease management based on four princip les originally developed by the FAO lntercountry IPM Program for Ri ce: • grow a hea lthy crop, • conserve the natural enemies of pests, • monitor fi elds frequently and, through training and hands-on experience, • fa rmers become expert practitioners of integrated pest management. In sweetpotato culture, the most impor- tant issue to be resolved in thi s region is that of sweetpotato use. The problem of use is close ly followed by the need to control the sweetpotato weevi l, the most important 1 6 ESEAP Region sweetpotato pest in Southeast As ia. Efforts to contro l the weev il w ith c lass ica l IPM methods have evo lved into a comprehen- sive package of integ rated crop manage- ment (ICM) si nce national programs w ill not, as a rule, implement an isolated IPM solu- tion that does not address other related prob- lems of raising a successful sweetpotato crop. As already indicated fo r potato, in As ia, the first principle of ICM is "grow a healthy crop," implying the central pos iti on of IPM within an approach of integrated crop management. Successful ICM depends on coupling the class ica l acti v ities of deploy- ing host-plant resistance, using biological contro l, reducing pesticide use, enhancing and protecting natural enemies, and deploy- ing sex pheromones w ith equally classical agronomic measures. Moreover, ICM can only be successful if indiv idual farmers and the farming commun ity understand and par- ticipate in these measures. Thi s has been achieved through a seri es of highl y success- ful Farmer Field Schools in Indonesia. The concept of ICM and Farmer Field Schoo ls is now spread ing to other countries in this region, particularly Vi etnam, where it will be applied to both potato and sweetpotato culture. A special project funded by Swiss Devel- opment Cooperation (S DC) concentrated on the co ll ection, c lass ifi cation , and use of sweetpotato biodiversity in Indonesia, com- bined w ith a pioneer effort to document the assoc iated indigenous knowledge of this germpl asm to facilitate its rapid use. Thi s project was designed to identify sweetpotato germplasm that would meet specific needs, initi ally of Indonesian sweetpotato cultivators and later of growers in other areas of Asia. Exploration of in situ conservation techniques for sweetpotato is under way in Indones ia and the Philippines. China's principal sweetpotato-breed ing inst itutions have worked close ly w ith CIP's regional sweetpotato breeder in the use of a w ide array of germplasm to develop cultivars w ith high dry matter content. A highly visible and successful program, using v irus-free sweetpotato planting material in Shandung Province, has been acknowledged to have benefited from basic v irus research and asso- ciated training activ ities conducted by CIP. Sweetpotato processing, as it relates to the use of starch for noodle production, is another area in w hich great advances have been made in China, parti cularly in Sichuan Province-- the largest area of sweetpotato production in China. Research on sweetpotato pig-feeding systems in that province is an area of enor- mous potential, with spillover possibilities for other provinces of China. A major CIP-UPWARD 1 initiative on the use of sweetpotato and canna in north and central Vietnam compl eted a deta il ed assess- ment of research needs, and sweetpotato is now being introduced as an alternati ve source of starch fo r noodles and as an improved source of pi g feed. Vietnam's strength in po- tato and sweetpotato research is in no small part the result of highly active ongoing CIP- sponsored and CIP-mediated training efforts during the past 15 years. CIP's impact in the Philippines has been achieved through four institutional structures that developed and implemented a series of projects to address the needs of potato and sweetpotato culture in that country: (1) the UPWARD network, which, although regional in scope, has been particularl y active in its host country; (2) the regional ASPRAD 2 net- work, which implemented a signifi cant num- ber of hi ghly succe ss ful pota to and sweetpotato projects through its acti ve coor- dinating offi ce in Los Banos; (3) the SANREM3 watershed-based, natural resources manage- ment activ iti es in Mindanao; and (4) potato seed resea rch and deve lopment, impl e- mented through CIP's seed unit in Baguio and through ADB-financed TPS research in Luzon and Mindanao. 1 UPWARD (Users' Perspectives W ith Agricu ltural Research and Development). 2 Thi s network used to be known as SAPPRAD (Southeast Asian Program for Potato Research and Development) before it was renamed ASPRAD (As ian Sweetpotato and Potato Research and Deve lopment). Its Australian fund ing terminated on December 31, 1996. 3 SAN REM (Sustainable Agricultural and Natura l Resource Management). Producti on- level impacts are promising through the diffusion of CIP-der ived potato germpl asm via fo rm al and info rm al seed channels in the conventional potato produc- tion areas, and through the use of TPS in Mindanao. Signifi cant takeoff is already evident in the use of Farmer Field School approaches to the integrated management of bacteri al wilt. Ur- ban home gardens research in northern Phil - ippines is now showing payoffs w ith large- sca le expansio n in th e use of improved sweetpotato germpl asm and other techno lo- gies. Institutional impact has been achieved through the adopt ion of parti c ipatory ap- proaches to research by many institutions, involving fa rmers, fa rming communities, and other end users of potato and sweetpotato in research. For example, participatory variety evaluation procedures developed by ASPRAD have been adopted in the Philippines and in other countries. lnteragency research partner- ships developed by UPWARD are now prac- ti ced by loca l institutions. Filipino agri cultural research insti tuti ons have highly trained in- dividu als and, through UPWARD and ASPRAD, Filipino sc ientists have been able to make a significant research impact in other countries of thi s region through short-term consultancies. In Indonesia, research has concentrated on the introduction and testing of new po- tato germpl asm, the use of TPS as an alter- native seed system, the integrated manage- ment of bacteri al wilt, the integrated crop management of sw eetpotato through devel- opment of the Farmer Field Schoo l model, and th e prese rv ation of sweetpo tato biodiversity and indigenous knowledge as- sociated w ith sw eetpotato biodiversity. CIP has also contributed in signifi ca nt w ays through PhD training to the pool of agri cul - tural research experti se in potato. In summary, CIP has responded fl ex ibly and with the prudent use of scarce human and financial resources to the needs of thi s region. Results and impact are apparent and the foundations for further adva nces have been laid . CIP Program Report 1995·96 l 7 PR OGRAM REPORT Production Systems Thomas S. Walker1 Since the last Program Report (for 1993-94), the Producti on Systems Program has contin- ued to emphas ize w ork in three areas: (1) characterization of constraints and produc- tion opportunities in potato and sweetpotato, (2) adaptation and integration of potato and sweetpotato production technologies, and (3) assessment of the impact and sustainability of potato and sweetpotato production tech- nolog ies. Research and training highlights in these three areas are presented for 1995-96, and imp I icat ions of the M edium-Term Plan 7 998- 2000 for program activiti es are described. In parti cular, the management of natural re- sources is an emerging area of emphas is as CIP responds to environmental concern s in shaping its portfolio of act ivities. 1 Program Leader, CIP, Lima, Peru. Pro Characterization of Constraints Characterization of production constraints and opportunities is an area of declining im- portance. Several studies were completed in 1995-96. In potato, lessoris from five years of parti c ipatory farmer se lection of late-blight- res istant varieties were drawn and reported in Bolivia. Di agnost ic research exp laining yield var iation in farmers' fi elds in Bolivia is presented later in thi s report. The compara- tive advantage in potato production across the three main producing regions of Ecuador was assessed. Surveys of national agri cultural research systems (NARS) provided data for documenting present leve ls of investment in potato crop improvement programs and for prescribing resource allocation, given the importance of potatoes in nat iona I CIP Prog ram Report 1995-96 l 9 production and assumptions on expected ef- fects. In sweetpotato, the last offive country char- acterization studies in eastern and centra I Africa was completed for Tanzania. Prospects for intensifying and sustai ning sweetpotato production in response to increasing popula- tion pressure were also examined at three benchmark sites in southwestern Uganda. The agroeco logica l characteri zation of sweetpotato product io n systems was re- viewed in Asia. Institutiona liz ing a participa- tory approach in root crop research was the foca l point for both research contracts and training workshops for NARS partners in China, Indonesia, Nepa l, the Philippines, and Vietnam. In the medium term, characterization re- search wi 11 be merged with impact assessment into one project for potato and another for sweetpotato. Priorities fo r potato include up- dating the perceptions, first elic ited in 1987, of NARS scientists on constraints and oppor- tuniti es; characterizing constra ints and oppor- tuniti es for production in the central Asian countries; and evaluating regional competi- tiveness in production. For sweetpotato, the globa l synthesis of information from invest- ments in characterizat ion research in the early and mid-1990s is still the main priority. Technology Adaptation Advanced varietal testing and release figured prominently in technol ogy adaptation . In potato, several virus-resistant, heat-to lerant varieties with good processing traits were re- leased for earl y and late planting in coastal Peru. These varieties w ill extend the length of the growing season and hence reduce sea- sonal price fluctuations. Steady progress was made in the advanced testing of sweetpotato var ieties. In Uganda, the most important producer of sweetpotato in sub-Saharan Africa, five varieties were re- leased by the national program. Although CIP was not intimately invo lved in the breeding 20 Progroml or se lection of these varieties, their release establi shed an institutional precedent for the distribution of promising CIP-related mate- ri al. In 1998, research on varietal adaptation w ill take place in self-contained projects in other CIP programs. Work on adapting v irus- res istant potato varieties will become part of the project on potato vi ruses. Similarly, adap- tive trials on advanced sweetpotato material w ill be subsum ed into the proj ec t on sweetpotato dry matter and adaptation. Impact Assessment Turning to the area of impact assessment, a monograph documenting the impact of nine CIP-related techno logies was published in 1996. Two priority-setting exercises, w hich used info rmation from the impact case stud- ies, w ere also carried out and reported . These laid a firm foundation fo r CIP's Medium-Term Plan 1998-2000. Among CIP-related techno logies, true po- tato seed (TPS) received the most scrutiny. Resu Its from several years of on-farm trials formed the building blocks for an ex ante as- sessment in Egypt, India, and Indones ia. In the future, selective case studies of ex post impact wi ll still be conducted . However, we wi ll place more emphasis on eva luating the effects of improved technology on pov- erty and the environment. Some of this work will be carried out in the framework of the recent CGIAR initiative on impact assessment. Examining the effects of cyclical price insta- bility on potato production and consumption is a pol icy- related area targeted for intensive study. Natural Resources Management Organizationally, Production Systems is the locus for much of Cl P's increasing investment in research and training in natural resources management. Selective investments in sys- tems modeling, soil fertility management, and natural resources economics supplied CIP with a criti ca l mass to tackle issues related to land use in the Andes and to intensive po- tato-cropping systems. Research addresses the common theme of interactions between land productivity, technologica l change, pol icy intervention, and environmental improve- ment or degradation. This work is highly complementary to CIP's potato research as potato cropping systems are usually pivotal to the susta inability of land use systems in tropica l mountain environments. Modeling approaches, featuring widespread participa- tion from developing-country NARS and men- tor institutes, are used to overcome problems of site specificity, whi ch erode the transfer- ability of research results. C IP 's contribution to research on the susta inability of land use in the Andes is con- ducted w ithin the CON DESAN network and is embodied in two research and training ac- tiv ities. The first activity, DME-Norte, marri es bio- physica l and economic models to examine techno logy, policy, and environmental sce- nari os. It draws on and extends models de- veloped in recently completed research on groundwate r con taminat ion as a conse- quence of the heavy use of pesti c ides in po- tato production by sma ll farmers. Potatoes rotated w ith pastures is the land use of inter- est for modeling. The second project, DME-Sur, combines satellite imagery with biophysica l models to assess frost, drought, and sa linity risk in crop and livestock production. We are also eva lu- ating techno logica l and policy options to mitigate the severity of these abiotic stresses in one of the largest and poorest regions of the Andes. Product ion Systems also houses natural resources management research carr ied out in the Global Mountain Initiative (GMI) con- vened by CIP, ICIMOD, and ICRAF. This ini - tiative st imul ates research and synthes is across the three most important tropical and subtropical mountainous regions of the world . Priority research areas include agroecologica l characteri zation, integrated nutrient manage- ment, agri cultural intensification in response to population pressure, the fate of investments in soil conservation, and dairy policy. The sustainability of intensive potato-crop- ping systems is the third area of natural re- sources management for program act iv ity. Thi s project begins in 1997 and focuses on sustainability of rapidly expanding potato pro- duction in the lndo-Gangetic Plain of South Asia. Initially, attention will be directed at the intensive rice-potato-rice system in regions that have witnessed spectacular growth in potato production since 1985. This project will be executed under the umbrella of the CGIAR's Rice-Wheat Program. Looking ahead, Production Systems is rap- idly evolving into a natural resources man- agement program. The name of the program will likely be changed to recognize this shift in emphasis. CIP P1ogrom Repo1t 1995-96 21 Prospects for Sustaining Potato and Sweetpotato Cropping Systems in Southwest Uganda Jan W. Low1 Many of the most productive agricultural sys- tems in sub-Saharan Africa are located in the densely populated highlands of Eastern and Central Africa. Kabale District in southwest Uganda is representative of the highlands in this part of the world. Potato figures promi- nently as a cash crop, and sweetpotato is an important food crop. Kaba le District provides an opportunity to examine the dynamic roles of potato and sweetpotato in an area of intensifying agri- cultural production in response to increasing population pressure and market demand. Moreover, CIP participated with the National Agricultural Research Organization of Uganda in re-establishing potato research and seed production activities at the Kalengyere research station in Kabale Distri ct from 1989 to 1994. This study has two main objectives: (1) to obtain baseline data on current production practices in representative sites so varietal change and other technology-related issues can be monitored over time, and (2) to assess the prospects for the intensification and sustainability of potato and sweetpotato pro- duction. To address these objectives, we undertook field research in three phases. In February 1995, we visited potential sites, defined as a distinct valley with surrounding hillsides, held group discussions with farmers, conducted in- formal market surveys, and made final site selections. 1 Formerly at CIP Sub-Saharan Afr ica reg iona l office, Nairobi , Kenya. 22 Program I Subsequent field work combining struc- tured household surveys with agronomic ob- servations on potato and sweetpotato plots was done in September 1995 and February 1996. In total, we interviewed 134 potato pro- ducers and 89 sweetpotato growers. All the sweetpotato growers were women; the ma- jority of the potato producers were men. Fig- ure 1 shows the geo-referenced survey data. Of 1 7 potential field sites, 3 were selected to provide a range of agroecological condi- tions in the production of potatoes in a val- ley or neighboring hillside where sweetpotato is also an important crop. Kalengyere is a high-altitude area (2, 100-2,500 m) of inten- sively cultivated, steeply sloped hills. Bukinda (at 1,700-1 ,900 m) is lower than Kalengyere and much warmer. Therefore, conditions are less favorable for potato production. How- ever, both potatoes and sweetpotatoes are widely grown on the hillsides and in the val- ley bottom. Kicumbi (at 1,800-2,100 m) is somewhat higher, and its most notable fea- ture is the huge, flat Katuna valley. Parts of the valley floor are dominated by larger fenced landholdings for pasturing dairy cattle. In other parts of the valley, wall-to-wall po- tato is seen during the dry season. Extensive sweetpotato ridges with deep drainage canals dominate the hillside landscape. Changing Roles of Potato and Sweetpotato as Income and Food Sources Growers were asked to assess whether the importance of potato and sweetpotato had been increasing or decreasing, or had re- mained the same, during the past five years, both as a food for home consumption and as Kalengyere (2, 100-2,500 m) • <> ~ Ao ·=·· Research•• Station Lake Mutanda RWANDA 0 10 20 30 Bu kinda (1 ,700-1 ,900 m) AOAo 0. Kicumbi (1,800-2, 100 m) o<>A 0 0 <> • 40 kilometers Survey sites: 1995/96 A Potato plot <) On-farm trial : sweetpotato o Sweetpotato plot t Kalengyere Station 0 District Capital N - - A - - Figure 1. Research sites in Kabale benchmark survey. a source of income (Tabl e 1 ). Over half the growers reported that potato w as increasing in importance as a source of cash and as a food for home consumption. They c ited two major reasons fo r thi s in- crease: increasing demand in the marketplace and the high productivity of the crop, in part because of new varieti es. Potatoes as a food are also becoming more popul ar. Greater home consumption is in part a refl ection of their greater avai labi I ity atthe household level as a by-product of planting more of the crop fo r cash sa le. In sp ite of the better market fo r potatoes, some fa rmers reported a dec line in the im- portance of potato as a cash source during the past five years. The overwhelming rea- son fo r thi s tendency w as declining yi elds. Farmers cited poor soil fertility or di sease as contributing factors. Yield decline also fi gured prominently in the minority percepti on that potatoes were diminishing in importance as a food crop. Sweetpotato, on the other hand, has never been an important cas h crop in Kabale Di s- tri ct. Therefore, the 32 households noting an increas ing significa nce of sweetpotato as a cash crop is of interest. This increase w as at- tributed primaril y to increasing demand fo r the crop, parti cularl y near urban centers such as Kaba le, and fo r feeding students in boa rd- ing schoo ls. M any more respondents stated th at the importance of sweetpotato as a cash crop w as declining because of decreasing market de- mand as well as declining yields. These fa rm- ers complain that because everyone grows sweetpotato and there is no market fo r the crop outside of the loca l area, it is demanded loca ll y only when other foodstu ffs are in short supply. Dec lining yi eld has also contributed to a reduction in the ro le of sweetpotato for home consumption. That is a disturbing trend, given th e traditi o nal yea r- round re li ance o n sweetpotato as a food security crop in Kaba le. CIP Progro m Report 1995-96 23 In higher-el evat ion areas, suc h as Kalengyere, sweetpotato is also losing ground to potato as a food for home consumption. In group discussions, Kalengyere mothers often remarked that children preferred potatoes to sweetpotatoes. Older household members re- tai ned their preference for sweetpotato as a staple food. In contrast, the erod ing role of sweetpotato in home consumption in Bu kinda and Kicumbi, where preference for sweetpotato over potato remains high, is al- most exclusively attributed to declining yields. Sustainability of Potato Production Resu lts from on-farm y ield assessment of po- tato production indicate that farmers in Kalengyere are, in general, obtaining excel- lent yields (29.5 t/ha). It would be highly de- sirable to sustain these y ields, albeit with lower production costs. At the other end of the spectrum, potato y ie lds for Kicumbi swampland farmers are already low (7.7 t/ha). The system is fragile and y ield increases are urgently needed. Sustaining mean y ield lev- els of 19.6 t/ha at the lower altitudes (1,800- 2, 100 m) of Bu kinda over time would mean that potato would remain an important cash crop within this area. In interpreting these y ield data, it is impor- tant to note that the mean potato plot size on these farms is small , ranging from a low of 560 m2 in Bukinda to a high of 812 m 2 in Kicumbi. Thu s, even the high yield of 29.5 t/ha in Kalengyere translates into 2.0 t (ap- Table 1. The perceived changing importance of potato and sweetpotato as cash and food crops in Kabale District, Uganda, 1995-96. Trend Increase in importance' Increasing yield Increasing market demand Increasing family size Increasing food preference Increasing area planted More land Other Decrease in importance Declining yield Declining market demand Decreasing area planted Declining food preference Decreasing family size Less land Other No change Cash 122 61 76 2 2 1 85 60 9 6 2 2 6 15 Potato Food 121 41 1 47 33 7 1 l 57 27 5 10 7 5 2 44 Sweetpotato Cash 32 10 23 77 20 35 6 10 2 113 Food 76 18 43 7 8 2 88 48 14 18 4 2 4 58 a. Number of respondents. A few respondents gave more thon one reason for a perceived trend; therefore, the totals for reasons exceed the total for trends. 24 Program l proximately 19 bags) of potatoes from an av- erage plot size of 699 m2 • Several encouraging signs point to sus- tained production of potato in Kabale Dis- trict. They include (1) high demand for pota- toes in Kampala, the capital city, (2) improved road infrastructure, (3) increasing entrepre- neurship, (4) increas ing input use, and (5) the ex istence of late-blight-tolerant varieties. The greatest threat to sustained potato pro- duction in Kabale District is the increasing incidence of two major diseases: late blight caused by Phytophthora infestans and bacte- ri a I wilt caused by Pseudomonas solanacearum. Farmers lack knowledge of recommended control practices so their abi l- ity to employ them is severely constrained. Although farmers can recognize the symp- toms of both diseases and know the associa- tion between high humidity and late blight, knowledge of the causes of bacterial wilt is extremely limited (Table 2). Moreover, their awareness of nonchemical approaches to dea ling with late blight is scanty (with the exception of delayed plant- ing to avoid heavy rains). Knowledge of how to prevent the spread of bacterial wilt and deal with an infected fi eld is marginal to nonexist- ent. The introduction of some agronomic practices, such as roguing volunteer potato plants, may encounter considerable resis- tance, in that many farmers perceive these re-emergent plants as an important source of food. Although the use of animal manures and crop residues on potato fi elds is increasing, steep terrain and limited availability make manure application a labor-intensive process. That encourages farmers to keep re-using plots for potato cultivation nearer to home. The preferred management strategy is to wait until yields have fallen to unacceptably low levels, then graze livestock on the fallow plot so that manure transport is not an issue. The unavailability of inorganic fertilizers is a major constraint. It is unlikely that given Uganda's landlocked status, and the distance of Kabale from Kampala, that fertilizer use could be undertaken on a large scale with- out being subsidized. Yet another serious constraint to imple- menting proper agronomic practices is the lack of sufficient quantities of c lean seed available to farmers. High-quality seed from Kalengyere is sold cheaply to a limited num- ber of contact farmers in the area at a subsi- dized rate. In theory, these farmers multiply seed and se l I it to other farmers in the area. In practice, most of the next harvest is sold off as ware potatoes for consumption or re- tained by the contact farmer for replanting. Only a few other farmers purchase the new varieties for seed. In addition to having insuffi cient quanti- ties of good-qua I ity seed to start with, farm- ers lack knowledge of how to select and prop- erly store seed. To ensure that the skin of the potato does not easily peel during storage, stems of the potato plant should be cut off a week or so before harvest. Many farmers fear doing this, however, as cut stems advertise to potential thieves that potatoes are ready for harvest. Without significant investments in farmer education and improved distribution of high-quality seed, current potato yields are not sustainable. Sustainability of Sweetpotato Production While farmers widely recognize declining soil fertility as a problem, little evidence exists linking defic iencies in swe2tpotatoes with poor soil fertility. Preliminary research on nutrient deficiencies in sweetpotatoes and concurrent so i I ferti I ity was conducted in Feb- ruary 1996. Leaves were evaluated for criti- ca l amounts of macronutrients and micronu- trients. Soil samples were also taken from the base of each deficient plant at a depth of 15- 30 cm. The major findings from the leaf analyses were: • Phosphorus deficiency was very common. • Nitrogen, potassium, and sulfur are also generally suboptimal. (IP Program Report 1995-96 25 Table 2. Farmer beliefs about late blight and bacterial wilt symptoms, causes, and prevention, Kabale District, Uganda, 1995-96. Symptoms, ca uses, and steps• Symptoms of late blight Tubers do not enlarge Leaves dry Leaves blacken Tubers usua lly rot Plant rots & breaks Low to no yields Causes of late blight Excessive rain Cold weather Loss of soil fertility Growing potato on some plot several times Lock of spraying Aphids Poor seed quality Does not know Steps token to avoid late blight Sprays chemicals Wonts to spray, but locks funds Adds manure Symptoms of bacterial wilt Plant dries up ("wil ts") Tubers rot Stunted growth Causes of bacterial wilt Does not know Infected seed Loss of soil fertility Overcultivotion Insects in soi l Planting potato ofter potato Soil unsuitable for particular potato variety Infected soil Seed left in rain Using damaged or cut seed Poor land management Planted in dry season Premature seed Too much ra infa ll Coldness Steps token to ovoid bacterial wi lt Uproots infected plants None a. n = 99 fa rmers. 26 Program! Crop rotation Following Harvesting in dry weather Sells infected seed and buys new seed Adds manure Proper seed selection Percentage of all farmers interviewed 70 61 37 11 3 6 86 18 7 5 2 2 1 1 97 2 l 90 76 l 38 17 13 9 9 6 6 4 2 2 3 l l 1 l 48 37 10 4 1 l l l • Boron concentrations were low to very low at al I but two sites. Soil analyses frequently showed deficien- cies in P, K, and Ca. Although these results are preliminary, because of the limited de- gree of sampling, they reinforce the percep- tion that sweetpotato is increasingly relegated to soils of dec lining fertility. Even though something is produced, yields continue to decl ine. M oreover, there is an assoc iation between increas ing severity of the fungal dis- ease Alternaria so/ani and decl ining soil fer- tility. The growth of potato in importance as a cash crop and for home consumption has, in the higher elevations of Kabale, displaced sweetpotato in the diet to a certain extent and pushed the crop into more marginal produc- tion areas. However, three facts assure that sweetpotato will continue to remain an im- portant food security crop in Kaba le: • Sweetpotato is the cheapest source of ca lo- ries available year-round. • Sweetpotato outperforms other crops on soils of decl ining fertility. • Sweetpotato is an important rotation crop for sorghum, maize, beans, and peas. There are warning signs that sweetpotato yields will continue to decline. In areas with littl e or no demand for co mm erc ial sweetpotatoes, farmers are unlikely to invest many resources in improving the fertility of marginal soils on whi ch they prefer to plant sweetpotato. Problems with Alternaria spp. are likely to increase, particularly at the higher elevations, because of the worsening nutri- tional status of the pl ant. Although periodic outbreaks of sweetpotato butterfly (Acraea acerata) are likely to continue, they tend to be loca l and farmers take group act ion when severe infestations occur. Prospects for market demand to increase beyo nd Kaba le District are limited. Sweetpotato is wide ly grown throughout Uganda and higher-quality production areas surround Kampal a and other major urban centers . Increasingly, valley bottom lands and swamplands are being targeted fo r potato or vegetable production during the dry season. In the past, sweetpotato was planted in the mo ist valley bottom lands to assure adequate supplies of planting material for hillside pro- duction the following season. The evidence suggests that sweetpotato will continue to be grown in large quantities on better so il s as part of rotation schemes. In addition, it will be grown on marginal soils where other crops fail, in spite of continual mean yield declines. Overall , yield decl ines can be reversed only if non-labor-intensive solutions to increasing so il fertility are found or sweetpotato becomes commercially i mpor- tant, hence increasing the benefit-cost ratio of increased input use. Research Implications Declining soil fertility is a major constraint of increasing importance to all crop production in Kaba le District. Pol icy research needs to be conducted on the cost-effectiveness of importing and subsidizing appropriate inor- ganic fertili zers. It is clear that the limited availability of organic manures and the labor to apply them are likely to remain significant constraints in the long term. Research efforts should also concentrate on testing green ma- nure crops that are adapted to higher altitudes and that could be pl anted on fallow fields. Unlike the major effort that has been made on potato, there has been no systematic in- volvement in adaptive testing of sweetpotato varieties in southwest Uganda. However, given that varietal tolerance of low P levels and Alternaria spp. exist, it wou Id make sense to test sweetpotato varieties bred by the Rwandan national program. During the past two decades, the Rwandan program has bred many varieties adapted to higher altitudes. Those should still be available. Finally, the investment in a baseline sur- vey is fully rea lized only if sites are periodi- cally revisited. Although the selected sites are excellent for monitoring input use, disease in- CIP Program Reporl 1995-96 27 tensification , and potato y ields, fa rmers at these sites do not possess a significant num- ber of plots w ith vo lcanic soils. An additional 28 Pmgmm l site should be added in the fo llowing round that would include these volcanic soil types that are highly suitable for potato production. Risk Analysis of Potato Production in the Altiplano: Quantifying Farmers' Beliefs R. Valdivia Alatrista1, R. Quiroz1, R. Valdivia Fernandez1, and V. Choquehuanca2 The Altiplano of Peru and Bo livia is home to many very poor farm families for w hom agri- culture is the main source of income. Pota- toes are the staple food crop, and production ri sk, usually in the fo rm of drought and frost, is pervasive. The incidence and productivity conse- quences of ri sk are difficult to assess because historica l data on the productivity of fields and fa rms are not ava il ab le. Nonadoption of improved technologies is often attributed to ri sk, but both subjective and more objective information is usually lacki ng to document the relative ri sk iness of competing technolo- gies. Nor has the effectiveness of different meth- ods to assess risk been estab lished. In thi s study, yield and net revenue risk of native and improved var ieties are exami ned in three ways: (1) estimated yields from samp les taken in potato fields, (2) elic ited information on fa rmer beliefs about varietal productivity, and (3) simulated yield estimates from process- based, crop-growth models. Site and Methods This research was conducted in Santa Maria, a community in the Department of Puno, Collao District, Peru. It is situated at about 3,900 m above sea level. Land is private ly and communally farmed. Potatoes are a dryland crop grown from October to April in the mai n growing season. 1 CIP-CONDESAN (Consortium for the Sustainable Development of the Andean Ecoreg ion). 2 CIRNMA-CONDESAN (Centro de lnvestigaciones de Recursos Natura les y M edio Ambiente, Peru). Mixing native and improved potato varieties in the same field is common. Native varie- ties include bitter (Ruqui , Pinaza, Lok' a, Luki, and Ajanhuiri) and non bitter (Ccompis, lmilla Negra, Sacampaya, lmilla Blanca, and Cho- quepito) cultivars. Improved varieties are rep- resented by And ina, Tomasa Condemayta, Mariva, and Mi Peru. Yield samp les in farmers' fields were co l- lected during six grow ing seasons from 1987 to 1992. Stratified sampling by elevation was used to generate estimated yie lds for native and improved vari eties. About 1 0 fields were sampled in each of the 6 yr. Farmers were interviewed in 1996 when information on the inc idence and conse- quences of risk was canvassed. Farmers rec- ogn ize four classes of years for potato pro- ductivity: bad, norma l, good, and very good. Beliefs on varietal yield were assessed using a tr iangu lar method where the respondent provided information on the highest, modal, and lowest observations. By multiplying the frequency of the four productivity events by the results from the triangular elicitat ion pro- cedure by varieta l type, we generated yield distributions for native and improved varie- ties. Daily weather data and soils info rmation figured as inputs into a process-based, crop- growth model to simulate potato yield . Ge- netic coefficients on potential leaf and tuber expansion rate, degree of determinacy, pho- toperiod sensiti vity, and temperature sensitiv- ity for lmill a Negra and Andina were used to simulate the performance of the bitter and nonbitter varieta l types. CIP Progrom Report 1995·96 29 Table 1. Measured , elicited, and simulated potato yields in Pu no, Peru, during the period 1987 through 1992. Variety Native Improved Mixed Measured 8.28 (0.85)0 12.04 (1.24) 10.03 (0.84) a. Numbers in parentheses ore stondord errors of meOf\S. Results We determined risk by comparing yield and net revenue of native, improved, and mixed varietal groups. Yield comparisons Productiv ity of the native, improved, and mixed varietal groups is compared for each of the three methods of estimation (Table 1 ). The yield for the mixed varieties was ca lcu- lated as the average fi eld proportions of na- tive and improved varieties across the 6 years. All methods give productivity estimates substantially higher than the departmental average yield of about 5 t/ha for Puno. The difference between estimated and officially reported yields partially reflects the impact of technological change introduced by the Proyecto de lnvestigac i6n de los Sistemas Agropecuarios Andinos in 1985. Hea lthier seed, improved fertilizer, and other crop man- agement practices played a rol e in this change. The simulated results in Table 1 are espe- c ially noteworthy because est imates from such models are often several orders of mag- nitude higher than yields in farmers' fields. In our case, the crop simulation accurately reflects rea lity. The measured and elicited estimates are similar. Indeed, regress ing the elic ited on the measured yield data for the 6 years accounts 30 Progrnm I Yield (t/ha) Elicited Simulated 9.77 (l.21) 8.7 6 (0.88) 13.23 (1.50) 9.80 (0.85) l 0.25 (0.61) 8.99 (0.85) for about 98% of the variation in the mea- sured productivity estimates. Improved varieties were also characteri zed by the highest mean yields. Differences be- tween the improved and nat ive varietal types were most apparent with the measured and el icited methods. In contrast, the simulated results gave more or less the same results for each varietal type. Figure 1 shows the tradeoffs in risk and productivity and presents the estimated cu- mulat ive di stribution function of yields by var ietal type for each of the three estimation methods. The cumulat ive probability reflects the likelihood of obtaining a yield less than or equal to a stipul ated leve l. For example, in Figure 1 A, the chance that native varieties produce less than 4 t is only about 3%; the expectation of this very low yield for the im- proved varieties is 10 times more likely at 30%. Crossing curves imply tradeoffs in risk, that is, native varieti es are less prone to both low and high yields (F igure 1 A). The crop-cut method gives results highlighting the stabil- ity of nati ve varieties and the yield respon- siveness of improved varieties to good weather. In contrast, the fa rmers' subjective beliefs (Figure 1 B) show almost no conflict between risk and expected productivity. For almost all y ield levels, improved varieties dominate. The cumulative probability for a stipulated yield level is everywhere higher for Probability (%) 100 90 80 Improved 70 60 A. Measured 50 40 30 20 10 12 16 20 24 28 32 Yield (Uha) Probability (%) 100 90 80 70 60 B. Elicited 50 40 30 20 10 12 16 20 24 28 32 Yield (Uha) Probability (%) 100 90 80 70 60 C. Simulated 50 40 30 20 10 10 11 12 13 14 15 16 17 Yield (Uha) Figure 1. Estimated cumulative distribution of yield by estimation method. CIP, 1996. CIP Program Report 1995-96 3 1 Probability (%) 100 90 80 70 60 A. Measured 50 40 30 20 10 0 O> O> O> O> co N N N N N N N N O> 0 ~ O> ..,. O> ..,. :JS ,_ N ,_ N ,_ N ,_ N :JS N ,_ ~ ~ c;; O> ... co "' ,_ N "' N N "' "' ... ... "' "' "' "' Net income (US$) Probability (%) 100 90 80 70 60 50 B. Elicited 40 30 20 10 0 "' "' "' "' "' "' "' "' "' "' "' "' "' "' "' "' "' "' "' "' ~ ;:': ~ co co co co co co co co co co co co co 0 0 0 0 ~ ~ N "' 0 ... co N "' 0 ~ co N "' 0 ... co N N "' "' "' ... ... "' "' "' "' .... .... .... Net income (US$) Probability (%) 100 90 80 70 60 50 C. Simulated 40 30 20 10 0 "' I'- I'- I'- I'- I'- "' I 0... <.) () () (") ::> I 0... <.) () () se lects accessi ons from each stratum to maxi - mize the genetic di stances in the samp les . We d iv ided the phenogram at 0.65 and 0.70 va lues of Jaccard's simil arity coeffic ient, whi ch resulted in 16 and 35 clusters from w hich one genotype was chosen at random. The diversity level in the se lected sample was measured by RAPD marker variance V m = S (npq)/(n-1) where n = number of indi v iduals, and p and q are the frequenc ies of the presence or ab- sence of a RAPD marker (F igure 4). Th e analys is shows that in both cases (sample size of 16 and 35), marker-ass isted sampling has higher mo lecular vari ance val- ues (i .e., geneti c d iversity) than random sam- pling (Figure 4). The va lue of V m increased by 17% w hen 16 accessions were sampled, and by 15% w hen 35 access ions were sampled. Th is resu It demonstrated th e adva ntage of using marker data to construct a core co llec- tion w hen the co llect ion can be ass imilated into a si ngle popul ation. 1.15 KB 1.09 KB 0.80 KB co C\J ::> I 0... <.) () () (") ::> I 0... <.) () () Figure 3. Morphological duplicates of the Solonum phurejo collection resolved by molecular assays : polymerase chain reaction amplification with primer OPR3 on the left and OPZ4 on the right. Arrows indicate differential markers. CIP Program Report 1995-96 81 TMV 14 13.5 13 12.5 12 11.5 11 10.5 10 Global 131 MS16 RS16 MS35 RS35 Figure 4. Total marker variance (TMV) for random (RS) vs. marker-assisted (MS) sampling of the Solonum phureia collection using 16, 35, and 131 genotypes. Conclusions To form a core collection, the large potato collection held at CIP should first be sampled and stratified based on molecular marker data. Then a sample should be taken from each stratum to represent each marker-differ- entiated group. Also, the number of core accessions should be decided based on the retention of the maximum number of alleles. This procedure will lead to a core subset covering the maxi- mum genetic diversity in the large collection. Based on th is subset, curators and breeders could add certain genotypes known to pos- sess important agronomic traits . 82 Progrom2 The designated core collection will then be systematically evaluated for major agro- nomic traits so that the gene variation in the collection will be better understood. These core accessions will then be cleaned and made available to germ plasm users. This core collection approach would allow us to better use our limited resources for rat ionalized management of the potato germ plasm col lec- tion. Selected Reading Hawkes, J.G. 1992. Biosystematics of the potato. In: The potato crop. P. Harris (ed.). Chapman & Hall , London . p.13-64. Nienhuis, J., J. Tivang, and P. Skroch. 1994. Analysis of genetic relationships among genotypes based on molecular marker data. In: Proceedings of the symposium Analysis of Molecular Marker Data. p. 8-15. Ochoa, C.M. 1990. Solanum phureja. In: The potatoes of South America: Bolivia. Cambridge University Press. p. 315-334. CIP Program Report 1995-96 83 DNA Markers for the lntrogression of Late Blight Resistance in Potato M. Ghislain, B. Trognitz, C. Herrera, A. Hurtado, and L. Portal1 Resistance to Late Blight in Potato A major challenge in breeding for resistance to late blight caused by Phytophthora infestans in potato is to fully exp loit tuber- bearing 5olanum germp lasm as a source for res istance to the disease. Indeed, germplasm used by breeders has a narrow base compared to th e potential ava ilable in 5olanum germplasm. Two types of resistance to late blight (LB) exist in this germplasm: vertical (governed by a specific gene-for-gene interaction) and hori- zontal (governed by several genes whose in- teractions are still large ly unknown). CIP 's breeding strategy is to use horizon- tal resistance from nove l sources to make re- sistance in potato to LB less dependent on pathogen races and hence more durable. Quantitative inheritance and the high load of unfavorable genes in wild and native pota- toes, however, have impeded the broad use of th is type of resistance to develop new po- tato cultivars. Unraveling the Genetics of Horizontal Resistance to Late Blight Researchers explained some of the genetics of horizontal res istance to LB in the early 1990s. Potato scientists have characterized horizontal resistance in several 5olanum spe- cies and in interspec ific hybrid progenies. A molecular genetics study has even been de- veloped with a preliminary geneti c map of quantitative resistance to LB from an intraspe- cific 5. tuberosum cross. This type of resis- tance is thought to be the result of several independent mechanisms that are constitu- 1 CIP, Lima, Peru . 84 Program2 tively expressed and induced upon pathogen infect ion . At CIP, we have been applying molecular techniques to unrave l the geneti cs of horizon- tal res istance to LB in tuber-bearing 5olanum spec ies , especially 5. phureja and 5. verrucosum. Developing detailed genetic linkage maps using molecular markers will help us explain the genetic control and ar- chitecture of this quantitative trait. Thi s knowledge and molecular tools will improve the efficiency of the i ntrogress ion of valuable resistance and defense genes from 5olanum germplasm into the cultivated potato. Materials and Methods We have deve loped populations of diploid potatoes that segregate for LB resistance from three accessions of a native cultivated potato, 5. phureja. This diploid species was selected for its high levels of horizontal resistance to LB (leaf resistance), favorable tuber charac- teristics and culinary quality, and crossabil- ity with dihaploid potatoes. We are analyz- ing the three populations (VP, PD, and PP) at CIP for phenotyp ic segregation of resistance using a complex race of P. infestans (Table 1). We have worked in close co llaboration with scientists Rhonda Meyer and Robbi e Waugh from the Scottish Crop Research In- st itute, and Susana Marcucci and Esteban Hopp of the lnstituto Nacional de Tecno logia Agropecuaria , Centro de lnvest igac i6n en Ciencias Veterinarias, in Argentina. Molecu- lar markers have been developed to analyze hybrid populations, parental heterozygosity, and simultaneous segregat ion with pheno- typic scores. These tool s are polymerase chain reaction (PCR)-based deoxyribonucleic Table 1. Solanum germplasm used to map horizontal resistance to late blight. Population Individuals Female parent Male parent code (no.) Identification Resistance Identification Resistance VP 102 S. verrucosum Vertical and S. phure;a Horizontal Pl-116163 horizontal CCC81 cv. Yema de Huevo PD 246 5. phure;a Horizontal 5. tuberosum Susceptible CHS-625 dihaploid PS-3 PP 305 S. phure;a Horizontal 5. phure;a Susceptible Pl ~225678 . 52 acid (DNA) markers-randomly amplified polymorphi c DNA (RAPD), amplified frag- ment length polym o rphi sm (AFLP) , and m icrosatel I ites--and have been used on the first two populations onl y. The markers have proven the hybrid nature of the VP and PD popul ati ons, a prerequisite for a breeding strategy based on molecular geneti cs. Th e segregation of these markers in the progeny is indicative of the level of heterozygos ity of the parents. Results Characterization of DNA markers DNA markers were scored separately for each of the parents, leading to the definiti on of two parameters: heterozygosity index (Hi) and a marker value (Mv) . The heterozygosity index of each parent is ca lculated here as the ratio b etwee n seg rega tin g band s and nonsegregating scorabl e bands th at co rre- spond to either one of the two parents in the progeny. The marker value is the average number of informative DNA markers (po ly- morphi c and segregating) per assay. Th e analysi s of more than 50 RAPD primers indi- cates by means of th e Hi valu e th at th e di haploid 5. tuberosum clone is th e most het- erozygous of the four (Tabl e 2). In contrast is the 5. verrucosum parent, whi ch barely shows any marker segregati on. Pl-320362 .57 The marker value helps us se lect an ap- propri ate marker system for a parti cul ar cross and all ows us to predi ct costs in terms of time and assays of building a geneti c map. Again, the di haploid (D) pa rent w ill on average pro- duce 2.7 scorable marker loci per RAPD as- say, w hereas more th an 20 are necessary for the 5. verrucosum (V) parent. Therefore, the DNA marker analysis led us to conclude that the geneti cs of LB res istance in the VP popu- lation w ill be studied more effi c iently w ith a subsequent backcross in which the female marker loc i will segregate. DNA markers in diploid hybrid potato populations The segregation of DNA markers in hybrid popul ations has been analyzed separate ly for markers that derive from each parent fol low- ing a backcross model. All DNA markers have been treated as dominant markers in which each segregating band in the hybrid popula- tion is expected to have a 1: 1 segregation ratio for presence-absence. This model allows us to analyze the data with MapMaker v3 .0 soft- ware and to develop two geneti c linkage maps, one for each parent. VP. In the VP (5. verrucosum x 5. phureja) popul ation of 102 individuals, we scored sev- eral RAPD and AFLP markers that deri ve from both parents (Table 3). Chi-square (X2) tests CIP Program Report 1995·96 85 for goodness of fit to the monogenic ratio 1 :1 revealed a high degree of distortion of segre- gation from the female parent of VP. Only eight markers (22% of the fema le parent mark- ers) from the female parent showed unskewed Mendelian 1 :1 segregation. Of the 122 RAPD and AFLP markers scored in this population from the male parent, 84 (69%) did not deviate significantly from the expected monogenic ratio. Linkage analys is between the 84 unskewed markers deriving from the male parent led us to define five link- age groups (usin g the defau lt val ues of MapMaker). The highly distorted segregation for markers deri ving from the fema le parent prevents a genet ic ana lys is of the res istance coming from the 5. verrucosum parent. We plan to analyze this source of res istance in the offspr ing of the resistant hybrids. PD. We also analyzed LB resistance ge- net ica ll y on the other diploid potato popul a- tion, PD (5. phureja x di haploid 5. tuberosum). We are now buildi ng genetic maps for the parents using 164 RAPD and 151 AFLP mark- ers, and 16 microsatell ites (Table 3) on 94 PD hybrids. In contrast to the VP population, 81 % of the fema le parent markers and 68% of the male parent markers show Mendelian 1: 1 segregation in this interspecific hybrid prog- eny. Linkage analysis of the unskewed mark- ers, 118 female and 127 male, led us to de- f ine 13 linkages fo r the female map and 16 Table 2. Characterization of diploid potato populations : segregation of dominant markers in hybrid progenies, heterozygosity index of parents, and marker value assessed by RAPD markers. A: presence of the band; a: absence of the band Monomorphic, no segregation Parent Primers V Pl-116163 58 p (((81 58 P CHS-625 63 D PS-3 63 (no.) Monomorphic, segregation of the band in 3:1 Total Mono- Poly- bonds morphic morphic (no. of bonds) 301 11 3 188 353 113 240 550 382 168 586 382 204 a. Hi = heterozygosity index, Mv = marker value. 86 Program 2 Polymorphic, no segregation Locus Locus AA Ao (no.) 181 7 130 110 50 118 34 170 Polymorphic, segregation of the band in 1 :1 Hi 0 Mv ' 0.037 0.121 0.458 1.897 0.702 1.873 0.833 2.698 Table 3. DNA markers scored in the offspring of diploid potato interspecific crosses and association with late blight resistance. Population Markers Female parent Male parent Total Un skewed markers markers' VP RAPD 7 4 AFLP 30 4 PD RAPD 57 50 AFLP 82 62 Micro- 6 6 satellite a. x1 test at p < 0.01 . b. Student's I-test at P < 0.05. for the male map. As a result of thi s analys is and earli er phenotypi c eva luations, thi s PD popul ati on displays the expected features of a genetic material adequate for mapping hori- zontal res istance to LB. DNA markers for resistance to late blight We applied singl e-m arker analysis to de- tect linkages between genotypic classes (pres- ence o r absence of the band) and their re- specti ve phenotypic va lues (LB resistance vs. susceptibility) using a two-sample Student's t-test. Thi s test detects whether the means of the ph enotypic values for two genotypi c c lasses differ signifi cant ly. VP. We have identified 7 DNA markers out of a tota l of 92 unskew ed markers in the VP popul ation that are assoc iated (0.01 N s: 0 co r-- co (') l!l "' (') (') co (') "' e e :::;: a:: a:: a:: a:: a:: a:: c c -' -' -' -' -' -' 8 0 0 0 0 0 0 0 0 Figure 2. RAPD revealed intervarietal variation of 15 Bolivian cultivars by oligonucleotide primer 814. Lanes l and 19 are a l kb DNA ladder. Lanes 17 and 18 are controls. 94 Program 2 -- Group 12 Group 13 Gr.14 Gr.15 Gr.16 r-----1.--~~~--. • -. --- - - - - - - w - ·- $£· w .... , ..•. ,., ••'.v·· ············· .... ,,,.. ' -- -- I'- 0 co C1) s co (J) LO C\I "'" co C\I C1) (J) co co (J) 0 co co C1) 0 "'" C\I (J) I'- co co LO LO LO I'- LO I - I'- I'- co C1) I'- 0 C\I co LO co co C1) "'" C1) LO LO co < MHS49 ]~ {:. ABS54 11; J 8J ARCH5435J N ABS5515 i~ ABS552G MHS47 l '."? ARB55.J.t ] ~ AR854415 ]::: ARB544i A8S549i' ]~ A8S5374 ABSS493 ABS5373 ABS5523 ]~ A8S5372 ];". A8SS493 j~ A8S552'.l ABSS516 "' A8SS511 ., ABS5521 a. ASl88 i~ >-ARB5381 0 ARCHS436 ) ..- .c AACH5434 ~ a. :~~~ ]s 0 :2 M.~M5148 l AMM5161 en .,; AAB5535 0 A8S549! -5 ABS549-\ A8S54?5 8 A85549G ~ ABS54~ ABSSSOl 0 c: ABS5502 0 ABS5503 ·:;;: ABS5492 i ~ :: AM'.15!3J ] "' AMM5232 w c.... AB55514 l"' 0 ABSSS.1? ~ ABSSS.'.P. c: A8S5571 ·~ ABS5527 ARBS\0:; ]~ § ASll.JJ AR85181 0 A8SS524 ]M '° ABSSS\1 JN '° A8$551 2 = c: A8SS51 ~i ABSSS'XI J~ ~ ::::> E ~ = 0 c: "' ...c: c.... N ~ ... ~ O> i.i: 114 Program 2 Table 3. Latent root and percentage of total variation of the first 18 principal component oxes for the material studied. CIP, Limo , Peru, 1996. Principal Oco Ulluco Mashuo Arracacha component Latent Variation Cumulative Latent Variation Cumulative Latent Variation Cumulative Latent Va riation Cumulative root (%) variation root (no.) variation root (%) variation root (%) variation (no.) (%) (no.) (%) (no.) (%) (no.) (%) 1 3.76 15.7 15.7 5.27 26.4 26.4 3.40 18.9 24.1 24.1 18.9 6.74 2 2.88 12.0 27.7 2.45 12.2 38.6 3.09 17.2 15.4 39.5 36.1 4.32 3 2.33 9.7 37.4 1.77 9.8 47.4 2.38 13.2 11 .3 50.8 49.3 3.16 4 2.20 9.2 46.6 1.67 8.4 55.8 1.62 9.0 8.5 59.3 58.3 2.38 5 1.73 7.2 53.8 1.56 7.8 63.6 1.49 8.3 6.7 66.0 66.6 1.88 6 l.33 5.6 59.4 l.20 6.0 69.7 1.21 6.7 6.2 72.2 73.3 1.73 7 l.25 5.2 64 .6 0.96 4.8 74.5 0.94 5.2 4.8 770 78.5 l.34 8 1.07 4.4 69.0 0.88 4.4 78.9 0.85 4.7 4.1 81.1 83.3 1.16 9 l.02 4.3 73.3 0.74 3.7 82.6 0.72 4.0 3.5 84.6 87.3 0.99 10 0.93 3.9 77 .2 0.63 3.2 85.8 0.61 3.4 3.0 87.7 90.7 0.84 11 0.75 3.1 80.3 0.61 3. 1 88.8 0.42 2.3 2.9 90.5 93.0 0.80 12 0.70 2.9 83.2 0.59 3.0 91.8 0.36 2.0 2.1 92.7 95.0 0.60 13 0.66 2.7 86.0 0.42 2. 1 93.9 0.31 1.7 1.8 94.4 96.7 0.49 =;; 14 0.60 2.5 88.5 0.37 1.8 95.7 0.27 1.5 1.1 95.6 98.2 0.32 ,g a 15 0.58 2.4 90.9 0.32 1.6 97.3 0.13 0.7 1.1 96.7 99.0 0.30 3 16 0.51 2.1 93.0 0.22 1.1 98.4 0.10 0.6 0.9 97.5 99.5 0.24 { ~ 17 0.38 1.6 94.6 0.1 6 0.8 99.2 0.06 0.4 0.6 98.2 99.9 0.18 ~ ~ 18 0.30 1.3 95.9 0.06 0.3 99.6 0.02 0.1 0.6 98.8 100.0 0.17 ~ - -U'1 ~ 0- ~ 3 ~ Table 4. Major characters' associated with the first three principal component axes of 227 accessions of oca, 149 of ulluco, 64 of mashua, and 66 of arracacha. Peru, 1994-96. Crop Oca Ulluco Mashua Arracacha Principal component 1 Petiole color (0.719) Ped. and pedicel color (0. 711) Sec. tuber flesh color (0 .659) Stem color (0.620) Predom. tuber flesh co lor (-0.593) Sepal color (0.551) Sec. tuber flesh color (0.888) Distrib. sec. tepal color (0.818) Distrib. sec. tuber flesh color (0.798) Sec. tepal color (0.792) Predom. tuber skin color (0.714) Tepal base color (0.555) Distrib. sec. stem color (0.770) Predom. stem color (-0.7 68) Sec. stem color (0.7 45) Twining (-0.617) Distrib. sec. tuber fl esh co lor (0.527) Mature leaf color (-0.502) Acumen terminal leallet(-0.769) Sec. stor. root surf. color (O. 707) Leaflet margin color (0 .687) Sec. propag. fl esh color (0.681) Distrib. propag. flesh color (0.681) Predominant petiole co lor (0.651) Principal component 2 Principal component 3 Predom. abaxial leaflet color (-0.644) Sec. abaxial leaflet color (-0.714) Stem color (-0.633) Distrib. sec. abaxial leaflet color (-0.707) Sprout shape (-0.522) Distrib. sec. tuber flesh color (0.522) Ped . and pedicel color (-0.476) Sec. tuber flesh color (0.443) Petal color (-0.463) Depth of eyes (0.386) Distrib. sec. abaxial leaflet color (0.457) Sepal color (-0.382) Sec. tuber skin color (0.540) Plant type (-0.449) Distrib. sec. tuber skin color (0.516) Leaf shape (-0.431) Tuber shape (0.513) Fol iage color (-0.422) Tepal base color (0.465) Sec. tuber skin color (-0.400) Depth of eyes (-0. 441) Intensity tuber skin color (-0.399) Predom. tuber flesh color (0.438) Leaf size (0.355) Intensity tuber skin color (-0.729) Predom. tuber skin color (0.839) Sec. tuber flesh color (0.685) Sec. tuber skin color (-0. 641) Distrib. sec. tuber flesh color (0.681) Tuber shape (-0.450) Predom. tuber flesh color (-0.670) Predom. tuber flesh color (0.423) Sec. stem co lor (-0.531) Sec. tuber flesh color (0.421) Predom. stem color (0.451) Distrib. sec. tuber skin color (-0.407) Foliage color (-0.7 43) Stor. root shape (-0.613) Sec. abaxial leaflet color (-0.682) Predom. propag. surf. color (-0.542) Di strib. sec. aboxial leaflet color (-0 .670) Predom. abaxial leaflet color (-0.538) Sec. star. root flesh color (0.621) Wax on petiole (-0.487) Predom. petiole color (-0.600) Distrib. sec. petiole color (-0.471) Sec. propag. flesh color (0.548) Leaflet margin (0.469) a. Distrib. = di stributi on, ped. = pedu ncl e, predom. = predom inant, propog . = propogule, sec. = secondary, star. = storage, surf. = surface. secondary tuber fl esh co lor to the third prin- cipal component (Table 4), which accounted for 9.7% of the va riance (Table 3). For ulluco, tuber and tepa l characters con- tribute to the first and second principal com- ponents, whereas plant and tuber characters are the important delineating characters in the third component. Tuber characters such as shape, predominant skin color, secondary skin co lor, distribution of secondary ski n co lor, intensity of skin co lor, predominant fl esh color, and secondary flesh color were reported by Finnish scientists to be stab le characters . A number of Andean invest iga- tors have also been using ulluco tuber char- acters along with plant type, foliage co lor, leaf shape, and leaf size to study variat ion of ulluco in the ecoregion, but only some of them considered tepal features. Mashua. Tuber and plant characters are im- portant determining characters in the first, second, and third principal components in mashua. As in ulluco, the characters identi- fied for mashua in the present work have been used to study morphological variation of the plant in the Andes, but have not been ana- lyzed or interpreted appropriately. Arracacha. Leaves, propagu les, and stor- age-root characters were shown to be impor- t ant determinants of morphotyp es in arracacha. Leaf characters were the most highly va riabl e in Ecuadorian arracachas. Four leaf characte rs and 3 storage root char- acters were also used by scientists in Ecua- dor to identify 1 7 morphotypes of arracacha. Four leaf characters w ere also used to char- acteri ze arracachas morphologically in north- ern Peru. Conclusions Morphological characterization by grouping access ions according to their morphologica l simil arities and by c luster analysis proved to be accurate and re li ab le tools to identify morphotypes in ARTC co ll ect ions . Molecu- lar characteri zation in arracacha indicated that grouping and cluster analysis gave con- sistent results. Our results sugges t th at Andean gene banks can use visual grouping alone to iden- tify morphotypes and duplicates in their ARTC collections at relatively low cost. Cluster analysis, if it could be afforded, would con- firm grouping results. Molecular characteri za- tion could be used only at an advanced stage following visual grouping and cluster analy- sis. The grouping patterns of morphotypes in oca, ulluco, mashua, and arracacha show different geographical origin. Accessions from different agroeco logical zones of Peru were grouped by principal components into the morphotypes indicated above. Therefore, it is likely that ethnic groups were responsible for disseminating, maintaining, and using these crops in the Andean ecoregion. The use of qua I itative morphological char- acters (Table 4) resulted in an excellent reso- lution of the morphotypes for each crop. These characters could be the basis of de- scriptor li sts for oca, ulluco, mashua, and arracacha for morphologica l characterizat ion to optimize their handling and use in Andean gene banks. Selected Reading Cardenas, M. 1969. Manual de plantas econ6micas de Bolivi a. lmprenta lcthus, Cochabamba, Bolivia. 410 p. Le6n, J. 1964. Plantas alimenticias andinas. Boletin Tecn ico No. 6, lnstituto lnterameri cano de Ciencias Agrfcolas, Zona Andina. Lima, Peru. 112 p. National Research Council. 1989. Lost crops of the Incas: Little-known plants of the Andes with promise for worldwide cultivation. Nat ional Acad- emy Press, Washington, D.C. 415 p. CIP Program Report 1995-96 1 1 7 PROGRAM Disease Management Luis F. Salazar1 REPORT The major activ ities of the Disease Manage- ment Program during 1995 -96 included re- search on the control of late blight, bacterial wi It, and virus diseases of potato, sweetpotato, and Andean root and tuber crops (ARTC). The program's strategy is to combine basic and app li ed research, and this is carr ied out at headquarters or in collaboration with sc ien- tists in developed countries when required. Late Blight Late blight (LB) research saw major advances in the improvement of breeding populations (population B) that carry significant levels of non-race-spec ific and durable (horizontal) 1 Program Leader, CIP, Lima, Peru. Pro resistance to the causa l fungus Phytophthora infestans due to their minor genes of quanti- tative inheritance . Three sets of population B (Bl, B2, and B3) are now being tested for improved LB resistance. Population Bl was developed through several recombin at ion cyc les of resistance sources of 5olanum andigena. Population B2 was obtained from crosses between 5. andigena and 5. tuberosum sources of resistance. Population B3, the most advanced source of horizontal resistance available at CIP, was selected from popula- tion A, a previous LB-resistant population carrying both quantitative and qualitative types of resistance. B3 contains mostly 5. demissum-derived horizontal resistance im- proved mainly in an 5. tuberosum germplasm background. CIP P1ogram Repo1t 1995-96 l l 9 Corre lation studies co ndu cted on LB eva luation results from Mexico and Colom- bia have show n that 83 res istance is stable under those two di verse phys ica l env iron- ments and pathogen populations. Widening adaptation of the 83 population to long days in Argentina (at 38°S latitude) and China (at 40°N and 26°N latitude) is in progress. Add it ionally, extreme resistance to potato vi- rus X (PVX) and potato virus Y (PVY) from CIP genotypes that carry the genes for virus resis- tance in triplex (XXXxYYYy) is being introduced into popu lation 83. Further improvement ofB1 and 82 popu lations continues. A new group of LB-res istan t materials (popu lat ion C) is at the prebreeding stage of deve lopment at CIP-Lima. Thi s breed ing stock now comprises 186 d iploid potato hybrids that carry factors of res istance to foliage blight from 13 species of wild and nat ive cultivated potato. The prebreeding effo rt aims at introgress ing durable res istance to late b light into population C breed ing stock. Population C wi 11 be a core co ll ection of a w ide range of diverse resistances in ready-to-use potato breeding clones. CI P' s major emphas is in LB research was until recentl y on building resistance to the fungus for use by NARS in integrated disease management (IDM). With the Global Initia- tive on Late Bli ght (G ILB) now in operation, the IDM approach w ill receive high priority. Research on and app li cat ion of other IDM components such as the use of appropriate agronomic practices, use of c lean seed, or biologica l control w ill be encouraged. Bacterial Wilt Research on contro lling bacter ial w ilt (BW) of potato, caused by Ralstonia (Pseudomo- nas) so lanacearum, was particu lar ly ad- dressed to IDM components. The use of clean seed tubers, identifi ed by the enzyme- linked immunosorbent assay, has proved an effec- t ive BW control measure in Indonesia. Be- cause it has proven diffi cu lt to bu ild genetic res istance to BW into acceptab le genotypes, we reduced efforts on th is approach. Remain- 120 Progrom 3 ing acti v iti es conducted in collaboration w ith some national agricultural research systems (NA RS) were mainl y directed to eva luat ing BW resistance of promising genetic materi- als generated in CIP. Some SW-res istant po- tato clones, also ca rrying acceptable leve ls of res istance to LB, have been selected in In- donesia and China. In other studies on IDM components, pre- 1 iminary experim ents in China suggested good possibi l it ies of using bacteri al suspen- sions of two strains of Bacillus for the bio- logica l control of BW. Farmer partic ipatory experiments in Uganda showed t hat im- proved IDM packages that inc lude a res is- tant cultivar, clean seed, adequate planting space, rogu ing of vo lunteer potatoes, sanita- t ion, and minimum cultivat ion increased yie lds up to 75%. Virus Diseases New virus diseases were found to threaten potato production in some countries . Among these, ye llow ve in in Ecuador and Co lombia, Saq'o in Bo liv ia, deforming mosaic in Brazi l, and rough dwarf or potato v irus P (PVP) in Argentina and Braz il deserve mention. Yel- low ve in is by far the most important because of its effect on y ield. Al though the v irus has not yet been identified, ev idence indicates it is a vi rus with unusual characteristics. The putative vi rus is apparentl y preva lent in weeds and other crops and moves into the potato crop through its w hitefly vector Trialeurodes vaporariorum. Studies on Saq'o suggested the invo lvement of an unusual strain of potato leafro ll virus (PLRV) . Other yet-unidentifi ed phytop lasma- li ke microorganisms also appear to be impli cated in the Saq'o disease. The other vi ruses ment ioned are still under study in co ll aborat ion w ith sc ientists in Brazi l and Argentina. Deform ing mosa ic from Braz il appears different from that reported years ago from Argentina. It is probabl y a ge miniviru s, whereas rough dwarf in Argentina or PVP in Braz il is a ca rl av irus resembling the w ell - known potato v irus S (PVS), w hich ca uses severe symptomato logy in potato. Phytoplasmas, formerl y known as myco- plasmas, have also sp read in recent years. They have caused severe damage, parti cu- larl y to seed tuber production in Peru and Mexico. Because of the importance of these diseases, we have diverted limited funds to study these pathogens. Our major activities are directed toward developing sensitive de- tect ion technology for all these diseases. Because pathogens usually interact in na- ture in a positive or negati ve manner, we are studying these plant-pathogen interactions in deta il. An example of a negative interact ion having epidemiologica l significance is trans- mission of the potato spindle tuber v iro id (PSTVd) by aphids when PSTVd and PLRV co- infect potato plants. Molecular experiments showed the transmission to occur through encapsidation of the small PSTVd molecule in particles of the aphid-transmitted PLRV. Spread of the v iroid through encapsidat ion has occurred in potato crops in China. An important interaction among patho- gens of the positive type is the signifi cant re- duction in the rate of P infestans (LB) deve l- opment in potato plants infected w ith PVX, PVY, or PVS. Viru s- infected pl ants showed red uced penetration of zoospores and re- duced sporulation of the fungus, thus reduc- ing the number and size of les ions. With v i- ru ses such as PVS that have no significant effect on yield, this interaction will be further explored as a component of IDM. Molecu lar approaches are seen as an im- portant complement to traditional breed ing for res istance to v irus diseases. In co ll abora- tion w ith the Sainsbury Laboratory in the United Kingdom, mo lecular characteri zat ion of v irus res istance genes in potato was con- tinued. For the Rx gene (conferring extreme res istance to PVX), a BAC (bacteri al artificial chromosome) library for the cloning of Ryadg has already been developed and iso lati on of the Rx locus is under way. For Ry (conferring extreme resistance to PVY), molecular char- acterization of the Ry510 locus is in progress at Sainsbury. Virus-free p lanting materi als were eva lu- ated for the control of sweetpotato v irus dis- eases. In China, virus-free planting materi als outyielded farmers' seed by at least a facto r of 2. Efficient virus detection technology is essential to produce v irus-free planting ma- teri als. We are continuing virus identifi cation and characteri zation studies. Severa l new ly recorded v iru ses have been characteri zed. Studies on sweetpotato v irus disease (SPVD) suggested that one of the two components, the whitefl y-borne closterovirus (WBV), va r- ies in its interact ion w ith sweetpotato feath- ery mottl e virus (SPFMV), depending on the region where SPVD occurs. O ur prev ious hypothes is that control of SPFMV was sufficient to control SPVD was not borne out in Uganda, w here culti va rs highly res istant to SPFMV degenerate with SPVD. Apparently WBV breaks the res istance to SPFMV and SPVD develops. O ur new strat- egy ca lls for developing WBV res istance. If host-mediated res istance cannot be found in CIP sweetpotato germplasm or elsewhere, development of transgenic resi stance w ill be attempted using v irus-deri ved genes. Andean Root and Tuber Crops Research on ARTC diseases all owed us to deve lop a manual on ARTC di seases to be published soon. In addition to ARTC-spec ifi c v iruses, there are some important potato v i- ru ses that also attack ARTC. From an epide- miological point of v iew, these findings are imµortant for developing viral control mea- sures in both potato and ARTC. CIP Program Reporl l 995-96 1 21 Developing Horizontal Resistance to Late Blight in Potato J.A. Landeo1, M. Gastelo1, G. Forbes1, J.L Zapata2, and F.J. Flores3 Late blight (LB) d isease in potato, caused by Ph ytophthora infestans, is one of the major diseases researched at CIP since its founding. It is a cause of seri ous potato production losses, particul arl y in less deve loped coun- tr ies around the world. CIP sc ienti sts have al- ways be li eved that host res istance could play an important ro le in the management of the di sease, pa rt icularl y in CIP's client countries. Therefore, breeding fo r res istance has been a major endeavor since its earl y days . During the earl y 1980s, CIP resea rchers began to exploit their own breeding popu- lation assembled from 5olanum demissum- deri ved advanced sources of res istance in- troduced into 5. tuberosum subsp. tuberosum, Neotuberosum, and 5. tuberosum subsp. andigena germpl asm, and four-way hybrids betwee n 5. aca ule, 5. bulbocastanum, 5. phureja, and 5. tuberosum (ABPTs). Through a testing strategy that included Peru , Co lom- bi a, and Mexico, by the end of the decade som e 300 clon es w ith va riou s leve ls of hori zontal resi stance to LB and good agro- nomi c attributes, inc luding qu ali ty, w ere selected. They were made available to client countri es on all continents. From 1990 to 1996, more than 30 LB-re- sistant varieties were released in 15 countries (Table 1 ). A parti cul ar fea ture of th is breeding popul ation is that hori zontal res istance w as improved in the presence of undesired, un- know n major (R) genes for verti ca l res istance (popul ation A). Their presence, rather than contribu ting to the overall res istance, made the recognition of true hori zontal res istance 1 CIP, Lima, Peru. 2 Corporac i6n Reg ional del lnstituto Colombiano Agropecuario (CORPO ICA), Rionegro, Co lombia. 3 lnstituto Nacional de lnvestigaciones Foresta I es y Agropecuarias (I NIFAP) , Toluca, Mexico. 1 22 Program 3 and effective gene frequency upgrad ing more d ifficul t. Starting in 1990, fo ll owi ng a LB strategy rev iew at CIP, a new breeding strategy was des igned to upgrade horizontal res istance in the absence of R genes (popul ation B) . The absence of R genes, o r removing their inter- ference in assessi ng true hori zontal res istance, has already resulted in a more effi c ient breed- ing scheme. We are w itness ing an increase in gene frequenc ies and higher leve ls of re- sistance along w ith traits of agronomic va lue. The new strategy includes the develop- ment of three independent subpopul ati ons to broaden the geneti c di versity fo r res istance and maximize the use of hori zonta l res istance from cul t ivated germpl asm. The first popula- tion is deri ved from a w ide sample of nati ve cultivars from 5. tuberosum subsp. andigena (B 1 ). The second is deri ved from the same source of 5. andigena, but crossed only once to R-gene-free 5. tuberosum culti vars to im- prove some agronomic traits lacking in 5. tuberosum subsp. andigena (B2). The third , w hich is the most advanced agronomica ll y, is deri ved from popul ation A (B3). Prog ress in deve lop ing popul at ion B3 is repo rted here since this breeding popula- ti on rece ives major emphas is in CIP's re- sea rch program. It has many des irabl e ag- ronomic characters and good leve ls of hori- zontal res istance obtained from sources of popul at ion A . Quantifying Horizontal Resistance in 83 Materials and methods Th e ex perim enta l m ater ia l in c lud ed progenies obtained follow ing a line x tester Table 1 . Varieties released fro m late-bl ight-resista nt breed ing population A, 1990-96. Region Country CIP number Local name Latin America and Bolivia 385240.2 Chaposa the Caribbean Colombia 3821 19.20 Unnamed Costa Rica 386040.9 Birris 386056.7 Floresta Ecuador 384638.1 INIAP-Sta. Rita 388790.24 INIAP-Fripapa 382119.20 B INIAP-Rosita 388749.3 IN !AP-Margarita Guatema la 382170.101 ICTA Xalapan Panama 381381.13 IDI AP 92 381390.30 IDIAFR IT 382171.10 PRECODEPA Peru 380389.1 Canch6n-INIAA 384866.5 Amari lis-1 NIA 377744.l Kori- IN IA 380496.6 Chagll ina-INIA Venezuela 380013.2 Andina Sub-Saharan Burundi 381381.9 Rukinzo Africa 381381.26 lngabire 3821 47.18 Jubile Cameroon 381381.13 Cipira 381406.6 Tubira Ethiopia 378501.16 Sissay Kenya 381381 .13 Tigoni 381381.20 Asante Rwanda 381381.3 Nderera 381395. l Ngunda 382120. l 4 Kigega 383140.6 Mugogo 386003.2 Mizero 387233.24 Gikungu Uganda 381379.9 Kisoro 381381.20 Victoria Zaire 378699.2 Kinigi 380583.8 Baseko 380606.6 Enfula 386022.22 Nuru la CIP Program Report 1995-96 l 23 mating design of a samp le of R-gene-free clones . Three clones were used as male testers. Quarantine-produced tuber fam i I ies of the progen ies obta ined were arranged in incom- plete simple lattice des igns w ith two rep li ca- t ions and sent fo r testing under loca l LB epi- demics to two sites in Cajamarca, Peru, and one site each in Rionegro, Colombia, and To luca, Mex ico, during the growing season. Percentage of leaf area infected was recorded for 6 consecuti ve weeks and average appar- ent infection rates per fa mily were calculated to quantify hor izontal resistance. Likew ise, fa mil y average tota l tuber weight per plant was used for y ield. Ana lys is of va ri- ance was determined for the line x tester mat- ing des ign, and the source of var iance due to clones was used to ca lcu late the genera l com- bining abili ty va ri ance, w hich is associated w ith the add it ive portion of the tota l geneti c varia nce. At each of the three sites, narrow- sense heritability for res istance and yie ld was estimated fo ll owing stand ard procedures. Combined anal yses were not attempted be- ca use different line samp les and testers were used for Peru , Colombia, and Mexico. Results The mean sq uares for clones at all si tes were highly significant for resistance and yield. Combining ability vari ances, ca lcu lated from the sources of var iation of the clones, and her itab i I iti es ranged from 0.41 to 0.71 fo r res istance and from 0.24 to 0.51 for yield (Tab le 2). These estimates indi cate that both characters are at mid to high levels for a quan- titat ive trait. The va lu es may be somewhat infl ated, however, because of the contribution of the source genotype x environment in teract ion, w hich was not iso lated from the main source of genet ic va ri at ion. In this sa mpled popu la- t ion, additive geneti c vari ances have magni - tudes significa nt enough to ensure progress in breedi ng and se lection for outstandi ng LB- res istant c lones w ith attributes to become va ri et ies. Correlation Studies for Horizontal Resistance to Late Blight in B3 In the process of continu ing improvement of population B3 at CIP, correlation studi es for LB res istance were conducted in samples of thi s popul ation fo r two co nsecutive years (1995 and 1996) at two LB endemic sites w ith contrasti ng eco log ies and pathogen popu la- ti ons. The sites chosen were Toluca, Mex ico, and Rionegro, Colombi a. Materials and methods The sample tested in 1995 was composed of 60 fam i I ies from i ntercrosses between the first lot of R-gene-free parents (47 cl ones) and contai ned approx imately 6,000 indi v iduals. The sample tested in 1996 included another 60 families w ith approx imate ly 8,000 indi - v iduals. They were both tested for LB res is- tance in the fie ld at Rionegro and To luca, the samples being exact dupli cates at each site. Families in the 1995 samp le were not rep li- cated, w hereas families in the 1996 sample were rep li cated and arranged in the fi eld in an in compl ete simpl e lattice des ign. Six weekly read ings on percentage of fo li age in- fection were taken for every indi v idual dur- ing the season at both sites in 1995; fou r Table 2. Estimates of heritability for horizontal resistance to late blight and total tuber yield in a sample of clones free of R genes from population 83. Trait Late blight resistance Total tuber yield 124 Prog ram3 Peru Site 1 Site 2 0.71 0.25 0.56 0.52 Colombia 0.41 0.24 Mexico 0.43 0.26 weekly read ings w ere taken in 1996. The area under the di sease progress curve (AUDPC) was ca l cu lated and used as a parameter for resistance. Ave rages of AUDPC per famil y were used to rank fa mili es at both sites and Spearman's rank corre lation was performed. Results Th e combined analysi s of vari ance for the 1996 sample (CV = 17%) indi cated that the family source of vari ation was highl y si gnifi- cant w hereas the environment source of varia- tion was not. However, the family x env iron- ment interact ion w as also signifi ca nt. Th is indi cates that famil y performance for res is- tan ce vari es sign ifi cantly and that some may have performed di ffe rentiall y at each of the two sites. After ca lculating the correlat ion coeffi c ient, however, whi ch was also signifi- cant and quite high (r = 0.79 for 1995 and r = 0.82 fo r 1996), it seems that there is a linea r association between resistance performances of famili es at the two sites (F igure 1 ). The coeffi c ients of determination (R 2 = 0.61 for 1995 and R2 = 0. 66 for1996) indicate that the linea r assoc iation between the family per- formances at these two sites can be explained by about 60%. That leads us to conclude at this point that the express ion of hori zontal re- sistance to LB present in this improved source is rather stabl e and was not affected signifi - cantl y by divergent env ironments. The geno- type x env iro nment interaction , although present, may not be as high as earli er sus- pected according to our results. On the bas is of these resu Its, whereby test- ing was conducted under LB disease pressure of two endemic locations, we can conclude that hori zonta l res istance, under improvement in the B3 population, is in fact expressed ef- fectively in a wide range of the pathogen and under contrasting environments. To luca is noted fo r containing the most diverse patho- gen popul ations of P in festa ns, including both asexual and sexual stages of the fungus; in Rionegro the disease pressure is more uni- fo rm and higher than in Tolu ca throughout the season, but less diverse. Thus, either site ca n be used to effectively test and select for stable hor izonta l resistance to LB. Conclusions Class ica l breed ing approaches appli ed to population Bas a continuation of popul ation A improvement have been quite effi c ient in upgrad ing gene frequencies for hori zontal re- sistance to LB , together w ith important agro- nom ic tra its . B3, the most advanced source of res istance, is at the stage of being readil y used by breeding programs. It is also be'ng further improved at CIP through recurrent se lection w ith progeny testing. Furth er improveme nt is fo re see n in achieving higher levels of stab le resistance, broadening its genet ic diversity, increas ing adaptabi I ity to a wider range of day lengths, and gradual combinati on with other impor- tant di sease and pest res istance or tol erance. Marker-ass isted se lection, as it becomes ap- pli cable and practical , is contemplated to fur- ther increase accuracy and effic iency in the overal I process of se lect ion for hori zontal re- sistance to LB. Selected Reading CIP (International Potato Center). 1973-74. Program pl anning conferences. Report of the program pl anning conferences held at CIP during 1973 and 1974. Lima, Peru. 384 p. CIP (International Potato Center). 1989 . Funga l diseases of the potato. Report of the planning conference on fun ga l di seases of the potato held at CIP, Sep- tember 21-25, 1987. Lima, Peru . 216 p. CIP Program Report 1995-96 l 2 5 A AUDPC (Mexico) 1400 1200 1000 800 600 400 • 200 • 0 200 B AUDPC (Mexico) 2000 1800 1600 1400 1200 • • • • 400 • • • # ••• • ••• • • ••• • •• • • ·i· . • • • * • • .. ,. •• • 600 • I • • • • 800 1000 •• • • 1200 r = 0.79 r2 = 0.61 1400 AUDPC (Colombia) •• • ••• • • • • • • • . ... .. • • • •• ¥ .. # • • • 1000 •••• • ••• • 800 • 600 • 400 • 200 0 0 200 .. • • • 400 600 800 1000 AUDPC (Colombia) 1200 r = 0.82 r2 = 0.66 1400 Figure 1. Simple correlation of AUDPC from Colombia and Mexico, 1995 (A) and 1996 (B) . 126 Prog rnm3 1600 1600 Resistance to Late Blight from Diverse Wild Sources B. Trognitz, M. Eslava, L. Portal, and P. Ram6n 1 Widening the gene pool of potato (So lanum tuberosum) to enhance the crop's res istance through the use of w ild relatives has become a routine strategy. Thi s is espec iall y true for resistance to Phytophthora infesta ns, the causa l agent of late blight (LB), for wh ich re- sistance of the common potato is limited. Several w ild relati ves of potato have been used, but efforts are concentrated on onl y a few sources of res istance. Some of them have been used repeated ly over the past 50 years. Because many improvement programs are in developed countri es w ith moderate cli- mates, the new hybrid potato stocks devel- oped there may be of limited use in the tropi- ca l and subtrop ica l environments of devel- oping countri es. To more systemat ica ll y exploi t the large co ll ections of w ild potatoes maintained at CIP and in other potato gene banks, we invest i- gated several lesser-known and underused potential sou rces of res istance to LB. We in- cluded South American w ild species in the experiment because it was believed until only recentl y that these species would not possess major (R) genes fo r race-spec ific res istance to LB. The South American species, unlike several native species of Mexico, have been exposed to P infestans relati ve ly recently in their evo lutionary hi story. But ev idence is growing that R genes, each one of them con- ferring res istance aga inst only a few strains of the pathogen, also occur in potatoes in- digenous to South America. Therefore, we must also test our hybrids for the presence of R genes. 1 CIP, Lima, Peru. Materials and Methods Samples of botanical seeds of several acces- sions of diploid w ild and native cu ltivated po- tatoes were grown. The accessions had been selected, based on their known or suspected res istance, from co llections held at CIP; the Potato Introducti on Station, Sturgeon Bay, Wi sconsin, USA; and the potato gene bank at Gross Lusewitz, Germany. To detect pos- sible race-specific interactions, detached leaf- lets were placed in petri dishes and inocu- lated w ith iso lates of P infestans of different v irulences. A quick and reli ab le test of fo li age res is- tance under contro ll ed conditi ons in the greenhouse was developed at CIP-Lima. For thi s test, stems with fully developed fol iage are taken from adult plants and placed in milk bott les in a screenhouse equipped w ith mist irri gation to produce high air humidity and favorable conditions for blight development. A randomized two-block des ign is used, in which the experimental unit is a samp le of three stems of a given genotype in a single bottl e. The fo li age is inoculated once after sunset w ith a suspension of 15 ,000 sporan- gia/ml of a virulent iso late representing a po- tato (5. tuberosum) race of P infestans of a lineage common in the Andean reg ion of Peru, Ecuador, and Co lombia. A polycyclic blight ep idemic deve lops within 2 wk. Three read ings of the percent- age of fo li age area affected are taken at 2- day interva ls, starting when the susceptible check, Yungay, has an average of 30% fo li- age area affected. The average of the three readings over the two repetitions is then taken as a measure of the degree of resistance of a genotype. Data are analyzed by ANOVA for each group of clones assayed at a time. In CIP Progrom Report 1995-96 l 2 7 thi s way, successful identification of LB resis- tance is poss ible throughout the w inter pe- ri od in Lima, from June to November. Bes ides Yungay, four other checks wi th di f- ferent levels of res istance are always included in the assay. A lthough the overall degree of infection varies between indi v idual inocula- tion dates, the relative differences in res is- tance leve ls expressed by the checks are highly reprodu cibl e, w ith a correlat ion coef- fi c ient of r=0 .70-0.99 between pa irs of indi- vidual inoculat ion cycles. The average area of b lighted fol iage of the checks observed in f ive inoculati on cyc les in 1996 was (from highest to lowest leve l of res istance): CIP breed in g cl o ne 38 138 1.2 6 (re leased as lngab ire in eastern Afr ica), 17. 1 %; Canchan INIAA (released in Peru ), 21.2%; Atz imba (Mex ica n va ri ety), 34 .6%; Yun gay (Peru ), 48.S%; and Pimpernel (Netherl ands), S0 .6%. The ranking and the relati ve d iffe rences be- tween the va ri eties' scores observed in th is detached fo li age test are the sa me as those observed in fi eld experi ments. Therefore, our res istance test using detached fo liage is re- produc ible and gives a good est imate of the leve l of fi e ld res istance of th e genotypes tested. Detached leafl et assays revea led that clone 38 138 1.26 does not sporul ate after inocul a- tion w ith the iso late used for the screen house inoculations. Thi s ind icates that 38 138 1.26 possesses spec ific (R-gene-med iated) res is- tance to thi s iso late. We concl ude therefo re th at R gene res istance is equi va lent to read- ings of an average of 17% affected foliage area in the tri al used. Therefore, the direct se lection of highl y res istant genotypes that do not express R gene res istance, by rely ing on the phenotype of a sing le c lo ne in th e screenhouse or f ield, appears to be impos- sible. Selected LB-res istant clones of access ions of di p lo id w i ld spec ies were c rossed w ith potato d i haplo ids. The resulting hybrids have also been subjected to a seri es of tests for the level and race-spec ific or -nonspec ifi c resi s- tance inherent in them, as described above. 128 Prog ram 3 Results O ne hundred an d three ind iv idu als of 28 single-cross progeni es were selected in Lima, and 70 ind ividuals of l S progeni es in Quito, fo r their high levels of fo li age res istance to LB . These se lected hybri ds incl ude the res is- tances of 16 accessions of 13 wi Id and na- t ive cu ltivated 5olanum species (Tab le 1 ). These hybrids are all d iploid, and most of them produ ce unreduced po ll en. Thi s w ill a ll ow th em to be used in mei o ti c tetrapl o idi zation through po llinat ions of tet- ra pl o id va ri et ies and breeding clones . For se lected clones, the mechanism of fo rmat ion of unredu ced po ll en w ill be elu c id ated through cyto logica l means. In addi t ion, we used 13 clones w ith high leve ls of quant itati ve LB res istance, w hi ch were se lected from a set of diplo id hybrids of the fo rmer bacteri al w ilt breed ing program at CIP, in crosses w ith potato varieti es to obtain tetrapl o id hybri ds. The tetraplo idized geno- types were propagated c lonally and tested fo r LB res istance. We have ident if ied severa l hi ghl y LB-res istant, ea rl y to medium-l ate maturing genotypes express ing agronomica lly acceptab le tu ber characteri sti cs . Segregation of res istance in ind iv idua l di p- lo id hybrid populat ions was analyzed using the data of both the detached-leafl et and de- tached-fo li age assays. In the inoculations of detached leaflets, sporu lation was taken as a criteri on fo r compatibili ty w ith th e iso late used. Several hybrid popul ations sc reened expressed sporul ation on leaflets of all geno- types w ith four iso lates used. O ne iso late had no av i ru lence genes; the others possessed different comb inati ons of av irulence genes interacting w ith the 11 known R genes of 5. demissum. For a summary, see Tabl e 1. Segregation of the level of resistance in re- lat ion to the resistant check in a progeny of thi s race-non-spec ifi c react ion type is given in Figure 1. The segregat ion in the progeny of t he cross of CIP dih ap lo id PSS x 5. ambos inum, OC H l l 86S, c lo ne 24 , i s unimodal and f its a normal di stributi on. A progeny obtained from a se lf of di haploid PSS also showed a pattern of segregation th at f it a normal distribution in the screen house test. All indi vidua ls expressed sporu lat ion after inocu lation of detached leafl ets with a ll iso- I ates used, suggest ing that PSS does not pos- sess race-spec ifi c res istance all e les for any of these isolates. Table 1. Accessions of wild and cultivated potato containing high levels of resistance to late blight that are used for introgression of resistance into potato. Solanum species Accession code, collector Resistant Resistance Race-specific number plants used {no.) level reaction observed acroglossum CIP 761070, OCH 11297 2 Medium Yes? 0 ambosinum OCH 11865 2 Medium No berthaultii CIP selection, KV Raman High Yes albarnazii CIP 761164, OS 11007 Medium No chiquidenum CIP 7 61588, OCH 13345 Medium Yes? chomataphilum CIP 761582, OCH 13325 Medium Yes? laxissimum CIP 761028, 0/S 11855 Medium No micradantum Pl 500041 High Yes phureja INIAP, Ecuador, BOM 540 High No phureja IN IAP, Ecuador, CHS 62 5 High No phureja Pl320376 2 High No phureja Pl225678 High No piurae CIP 761072, 0/S 11615 2 High Yes? paucissectum CIP 761243, OCH 11630 2 High Yes? santolallae CIP 761691 , OCH 13640 Med ium Yes urubambae OCH 11059 Medium No a. The question mark indicates that race-specific interactions observed in preliminary experiments must be confirmed in subsequent studies. A precondition is the identification of differential isolates for these materials. Individuals (no.) 8 6 4 2 0 - I 2 3 4 5 Level of resistance (1 = level of the resistant check, clone 381381.26; 5 = low level) Figure 1. Histogram showing the segregation of resistance to late blight in a progeny of 18 individuals of the cross of dihoploid PSS x S. amhosinum, OCH 11865, clone 24. CIP Program Report 1995-96 l 2 9 Several hybrid progenies, howeve r, had a percentage of indiv iduals that reacted with hype rse nsiti v it y o r no t at a ll t o the inocul ati ons of detached lea fl ets w ith a parti cular isolate. These popul ati ons also expres sed segregati o n of th e deg ree of res istance, as measured in the foli age assay in the scree nhouse . An exa mpl e is the progeny of the cro ss of d ihapl o id PSS x 5. berth au/tii, clon e 1 (prov id ed by K. V. Raman, CIP, 1991 ), shown in Fi gure 2. Of 44 indi v iduals tested in the sc reenhouse, 22 either had no infect ion or up to twice the affected foli age of the res istant check (clon e 38 1381.2 6), res istance leve ls 1 and 2 in Fi gure 2. Th e sa me resistant indi v idu als did not sporul ate after inocul ati ons of leafl ets with thr ee indi v idu al i so lates . So me o f the susceptibl e indi v iduals (Figure 2, res istance levels 3-7) w ere tes ted in th e detached- leafl et assay. All showed sporul ati on with at least two of the three iso lates used. The obse rved segregati o n into res istant and susceptibl e phenotypes perfectl y fits a 1: 1 rati o indi cati ve of a single gene. However, the frequency di stributi on shown in Fi gure 2 is not bimodal , as would be expected for segrega tion of a sin g le facto r, but is unimodal and skewed. Therefo re, a model of seg regation of several dominant, race- spec i f i c resistan ce ge nes , eac h on e Individuals (no.) co ntri b ut in g a di stinct leve l of fo li age res istance, ca n also be envisaged. Further analys is of seg regati on in F? o r backcross popul ations deri ved from thi-s 5. berthaultii hybrid progeny is necessary to elu c idate th e number of putati ve R genes in vo lved and th eir type o f ac ti o n. It is poss ibl e th at th ese genes co ntribute a durab le, res idu al effect to the res ista nce, even w hen they are broken down. Seg rega ti o n into di stin ct res i stance c lasses w as obse rved in progeni es ca rry ing th e res istances o f the South A meri ca n spec ies 5. acroglossum, 5. berthaultii, 5. c hiquide num, 5 . chom atophilum, 5 . microdontum , 5. pau cissec tum , and 5. santolal/ae (Tabl e 1 ). Hybrids into whi ch the race-spec ifi c res istances of th ese South Am eri ca n spec i es w ere in co rp o ra ted expressed interm ediate to hi gh leve ls of res istance in the screenh ouse test. Of th e spec ies contributing race-n on- spec ifi c res istance, comparabl y high leve ls of res istance of hybrids were obtained only w ith 5. phureja. This native culti va ted po tato reacts to infection by droppin g di seased leaves throu gh absc iss ion at th e peti o l e base . A s such it represe nts a res i stance phenotype di ffe rent fr om 5. tub ero sum , in w hich di seased fo li age 20..---~~~~~~~~~~~~~~~~~~~~~~~~~~~~---. 15 -t--~-~--tt·:~&Jl--~~--------------~~~~--1 10 -·1-- ----f· 5 -r- ----; O+-_.__"""".._-+-_._~_._--+__.......__._---;~..___.~+--"'-~...._-+----"'""-"'u.--+---'~-"'---1 2 3 4 5 6 Level of resistance (1 = level of the resistant check, clone 381381.26; 7 =low level) 7 Figure 2. Histogram showing the segregation of resistance to late blight in a progeny of 44 individuals of the cross of dihaploid PSS x S. berthaultii, clone l . 1 30 Progrom 3 remains on the plant. Abscission of infected leaves was observed in the field and on detached foliage tested in the screenhouse. This resistance type is also expressed by 5. urubambae, and it is transmitted to the hybrids. Conclusions The total of 186 selected diploid hybrids described above represents the basic pool of LB resistance genes of a wide range of genetic resources. We are currently using this cohort of selected clones to develop tetraploids in i nterploidy crosses with advanced potato breeding clones. The tetraploid hybrids will then complement a set of progenitors- Cl P's group C or population C-that carry the new resistances. We envisage group C as a core collection of a wide range of diverse resistances, in a ready-to-use form of potato clones that are acceptable to breeders because of their improved agronomic characters. The hybrids will also be used for studies of the genetics and inheritance of resistance to LB. Segregating populations for genetic mapping of resistance factors will be obtained through F2 or backcrosses for efficient intro- gression of these factors into potato breeding stocks and to isolate candidate resistance genes for plant molecular transformation. Despite our observation that one-half of the wild potato species investigated possess race-specific resistance conferred by R genes, we will explore further the feasibi I ity of breed- ing using R-gene-free materials. Nevertheless, wild species may harbor stable R genes that could provide durable protection against LB. We will survey the genetic material studied in this paper for its resistance genes through inoculations with Phytophthora isolates that represent a wide range of virulences and hosts. Selected Reading Toxopeus, H.J. 1964. Treasure-digging for blight resistance in potatoes. Euphytica 13:206-222. Ross, H. 1966. The use of wild Solanum species in German potato breeding of the past and today. Am. Potato J. 43:63-80. Ross, H. 1986. Potato breeding: Problems and perspectives. Advances in Plant Breeding, 13. Paul Parey, Berlin and Hamburg. 132 p. CIP Program Report 1995·96 l 31 Evaluation of Late Blight Resistance in Populations of Diploid Potato Hybrids for Genetic Mapping B. Trognitz, M. Ghislain, G. Forbes, P. Oyarzun, M. Eslava, R. Herrera, L. Portal, P. Ramon, and G. Chac6n1 Wild and cultivated relatives of the potato (5o lanum spp .) carry valuable resistances to late blight (LB) that, when introgressed into potato, are thought to cons iderab ly reduce the crop's vulnerabi lity to this devastating dis- ease. To increase the durability of resistance to LB, researchers seek forms of resistance that are effective against a broad range of patho- gen ic strains of Phytophthora infestans. Poly- gen ic, additive resistance is the most prom is- ing. This resistance can be masked by mono- genic, race-specific major (R) gene-mediated resistance when no races compat ible w ith such an R gene are available. This is most important when the identity of R genes is not known. Twelve R genes for res istance to LB have been identified. Differential potato clones to distinguish 11 of them are ava il ab le. These 11 R genes originated in Mexican wi Id 5. demissum, and this and other Central Ameri - can potato relati ves are assumed to have de- veloped many more R genes. The Mexican wi ld potato 5. verrucosum is reported to pos- sess high leve ls of quantitative res istance be- sides R gene-mediated resistance. The suc- cessful use in resistance breeding of this and other wi ld and native cu lti vated potatoes therefore depends on knowledge of the oc- currence and identity of R genes. One strategy to avoid interference with R gene resistance cou ld be to choose genotypes w ithout R genes. Previously, it was hypoth- 1 CIP, Lima, Peru. l 32 Program 3 es ized that potato relatives indigenous to South America would not possess R genes; all resistance expressed by them wou ld be polygenic and quantitative. However, ev i- dence is emerging that many South Ameri- can species also emp loy R genes. Therefore, it is desirable to test every potentia l source of resistance fo r the occu rrence of race-spec ifi c resistance genes, and it is reasonable to ex- pect that additional R genes are yet to be iden- tified. Two sources of resistance to LB were investi gated in two segregating populations of diploid potato hybrids. One population was a cross between highly resistant 5. ve rru cosum and susceptib le 5. phureja, designated population VP. The other was a cross between resistant 5. phureja and a susceptib le 5. tuberosum dihaploid , designated population PD. The PD popu lation was selected for the study from nine diploid hybr id populations carrying high levels of resistance to LB . This invest igation had two objectives. One was to characterize the level of resistance of every individual of the two populations as a precondition for ge netic mapping. Th e second was to test the popu lations for the occurrence and segregation of race-specific R genes. Population VP This population comprises 102 indi viduals, all of them late-m aturing under the short daylength of Cajamarca, Peru . All genotypes form sma ll, pear-shaped tubers of creamy- w hite fl esh that sprout earl y. The entire popu- lat ion expresses a cytopl asmic male sterility (CMS) phenotype that is known as eclipse ste- rili ty. Although preliminary in trapopulat ion crosses were unsuccessful , it may be poss ible to use the VP indiv iduals as females in crosses w ith poll en-fertil e genotypes because female ferti I ity is not affected by CM S. VP field resistance VP indi v iduals were eva luated fo r res is- tance in the fi eld at Cajamarca, Peru, under high infection pressure, in 1994, 1995, and 1996. The parents did not grow in the fi eld or were not ava il abl e and could not be in- c luded in th e experim ent. A randomized block des ign w as used and the area under the di sease progress curve (AUDPC) w as ca l- cul ated from weekly readings of the percent- age of di seased fo li age in pl ots of 10 plants per cl one and bl ock. The VP population had a high average fi eld res istance (progeny mean, AUDPC=303, range 168- 999; res istant stan- da rd Perri cho l i, AU DPC=41 O; susceptibl e stand ard Yun gay, AUDPC= 997). Its fre- quency di stribution (Fi gure 1) devi ated from the desired normal distributi on typi ca l for a quantitati ve trait. Individuals (no.) 50 40 30 20 Race-specific resistance of VP We perfo rmed detached-l ea fl et tests to an alyze the segregation of the di screte trait res istance to sporul ati on. The phenotype of R gene res istance in this test is a hypersensi- tive response of leaf tissue. Sometimes no symptoms or only weak infection is observed. Sporulation occurs rarely. We cl ass ified in- di viduals as susceptible or resi stant based on this va ri abili ty of the expression of res istance. Indi viduals on w hich P. infestans sporul ated were cl ass ified as susceptible; those that did not all ow the pathogen to sporul ate in any of the repetitions were considered res istant. Ninety-nine indi v iduals were inocul ated with five iso lates of P. infestans, each possess- ing a different level of virulence (Table 1 ). The experimental unit w as a petri di sh contain- ing four lateral leafl ets from di ffe rent full y deve loped top leaves of one to three pl ants in bud or fl owering. The pl ants were grown in pots in a greenhouse at Lima during the winter of 1995 . Humidity was maintained by adding a sheet of mo ist filter paper to each petri di sh. Readings of sporul ati on and area affected were taken 5 dafter inoculati on, or wh en th e control s showed th e expected symptoms. Tests were repeated one to three 125 213 300 387 475 562 649 737 824 911 More AUDPC value Figure 1. Population VP: frequency distribution of AUDPC measured on 99 individuals in 3-yr field trial, Cajamarca, Peru. CIP P10gram Report 1995-96 l 3 3 Table 1. Inoculations of detached leaflets of VP individuals with 5 isolates of Phytophthora infestans. Isolate Virulence Resistant progenies (no.) 275 36 P2 l.3.4.6.7.10. ll 30 50 1.3.4.7. ll 15 8 0 260 l.2.3.4.6.7.9.10.11 ° 0 a. Not tested against RB. times and the accuracy of a test resu lt was estab lished by comparing it w ith the reaction of resistant and susceptible control cu ltiva rs. The inoculum was applied w ith a spray bottle. The inoculum concentrat ion was 5,000-15,000 sporangia/ml washed from mycelium grown on tuber slices. Thi s high inoculum concentrati on was chosen to en- sure that all susceptible individuals became infected. Of 99 plants tested, none was resistant to isolates 8 and 260. With the remaining three races, segregation into res i sta nts and suscept ibles was obtained (Tabl e 1). The ra- tios of resistant:susceptible individuals ob- tained with either race significantly (P<0.001) diverge from ratios expected for a model of single dominant genes of res istance. More feasible models were the complementary ac- tion of two or more res istance genes, or a re- sistance gene and a suppressor gene. In doz- ens of inoculations done over 5 yr, our con- trols ca rrying R genes always responded re- sistant to the respective av irul ent iso lates. None developed sporulating mycel ium. Pos- sibl y the resistance genes of 5. verrucosum break down under particular env ironments, thus resu lting in an excess of susceptibles. Segregation results must be confirmed by test- ing the V and P parents and by analyz ing l 3 4 Progrom 3 Susceptible Segregation Segregation progenies observed expected for single- (no.) gene resistance 63 l:l .75 l:l 69 1:2.3 l:3 84 1:5.6 l :3 99 0:1 l:l 99 0:1 0:1 backcrosses to the susceptible 5. phureja par- ent. A host-pathogen interaction scheme is pre- sented in Tab le 2. The eight res istance pat- terns observed in the population indi cate that the v irulences of the isolates 275 , P2, and 50 are different from each other. At least three facto rs of race-specific res istance mu st be assumed to segregate independentl y in the VP population. Iso lates 8 and 275 share the av rl gene for v irul ence o n potato (5. demissum) gene R 1 (Table 1 ), yet they have differing patterns of compatibility w ith the VP individuals. That result indi cates that iso late 8 possesses more virulences either to R gene 8 (for w hich no differential was ava il able) or to some other unknown res istance genes. After 3 y r of res istance testing in the field, th e AUDPC means of genotypes dif- fering in their res istance to 0, 1, 2, or 3 iso lates were co mpared by a series oft- tests (Fi gure 2) . None of the mea ns cou ld be c lea rl y separated. But the tendency of sma ll er AUDPC val ues to be associated w ith a hi gher number of iso late-specif ic resista nces indi ca tes a small but favorable residual effect of these fa cto rs on the ex - pression of resistance in the field. A lso , the m ea n AU DPC va lue (mean AUDPC=2 80) of all 36 indi v iduals that were Table 2. Response' of 99 VP plants to inoculation of detached leaflets with five isolates of Phytophthora infestans. Isolate Individuals resistant 275 P2 50 8 260 (no.) s s s s s 43 susceptible R s s s s 20 s R s s s 13 s s R s s 4 R R s s s 8 s R R s s 3 R s R s s 2 R R R s s 6 a. R = resistant to development of sporulating mycelium, S = susceptible. AUDPC 350 300- 250 .. 200- 100 50 0 I ,. - I Individuals (no.) 43 6 0 2 3 Resistant to no. of isolates Figure 2. Mean AUDPC values of classes of VP individuals resistant to 0-3 isolates. resistant to isolate 275 was smaller than that of the individuals susceptible to that iso- late (mean AUDPC=317, 63 individuals). Population PD PD is the result of a seedling family selection process. A sample of accessions of wild and native cultivated potatoes as candidate sources of resistance to LB was screened at CIP-Quito, in 1992 and 1993. Resistant clonal selections were crossed with potato dihaploids to produce diploid hybrid prog- enies. A sample of nine early-maturing prog- enies was subjected to a 2-yr field trial for resistance. The cross CHS-625 x PS-3 per- formed best of all progenies. Its progeny mean of resistance (AUDPC=l 52, 50 individuals) CIP Program Report 1995·96 135 was, of all nine progenies, closest to the va lue of the res istant sta ndard , va ri ety Cata lina (AUDPC=98). The indi viduals had AUDPC va lues between 65 and 240, d isp lay ing a norma l distribution. Twenty percent of the indi v idu als had the same resi stance as the standard, or a higher leve l. All individuals of this progeny are male- and female-fertile and 65% of them produce more than 2% unre- duced po ll en grain s. This progeny also has smooth round and ova l-shaped tubers with shallow eyes and ye llow skin. The ye llow- fl eshed sta rchy tubers have good cu Ii nary quality. The cross of the parents was repeated in Peru to produce the PD mappi ng popula- tion . All individuals of population PD are male- and female-fertile and flower profuse ly in the greenhouse at CIP-Huancayo, Peru . Plants are vigorous, but seem to be vulnerab le to infec- tion by mosaic v iruses-a feature frequently observed in w i Id and native potatoes. Intro- gression of v irus res istance in subseq uent crossing generations w i 11 be necessary. PD resistance in a controlled-environment test PD plants and their parents were grown in pots in the greenhouse at Lima, during the 1996 w inter season. Stems with comp lete foliage were used in a resi stance trial in a screenhouse equipped w ith mist irri gat ion to constantl y maintain high relative humidity Individuals (no.) so ~-- 40 ~ and low temperatu re. A randomized 2-block design was used. The experimental unit was a milk bott le containing th ree stems of a PD genotype. The plants were inoculated at night w ith iso late 260 (complex v irulence to 5. demissum R genes, see Table 1 ), at a concen- tration of 13,000 sporangia/m l. The ep idemic developed after 4-7 d and visua l readings of the percentage of diseased fo li age were taken three times at 2-d intervals. The experiment was repeated after 4 wk, and a two-factor ANOVA (facto rs PD genotype and bl ock, nested in repetition) was run on the mean percentage of diseased foliage ca lcu lated over the three read ings. There were significant differences between the PD indiv iduals, al- though no groups could be separated by multiple comparisons of means. Based on the conditi ons of the screening fac ility, the overa ll leve l of disease was different for each block w ithin each repet iti on. Figure 3 shows the frequency distribution of diseased fo li age fo r the PD individuals, wh ich fits a norma l distribution. The overall leve l of resistance of the PD popul ation (27% diseased foliage) observed was higher th an that of the resistant control, Canchan (4%). But it was much lower than that of the sus- ceptib le check, Yungay (33%) . Overall, high leve ls of resistance that vary graduall y between sister indi v iduals depi ct the im age of tru e quantitat ive resistance p D 1-- 1.7 5.4 9.2 13 16.8 20.6 24.3 28.1 31 .9 35.7 39.4 43.2 47 50.8 54.6 More Diseased foliage area (%) Figure 3. Population PD: frequency distribution of diseased foliage area(%) measured on 240 individuals and the P and D parents in 2 repetitions, Lima, 1996. 136 Program 3 caused by many additive genes. Thus, the PD population is useful for molecular mapping of quantitati ve trait loc i for resistance. More re- sistance studies must be done under actual field cond iti ons and in different environments to elucidate the stability and sustainability of the res istance detected. Testing for R genes in population PD Most of the PD indi v iduals developed foliage damage of more than 15-20% after inocul at ion with a virulent iso late of P. infestans (F igure 3). A sma ll fraction of in- div iduals exhibited less than 10% damage and preponderantly nec rotic les ions. There- fore, it is preferable to test th is population for the segregation of putative, isolate-spe- c ifi c res istance as well. Two repetitions of a test for sporu lation on detached leafl ets of 218 PD individuals were done using the isolate P2 (Table 1 ). Of 218 indiv iduals tested, 9 did not sporul ate in one of the two repet itions, and 13 did not sporu- late at all . No necroti c les ions developed, and the size of the lesions that did develop on these nonsporulating leafl ets was similar to that on leafl ets of other individua ls on which the pathogen sporulated. Further tests w ith this and other iso lates of P. infestans on the PD indiv iduals and on progenies from crosses of sisters and backcrosses to the D parent are necessary to elucidate whether race-spec ific resistance is also segregating in thi s quantita- ti ve ly LB-res istant material. Conclusions Populat ion VP expressed high leve ls of res is- tance in the field, with little va riati on among genotypes . Besides the population 's overall field res istance, race-specifi c interact ion of individuals was detected in detached-l eafl et assays. These race-spec ifi c res istances had a small favorable effect on res istance in the fi eld . We will use male-steril e individuals of the VP population as females in backcrosses with the susceptibl e P parent and with a di haploid tester clone to obtain advanced backcross populations for further anal ys is of the resis- tance and use in breedi ng. Population PD had a wide range of gradu- ally different res istance levels in a repeated control led-environment, foliage res istance assay, thus allowing the separation of most- res istant from most-susceptible individuals by statisti ca l means. High leve ls of quantitative res istance to LB as we ll as good fertility and the production of unreduced (2x) gametes, earliness, ye llow tuber fl esh, and other good agronomic characteri st ics of the PD popula- tion make it a valuable material for breeding. lntrogress ion of res ista nce to LB from this populat ion into culti vated potato appea rs to be possible. To carry out this introgression efficiently, we w ill need to genetically map quantitative res istance loc i as a precondition to marker-assisted se lection . Selected Reading Malco lmson, J.F. and W. Black. 1966. New R-genes in Solanum demissum Lindi. and their complementary races of Phytophthora infestans (Mont.) de Bary. Euphytica 15: 199-203. Ramanna, M.S. and J.G.Th. Hermsen. 1974. Unilateral "ecl ipse sterility" in reciprocal crosses between Solanum verrucosum Schlechtd. and diploid 5. tuberosum L. Euphytica 23 :41 7-421. Stewart, H.E. 1990. Effect of pl ant age and inoculum concentrat ion on express ion of major gene resi stance to Phytophthora infestans in detached potato leafl ets. Mycol. Res. 94:823 -82 6. Trognitz, B.R. , G. Chacon, H. Pinedo, and M. Eslava. 1995. Screening for R genes causing race-spec ifi c resistance to late bli ght in wild potato spec ies . Am. Potato J. 72:662. CIP Program Report 1995-96 l 3 7 Host Specificity of Late Blight Pathogen on Potato and Tomato in Ecuador G.A. Forbes, P.J. Oyarzun, A. Pozo, and M.E. Ordofiez1 In Ecuador, potato and tomato are cultivated year-round. Late blight caused b y Phytophthora infestans occurs in potato and tomato at any stage of plant growth after emer- gence, because of the continuous presence of inoculum. Potato and tomato are not gen- era ll y grown in the sa me area, but produc- tion zones for the two crops ca n be w ithin a few kilometers of each other or even contigu- ous. Occas ionally, we have seen potato and tomato grown on the same farm , and on rare occasions in the same field. Host spec ificity has important impli cations for management of the disease, espec ial ly in areas where both crops are grown in the same vic inity. Farm- ers need to know w hether their crops are threatened by nearby alternative hosts that are badly infected. Researchers and extensionists working on potato and tomato must coordi- nate efforts if the pathogen is equa ll y aggres- sive to both hosts. Thi s study was begun to eluc idate two as- pects of P infestans host specificity: w hether potatoes and tomatoes are attacked by the same population of P infestans in Ecuador, and what rol e specific v irulence plays in de- termining host specific ity. Materials and Methods Several markers were used to characterize both potato and tomato populations. A total of 120 iso lates, 60 from potato and 60 from tomato, were co llected from central and northern Ecuador, between November 1994 and January 1996. Approximately equal numbers were taken from the two zones. 1 CIP, Quito, Ecuador. 138 Program3 All iso lates were inoculated on 11 potato differenti al plants co ntaining 1 each of 11 known R genes fo r res istance and 4 tomato differential cultivars. Detached-leaf tests of aggress iveness were performed on 3 potato culti va rs and 3 tomato cultivars using 7 iso- lates from each host. Iso late res istance was assessed on 10% V8 agar amended wi th 5 and 100 ppm meta laxyl. The matin g type test was performed by pairing iso lates w ith known A 1 and A2 geno- types on clarified rye A aga r. The presence or absence of oospores was recorded after 15 d. Mitochondrial hap lotype was determined w ith the polymerase chain reaction procedure using primers deve loped and provided by G.W. Griffith, Univers ity of Bangor, Bangor, Wa les, U.K. Iso lates were cultivated in still culture of pea broth or on rye B agar for 7 d and gels were sta in ed for g lucose - 6-ph osp hate isomerase (G PI) and peptidase (PEP) . Eleven randomly selected isolates from tomato were analyzed for restri ction fragment length poly- morphi sm (RFLP) fingerprints using the mod- erate ly repetiti ve probe RG 57 developed in the laboratory of W.E. Fry at Cornell Univer- sity. Results Genetic description of P. infestans All iso lates from potato had the al lozyme genotype of 96/ 1 00 fo r Gpi and 100/ 100 for Pep (Tab le 1 ), and the mitochondrial DNA hap lotype of llA. Thi s multi locus genotype is identica l to that found for the lineage EC-1 , w hich was described as the dominant lineage on potato in Ecuador. Thirty genotypes were fingerprinted w ith probe RG57. Of these, 28 Table 1. Genotype characterization of populations of Phytophthora infestans on potato and tomato in Ecuador. Host Number Gpi " Pepb RFLP fingerprint' Lineage Tomato l 86/l 00 92/l 00 101010QOQ10011010 US-1 Tomato 2 86/ l 00 92/100 10101010Ql0011010 US-1 Tomato 8 86/ 100 92/ 100 10101010110011010 US-ld Tomato 48 86/l 00 92/100 Nottested US-1? Tomato 96/ l 00 100/ 100 Not tested EC-1? Potato 2 96/100 100/ 100 lllllOlOOQOOllOlO EC-1 Potato 28 96/100 100/100 11111010010011010 EC-1 d Potato 30 96/ 100 100/100 Not tested EC-1? a. Glucose-6-phosphote isomerose. b. Peptidase. c. Underlined bonds indicate polymorphism within lineages. d. Indicates published RFLP and isozyme genotype, which define that lineage. had the fingerprint of EC-1 and the fingerprints of the other 2 were similar to EC- 1, but lacked band 10. A ll potato iso lates were llA mito- chondrial hap lotypes and A 1 mating type. A ll iso lates from tomato, except for one, had the allozyme genotype of 86/100 for Cpi and 92/100 for Pep (Tab le 1 ), which is defini- ti ve fo r the globa lly distributed lineage US-1. Eight of 10 isolates fingerprinted had the US- 1 genotype. Two were simil ar, but lacked band 9, and one iso late lacked both band 9 and 7. All tomato iso lates but one were I B mitochondrial hap lotypes, and all were A 1 mating type. Specific virulence The spec ific v irulence patterns of potato and tomato populations of P. in festans differed greatly. Iso lates were generally highl y viru- lent on their own host d iffe rentials, but not on the alternative host differentials (Table 2). In several cases, isolates from one host did not infect any of the other host d ifferentials, even those free of known major genes for re- sistance. Two race-spec if ic genes in tomato were reported previous ly, but our data indicate that there is at least one more in the four differen- ti als we used. Pieralbo, reported ly a major- gene-free iso lin e of the Ph 2 diffe rent ia l Pi eraline, shou ld have behaved similarly to FMX-93, also putative ly free of major genes. However, these two tomato cu ltivars inter- acted differentiall y with some iso lates, thus indi cating the interaction of a hitherto un- known major gene (Tab le 2). Some isolates from potato did not in fect any of the four tomato differentials. Similar ly, some iso lates from tomato did not infect any potato differentia ls, includi ng those free of known R genes. No rel at ion between tomato and potato avirulence genes in P. infestans could be de- duced from our resu Its . Iso lates that were avirulent on all potato differenti als were viru- lent on all tomato differenti als (Tabl e 2). Therefore, all av irulence genes for tomato expressed in th is study appear to be indepen- dent of all known potato avirulence genes. Pathogenic aggressiveness Resu Its are presented as the average over all cu lt ivars for each host and all iso lates from each origin (tomato or potato). There was a very c le ar and stati st i ca ll y signifi ca nt (P<0.0001) interacti on between ori gin of iso- CIP Progrom Report 1995-96 139 Table 2. Specific virulence patterns of P. infestans isolates coming from potato or tomato and inoculated on major-gene differential plants of both hosts. ~ _,,. 0 ~- Isolates collected from potato Isolates collected from tomato 3 ~ Potato differentials' Tomato differentialsb Pathotype frequency Potato differentials' Tomato differentialsb Pathotype frequency 0, 1,3,4,7,8, l 0, 11 0,1 23 0,3,7 0, 1,2,3 15 0, 1,3,4,7,8, l 0, 11 0, 1,3 11 0,3 0, 1,2,3 12 0, 1,3,4,7,8, 10, 11 0 8 0 0, 1,2,3 8 0, 1,3,4,7, 10, 11 0,3 3 0,2,3 0, 1,2,3 6 0, 1,2,3,4,6,7, 10, 11 0,1 2 No infectionc 0, 1,2,3 5 0, 1,2,3,4,6,7,8, l 0, 11 0,3 2 3 0, 1,2,3 3 0, 1,3,4,7,8, 11 No infectionc l 0,2,3,7 0, 1,2,3 2 0, 1,3,4,7,8, 11 0 1 No infectionc 2 0, 1,3,4,7,8, 11 0,3 1 No infection' 0,1,2 0, 1,3,4,7, 11 0,1 1 0,4 0, 1,2,3 0, 1,3,4,7, 10, 11 0 l 0,11 0, 1,2,3 0, 1,3,4,7,8, 10, 11 No infection' 1 0,3,7 0,1,3 0, 1,3,4,6,7, 10, 11 0,1 1 0, 10, 11 0,1,3 0, 1,2,3,4,6,7, 10, 11 0,1,3 1 0, 1,3,7 0,1 ,3 0, 1,2,3,4,6,7,8, 10, 11 0 l 0, 1,3,7 0, 1,2,3 0, 1,2,3,4,6,7,8, 9, 10, 11 0,3 1 0, 1,3,4,7,8, l 0, l ld 0,3 0, 1,2,3,4,6,7,8, 9, l 0, 11 0, 1,3 a. Numbers represent major genes overcome by that pathotype. b. Numbers represent four tomato differentials: 0 = FMX-93, l = Pi era Ibo, 2 = New Yorker, and 3 = Pieraline. c. These individuals did not infect any of the differentials, including those considered free of major genes (O). d. This tomato isolate that was highly virulent on potato belongs to the EC-1 lineage. late (tomato or potato) and les ion diameter caused on either host (Figure 1 ). On the aver- age, iso lates from tomato ca used lesions about 2 cm greater in d iameter on tomato than on potato. Iso lates from potato caused lesions about 1 cm greater in diameter on potato than on tomato. Metalaxyl There was a marked and statisti ca lly signifi- cant difference in three levels of sensitivity to metalaxyl between tomato and potato isolates (Table 3). More than half the iso lates tested from potato (30 of 59) were resistant to metalaxyl , but on ly 3 of 43 isolates from tomato were re- sistant. In contrast, on ly 5 of 59 iso lates from potato were intermediately res istant; 17 of 43 from tomato were intermediate. Lesion diam (cm) 4 Discussion Our data amply add ress the first object ive of the study, that is, to determine w hether popu- lati ons of P infestans attacking potato and tomato differ in Ecuador. All our data indi- cate th at the two populations are distinct. All iso lates co llected from potato fo r thi s study belong to the EC-1 lineage. All but one of 60 iso lates from tomato belonged to the globall y distributed lineage US-1. That single var iant has the same dil ocus all ozyme genotype as EC-1 and probably belongs to that lineage. We do not know whether the one EC-1 iso- late found on tomato indi cates the in itiat ion of evoluti on toward aggress iveness on to- mato, or simply represents a wea kl y patho- Potato D Tomato ~ 3 - 2 - - ----- 0 ---- Isolates from Isolates from potato tomato Figure 1.Pathogenic aggressiveness of Phytophthoro infestons isolates from potato and tomato, inoculated on both hosts. CIP Program Report 1995-96 l 4 l Table 3. Level of sensitivity to metalaxyl of isolates of P. infestans from potato and tomato in Ecuador. Origin of isolates Potato Tomato Resistant 30 3 Sensitivity level Susceptible 24 17 Intermediate 5 2l Chi-square value for independence = 28.81, P = 0.001 , 2 degrees of freedom. genie iso late found by chance. Nor do we know w hether the les ion was large or sporu - lati ng when take n from the f ield. Further moni toring is requ ired to determine the sig- nifi cance of thi s f inding. In general, however, management of late bli ght is sim p li f ied in Ec uado r. For a ll practi ca l purposes, and until popul at ions of P. in festa ns change, fa rmers do not need to wo rry about cross in fect io n from nearby altern ati ve hosts, at least until EC- 1 is shown to be an agg ress ive pa th ogen of tomato. Th i s m anage ment reco m me nd at i o n is supported by f ield obse rvat ions we made w here ep id emi cs occurred in one of th e hosts w hil e nearby f ields of th e altern ati ve host remai ned c lean. The second object ive of the study, to elu- cidate the ro le of v irulence in host spec ifi c- ity, was less we l I addressed by our resu Its, but we believe that a working hypothes is can be deve loped. O ur best interpretation is th at host spec if ic ity is not determined by v iru- lence, but rather by pathogenic aggress ive- ness. None of the EC-1 potato iso lates infected New Yorker, the tomato cul t ivar containing the Ph l gene (Tab le 2) . The existence of an avirulence gene fo r Ph l , w hich is costl y fo r the pathogen to lose, would be an attracti ve model to expl ain host spec ifi c ity. Unfortu- nately, we were not able to survey cul t iva rs being used in Ecuador and we do not know whether Phl is common. We do not be lieve, however, that it occurs at a rate that would expla in host spec if ic ity as seen in Ecuador and elsewhere. 142 P1ogram 3 We sampled extensi ve ly in the central and northern pa rts of the country and it is highl y unli ke ly th at all these fields had cultivars w ith Ph1. Separate popul ation s of P. infestans are found on potato and tomato in Braz il , the Phi I ipp ines, and pa rts of the Nether lands. Ph 1 is also unlikely to be responsib le fo r host spec ifi c ity in those countries . Pathogenic aggress iveness appea rs to be a more importa nt factor th an spec if ic viru- lence in determining host spec ifi c ity. There was a clea r interact ion between or igin of iso- lates and their abi lity to infect tomato and potato (Fi gure 1 ). The iso lates from potato in fect potato more aggress ive ly than tomato, w hereas the oppos ite is true for iso lates from tomato. That f inding supports other stud ies that showed that host spec if icity was deter- mined by quantitative rather than qualitat ive facto rs. O ur study also leads to other interesting observat ions regarding spec ifi c v irulence. A pparent ly, P. infestans has di ffe rent avirulence genes for potato and tomato, at least among those that can be ident if ied w ith existing differenti al culti vars of both hosts. We found that severa l iso lates, w hich possess all known avirulence genes fo r potato, in fect all four tomato di ffe renti als (Tab le 2). Therefore, av irulence genes fo r potato must not eli cit a hypersensiti ve response in potato. Simil arl y, several iso lates, w hich did not infect any to- mato di ffe renti als, were highl y v irulent on potato (Table 2). O ne aspect of host spec iali zati on that re- mains unc lear is w heth er P. infestans ca n evolve into ~ n aggressive tomato pathogen with no loss of aggressiven ess on potato. There seems to be evidence for both sides of the argument. One study done recently in North Ameri ca demonstrated th at ce rtain genotypes of P infestans can infect both hosts with a high level of aggressiveness. It is not evident, however, that this phenomenon is universal. If dual aggressiveness were an in- herent capability of P infestans, and if adap- tation to tomato aggress iveness occurred rela- ti ve ly quickly (> 10 vegetati ve cycles), then it seems logi ca l that the dually aggress ive geno- types would quickly dominate both hosts. That is not the case in Ecuador and other parts of the world , where distinct genotypes are found in close assoc iation with each host. Our data do not conclusively demonstrate th at aggress iveness to tomato results in a loss of aggressiveness on potato, because the two populations represent different lineages of the fungus. We could have addressed thi s point more clearly if we had studied tomato and potato populations with one lineage. In 5 years of sampling from potato in Ecuador, however, we have found only one isolate belonging to US-1, and it was iso lated from the diploid potato spec ies Solanum phurej a. One plausible explanation for the incon- si stency between obse rvati ons in North Ameri ca and in other parts of the world is that genotypes outside North Ameri ca do not have the genetic potential for develop ing high levels of aggress iveness on both hosts. The genotypes th at do attack both hosts equally in North Ameri ca were recentl y introduced from Mexico, the center of origin of the patho- gen. If that is the case, the introduction of the North Ameri can genotypes to other parts of the world will have major impli ca tions fo r management, especiall y in areas where the two crops are grown in c lose assoc iation . Selected Reading Legard, D.E., T.Y. Lee, and W.E . Fry. 1995. Pathogenic spec ialization in Phytophthora infestans: Aggress iveness on tomato. Phytopathology 85: 1356-1 361. Turkensteen, L.J . 1973. Partial res istance of tomatoes aga inst Phytophthora infestans, the late bli ght fungus. Thesi s. Uni versity of Wageningen, Netherl and s. CIP P1ogrom Repo1t 1995·96 143 Use of Natural Resistance Genes for Transgenic Resistance to Potato Viruses M. Querci1, G. Brigneti2, J. Garcia-Mas2, and D .C . Baulcornbe2 O ne of t he p rin c ipa l t hreats to pota to (So lanum spp.) cul t ivati on is the susceptibil- ity of potato to pests and diseases. Viral di s- eases in parti cu lar cause signifi cant qualita- t ive and quanti tati ve crop losses. The most w idely used strategies fo r contro l of v irus dis- eases focus on methods to prevent infection or on geneti c res istance. Resistance describes the genera l response of the pl ant in w hich the effect of virus infection is reduced or elimi- nated. Thi s ranges from to lerance of or hy- persensiti v ity to the most durabl e extreme resistance or immuni ty. Few sources of ex- treme res istance prov ided by dominant genes ex ist fo r some potato v iruses. Examp les of durable resistance genes so fa r include Rx and Ry genes conferring extreme res istance to potato v irus X (PVX) and potato v irus Y (PVY), respecti ve ly. Breeding for virus resistance, a major com- ponent of most breeding programs, is gener- all y regarded as being the best strategy for long-term virus control. This approach is ham- pered, however, by the limitati on of suitable sources of res istance. In the past, the intro- gress ion of genetic sources for plant res istance has been successfully applied to develop a limi ted number of v irus-res istant cultiva rs . A lthough pl ant breeding fo r v irus res istance still has great potenti al, there are limi tati ons to thi s conventi onal approach. An appropri - ate source of resi stance may not be ava il abl e, the res istance may be t ightl y linked to unde- sirab le traits, or it may be mu lti geni c and as such di fficult to use in breeding programs. Plant geneti c engineering is an alternati ve to conventional breeding for res istance. One 1 CIP, Lima, Peru. 2 Sa insbury Laboratory, Norw ich, U.K. 144 Program3 approach to genetica ll y engineering virus re- sistance is pathogen-deri ved res istance in- vo lving express ion of transgenes that contain v irus-deri ved sequences . Th is has proven ef- fect ive against many different types of pl ant v i ru ses. The protection is expressed as pre- vention of infect ion, a delay in the onset of virus accumulation, or symptom development in systemica lly infected leaves. There are, however, concerns about the biosafety of thi s approach to v irus resistance. In additi on, pathogen-deri ved res istance may break down under high inoculum pressure or could promote evo luti on of res istance- breaking v iral iso lates. Therefore, we are at- tempting to deve lop a molecular breeding approach to vi ru s res istance in potato that does not provoke these concern s and that does not have the limitat ions of conventional p lant breed ing. "M o lecular breed ing" is the transfer of indiv idual genes between plants using transformat ion rather th an intercross- ing. It has been demonstrated recentl y that molecu lar breeding could be appli ed to v i- ru s res istance. The N gene and its phenotype of res istance aga inst tobacco mosa ic v irus was successfull y transferred by transforma- t ion from tobacco to tomato. Although the technology to iso late di sease res istance genes was developed only recently, it has now been refined to the leve l that it ca n be applied even in crop pl ants such as potato. The most important components of thi s tec hnology (Figure 1) include the deve l- opment of geneti c maps in whi ch there is a high density of molecular markers. Using the markers that are c lose ly linked to and fl ank the res istance gene of interest, it is possible to identify deoxy ri bo nuc le ic ac id (DN A) c lones th at span the interva l between the RxS /~-CT182 / •o CP54a CT1 82 CP58 M39B, M33 (M38) GP230 ~ Ry, M45, MS l Mapping --- --·~ GP125g:;io\ M1? CP117 M6 GP198a M41, M53 GP162a M43 CP137c GP259 F1 population TG105a GP1 63 GP185 GP222 GP38 4.0 GP16 1a - M35 Ry Transformation ------~ Figure 1. Schematic representation of the steps involved in the transfer of individual genes between plants using transformation. markers, and therefore include the resistance gene of interest. This high -resolution genetic mapping is useful not only for molecular breeding but also for marker-assisted transfer of desirable genes among varieties from related wild spe- cies. It will make it possible to analyze and transfer complex polygenic characters as well. Materials and Methods Virus diseases in potato In potato, PVX, PVY, and potato leafrol I vi- rus (PLRV) are the three most important viral pathogens. PVY, the type member of the potyvirus group, is an aphid-transmitted vi- rus and probably the most difficult to control in potato fields. Genes triggering hypersensi- tive response and field immunity against PVY are present in several potato species. There are also Ry genes conferring extreme resis- tance to PVY in 5. tuberosum subsp. andigena (Ry d ) and 5. stoloniferum (Ry ). Already a g sto these loci have been introduced into several commercial cultivars by conventional breed- ing. PVX is the most common virus infecting potato worldwide. Strains of PVX have been CIP Program Report 1995-96 l 4 5 classified into four groups on the basis of their interactions with the dom in ant res istance genes Nx and Nb, w hich contro l a hypersen- sitive response, and the extreme resistance gene Rx. The Rx res istance is effective against all known iso lates of PVX, w ith the excep- tion of PVXH8, w hich can overcome all known res istance genes. PVXHB is found in Bolivia and a limited number of sites in the Andean region. Severa l 5olanum spp . carry Rx genes for extreme resistance. Thos e from 5. tuberosum subsp. andigena and 5. acaule are the most common ly used in breeding pro- grams. The potato has a good regeneration and transformat ion ab ility and, in genera l, pro- v ides a good background for the application of recent deve lopments in molecul ar biology. To exploit this potential , the molecular v irol- ogy unit of CIP is col laborat ing w ith the Sainsbury Laboratory, U .K., on the charac- terization, iso lation, introduction, and expres- sion of plant-deri ved resi stance genes in transgeni c potato pl ants. The immediate tar- gets of this program are the Rx and Ry genes of potato, for the reasons described below. First, res istant plants challenged w ith ei- ther virus do not develop v isible symptoms, and virus accumulation cannot be detected by either enzyme- linked immunosorbent as- say (ELISA) or ri bonucleic acid (RNA) hybrid- ization. There is no evidence that these resis- ta nce tra its can be overcome by high inocu- lum pressu re. A second considerat ion was the durability of the resistance conferred by Rx and Ry. In the case of Ry, the resistance is effective against all known iso lates of PVY, w hereas Rx is overcome by a res istance-breaking iso- late (HB), although it has not become a prob- lem in most potato-growing areas. The potential to carryout molecular breed- ing w ith both genes simultaneously wou ld have agronomic benefits, because these v i- ru ses interact to produce severe crop losses. In add iti on, studies have shown that PVX or PVY reduce the durabi I ity and the degree of res istance to PLRV, for which no immunity 146 Progrnm3 genes have been found so far. The estab li sh- ment of stable v irus resistance is of high pri- ority, and the development of cu ltivars res is- tant to the three v iruses has always been a great chal lenge fo r breeders . Mo lec ul ar breed ing of res istance to PVX and PVY wou ld be an important and signifi cant step in that direction. ~' : _ _ )}!§ _ _ ~,M33 Ry m--++-a1-+-+-____ __,,,,___,H-- M41, M53 ll=::+:t:=ll=ll=t:t==B=ll~l=l== ~~~59 GP163 Figure 2. Detail of an AFLP-gel autoradiograph showing resistant (R) and susceptible {S) bulk DNA samples {each consisting of DNA from l 0 individuals) and six individuals from the F1 population {top) , and schematic representations of the corresponding chromosomes {bottom). The segregation of marker M39 is indicated on the autoradiograph . The location of additional markers and recombination events in the F1 lines is indicated in the bottom panel. Mapping Ry A chromosomal assignment of Ry510 was the first step toward the molecu lar cloning of Ry510 from potato by a map-based approach. Mo- lecul ar linkage maps of potato and tomato genomes were readily available, so we used amp lifi ed fragment length polymorphism (AF LP) technology as a source of molecular markers linked to Rysto (Fi gure 2). Ry from 5. sto lonife rum has been mapped at th e Sainsbury Laboratory on the top arm of the potato chromosome XI. A second step involved constru ct ing a high-resolution genetic map around the Ry locus. That was obtained using an F1 segre- gating population of 360 plants. Several AFLP primer combin ations were tried on DNA pools of resistant and susceptible plants. Two AFLP markers (M l 7 and M6) at either side of the gene (separated by a single recombina- tion event) and two more markers (M4S and MS) cosegregat ing w ith Ry were found. DNA markers linked to the resistance gene have been used to sc reen 22 CIP crosses that carry Ry 510 and segregate for the res istance. To increase the map resoluti on of the mark- ers relat ive to the res istance gene, an add i- t iona l 3,079 F1 segregating progeny (1,779 progeny from the 1st year and 1,300 from the 2nd year) have been screened. The resistant parent {1-1 039) of thi s new cross was the same as the one orig inall y used at the Sainsbury Laboratory for chromosoma l ass ignment of Ry 510 and construction of the fi rst genetic map. Results The screen ing has been carri ed out in sev- era l stages. First, polymerase chain reaction (PCR) markers were used so that p lants w ith recombination events close to the res istance gene cou ld be identifi ed (i.e., plants w ith re- combination between the markers). The flank- ing PCR markers used in the screening of the new cross were M39b and GPl 63. A few hundred recombinants were found in this in- terval. Using the GP2S9 PCR marker, the number of recombinants was reduced. From these data it was also possible to confirm the order of the PCR markers in the genome rel a- tive to each other and, following res istance testing of the recombinant plants, relative to Ry 510 • Th e recombin ants identified were ana- lyzed by AFLP using the primer combinations for the c losest markers: Ml 7, MS, M4S, and M6. The new map obta ined with this second mapping population was compared w ith the previous map. Th e order of the markers is conserved although th e genet ic distances have increased slightl y. We have been ab le to separate M39b from M33, but Ml 7 is now cosegregating w ith MS and M4S (F igure 3). To obtain an est imate of the physical dis- tances around the Ry locus, we began screen- ing bacteri al artifi c ial chromosome (BAC) and cosmid librar ies. We identified a single BAC clone (cBAC-1 S) of 110 kb containing three of our markers (Ml 7, MS, and M4S) although it is from a chromosome that carri es the ry (rather than the Ry) all ele. We are now at- tempting to map the ends of cBAC-1 S to ob- tain new markers in the vici nity of Ry. If the cu rrent ly ava il able librari es do not yield Ry, we w ill prepare new librari es from dihap lo id plants that have been generated from the tetraploid parent 1-1039 (which ca r- ries Ry in the simp lex co nditi o n). When clones th at span the Ry locus have been pro- duced, we w ill identi fy the precise locat ion of the gene by comp lementati on. Different fragments of DNA from the Ry interva l w ill be transformed into susceptible plants and the transgen ic Ii nes wi 11 be tested for resistance to PVY. Discussion For severa l years CIP has invested a great deal of effort in searching for sou rces of res istance to severa l v iru ses in the w ide range of potato genotypes in the germpl asm co llection held at CIP. Breeding for v irus res istance was fo- cu sed on the th ree major v iru ses, PVX, PVY, and PLRV. The main goa l was the develop- ment of advanced materials with comb ined resistance to PVX, PVY, and PLRV to al low farmers to keep their own seeds season after season, in this way reduc ing their cost of po- tato production. CIP Prngrnm Rep01t 1995-96 l 4 7 M33 M39b M39b M17 0.56 0.34 M33 J 6/1779 M45, M5Ry,,0 0.27 0.28 M17 M45 M5 J 511779 0.27 Ry,,"? M6 0.78 14/1779 0.56 M53 M41 M6 1.11 M43 0.27 3.93 77/1779 GP259 1.39 GP163 GP259 2.25 40/1779 GP163 Figure 3. High-resolution map of the Rr,,, region on the top arm of potato chromosome XI. The genetic distances between markers were inferred from an F1 population of 360 plants (chromosome on left) and for a population of 1,779 plants (chromosome on right) . The number of recombinant progeny obtained for different intervals is indicated on the right. In the larger population, markers M39b and M33 were separated. An additional l ,300 plants are being analyzed to refine the relative positions of markers Ml7, M45, and MS. The strategy fol lowed was to first introduce PVX and PVY extreme resistance genes from Solanum tuberosum subsp. andigena into advanced breeding populations, then to com- bine PVX and PVY extreme resistance with res istance to PLRV and possibly other patho- gens. Genes control I ing resistance often origi- nate from wild species; undesirable traits may be closely linked to the resistance gene and may be difficult to eliminate. In our transformation approach, we will transfer only isolated and characterized genes 148 Program 3 (not a large piece of the whole genome) so that the recipient genotype remains largely intact. Transformation with natural resistance genes is potentially the shortest and most pre- cise procedure for introgressing resistance genes into a crop species. The ultimate aim of this work is to support increased potato production in developing countries. But we also expect the project to generate knowledge of the mechanisms in- volved in virus resistance as well. We per- ceive a practical application of this new ap- proach over the medium term and long term. In the medium term, practical applications wi 11 be achieved through the identification and isolation of natural genes that confer ex- treme resistance to PVX, PVY, and PLRV, and their insertion into available susceptible but agronomically important potato cultivars. Over the long term, we plan to integrate this nonconventional approach to optimize advances on resistance to other pests and pathogens already achieved through conven- tional breeding. Of particular interest will be the combination of conventional and nonconventional breeding methods to in- crease levels of PLRV resistance. In addition, molecular breeding methods can be used to add resistance to viruses (Rx and Ry genes) to advanced materials with resistance to other pathogens . The concept of plant-derived transgenic resistance provides an attractive strategy to produce a novel but genetic form of virus control by transforming crop plants with nucleotide sequences derived from the plant genome itself. Natural resistance genes for plant protection would eliminate most of the biosafety concerns associated with the use of virus-derived sequences in transgenic plants. Selected Reading Bendahmane, A., K. Kanyuka, and D.C. Baulcombe. n.d. High resolution geneti- cal and physical mapping of the Rx gene for extreme resistance to potato virus X in tetraploid potato. Theor. Appl. Genet. (In press.) Brigneti, G., J. Garcia-Mas, and D.C. Baulcombe. 1997. Molecular mapping of the potato virus Y resistance. Theor. Appl. Genet. 94:198-203. CIP Program Rep ort 1995-96 l 4 9 Encapsidation of PSTVd in PLRV Particles and Its Transmission by Aphids M. Querci1, R .A. Owens2, and L.F. Salazar1 Like al l known viroids, potato sp indle tuber viroid (PSTVd) is an independent ly replicat- ing agent, w hich completes its infection cycle without generating either a caps id or other viroid-spec ifi c proteins. Its genome is a sma ll (359 nt), single-stranded, covalent ly closed c ircular ribonucleic acid (RNA) molecule whose extens ive reg ions of intramolecular complementarity are responsible for its un- usual stabi I ity in vivo. The natural spread of PSTVd in potato was known to occur only by fo li ar contact or bo- tanica l seed. In contrast to PSTVd, potato leafroll v irus (PLRV), a member of the genus Luteovirus, is read ily transm itted by aphids, with Myzus persicae being the most effic ient natural vec- tor. Like all luteov iru ses, PLRV is restri cted to the phloem tissue of infected plants, and aphid transmission is both pers istent and non propagati ve . PLRV has a narrow host range, and the v irus is commonl y found wher- ever potatoes are grown. Earlier we confirmed that M. persicae cou ld transmit PSTVd to potato and other test plants only when the source plant was dou- bly infected with PLRV and PSTVd . In some experiments , PSTVd transmission reached 100%; but no transmi ss ion was observed from source pl ants infected w ith the v iroid alone. Transmission by Aphids Occas ional PSTVd conta mination of PLRV isolates maintained at CIP prov ided the first indi cation that PLRV might fac ilitate aphid 1 CIP, Lima, Peru. 2 Beltsv ille Agri cultu ral Research Center, Beltsv ille, M ary land, USA. 150 Program 3 transmi ss ion of PSTVd. Addition al epidemio- logical ev idence for an association between PSTVd and PLRV in field-grown potatoes was subsequently obtained by testing plants grow - ing at severa l sites in China for the presence of PSTVd and PLRV. A total of 880 plants from three sites were tested by nucleic acid spot hybridization (NASH) in 1994, and Tabl e 1 summarizes the results from these anal yses. Rates of infection for both PLRV and PSTVd vari ed considerab ly, ranging from 3.6% to 86% and 0.9% to 27.2%, respecti ve ly. Vec- tor pressure was highest at the Bashang Insti- tute (Hebe i Prov ince). As expected, the incidence of PLRV at that site appeared to be somew hat greater than at the others. Most (but not all) PSTVd-infected plants at al I three sites were also infected w ith PLRV. No effect on either host genotype or si te was identified for aphid transm iss ion of either PLRV or PSTVd, and the pooled data were ana lyzed to test the independence of PSTVd and PLRV d istribution. The x 2 va lue (29.68) was highl y significant, strongly sug- gest ing th at PLRV facilitates th e sp read of PSTVd under field conditions . Additi onal aphid transmission exper iments were ca rri ed out to more precisely charac- ter ize the mode of PSTVd transmission by M. persicae. Materials and Methods Potato clone DT0-33 (C IP No. 800174), a Solanum tuberosum x 5. andigena hybrid sus- ceptible to PLRV and PSTVd, was used as both inoculum source and test plant. Pl ants dou- b ly infected w ith PLRV and PSTVd were ob- tained by initi ally inoculating them w ith PLRV Table 1. Incidence of PLRV and PSTVd at three potato-growing sites in China, 1994.' Potato variety Site Samples Infection rate(%) Co infection rate(%) PLRV PSTVd Tiger Head Bashang Inst. Hebei Province 132 66.7 (88) 2.3 (3) 2.3 (3) 88-1-19 Bashang Inst. Hebei Province 114 86.0 (98) 2.6 (3) 1.8 (2) Bashu No. l 0 Bashang Inst. Hebei Province 106 63.2 (67) 0.9 (l) 0 (0) Purple Flower White Wachuan County Inner Mongolia 241 17.8 (43) 1.7 (4) 0 (0) Desiree Wachuan County Inner Mongolia 56 3.6 (2) 0 (0) 0 (0) Tiger Head Inner Mongolia University 103 68.9 (71) 27.2 (28) 25.2 (26) Purple Flower White Inner Mongolia University 128 36.7 (47) 18.8 (24) 18.8 (24) a. Fie ld surveys were conducted in 1994. Samples from the Bashang Institute and Wachuan County were obtained from farmers' fields. All samples were randomly selected, and the presence or absence of PLRV and PSTVd was determined by NASH. Numbers of singly and doubly infected plants are shown in parentheses. Analysis of all data (including a Yates correction) yielded a x 2 value of 29.68. Assuming no effect of host variety or si te on aphid transmission, a x 2 va lue of 6.64 indicates significance at the l % level. by means of its vector, M. persicae, and then manually inoculating them with sap from PSTVd-infected tomato (Lycopersico n escu lentum cv. Rutgers). After inocu lation , plants were maintained in a growth chamber and periodically tested for PSTVd and PLRV infection. Apterous aphids from a nonviruliferous co lony of M. persicae were raised on Chi- nese cabbage (Brassica pekinensis (Lour.) Rupr.). One month after PLRV or PSTVd in- ocul at ion, apterous aphids were allowed a 3-day acqu isit ion access period on either si n- gly or doubly infected source plants. Aphids (5 aph ids/p lant) were then transferred to young uninfected potato plants by placing the insects on filter paper squares using a sterile brush and then allowing them to wa lk onto a plant leaf. A final 2-day passage on healthy Chinese cabbage was included in one exper iment. After a 3-day transmission access period (TAP), aphids were killed by spray ing plants with a contact insectic ide. Nature of PSTVd and PLRV Association Fifteen and 45 days after the TAP, inoculated plants were tested for the presence of PSTVd and PLRV using a comb ination of the NASH test and enzyme- linked immunosorbent as- say (ELISA). M. persicae was able to transmit PLRV under all cond itions tested (Table 2). PSTVd transmission, in contrast, was ob- served on ly when the aphids were all owed to acquire the vi roid from doubly infected source plants. Comparison of data presented in rows 3 and 4 shows that aphids allowed to feed sequentially on plants singly infected with PLRV or PSTVd were unable to transm it PSTVd . Feed ing on hea lthy cabbage before trans- fer to the test plants reduced, but did not elimi- nate, the abi li ty of aphids to transm it PSTVd from doubly infected sou rce plants. Thu s, PLRV-med iated transm ission of PSTVd by M. persicae appears to be of the persi stent type. To further characterize the nature of the association between PSTVd and PLRV, a se- CIP Progrom Repo1t 1995-96 l 51 Table 2. Aphid transmission of PLRV and PSTVd to potato variety DT0-33. lnocolum source(s) Passage PLRV None PLRV Chinese cabbage PLRV --7 PSTVd b None PLRV --7 PSTVd b Chinese cabbage PLRV + PSTVd None PLRV + PSTVd Chinese cabbage Successful transmission ° PLRV Yes (4/5) Yes (2/ 5) Yes (2/ 3) Yes (1/5) Yes (4/5) Yes (1 / 5) PSTVd No (0/5) No (0/5) No (0/3) No (0/5) Yes (3/5) Yes (1/5) a. Data presented as no. of infected plants/no. of inoculated plants. b. Aphids were allowed two acquisition feedings, on PLRV-infected plants followed by the second on PSTVd -infected source plants. ries of virus purifications were carried out using leaf tissue collected from singly and doubly infected potato plants. The first ex- periment compared the amounts of PSTVd RNA associated with virions purified from doubly infected plants with those associated with particles isolated from a pooled sample of leaves collected from singly infected plants. PLRV was purified, and encapsidated RNAs were recovered by phenol/chloroform extrac- tion and ethanol precipitation. RNA pellets were resuspended in sterile nuclease-free water and reverse-transcribed using a random hexanucleotides mixture (pdN6). PSTVd- and PLRV-specific cDNAs were then amplified separately by polymerase chain reaction (PCR) using the appropriate pairs of oligonucleotide primers. Amplifications (40 cycles) were carried out using a 94°( (1 min), 55°( (2 min), 72°( (1 min) profile followed by a 5-min final extension at 72°(; PCR products were visualized by elec- trophoresis on 5% acrylamide gels. Predicted sizes for the PSTVd- and PLRV-specific PCR products were 226 and 534 bp, respectively. As shown in Figure 1 A, large amounts of a PLRV-specific product of the appropriate size 152 Program3 were present in PCR reactions containing RNA derived from either singly (lane 3) or doubly infected (lane 2) plants. The corre- sponding PSTVd-specific product was present only in reactions containing RNA derived from doubly infected tissue (lane 5) . When virions were isolated from a mixture of leaves collected from singly infected plants, no PSTVd-specific PCR product was produced (compare lanes 5 and 6). To determine whether transencapsidation of PSTVd by PLRV was responsible for its observed aphid transm issibi I ity, two types of samples were treated with micrococcal nu- clease: PLRV virions isolated from doubly in- fected leaf tissue, and mixtures of virions iso- lated from singly infected plants plus suffi- cient PSTVd RNA to produce a comparable ratio of PSTVd/ PLRV. Trial experiments showed that 15-min in- cubation at 30°( with 1 x 10-3 units/µ1 micro- coccal nuclease completely destroyed con- centrations of PSTVd similar to those found in PLRV virions isolated from doubly infected tissue (results not shown). Parallel incubations were carried out, each containing the same amount of virus but only (A) (B) (C) PLRV PSTVd M 123456 587 bp ~ - -· = 359 bp~ - 267 bp~ 184 bp~ PLRV -- 1 2 3 4 5 6 7 8 9 IOM L - nucl ease _J L +nuclease _J PSTVd 123 4567 8910M $ ' ~; ~ -~ L - nuclease _J L +nuclease _J I ..... 587 bp ..... 359 bp ..... 267 bp ..... 184 bp ..... 587 bp ..... 359 bp ..... 267 bp ..... 184 bp Figure 1. PAGE (polyacrylamide gel electrophoresis) analysis of RT-PCR products. (A) Association of PSTVd with PLRV virions purified from different sample types : PSTVd RNA plus PLRV virions purified from singly infected plants (lanes l and 4); PLRV virions purified from doubly infected plants (lanes 2 and 5); PLRV virions purified from leaves from singly infected plants combined before PLRV purification (lanes 3 and 6). (B-C) Relative sensitivity of PLRV and PSTVd RNAs to digestion by micrococcal nuclease. RT- PCR analyses were carried out using pairs of (B) PLRV- or (C) PSTVd-specific primers. Virions purified from doubly infected potato plants were incubated for 15 min at 30°( in either the absence (- nuclease) or presence (+nuclease) of micrococcal nuclease before extraction of RNA. Lanes l-5 and 6-10 contain serial threefold dilutions of the mixture of randomly primed PLRV and PSTVd cDNAs. M =DNA size markers. CIP Program Report 1995-96 153 one containing micrococca l nuclease. After incubation, the undigested RNAs were iso- lated by phenol-chloroform extracti on, re- verse transcribed, and analyzed by reverse transcripti on pol ymerase chain reacti on (RT- PCR). Results from these analyses are pre- sented in Figures 1 B and 1 C. Each set of fi ve amp I ificat ions (l anes 1-5 and 6-1 0) contained a seri es of threefo ld di- luti ons of the initi al preparation of randomly primed PLRV and PSTVd cDNAs . Samples in lanes 1-5 were deri ved from untreated vi ri ons, w hereas those in lanes 6- 10 were treated w ith micrococcal nuclease before RNA extraction. In Figure 1 B, compari- son of lanes 1-5 w ith 6-1 0 shows that, as ex- pected, the v iral genomic RNA w as res istant to micrococca l nuclease digestion. The cor- respo nding PSTVd -spec if ic analyses are shown in Fi gure 1 C. Because the v irions con - tain relati ve ly low leve ls of v iro id RN A, the amount of PSTVd-spec ifi c PCR product de- creased as the amount of randomly primed cDNA tem plate was reduced. Micrococcal nuclease digestion , however, had no effect on the amount of PSTVd-speci fi c PCR prod- uct synthes ized. Thi s experim ent was repeated seve ral times with d ifferent v irus preparations, and though the relative amount of PSTVd in the different PLRV preparations vari ed, the results obtained were consistent and comparabl e. The explanation for our results is th at the PSTVd associated w ith v iri ons iso lated from doubly infected plants is, like the PLRV ge- nomic RNA, located within the virus part icle. Unencapsidated PSTVd RN A w ould have been degraded by the micrococcal nuclease digestion . Expressed on a molar bas is, the amount of PSTVd from doub ly infected plants is approx imately one molecule of PSTVd fo r every 3,000-5,000 molecul es of PLRV RN A. Discussion In nature, pl ant viruses and other subviral pathogens (satel lite/defective interfering RNAs or v iro ids) have many opportuni t ies fo r inter- action. Multiple v irus infections commonl y 154 Program 3 occur in crop and weed hosts. That may re- sult in vi ru s particles in w hich (1) ind ividual part icles conta in structural prote ins deri ved from more than one virus or (2) the genome of one v irus is encapsidated in the structural proteins of another. Thi s phenomenon, va ri- ously known as transencapsidation, genomic masking, or phenotypic mix ing, has been observed many times fo r members of th e luteov iru s and potyv irus groups . Several sobemoviruses are know n to support the rep- licati on of small v iroid-li ke satelli te RNAs, as in the case of ve lvet tobacco mottl e v irus (VTM oV), w hi c h ha s bee n show n to encapsidate PSTVd. Even more recent ly, a small , ri bozyme-containing satelli te RNA has been reported in associ ation with certain iso- lates of barl ey ye llow dwarf v irus-RPV sero- type. The size of PSTVd is simil ar to th at of barl ey ye llow dwa rf virus satellite RNA (i.e., 359 vs. 322 nt). Our data show that it too ca n be encaps idated by a luteov irus, PLRV. Transencapsidation of PSTVd by an ass istor v irus and the resulting acquisiti on of aphid transmi ss ibi I ity have important epidemiologi- ca l impli cati ons. A lthough v iroids are gener- all y thought to be mechanica lly transmitted, some workers see a di ffe rent poss ibility. In • th at v iew, a vi ro id originat ing in a p lant spe- c ies in w hich it is symptomless could be encapsidated by a v irus and vectored to an- other plant spec ies. Even though that p lant might not be a host for the v irus itse lf, it could be one in w hich the v iro id becomes patho- geni c. O ur survey data from China indi cate that, even where the overall level of PLRV infection was relati ve ly moderate, all or al- most all plants found infected wi th PSTVd were also infected w ith PLRV. VTMoV greatl y suppresses PSTVd repli cation in Nicotiana c/eve landi. Nevertheless, tomato pl ants, w hich are immune to VTMoV, became in- fected w ith PSTVd w hen inoculated w ith v i- ro id-containing VTMoV preparat ions. Addi- t ional data w ill be required to determ ine w hether PLRV has a simil ar effect on PSTVd repli cati on in potato, but transmi ss ion by aph ids or other insect vectors prov ides a pl au- sible exp lanation for the presence of PSTVd in such atypica l hosts as avocado, a phenom- enon occurring in Peruvian avocados . Aphid transmission of PSTVd may be most important in long-d istance dissemination of the viroid, thus fac ilitating its spread and greatly reducing the possibility of its control. Conclusions An important step in breeding for virus res is- tance is the identification of spec ific factors that tend to break down or overcome that re- sistance. In the case of PLRV, no genes con- ferring immunity have been described, and the vari ous res istance genes known appear to act by different mechanisms. Compared to the leve ls of geneti c res istance to potato vi- rus X (PVX) or potato virus Y (PVY), the leve l of PLRV res istance conferred by indi v idual genes is rather low. In prev ious studies we found that the pres- ence of PSTVd ca n lead to a decrease in the level of res istance to PLRV. Infect ion rates of 100% were observed for 5. acaule accessions OCH 13823 and OC H 13824, fo ll owing aphid-mediated transmission of v irus from doubly infected source plants. In the absence of PSTVd, infection rates were fourfold or five- fold lower. These two genotypes had been shown to be res istant to both PLRV infection and multiplication . A similar decrease in PLRV res istance is observed when res istant genotypes such as cv. Mariva are infected by either PVX or PVY. Knowl edge of how other pathogens such as PSTVd might interfere w ith the express ion of genes conferring PLRV res istance wou ld help in se lect ing the type of res istance to be incor- porated and predicting its expected durabil- ity. We are work ing to identify the different components of res istance to PLRV as we ll as to eluc idate the rol e of PSTVd in decreas ing such res istance . Selected Reading Querci, M ., R.A. Owens, I. Bartolini , V. Lazarte, and L.F. Salazar. 1997. Evidence for heterologous enca psidation of potato spind le tuber v iroid in particles of potato leafroll virus. J. Gen. Viro l. 78: 1207- 1 211 . (IP Program Report 1995-96 155 Progress in Identifying Viruses Infecting Andean Root and Tuber Crops C. Lizarraga, M. Santa Cruz, and L.F. Salazar1 The first step in the control of any disease is the isolation and study of the pathogen. Be- cause virus diseases can reduce yields in many crops and are easily transmitted in veg- etative propagation, identifying viruses that infect Andean root and tuber crops (ARTC) is a major priority in the effort to produce plant- ing materials of high quality, and to facilitate the international exchange of valuable virus- free germplasm. The Andean tuber crops include ulluco (Ullucus tuberosus Caldas) , oca (Oxalis tuberosa Mol.), and mashua (Tropaeolum tuberosum R. & P.). Two root crops are arracacha (Arracacia xanthorrhiza Bancroft) and mauka (Mirabilis expansa R. & P.). Viruses in these crops have not been well studied. Previously, only four viruses were reported infecting ulluco: ullucus mosaic virus (UMV), ullucus virus C (UVC), papaya mosaic virus (PapMV-U), and tobacco mosaic virus (TMV- U); three infecting oca: potato black ringspot virus (PBRV), arracacha B, and PapMV-0; and three infecting arracacha: arracacha A (AVA), arracacha B (AVB), and arracacha potyvirus 1 (AP-1 ). Their effect on ARTC has not been determined. Materials and Methods Virus identification Standard virological techniques applied to studies of potato viruses were used. Viruses were isolated by mechanical inoculation or graft inoculation, or were insect-transmitted to indicator plants and identified by serology using antisera available for known viruses. Purified preparations of the isolated viruses 1 CIP, Lima, Peru. 1 5 6 P1og1om 3 were injected into rabbits to produce antisera for detection and for determining serological relationships to wel I-known plant viruses. Virus effect on yield Two ulluco accessions, MH-290 and MH- 296, from the germplasm collection main- tained at CIP and freed of virus infection by thermotherapy and meristem culture were provided by C. Arbizu (CIP). The in vitro plants were multiplied by cuttings in the screenhouse in CIP-Huancayo (3,200 m above sea level) and the tubers were collected. To study the effect of primary infection, the tubers were planted in October 1994 in CIP- Huancayo. The young ulluco plants were mechanically inoculated in December 1994 with UMV, UVC, and PapMV-U individually, and with the three viruses simultaneously. Control plants were mock-inoculated with disti I led water. The tubers were harvested in June 1995 and stored until October 1995, when the MH-290 tubers were planted in the same field to study secondary infection. Tubers from MH-290 were harvested in June 1996. Results Table 1 shows the viruses isolated and the ones now known to infect five ARTC. No significant differences (P::;0.05) were found for the yield of ulluco plants with pri- mary viral infection and the healthy control. Table 2 shows the effect of three different vi- ruses on the yield of ulluco plants (MH-290) with secondary viral infection in CIP- Huancayo. As expected, the combined infec- tion by the three viruses caused a higher yield decrease. Table 1. Viruses known to infect ulluco, oca, mashua , arracacha, and mauka , CIP, 1996. Crop Vi ruses identified Vi ruses identified Viruses in process of previously in this study identification Ulluco UMV PLRV uvc APLV PapMV-U AVA TMV Oca PBRV PVT 0-2 AVB PapMV-0 Mashua SoMV M-1, M-3, M-4, M-5 Arra ca eh a AVA PB RV-A AV-3 AVB AP-1 Mauka Mir-1 Table 2. Average yields (kg) of ulluco plants MH-290 with secondary infection of three viruses, CIP-Huancayo, 1996. Treatment Total yield 0 First category tubers Yield reductionh(Ofo} UMV 19. l be 10.5 b 29 uvc 19 .7 be 10.7 b 27 PapMV 24.2 ab 14.5 ab 10 UMV + UVC + PapMV 16.7 c 10.5 b 38 Healthy control 26.9 a 16.0 a a. Inner rows of 66 plants. Means within co lu mns followed by the some letters ore not significantly different at the P::;: 0.05 leveL b. Compared with healthy control total yield. The reported incidence of viruses in ulluco, oca, and arracacha in the Andean reg ion (Tab le 3) ind icates the poss ibl e re lative im- portance of the d iffere nt viru ses that infect ARTC. Discussion In the Andes, ulluco, oca, and potato are of- ten intercropped and the presence of potato viruses in ulluco and oca supports the premise that plant viruses can adapt to new hosts sub- sequent to long-term assoc iations. Although the importance of these viruses in ulluco and oca is unknown, PLRV is the most important virus in potato. PLRV-infected ulluco plants grown near potato seed stocks could have seri ous epidem iological implications, parti cu- larly because PLRV has a reported incidence of 37% in ulluco. CIP P1ogrom Repo1t 1995-96 l 5 7 Table 3. Reported incidence of viruses in ulluco, oca, and arracacha in the Andean region . Crop Virus Ulluco UMV Ulluco uvc Ulluco PapMV-U Ulluco TMV Ulluco PLRV Ulluco APLV Ulluco AVA Oco PBRV Oco AVB Oco PapMV-0 Arra co cha AVA Arracocha AVB Arra co cha AP-1 Arra co cha PB RV-A Arracocha AV-3 The practi ca l impact of ARTC virus identi - fi cation studies is the produ ct ion of v irus-free planting materi als that w ill allow farmers to obtai n higher y ields and reduce degeneration of culti va rs . The results of fi eld experiments in Huancayo have shown that clean ulluco planting materi als can have higher yields than plants secondarily infected w ith viruses. UMV and UVC have reported inc idences of 75% and 87%, respectively, and apparently are the most important v iruses in ulluco since they cause yield reductions above 27%. The antisera produced in these studi es can be used in the sensitive serologica l technique ca ll ed enzyme-linked immunosorbent assay (ELISA) to prov ide reli able v irus detection for institutions invo lved in cleaning up ARTC and produc ing healthy planting materi als. Conclusions Although these resu Its indicate that v irus-free ulluco can have higher yields, further work must be done in farmers' f ields to confi rm this and to determine the rate of reinfect ion to 158 Pmgmm 3 Reported incidence(%) 57 87 85 34 37 44 3 0.2 0.2 22 6 25 51 7 52 justify using resources in the production of v irus-free material s. The same studies should be done for all ARTC, beginning with the most economical ly important ones. M any v iruses infect ARTC; determining which ones cause major y ield reductions must be a high prio rity in the continuation of this work. Informat ion on viral incidence and di stribution has a co ll atera l ro le in ascertain- ing the importance of v iruses. Selected Reading Li zarraga, C., M . Santa Cruz, and U. Jayas inghe. 1996a. Detection of an iso late of Andean potato latent v irus in ulluco (U/lucus tuberosus Caldas). Plant Di s. 80:3 44. Lizarraga, C., M. Santa Cruz, and L.F. Salazar. 1 996b. First report of potato leafro ll v irus in ulluco (U/lucustuberosus Ca ldas). Plant Di s. 80:344. Sa laza r, L.F. 1996. Potato v iruses and their contro l. CIP, Lima, Peru . 226 p. PROGRAM REPORT Integrated Pest Management Fausto Cisneros1 Six premi ses guide the work of CIP's Inte- grated Pest Management (IPM) Program within the overa ll ecologica l context of IPM for crop protection without affect ing human hea lth and the env ironment. First, sma ll- and medium-size fa rmers with limited resources are defined as the clientele. Second, IPM re- search and implementation are considered a continuum that culminates in fa rmers' fi elds. Third, IPM components are offered to fa rm- ers not as a technologica l package but as a menu of options appropriate to their situa- tions. Fourth, IPM is implemented in the broader context of integrated crop management in col- laboration w ith farmers' organizations, NARS2, and NGOs. Fifth, IPM programs are susta inable only insofar as they are simple to implement 1 Program Leader, CI P, Lima, Peru. 2 Acronyms cited in thi s section ca n be found written ou t in the section Acronyms, p. 320. Pro and maintain . Finally, training activiti es and training materials are oriented toward improv- ing farmers' capacity to understand and better select the IPM options. To develop our work along these premises, we fo llow a strategy th at has five develop- menta l phases: (1) pest assessment, (2) devel- opment of IPM components, (3) integrat ion of these components, (4 ) establishment of IPM pilot units in farmers' fi elds, and (5) large-sca le IPM implementation. In add ition, the FAO's fa rmer field schoo l model, developed for implementing IPM on ri ce, is bei ng adopted by CIP's reg ional of- fice in Southeast Asia to manage sweetpotato pests. The objective is to improve analytical and decision-making sk ill s of farmers. This experi ence is in the initi al stage. CIP P1og1om Report 1995-96 159 Target Pests Selected key pests of potato were the com- plex of potato tuber moths (PTM) that in- cluded the common PTM (Phthorimaea opercule //a), wide ly distributed in the warmer potato-producing areas; the spotted or Andean PTM (Symmetrischema tangolias), present at mid-elevations in the Andean coun- tries; the Central American PTM (Tecia solanivora), w hi ch is rapidly spread ing to South America; and the Andean weev i l com- plex (APW), which includes Premnotrypes spp. and other related genera, in the high mountains (> 2,800 m) of the Andean reg ion. These pests ca n ca use damages above 50% in the field (APW) or in stores (PTM). Another selected ke y pest is the leafm in er fl y, Liriomyza huidobrensis, and related spec ies. They are widespread and responsible fo r the heavy use of insectic ides in potato and other vegetab le crops. Weevils were the se lected key pests for sweetpotato. Th e As i an spec ies (Cy/as fo rmica riu s) i s th e most ser iou s pest of sweetpotato in the Caribbean and South Asia. The Afri ca n species (C. brunneus and C. punctico//is) that occur in sub-Saharan Africa are equal ly injuri ous. Research Achievements Progress in IPM for key pests under study is reported here in the IPM phases at wh ich most of the work was done during the report pe- riod , 1995 -96 . The economi c importance of each pest was assessed before biologica l re- sea rch started. Development of IPM components IPM components in the deve lopmental phase are those for contro l of the African spe- cies of sweetpotato weev il in sub-Saharan Africa, leafminer fly in Latin Ameri ca and the Caribbean, and PTM in Morocco and Yemen. Research on the Africa n weev il s was con- ducted in Uganda and included the study of the I ife cycles and seasonal hi sto ry of the two spec ies, the effects of cu ltural practi ces on weev il damage, and the identifi cat ion and l 60 P1og1om 4 synthes is of sex pheromones fo r both. Phero- mones may become a key IPM component for manag ing the weev i Is. These components are beginning to be tested in pilot uni ts. Resea rch on leafminer fli es included the study of the life cyc le, the seasonal history, p lant-insect interactions (with spec ial atten- tion to the egg-extrusion phenomenon), de- ve lopment of resistant potato c lones, the oc- currence and effectiveness of the comp lex of natural enemi es, the se lective effect of larvi- c ides, and the effects of cultural practices. In Morocco and Yemen, the development of IPM components for PTM co ncentrates on biological contro l. In addition to using granu- losis virus (GV) and Bacillus thuringiensis (Bt), four parasitoids are being reared for release in Yemen and Tuni sia: Copidosoma koehleri, Diadegma mo/lip/um, Che/onus phthorimaea, and Orgilus lepidus. Integration of IPM components into pilot units After severa l years of in vestigat ion, the components for the management of South Ameri can PTM and APW, and the common PTM in northern Africa, have been integrated into pilot units. In some areas, implementa- tion is already extending beyond the pilot unit boundar ies . Latin America and the Caribbean. In Peru, pilot uni ts have been established in six farm communiti es: two in the south (Cusco) w ith the partic ipation of ARARIWA (an NGO) and INIA, three in th e central mountains (Junin) w ith the part icipation of TALPUY (a n NGO) and INIA, and two in the north (Cajamarca) w ith the participation of INIA. Weevil damage in the north was reduced from an initial level of 60% to a current level of 10%, and from 44% to 7-10% in the south. The areas of influence of these pilot units are in- creasing steadily as farmers of neighboring com- munities participate in field days when fa rmers in the pilot units exp lain their experiences. The NGO CARE-Peru created an IPM pro- gram for the management of APW and PTM with the financial support of USAID (1994- 96) and the technical support of CIP. The pro- gram, called MIPANDES, proved very suc- cessful and benefited 3,200 peasant families in the high mountains. Similar pilot units have been establi shed in Bolivia (at Mizque for PTM and at Koll ana for APW) with direct parti c ipation of PROINPA and IBTA. In Ecuador, pilot units are situated in Chimborazo and Cotopaxi with the participation of INIAP. In Co lomb ia, pilot units at Ventaquemada and Motavita have the co ll aborat io n of CORPOICA, the local UMATAs, the Secretaria de Agricultura, and SENA. These pilot units in the Andean coun- tri es are supported financially by the Inter- American Development Bank. Finally, in the Dominican Repu blic we es- tablished pilot units for PTM in Constanza with the collaboration of MIP/JAD. Sprayings for PTM control have been reduced from six per cropping season to only one. Middle East and North Africa. Pilot units for managing PTM in stores were establi shed at 13 sites in Egypt to introduce the use of GV and Bt as alternatives to the toxic insect i- cide fenitrothion, which has been officially banned for use on ware potatoes. Five sites are producing GV and two are producing Bt. More than 250,000 pheromone traps are used to monitor PTM populat ions in the field to determine when to begin control measures and to reduce pesticide use. Three govern- ment agencies are invo lved in the program: PPRI , AGERI, and ARC. Pilot units have been established in three major potato areas in Tunisia: Jendouba, Cap Bon, and Bizerte. They are demonstrat ing the effic iency of GV for PTM control as an alter- nat ive to deltamethrin and malathion, which are commonly used. Large-Scale Implementation A case of extraordin ary growth of IPM with CIP's technical support is the management of the sweetpotato weevil in Cuba. Damage to the Cuban sweetpotato crop had been held to about 10%, with 10-12 sprayings a sea- son, when Cuba was obta ining insectic ides from the Soviet Union. But damage ballooned to about 50% when the insecticides were no longer ava il ab le, beginning in 1991-92. In just three years (1993-96), the first two pilot units covering 230 ha in two provinces (C ienfuegos and Villa Clara) expanded to about 10,000 ha in all 13 provi nces of the island. The Cuban organization INIVIT simul- taneously developed an act ive resea rch pro- gram and an implementation ca mpaign. Pheromone traps, predatory ants, the para- siti c fungus Beauveria bassiana, and strict cul tura l practices were the main components of the program. As a resu lt, weevil damage is back down to 5- 10%, without any use of in - secticides. CIP Progmm Report 1995-96 l 6 l Integrated Management for the Potato Tuber Moth in Pilot Units in the Andean Region and the Dominican Republic M. Palacios and F. Cisneros1 The potato tuber moth complex is the most damaging pest of potato (5o lanum tuberosum) under the warm and dry env ironments of fie lds and stores. It is formed by three main spec ies: th e common moth (Phtho rima ea operculel/a), distributed wo rldwid e; th e Andean species (5ymmetrischema tangolias) , in the Andean va ll eys of Boli via, Co lombi a, and Peru; and the Central America n moth (Tecia so lanivora) , w hich has spread from Central Ameri ca to Co lombia, Ecuador, and Venezuela in recent years. Damage ca used by these pests is around 30% in the fi eld and above 50% in stores w hen not contro lled. Farmers commonly use highly toxic chemi- ca ls to control tuber moths . As a result, pro- duction costs increase, fa rmer health is com- prom ised, moths deve lop resistance to most treatments, and in some places whiteflies (Aleyrod idae) appear as induced pests. Sur- veys show that farmers do not distingui sh dif- ferent moth species or moth behavior in fie lds or in stores, nor do they understand the sea- sonal hi story of the pest. That lack of knowl- edge results in inadequate control methods. CIP has developed a seri es of control meth- ods considered as in tegrated pest manage- ment (IPM) components. They are part of a menu of options being offered to fa rmers in pilot units di stributed in the Andean coun- tries and the Domini can Republic. Rationale of IPM Components The I ife cyc le and control measures of the common potato moth (P. opercu lella) are well known . But l ittle is known of 5. tangolias. Our 1 CIP, Lima, Peru. 162 Program4 studies have shown important differences be- tween these spec ies in terms of duration of li fe cyc les (Tab le 1 ), seasona l history, behav- ior in fields and in stores, favorab le eco logi- ca l conditi ons, and control methods. In most Andean va ll eys, P. ope rcul e /l a and 5. tango lias have three or fo ur generat ions a yea r. 5. tangolias is not a pest at low alti tudes, w hereas P. operculella develops 6 to 10 gen- erations a year. T. solanivora may have up to six generati ons a year at medium to high alti- tudes. Sources of infestation are potato stores and potato fi elds, especia ll y those f ields where potato residue is allowed to remain after har- vest. Moths commonly move in both direc- tions.Field-infested tubers, or tubers infested at harvest, ca rry initial infestations to stores. The var ious IPM components are des igned to reduce pest movement between field s and stores and to lower larva l and adult moth densities. Pheromone traps to capture adult mal es and the use of a Baculovirus to protect stored seed tubers are key IPM components. These measures have impressed farmers and prompted them to adopt other IPM measures as well. Field control measures during crop growth Measures to protect the potato crop from pl anting to harvest are primaril y cultural con- tro l methods. Good soil preparation. Adequate so il preparat ion not only ensures vigorous plant growth but also helps to destroy the remain- ing stages, mostl y pupae, of the tuber moth before pl anting. Table 1. Life cycle in days for the three species of the potato tuber moth complex under conditions of pilot units in Peru, Colombia , and the Dominican Republic, 1996. Species Phthorimaea operculella Symmetrischema Tecia solanivora tangolias Country Dominican Peru Peru Colombia Republic Department/ Province Constanza Cajamarca Cajamarca Antiaquia Boyac6 Av temp (°C) 20.6 15.7 Stage: Egg 5 12 La rva 19 21 Pupa 8 24 Total 32 57 Longevity 12 23 Oviposition period 7 8 Eggs/female 150 150 Timely planting. Based on seasona l oc- currence stud ies, optimum planting dates were determined to co incide with lower tem- peratures and the onset of rains. Farmers who plant in the dry period face high moth in- festations. Deep planting. Covering tuber seed to a depth of 5-1 0 cm prevents female moths from ov ipos it ing in seed tubers and keeps larvae from migrating to tubers from infested aboveground sprouts. Neonate larvae of 5. tangolias can burrow to a depth of 5 cm; those of P opercule lla as much as 10 cm to the seed tubers. High hilling. High hilling of growing plants protects the developing tubers from ovipos it- i ng females and reduces the poss ibi I ity of lar- vae reaching the bulking tubers. High hilling can reduce damage by 30%. Frequent irrigation. Ad equate water ing and cultivation prevent cracks from fo rm- ing in the soil. Soil cracks all ow female 13.0 13.0 16.0 13.0 21 17 10 14 71 57 20 30 36 31 20 23 128 105 50 67 32 33 22 22 24 25 13 13 138 185 101 290 moths to reach the potato tubers for oviposi- tion, and provide shelter for adu It moths. Spri n- kl i ng irri gation alone reduces damage by 30%. Pheromone traps . Commercia l phero- mones are avai lable for P opercu/ella and T. solanivora. M ass trapping of male moths re- duces the probabi liti es of moth mating, thus causing a drop in egg fertility. Pheromone traps can reduce infestation by 50%. Control measures at harvest The two most important control measures at harvest are protecting harvested tubers from ovipositing fema les and removing crop res i- dues from the fi eld . Timely harvesting. During the last phases of the crop (tuber filling and plant senes- cence), the infestation rate accelerates. De- laying harvest by 1 or 2 mo can increase dam- age by as much as 70%. CIPProgrom Reporr 1995-96 1 63 Storing healthy tubers. Only healthy tubers should be stored. Infested tubers should be buried under at least 10 cm of soi I. Covering tubers. Female moths become acti ve in the evening and most eggs are laid at that time. Harvested tubers shou ld not re- main exposed to ovipos iting fema les over- night. If they cannot be stored immediately, tubers shou Id at least be covered; otherw ise the infestation level cou ld reach 60% w ithin a few days. Destroying harvest residues. P. operculella and, to a larger extent, 5. tangolias pupate in tubers and dry stems left in the field. Moths from these pupae infest the crop the fo llow- ing season (Table 2). Also, tubers left in the field become volunteer plants in the rotation crop . For these reasons, all harvest res idues must be destroyed. Measures to avoid damage in stores Protecting tubers in stores can reduce moth damage by 70% to 95 %. Cleaning stores. Cleaning floors, walls, and ce ilings of rustic stores before storing healthy tubers destroys pupae and other I ife stages of the moth. Storing healthy tubers. Tubers should be sorted and infested ones discarded before stor- ing. Tubers exposed to moth oviposition shou ld not be stored, as eggs are commonly over looked during sorting. Storing infested tubers or those that have been exposed to moth oviposition along with hea lthy tubers may result in infestation of the entire store in as li tt le as 4 mo. Using Baculovirus. A dust formulation of Baculovirus phthorimaea, containing 20 dis- eased larvae/kg at a dose of 5 kg/t tubers, is the most effective insecticide against P. operculella. Protection is somewhat lower in the case of the other two moth spec ies. Tu- bers shou ld be treated before larval infesta- tion begins. Repellent plants . Th e fo li age of some plants, ri ch in essentia l oils, such as Eucalyp- tus spp., Lantana camara, and the native spe- cies Schinus mo/le and Minthostachys spp. , repels potato moths. The leaves are dried under shade, crushed, and then used to cover the stored tubers. On the average, protected tubers are 80% less damaged than nontreated tubers. Pheromone traps. Commercial phero- mones disrupt mating of P. opercule lla and T. solanivora during storage. Tubers stored with pheromone traps are about 95 % less infested than the control stores. Diffused-light stores. Potatoes stored in diffused light are generally about 70% less infested than those stored in the dark. 111 um i- nation results in green ing of the tubers (w ith glycoa lkaloid formation), which is unfavor- able fo r the moth . In addition, the arrange- ment of tubers in diffused-light stores fac ili- tates the periodic elimination of damaged tubers. Table 2. Number of moths emerged per hectare' from harvest residues (tubers and stems) and volunteer plants. Pilot units, Peru and Colombia, 1996. Species S. tango/ias (Peru) P. opercu/ella (Peru) T. so/anivora(Colombia) a. Sampling in pilot units (25 sampl es of l m1 each) 164 Program 4 Tubers 49,000 43,000 32,000 Moth larvae (no./ho) Stems 118,000 Volunteer plants 481,000 Selection of Pilot Units IPM components are tested in the fi eld for fi- nal adjustments and eva luation of fa rmers' acceptance. Th at is done in pil ot units of par- ti c ipating fa rmers who perce ive the impor- tance of the pest and are interested in trying new methods for its contro l. The parti cipa- tion of governmental agencies and nongov- ernmental organizations (NGOs) is cruc ial, because the pil ot unit is considered as a dem- onstration area for IPM implementation at the nati onal level. The foll owing acti v ities take pl ace in the pil ot unit: evaluation of the problem, training of fa rmers and local counterparts, implemen- tati on of IPM components, adaptive adjust- ments of components, and identi fication of new resea rch opportunities. Training and field days are essential in the pil ot units. They are designed to motivate all loca l stakeholders (including fa rmers), tech- nicians, rural schoolteachers, uni versity stu- dents, and community authoriti es to ensure the stabi I ity of the program. CIP provides tech- ni ca l support, parti c ipates in training, and provides bas ic instructional materi als. Moths (no./trap/wk) 600 500 400 '1 300 200 100 0 \ Month 0 N D J F M A M J J A S CIP set up IPM pilot units at two sites in Peru and Co lombia in 1995 and one site in the Domini ca n Republi c in 1993. The IPM components adopted by fa rmers in the pilot units and the results they obtained are re- ported here. Pilot units in Peru Two pilot units have been establi shed in Peru : one in the southern community o f Urquillos, Cu sco; and one in central Peru at Carhuapaccha, Huancayo. In those areas, two spec ies of moth s occur: P. opercule lla in the w armer va ll eys, and at higher altitudes 5. tangolias, which is rapidly increas ing in im- portance. No commercial sex pheromone is avail able for the contro l of 5. tango!ias. The most important bi o logica l control agent is a Baculovirus. Urquillos. Potato is grown year-round in Urquillos. During the rainy season, potatoes are grown at elevations of 3,600-3,800 m. In the dry season, they are grown under irriga- tion at lower elevati ons (2 ,800 m) where the crop is stored under constant pest pressure. The dominant spec ies is 5. tangolias, although P. opercule lla is also present (Fi gure 1 ). The pest problem is most serious in stores. Prac- J 111 P operculella S. tango/ias DJFMAMJ A S 0 N D Potato I Vegetablej I Potato I Vegetable I Before IPM I :=I ======~l...!oPM=im=p=le=m=e=nt!.._at-io-n========'. Potato 1996 Figure 1. Potato tuber moth field populations in two consecutive years showing the effect of IPM implementa- tion in Urquillos, Cusco, Peru. CIP Prog ram Reporl 1995-96 l 65 tices adopted by farmers (largely by women of the community) are cleaning of stores and use of repe l lent plants and pheromone traps agai nst P operculella. Clean ing alone re- duced infestation in rustic stores to onl y 10% of th at experienced previously. Field practices adopted are deep planting, high hilling, irri- gation, and timely harvest. Before the IPM program began in 1994, average damage in stores was 65% despite the use of highly toxic insecticides such as parathion. In 1996, average damage for farm- ers participating in the IPM program was 3.4% (0-12%), w hereas other farmers experienced 40-80% damage. Farmers in Urquillos have formed an IPM committee to expand the use of IPM in the area, and the Farmers Federation of Cusco now promotes IPM. Carhuapaccha . Potato production in Carhuapaccha is com merciall y oriented and insect ic ide use is intensive. Potato is pro- duced at the higher elevat ions of the com- munity (3, 160-3,800 m). Irri gat ion is avail- ab le at lower elevations and fa rmers produce other vegetab les. Storage is limited to seed tubers and tubers for se lf-consumpt ion. The bulk of production is so ld in the loca l mar- ket. Practices adopted in stores to control 5. tangolias were cleaning stores, sorting, use of Baculovirus to protect tuber seed, and re- pel lent plants to protect potatoes for home use . About 50% of the tubers were infested in stores before the program started in 1995; damage in 1996 was reduced to 10%, w hereas damage o n ne ighbo rin g farms ranged from 65% to 85%. Pilot units in Colombia The most damaging pests of potato in Co- lo mbi a are the Andean potato weev i I (Premnotrypes spp.) and T. solanivora, which was introduced from Venez uela to Co lombia in 1985. At first it was restricted to the Norte de Santander Department, but by 1996 it had 1 6 6 P1og1om 4 spread to the south of Colombia (Narifio De- partment) close to Ecuador. The aggress ive- ness of th is pest, w hich infests tubers but not abovegro und plant parts , c reated panic among farmers, w ho then started heavy spray- ing of fie lds and stores. CIP' s ass istance in dealing w ith the pest was requested in 1994. A work program ori- ented toward implementing IPM began in two pilot units, one in El Santuario, Antioquia, and the other in Ventaq uemada, Boyaca. El Santuario. The Corporaci6n del lnstituto Co lombiano Agropecuario (CORPOI CA) was CIP's counterpart, leading in the implemen- tation of Cl P recom mendations fo r the area. The first step was an intensive ca mpaign to teach farmers and technicians how to recog- nize this new pest and its life cycle, behav- ior, and movements. The effecti veness of sex pheromone trappi ng was demonstrated, the effect iveness of Baculovirus was tested, and training materia ls were produ ced. Farmers adopted the following practices: pheromone traps in fields and stores, use of Bacillus thuringiensis (Bt) and Baculovirus in stores as alternati ves to tox ic insecticides, use of healthy seed, high hilling, timely harvest, and c leaning stores. In 1994, 50% of the area planted to po- tato was in fested w ith T. solani vora . Thirty percent of tubers were infested at harvest. In stores, 100% tuber infestat ion was common. In 1996, the average number of infested tu- bers at harvest for the whole Anti oquia De- partment, including the pilot units and differ- ent leve ls of IPM influence, dropped to 4.4% (0-43%), and average damage in stores was 11.4% (0-37%) . During the last season, about 23 ,300 pheromone traps were used. Ventaquemada. Until 1993, the onl y po- tato tuber moth spec ies reported damaging potato was P opercule lla. Average infested tubers at harvest was low, 3.5%, because of the intensive use of insecticides (12 sprays per season). In 1994, the presence of T. so lanivora was confirmed in Boyaca and heavy losses were reported-56% infested tubers at harvest. The most affected munici- pality was Ventaquemada. A program deve l- oped included research on seasonal occur- rence, studi es on the effectiveness of Baculovirus and sex pheromone trapping, and the implementat ion of sanitary measures. An intense training program was developed for farmers and technicians. In fi eld s, farmers adopted destruct ion of crop res idues, sex pheromone trapping, high hilling, and timely harvest. In stores, they adopted c leaning, use of Baculovirus to pro- tect tuber seed, pheromone traps, and, to a limited extent, diffused-light stores. Farmers that adopted IPM measures had only 1 % infested tubers at harvest, whereas other farmers had 40% damage desp ite the use of insect ic ides (Figure 2). The selling price of IPM potato was 35% higher than that of insect icide-treated potato. A total of 14,800 pheromone traps were used in 1995 over an area of 925 ha. The Secreta ria de Agricultura de Boyaca has started producing Baculovirus to cope with the increas ing demand. Moths (no./plot/wk) Pilot unit in the Dominican Republic In the Dominican Republic, the main po- tato-producing area (85% of national produc- tion) is the high Constanza plateau, w here other vegetab les are also culti vated . Intensive spraying is a characteristic of the area. Six to eight sprays per season (at 6-8-day intervals) to control P operculefla are common. Under heavier infestat ions, potato fields are sp rayed at 3-4-day intervals. Th e lo ca l se mipri va te organization Programa Nac ional de Manejo lntegrado de Pl agas (MIP) started a program to reduce the excessive use of insecticides in 1991 . In 1993 , CIP coordinated a program with MIP to imple- ment IPM on potato. Farmers adopted pl anting hea lthy tubers, deep pl an ting, sex pheromone trapp ing, fre- quent irrigat ion, timely harvest, and the use of Bt. Action thresho lds were established for determining the timing of Bt applications. Field popul at ions of the moth have been reduced from an initial catch of 136 moths/ trap/week to 19. In 1996, infested tubers at harvest in IPM fi elds were 0.1 % without the Damaged tubers (%) 2,000 .-------------------------- . 50 D Inside plot [] Outside plot 40 1,500 ------------------------- 1,000 500 Apr Total capture 5,422 May June July Month Conventional ----+ 30 20 10 0 Aug Sept Figure 2. Population variation of Tecia solanivora in IPM plots and percentage of damaged tubers at harvest. Ventaquemada, Boyaca, Colombia, 1996. CIP Program Report 1995-96 l 6 7 use of insect icides. Fields with conventional control programs had 15% infested tubers. Average yield in MIP fields was 13.8 t/ha com- pared with 8.9 t/ha in fields w here conven- tional control practices prevailed. Conclusions The implementation of IPM for the manage- ment of the potato tuber moth in pilot units in Peru , Colombia,. and the Dominican Re- public has demonstrated that Cl P's approach is viable, well accepted by fa rmers, and cata- lytic for the parti cipat ion of loca l institutions, both governmental and NGO. Net benefits were evident: pest damage decreased sign ifi- cantly and the use of insecti c ides dropped to a minimum. The acceptance of IPM by farmers was due not only to the effective components devel- oped by CIP's program but also to the ad- equate training programs for farmers, techni- 168 Program 4 cians, and other stakeholders, and the par- ticipation of local inst itutions. In this case, a total of 9,728 persons recei ved some kind of training in 1,207 act ivit ies, most of them car- ried out by local institutions. Technologies such as mass trapping w ith sex pheromones and the use of Baculovirus have stimulated the interest of various private and official organizations in commercially producing these biological control products to make them w idely available to farmers. In Peru , the participation of CARE-Peru, an NGO, expanded the influence of the IPM pi- lot units to 3,500 families in the poorest ar- eas of the mountains. This successful program is now being extended to 10,000 families by a collaborative effort of CARE and the Programa Nacional de Manejo de Cuencas Hidrograf icas y Conservaci6n de Suelos (PRONAMACHCS), a government-related organization. Integrated Management for Andean Potato Weevils in Pilot Units J. Alcazar and F. Cisneros 1 Potato (So lanum spp. ) that is culti vated in the high Andean mountains (2,500-4,750 m) of Bo li v ia, Peru, Ecuador, Co lombia, and Ven- ezuela is severely damaged by the Andean potato weevil or white gru b. This is a com- pl ex of spec ies, most of them belonging to the genus Premnotrypes (Curculi onidae) . The dominant spec ies are P. latithorax (Pierce) in Bo li via and southern Peru , P. suturica llus Ku sc he l in ce ntra l Peru , and P. vo rax (Hustache) in northern Peru, Ecuador, Colom- bia, and Venezuela (Figure 1 ). Species of other genera, Rhigopsidius tucumanus Heller and Phyrdenus muriceus Germar, are largely re- stri cted to Bo livia. 1 CIP, Lirn a, Peru. Premnotrypes suturicallus _______. Premnotrypes latithorax M ost inhabi tants of the high Andes are poor peasants, w ith an extremely low level of li teracy, who receive li ttl e, if any, techni- cal ass istance. Th ey accept th e da mage caused by weev il s (commonly above 50%) as inevitable and occas ionally abandon fi elds because of high infestat ions. The divi sion of land into small uni ts has rendered impracti- ca l the old system of large communal land rotati ons that was once used to effi cientl y control weev il s and other pests and diseases. Farmers cl ose to c it ies are more commer- c iall y ori ented and use tox ic insect icides (ca rb ofur an, pa rat hi on, aldi ca rb , and methamidophos) to control weevils . Despi te ARGENTINA Figure 1. Distribution of the main species of Andean potato weevil. CIP Progrom Report 1995-96 l 6 9 the use of insecticides, 20-30% tuber infesta- tion is common . Although all farmers are fa- mili ar with the larvae that bore into tubers, they know littl e else about the insect, includ- ing its life cycle and seasonal hi story, w hich would help them improve their pest control practices. Rationale of I PM Components The three common spec ies of weev il s have similar life cycles (Tab le 1) (one generation a year, except for P. vorax, which presents two generations in some parts of Ecuador and Co lombia) , behavior, and seasonal hi story (Figu re 2). Research conducted on these characters led to the deve lopment of a series of control measures, most of which could be inc luded as components in integrated pest manage- ment (IPM). The absence of parasitoids, limited preda- tors, and only a single pathogen (Bea uveria brongniartii (Saccardo) Petch) made it clear that cultural practices would be major con- trol components. Even the pathogen had a scattered distribution in the field. Other im- portant considerations were: al I potato culti- vars are susceptible to weev i I attack; the wee- v i I invades potato f ields by crawling from overw intering places because it is unable to fl y. Sources of infestation are abandoned po- tato fields, fields harvested the previous year, places w here potatoes were piled up during harvest and sorting, rustic stores, and vo lun- teer plants in rotat ion crops. Most control measures aim at destroying the overw inter- ing population, interrupting migration of the weev i I to new fields, and reducing the wee- vi l population in infested fi elds. Reducing weevil infestation in the field There are several ways of reduc ing in-fi eld populations of potato weevil. They include earl y planting and use of early-maturing va- ri et ies, timely harvest, use of healthy tubers, handpick ing adult weev ils from the crop, and destroying vo lunteer potato plants in rotation fields . Early planting. Emergence of overwinter- ing adults lasts 8-14 wk and coincides w ith Table 1. Life cycles of the three most injurious species of Andean potato weevils in Cajamarca (Premnotrypes vorox), Huancayo (P. suturicollus), and Urubamba (P. lotithorox), Peru. Stage Egg Larva I Larva II Larva Ill Larva IV Larva V Pupo Wintering adult Lo ngevity Total cycle Ovipositing period Egg~female 170 Program4 P. vorax 43 16 14 17 18 58 50 66 181 463 119 374 Life cycle (days) P. suturicallus P. latithorox 33 48 11 9 9 8 12 8 57 35 54 29 115 110 143 179 438 424 106 101 631 162 Weevils/1-m trap or eggs/female 50 20 10 0 (%) 100 80 60 40 20 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Months D overwintering adults(%) D tuber infestation(%) ~ pupae(%) ~ oviposition • prepupae (%) II immigration to potato field II adult emergence Figure 2. Seasonal history of the Andean potato weevil (Premnotrypes latithorax) in Urubamba, Peru. Average of two years (1994-96) . the rainy season. Females lay eggs for 12-14 wk. As more adults emerge, there is a con- tinuous increase in the number of females that find later plantings as ovipositing places. Early-planted fields are 4-5 times less infested than those planted late. Th is measure reduced weevil infestation and damage by 82.2% (Table 2). Early-maturing varieties are less exposed to weevil infestation than late vari- eties and accomplish the same resu It as early planting. Timely harvest. Delayed harvest extends exposure of tubers to weevil infestat ion, al - though farmers sometimes delay harvest in the hope of getting higher prices. Occasion- ally, early harvesting is recommended. A de- lay of 10 d in harvest resulted in an increase in tuber damage from 6.8% to 29.5% in one field check. Healthy planting material. In general , healthy tuber seed favors vigorous growth of potato plants. In particular, seed tubers dam- aged by Andean potato weevils rot readily, which results in poor stand establishment. Planting seed that had been bored by the Andean weevil reduced yield by 31 % com- pared with planting non infested seed (Table 2). Handpicking weevils. Adult weevils in the potato field become active in the evening; they climb the foliage for feeding and mat- ing. They can easily be captured by shaking the foliage over any container. Farmers have readily accepted this practir.e. Collecting weevils six times in a growing season reduced damage by 34% (Table 2). Destroying volunteer plants. Volunteer potato plants in rotation fields are important sources of weevil infestation and should be destroyed as early as possible. Initially, some farmers were reluctant to accept this practice because they cou Id harvest some early pota- toes from volunteer plants. But more than 30,000 weevils infested 6,500 volunteer plants/ha in a rotation field in Huancayo. CIP Program Report 1995-96 l 71 Table 2. Experimental results showing the benefits of measures recommended to reduce weevil infestation and damage' (experiments conducted in Cusco and Huancayo, Peru) , 1993-96. Measure Check Tuber domoge {%) Decrease in or yield {MT /ho) domoge {%) Early planting (October} 3.2% 82.2 Late planting (December} 18.0% Timely harvest (l 0 April} 6.8% 77.0 Delayed ha rvest (20 April} 29.5% Healthy tuber seed 9.6 MT/ha 31.3 Infested tuber seed 6.6 MT/ha Handpicking weevils (6 times) 23.8% 34.l Handpicking weeevils {none} a. Expressed as percentage of infested tubers at harvest. 36.1% Interrupting adult migration and larval movement Because adult weevils ca nnot fly, physi- ca l barriers effect ive ly halt their migration from field to field as do peripheral trenches, which allow migrating weevils to be captu red. Larva l movement from tuber to soil can be prevented by using plastic sheets at harvest and so rting or storing tubers in diffused-l ight stores. Weevil larvae and pupae are often fed to chickens. Barriers. Migratory weev ils can be inter- cepted by digging field peripheral trenches and capturing or killing them w ith insecti- cides, or by simply spraying a 3-m-w ide band of in secticide around the perimeter of the field. Bordering fields with non host plants also disrupts the migratory process, and is prefer- ab le to chemical control w here it can be prac- ticed. Table 3 shows the reduction in tuber damage resulting from these practices. Shelter traps. Straw bundles, pieces of si- sal , or plastic sheets and other materials pro- vide she lter to weev ils during the day w hen they can be captu red. An alternative is to put insect icides in the shelter. Better effects are obtained w ith insecticide-treated potato fo- li age placed in new potato fields. Thi s prac- ti ce has been stud ied and well accepted in Ecuador. 172 Program4 Ground sheeting. At harvest, many full- grown larvae abandon the tubers and dig into the soi I to pupate. Larvae ca n be intercepted on their way to the soil by barriers of plastic sheets or other materi als. Diffused-light stores. Seed tubers kept in diffused-light stores maintain better quality. Because the tubers are not in contact w ith the soil , the larvae abandoning the tubers can be destroyed or fed to chi ckens. Reducing overwintering populations Measures to reduce overw inter ing weevil populations can be carried out in the field or in stores (Tab le 4). H arvested fields are w in- ter-plowed to destroy larvae that move from tuber to soil. Breaking the soil where pota- toes were piled at harvest or for sorting also destroys underground larvae and pupae. Plowing abandoned fields is effective in de- stroying overwintering insect stages. Winter plowing. Many larvae reach ma- turity before harvest and leave the tubers to overwi nter underground as pupae. Most of these pupae are destroyed by plowing the field 2 or 3 mo after harvest. At least 50% of the larvae and pupae in the so il are destroyed w hen plowing is done w ith oxen . Chickens help destroy exposed larvae and pupae. Table 3. Experimental results showing the benefits of measures recommended to intercept migration of adult weevils and movement of larvae from tubers to the soil (average of experiments conducted in Urubombo and Huancayo, Peru), 1993-96. Measure Trenches in store Trenches in field (insecticide treated) Non host plant barrier and picking weevil Sprayed border Sheets at harvest Sheets at pre-store sorting Intercepted population/damage at harvest Reduction in damage (%) 23,430 weevils (stores: 70 m2) 134 weevils/m of trench 68.9 12.7% vs. 35.0% (check) 63.7 6.7% vs. 19.9% (check) 66.3 1,515 lorvae/l 00 kg potato (2 4 hours) 23,450 lorvae/l 00 kg/90 days Table 4. Preferred adoption ( > 50% of target population) of IPM components by communities integrating IPM pilot units in Peru. Pravince Community Harvest on time Handpicking weevil Volunteer plants Trenches Plant barriers Border sprays Shelter traps Sheeting at harvest Diffused-light stores Winter plowing Breaking soil Chickens Parasitic fungus Urubamba Huatata x x x x x x x x Tayacaja Aymara x x x x x x x Huancayo Casabamba Chuamba Cajamarca Chilimpampa x x x x x x x x x x x x x x x x x x x x x x CIP Prog ra m Report 1995-96 l 7 3 Abandoned fields are another major source of weev il reproducti on and migration. Dur- in g w inter, aba ndo ned f ie ld s should be plowed at least twice. Breaking the soil. When recentl y har- vested potatoes are not pi led on sheets, large numbers of larvae d ig into the so il to pupate. The soi l shou ld be broken up in wi nter to destroy underground larvae and pupae. More than 90% of overw inter ing larvae are de- stroyed by th is practice. The paras itic fungus Beauveria brongniartii can be used to contro l larvae digging into the floor of ru stic stores. Mortality surpasses 80%. Selecting Sites as IPM Pilot Units Select ing the pilot unit is a crucial step in the implementation of IPM. Here, the scientifi c development of IPM components meets the rea l world of farmers- the ir risks, interests, economics, culture, cropping practices, and other facto rs that can determine the success or fa ilure of the program. Thi s precedes the large-sca le implementation of IPM. In addi- tion to farmers, other key players are mem- bers of nat ional agri cul tural research systems (NARS), nongovernmental o rga ni zat ions (NGOs), and other groups that will eventu- ally assume responsibility for expanding the program. Two condi ti ons had to be met fo r a com- munity to qualify as a pilot unit. The first was that farmers had to recognize the Andean potato weev i I as the most ser ious pest prob- lem. Second, there had to be a nucleus of farmers wi 11 i ng to try new methods to reduce tuber damage and insecticide use. Selected sites in Peru we re Hu atata, Urubamba Province; Chu amba and Casabamba, Hu ancayo Prov ince; Aymara, Tayacaja Prov in ce; and Chilimpampa, Cajamarca Province. Four other sites were chosen outside Peru: La Paz Department in Bolivia, Cotopaxi and Chimborazo provinces in Ecuador, and Boyaca Department in Co- lombia. l 7 4 Program 4 Description of Main Pilot Units W ith p ilot units, CIP's IPM staff works close ly with farmers to eva luate the effect iveness of pract ices and the reaction of fa rmers to the IPM program. Farmers from neighbor ing com- munities, d irect ly or by influence of loca l or- ganizations, commonl y adopt practices they have seen demonstrated in the pilot units. These anc illary "areas of influence" provide some assistance in diffusing IPM co ncepts. Farmers in the pilot units were trained on the basics of the program to make them fully aware of the options presented as IPM com- ponents. We organ ized short courses, field days, and w orkshops and produced training material s such as bulletins, posters, samples, slide sets, flip charts, and v ideos. Train ing was extended to selected collaborators from sev- eral inst itutions w ho would take over field work in the large-sca le implementation phase. In a four-year period (1992 -96), 3,984 fa rm- ers took part in field days in Jun in, 2,058 heard talks in Cajamarca, 1,627 parti c ipated in weev il-picking contests in Cusco, and 1,835 rece ived short courses in Cajamarca. Pilot units in Peru Pil ot units in Peru were se lected to opti- mize the evaluati on of the IPM components under different condition s. • In the south (U rubamba), w ith the weev il species P latithorax and traditionally ori- ented potato producti on. • In central Peru , with P suturica llus, in more commercially o ri ented Huancayo and in more tradition all y oriented Tayacaja. • In the north (Cajamarca), w ith P vorax and less traditional varieti es. Huatata. Practices that have more than 50% acceptance are t ime ly harvest, use of chickens as predators, breaking the so il o r using sheeting, handpicking weev il s, destroy- ing volu nteer plants, and usi ng 8. brongniartii in rustic stores. Although some farmers have stopped using insectic ides, many others sti 11 use one spray of a "safe IPM compatible" product, as it is described by insecticide dealers. Before the program began in 1990 and despite heavy use of insecti cides, damage(% of infested tube rs at harvest) was 44%. In 1996, damage va ri ed from 8% to 12%. Chilimpampa. Most fa rmers have adopted handpick ing weevil s, t imely harvest, use of tarw i or barley as nonhost field edge barri - ers, use of sheet ing at harvest and so rti ng, w inter plowing, use of ch ickens as predators, breaking of soi l, and shade traps. Farmers have adopted anywhere from three to f ive practices. When the program began in 1993, farm- ers had an average damage of 61 % of infested tubers at havest. During the 1996 harvest, damage dropped to 13%. Farmers in the com- munity have a fairly good knowledge of the biology and seasonal history of the pest and its control. Aymara . Most fa rmers adopted the use of sheet ing at harvest, eliminat ing vo lunteer plants, timely harvest, use of chi ckens as predators, w inter plowing, and border spray- ing. Some practices such as destroying vol - unteer plants and handpick ing weevil s are organized and conducted by the who le com- muni ty. Damage in 1993 was 39% of tubers at harvest despite the use of insectic ides. During 1996, damaged ranged from 15% to 20%. Casabamba and Chuamba. Most farmers adopted handpicking of weevils, use of sheet- ing at harvest, t imely harvest, high hilling, use of diffused-light sto res, use of chickens as predators, destroying vo lunteer plants, w in- ter plowing, and border spraying. Damage in Casabamba dec reased from 45% in 1994 to 19% in 1996. In Chuamba, the program was adopted only in 1996. Damage in demonstra- tion fields was 13%, w hereas damage in the rest of the area was 41 %. Pilot units in Bolivia, Colombia, and Ecuador The Peru vian experi ence was extended to Bolivia, Co lombia, and Ecuador fo ll ow ing similar criteria in the se lection of the pilot uni ts. In t hi s resp ec t, PROINPA-IBTA , CORPO ICA, and FORTIPAPA-INI AP played key roles. Bolivia. Farmers adopted f ield border trenches, winter so il breaki ng, use of chick- ens as predators, timely harvest, use of sheet- ing at harvest, border sp ray ing, and destroy- ing vo lunteer plants. Infected tubers at har- vest in the 1994-95 cropp ing season reached 44%. This was reduced to 30% in 1996, the first year of the program. Ecuador. Most farmers have adopted the use of shelter-treated traps, treated potato plant-ba its, crop rotation, destruction of vo l- unteer plants, breaking the soi l in sorting and stor ing pl aces, and se lect ive chemi ca ls (Orthene instead of carbofuran). Damage at harvest in IPM fields in 1996 var ied from 6% to 22%; in nonparti c ipat ing fi elds, damage vari ed from 33% to 97%. Colombia. Farmers in 1995-96 concen- trated on reducing spraying, which is com- mon in the area. Instead of broadcast treat- ments, spraying was limited to edge bands with similar results . Complementary practices include destroying crop res idues and moni- toring the weevil populat ion near field edges. More recently, diffused-light stores are being adopted to reduce damage from the weevil and the potato tuber moth Tecia solanivora. At the beginning of the program, damaged tubers at harvest vari ed from 20% to 60%, depend ing on the intensity of chemical use. During 1996, fields w ithin the program had 2-12% damage, wh ereas fields with trad i- tional (most ly chemical) protection susta ined 8.5% to 60% damage. Conclusions Working w ith farmers in the high Andes requires more resources than CIP has available. At the same time, there are governmental organiza- tions and NGOs in the area working to improve farmers' productivity and general welfare. These institutions benefit by taking advantage of ex- isting technologies such as those developed by CIP for the management of the Andean potato weevil. CIP in turn gains va luable research part- CIP Program Report 1995-96 l 7 5 ners. Figure 3 shows CIP cooperators in the Andean region IPM program. Most of the resea rch needed to develop IPM co mponents for the Pre mnotry pes Andean potato weev il s has been completed. There sti ll remain s the task of improving the use of the parasiti c fungus B. brongniartii and deve lopin g m easures for the co ntrol of Phyrdenus and Rhigopsidius weevi ls w hose behavior differs from that of the other spec ies. Two or three more years of implementa- ti on in the p il ot units w ill give a c lea rer pie- I Cajamarca I San Pablo Contumaza (APW-PTM) Otuzco Julcan (APW-PTM) Huaraz Aija Recuay (APW-PTM) Pu no Lam pa San Roman Chucuito Yunguyo (APW) ---- APW : Andean potato weevil PTM : Potato tuber moth C:=J CIP supervised pilot units ture of the year-to-year variations in climate and pest incidence. Thi s time w ill also allow us to consolidate the training of personnel from co llaborating instituti ons that wi ll be invo lved in the large-sca le implementat ion of IPM in areas w here the Andean weev il is the key pest. Future resea rch shou ld focus on the man- agement of the fl eabeetle Epitrix spp., w hi ch is the second most important pest for potato in the high Andes. Boyaca (APW-PTM) Figure 3. Integrated management of potato pests in the Andean region; location (provinces) of pilot units. Collaborating institutions are Corporaci6n Colombiana de lnvestigaci6n Agropecuaria (CORPOICA) , Colombia ; lnstituto Nacional Aut6nomo de lnvestigaciones Agropecuarias (INIAP), Fortalecimiento de la lnvestigaci6n y Producci6n de Semilla de Papa en el Ecuador (FORTI PAPA) , Ecuador; lnstituto Boliviano de Tecnologia Agropecuaria (IBTA), Programa de lnvestigaci6n de la Papa (PROINPA), Bolivia; lnstituto Nacional de lnvesligaci6n Agraria (INIA), Grupo de lnvestigaci6n y de Desarrollo de la Ciencia Andina (TALPUY), CARE-Peru , and Asociaci6n ARARIWA, Peru. 176 Progrnm4 Developing I PM Components for Leafminer Fly in the Canete Valley of Peru N. Mujica and F. Cisneros1 The leafmin er fly Liriomyza huidobrensis Blanchard is a ser ious pest of potato in many places where potato is intensive ly cultivated. It is the most damaging pest in the coastal va ll eys of Peru . Farmers in the Canete valley try to control the pest by spraying 8-1 3 t imes per season on a ca lendar bas is. Insecticides are the highest input cost (an average of US$600/ha) fol lowed by ferti I izer, fungicides, and manure. Noncontrolled infestat ions com- monly reduce yie lds by 50% or more. In most pl aces, leafminer fli es have devel- oped signifi ca nt leve ls of resistance to most carbamate, organophosphate, and pyrethroid insecticides commonly used to ki 11 adult fl ies. Rapid deve lopment of resistance and natu- rally occurring pest to lerance of many chemi- cals seem to be a genera li zed characteristi c of L. huidobrensis. For these reasons, this spe- cies is commonly rega rded as the most diffi- cult-to-kil 1 leafminer fly wor ldwide. Frequent spraying in the coast of Peru has created additional prob lems such as severe infestat ion s of th e wh ite mite, Poly- phagotarsonemus latus (Banks), and the budmidge, Prodiplosis longifila (Gagne). Although nat ive to the neotropi cs (i nfesta- tions are reported from Peru, Chile, Argen- ti na, Braz il , Central Ameri ca, and Mexi co), L. huidobrensis is rapidl y spreading to other areas w here it readily becomes a ser ious pest. It has recentl y been reported in Europe, sev- eral countri es in Africa, and in Indones ia, Malaysia, and Israel. 1 CIP, Lima, Pe ru. Life Cycle, Behavior, and Seasonal Abundance The li fe cycle of the leafminer fl y has been studied in the laboratory and the greenhouse in Lima and Canete, using potato and bean as host plants. We have determined the du- rat ion of each deve lopmental stage, ovipos i- t ion capacity, sex ratio, and adu It longev ity. Tab le 1 summarizes thi s information. The adult female punctures the upper or lower surfaces of tender leaves w ith her ov i- positor, or egg- lay ing structure. Males and fema les feed on the exudates produced by the lesion s. Sim i lar punctures are used by the female to encrust the egg in the leaf tissue. Neonate larvae start tunne ling th rough the chloropl ast-containing spongy mesophyll leaf layer. Tunnels increase in diameter as the larva grows. Although any part of the leaf bl ade can be tunneled, full-grown larvae tend to stay c lose to the midrib. Full-grown larvae pupate inside barre l-shaped pupar ia on the leaf surface. Puparia fa ll to the ground and remain there until the emergence of the adult fli es. Mating occurs 2-3 days after emergence. There is a clea r seasona l vari ation of fly popul ati on densiti es in the Canete va ll ey where ear ly potato pl antings take pl ace in March (end of summer) and harvest of late plantings occurs in December (end of spr ing). Higher popul ation densities occur during the w inter months- June, Jul y, August, and Sep- tember. Fly populat ions are low during the warm months (November-March). Within a given field, the fly popul ation is assoc iated with the phenology of the potato CIP Program Repo rt 1995-96 l 7 7 Table 1. Life cycle (days) of the leafrniner fly Liriomyza huidobrensis under screen house ambient conditions, Canete, Peru , 1996. Stage August (winter) min max Egg 3 4 Larva 3 3 II 3 3 Ill 4 4 Pupo 14 18 Tota l immature stages 27 32 Longevity Female 4 30 Mole 3 6 Total life cycle Female 31 62 Male 30 38 Eggs/female 40 285 plant. For example, there is a relatively slow increase during the vegetat ive growth and a rapid and susta inab le increase during flow- er ing, fo ll owed by a decline as p lants enter into senescence. This trend is more notori - ous in the case of the larva l popu lation . Host-Plant Range The leafminer fly is a po lyphagous insect that infests a large number of crop and ornamen- ta l plants . The lo ng li st includes potato, beans, peas, alfa lfa, and most vegetab les (to- mato, ce lery, lettuce, peppers, spinach, cu- curbits, and others) grown in Canete and other va lleys of the Peruvian coast. Thi s has impli cations w hen crop rotat ion is being con- sidered . In addition , leafm i ner flies infest many weed spec ies . l 78 Pmgmm 4 October (spring) av min max av 3.6 3 4 3.3 3.0 3 4 3.1 3.0 2 3 2.5 4.0 3 3 3.0 16.3 12 17 14.0 29.9 23 31 25.9 15.7 3 28 25.9 4.7 2 4 3.0 45.6 26 59 39.9 34.6 25 35 17.0 116.6 11 704 161.4 In a study to determine the role of weeds as sources of leafminer fl y infestat ions and as refuges of parasitoids, 24 weed spec ies were identifi ed cor respond ing to 13 fam ilies . Sig- ni f icant levels of fly larva l parasitism (40-70%) we re found in Trianthema portu lacastrum (A izoaceae) , Stachys arvensis (Lam iaceae), Che n opodium mura le and C. album (Chenopod iacea e), Datura stramonium (So lanaceae), Bidens pilosa and Ca linsoga parviflora (Asteraceae) , Diplotaxis muralis (Brasicaceae), Ma lva parviflora (Ma lvaceae) , Ricinus communis (E up horbiaceae), and Stellaria media (Caryophy laceae). The iden- tif ication of the leafminer fli es and their para- sitoids at the species leve l is under way. Insect-Potato Plant Interactions It is common to see adult flies in potato fields from when potato plants emerge unti I they becom·e senescent. Feeding punctures ca n often be seen all over a growing plant, giving the impression that a generalized outbreak of larva l infestation is in process. But the de- velopment of the larval damage fo ll ows a rather f ixed pattern, somewhat different from that of the adult fly population. First, the ini- tial larval infestat ion and the correspond ing damage occur in the lower third of the plant. As infestat ion increases, the medium part of the plant is affected and, finally, the top of the plant becomes infested (Figure 1 ). At this time, practically the whole aboveground part of the plant becomes necrotic and dies. The rate and leve l of infestation of plants are af- fected by varietal differences, age, and the physiological state of the plant. Damage(%) 100 0 100 0 100 0 Although potato plants can be infested by leafminer flies once they emerge from the so i I surface, larva l damage is consistently less severe during the vegetative phase of the plant. At this time, the first leaves to show larval damage are those of the lower part of the plant. In con trast, w hen the plant is fully grown, from the flowering phase onward , foliar in- festation develops at a much faster rate al I over the plant, giving the impress ion of a sud- den outbreak. The occu rrence of egg extru- sion might exp lain this phenomenon. When the leaf is mature, most of the egg remain s encrusted in the leaf tissue until hatching. Then, the neonate larva immed i- ate ly starts tunneling the foliar pallisade tis- sue. Larva l surv iva l is close to 100%. But soon after the egg is laid, on a still-growing leaf, a 3 10 17 24 31 7 14 21 28 6 12 Aug Sept Oct Figure 1. Occurrence of mining (Ill) in potato foliage caused by leafminer fly larvae, in the bottom, medium, and top parts of the potato plant (cv. Revoluci6n), Canete, Peru, 1995. CIP P1ogram Repo11 l 995-96 l 7 9 pro li ferat ion of ce ll s occurs in the ti ssue sur- round ing the egg. That results in the extru- sion of the egg to the leaf surface. There, the egg rema ins exposed to the act ion of adverse phys ica l facto rs (ma inl y dehydrat io n) and predators. The neonate larvae that are able to hatch are also exposed to dehydration and preda- tion before starting to tunnel into the leaf ti s- sue. More than 90% of the eggs were extruded and around 60% of the eggs and neonate lar- vae died by dehydration under laboratory co nditions. Under these conditi ons, lower larva l infestation during the vegetati ve growth of the potato plant wou ld be expected, de- spite a high adult fly popul ation. The prob- ability of successful larva l deve lopment from an egg in mature foliage is much higher than in a growing leaf. Developing Management Components Farmers' on ly measure to control leafminer fli es is the heavy use of chemica ls aga inst (t/ha) adult fl ies and larvae, usually w ith mediocre results. The alternative methods investigated included the eva luation of susceptibility/ tol- erance of commercial cu lt iva rs, the deve lop- ment of to lerant potato clones, the ro le of natural enemies, effects of cultural practices, trapping devices, and the se lecti ve use of lar- v icides. These techniques, along w ith meth- ods for monitoring the fl y populati on, are the basis for structur ing the integrated manage- ment of this pest. Differences in susceptibility Experiments to test the degree of suscepti- bi I ity among the most common var iet ies cu l- tivated in Canete showed differences in the area of fo liage mined, the necrosing rate of the mines, and yield redu ct ion. Effects of fly in festations on yie ld reduction were deter- mined by comparing plots with and without contro l measures (Figure 2). Developing resistant/tolerant clones Breeding work at CIP resulted in genotypes w ith reaso nab le leve ls of tol erance of 45 1 ............................... .................................................... .. .......... .. ...... .. . 40 f···· ..... .. .. ... .......... ........................................................ . 35 f .. .... . j 30 ~ ....... ! 25 f ...... I 20 t""" ! 15 10 5 0 Amarilis Canchan Gica Tomasa Revoluci6n Perricholi Yungay Figure 2. Yields of treated (D) and nontreated (•)plots to control leofminer fly infestation in the most common potato cultivors in Coiiete, Peru (overage of two years: 1994-95). 180 Program 4 leafminer fly infestati ons, high yields, and good quality characteristics. Farmers of the Tambo va lley participated in the final se lec- tion of a group of these materials. Farmers se lected c lone CIP-282 after eva luating its growing peri od, harvest, yield, cooking quali- ties, process ing qua I iti es, and marketabi I ity. Yields of marketable tubers were double those of the leading loca l va ri ety Canchan at me- dium elevation in the Canete va lley, 30.4 ti ha vs . 15.2 t/ha. At high elevation, CIP-282 yielded 51.6 t/ ha compared w ith 33.5 t/ha for Canchan. The clone has been released as Maria Tambena by the lnstituto Nacional de lnvestigac i6n Agraria (INIA). The good quali - ties of thi s new variety are in great demand by farmers. Biological control: occurrence and role of natural enemies An important complex of natural enemi es attacks leafm iner fly larvae in the Canete va l- ley. Intensive sampling of infested leaves of potato and bean plants rendered 1 0 spec ies · of paras ito ids whose seasonal occurrence and relati ve effectiveness aga inst leafminer fly have been determined. Endoparasitoids and ectoparasito ids were recorded . The occur- rence of high parasitism (close to 100%) dur- ing the warmer months of the yea r seems to be the factor responsible for the extremely low populations of the fly during this period. Ectoparas ito ids recorded are Oiglyphus websteri (Craw.), 0 . begini (Ash.), Oiglyphus sp., Closterocerus c in ctipennis Ash. , and Zagrammosom a multilinea tum (Ash.). The female ectoparasitoid first pa ralyzes the fly larva in the leaf mine and then lays its eggs in the vicinity of the larva. The parasitoid larva feeds extern ally on the fly larva l body and pupates inside the leaf. In the case of the endoparas ito ids, the female lays an egg in- side the body of the fly larva where the larva of the parasitoid develops. Pupation of the parasitoid occu rs inside the fly puparium, which is formed outside the leaf tunn e l. Endoparasitoid s reco rd ed are Halticoptera sp., H. arduine (W alker), Chrysocharis sp., C. phytomizae (Bre.), and Canaspidium sp. Ectoparas ito ids were abundant in bean plants. 0 . websteri showed the highest per- ce ntages of paras i tism fo ll owed by C. cin ctipennis. Endoparas itoids were more abundant in potato pl ants. H. arduine was the most common parasitoid in potato fo ll owed by 0 . websteri. The va lue of weeds as sources of paras itoids cannot be disregarded as we have recovered from weeds a ri ch fa una of paras ites that attack leafminer fli es. Average paras itism in 18 weed spec ies was 44%, w ith a max imum of 71 %. Not all pa ras ito ids have been identified. Predatory f li es of th e fam ili es Doli chopod idae and Empididae were found capturing and killing leafminer fly adul ts. Al- though these voracious predators have been observed occasionall y in large numbers, their importance has been difficult to assess quan- titatively due to the lack of adequate sampling methods. Cultural practices Hea lthy, vigorous-growing potato p lants are ab le to counteract the effect of leafmi ner infestation, particularly during the vegetat ive phase. The fast-growing fo li age enhances the egg extrus ion reaction of the fo li ar tissue. Plants defic ient in irri gation water and fert il- izer, or coming from low-quality seed (e .g., virus-infested seed), show damage earli er and their tunneled leaves dry more rapidl y. Un- der these circumstances, low yields due to inadequate agronom ic conditi ons or low- quality seed are further reduced by the syn- ergistic effects of leafm iner fly infestat ions. Trapping devices The attractive effect of ye llow surfaces on leafm iner fli es was reported in the ea rl y 1980s. Yellow surfaces covered w ith a st icky substance were used fo r trapping adult fli es to monitor their populations. Years later CIP entomologists verifi ed the trapping effects of ye llow sti cky traps and expanded the ir use for mass trapping leafmi ner flies in control programs. By using 60 fi xed traps/ha, the num- ber of insecticide treatments was reduced from 4-6 sprays to 1-2 sprays in the Tambo va lley, Peru, where fly infestations are moderate. CIP Program Report 1995-96 l 81 Leafm iner fl y infestations in the Canete va l- ley are much more severe than those of the Tambo va lley. Ye llow traps became saturated and needed to be changed before the end of the cropping season despite the use of 80-100 traps/ha. Close to 5 million fli es/ha/season were captured in eva luated plots. Half that amount was enough to saturate 1 00 traps. The increase in the number of traps required and the need to change them in the middle of the season made their use too expensive. The original traps were manu factured with expensi ve, impo rted sticky materi als (cost of the ye llow plastic plus the sticky materia l was about $1 per trap) that made them uneconomica l under severe sus- tained fly infestat ions. Therefore, new, cheap, loca ll y ava ilable materials were tested. The most cost-effective alternat ive was automo- ti ve motor oi l SOSAE, wh ich reduced the cost from $234/ha to $66/ha. A further develop- Adult flies/plant ment was the use of mo bi le traps-ye I low plast ic sheets that are passed over the plant canopy covering four or more rows at a time. Farmers adopted them read il y and modifi ed them accord ing to their ingenuity. Figure 3 shows the effect of trapping on the fl y popu- lation compared with a check field. Trapping ca n rep lace the use of insectic ides aga inst adult flies. Experiments were also conducted to im- prove trapping effic iency. Aspects studied were the effective attractive distance of the traps, the number of traps required per hect- are, orientation of the traps in relation to w ind directi on, spatia l distributi on of the traps, opt imum timing for the insta ll at ion of t raps w ith respect to crop growth, trapping effi - ciency and in relation to the field fly popu la- 50 . ................... . . .. ........ . . . ........ .. .... . ................ •.................... ..... . ... .. ... .. 45 ...... ~~~~.~ .!~~P.~ ............ .... .... ........... f ..................... t ... ...................... . 40 38 3 ~ Wth t . ...... M.~bii~· ;;~·~~······· ··,- ··+ftt., ...... r .................... .. ... -+- c~n1r:~PP'"9 · 35 .... .... . .. ........ . .................... ............ .. ..................................... . . .. .. . . . 31.5 31.1 30 25 20 15 10 5 4.7 1.7 0 1.1 25 23 31 7 14 21 28 5 12 19 26 2 9 16 July Aug Sept Oct Nov Days after 0 29 37 44 51 58 65 72 79 86 93 100 107 114 planting Figure 3. Effect of trapping on leaf miner fly field population, comparison of fields with and without trapping, Canete, Peru, 1996. 182 Program 4 tion, and comparisons of efficiency of fixed traps vs. mo bi le traps. Selective use of larvicides A critical aspect of leafminer fly manage- ment is the reduction in the use of insecti- cides. Farmers tend to spray wide-spectrum insecticides as soon as they see the first adult flies on the foliage. Nothing is more inappro- priate than this. Flies are not only tolerant of many compounds, but readily develop resis- tance to them. Even if affected by a spray, the treated population recovers rapidly because of immigration. Worse, early sprays against adults destroy natural enemies of leafminer flies and other pests early in the season. As a result, resurgence of the fly occurs and new pests are encouraged, as in the cases of the white mite and budmidge in Canete. The combined effects of egg extrusion by the growing foliage, which can be further stimulated with good cultural management practices, and the yellow sticky traps can ef- fectively counteract the apparent need to use insecticides against adult flies. Nevertheless, potato plants, after vegeta- tive growth stops, might need protection for their foliage against the mining larvae if the level of control by natural enemies is not enough to avoid yield reductions. When these sprays are required, they should be as selec- tive as possible. Cyromazine is effective against fly larvae and is largely compatible with natural enemies. Monitoring Methods for Fly Damage and Population Levels Monitoring pest populations and damage has two purposes: to study the population dynam- ics of the pest and to make decisions about control measures. Monitoring adu It fly populations by count- ing the number of flies per plant or the num- ber of flies captured by sticky traps is useful for population dynamics studies. Both meth- ods have been used in our studies and we have found a fairly good correlation between them. For many years, monitoring fly infestation for deciding spray treatments was based on adult fly counts. But we have demonstrated the inconsistent relation between adult and larval populations. In most cases, adult counts lead to unnecessary sprays. The alternative is to count the number of larvae or fresh tun- nels per leaflet by sampling the bottom, middle, and top parts of the plant. This has proved to be an efficient method for decid- ing when to use larvicide sprays. Unfortunately, direct observation of larvae in the field does not provide information about the incidence of natural enemies. To get this information, infested leaflets must be taken to the laboratory and kept for 21 days. By that time we know the number of adult flies that developed from healthy larvae, the number of larvae killed by insecticides, and the level of parasitism based on the number of parasitoids recovered (Figure 4). But a 21- day delay in obtaining this information is too long for deciding the timing of a spray. In sev- eral of our experiments, we found that a sec- ond larvicide spray could be avoided if we had the information on parasitism at the time of sampling. More research is needed to solve this problem. Conclusions The amount of research conducted so far on the development of IPM components for leafminer flies allows us to offer a prelimi- nary menu of options to farmers. Some pre- 1 iminary trials envisage the possibility of re- ducing the number of sprays currently used (8-13) against leafminer flies to one or two sprays per season. The integration and imple- mentation of IPM programs in farmers' fields will be completed with economic evaluations of the practices adopted. CIP Program Report 1995-96 183 Larvae/leaflet 0 3 .......... ................................................................................................ . 2.5 2 1.5 0.5 Larvae/leaflet 0 6 .. .................................................... ........ ............... ..... . .. .. .. .... .......... .. .... . . . 5 ..... ... ..... ... ..... ........ ...... .. .. ............. . . 4 .. ... ............ .......... ... ................ ..... .. . 3 .............................. . .. .............. ... . 2 .. ... ... ... ........... ........................ .... . 0 D Dead larvae • Parasitoids ·· · · ... ··· .... · ·· · .... 1111 Adult flies Bl Closed puparia 18 25 1 8 15 22 29 6 13 20 11 17 23 31 7 14 21 28 5 12 19 26 2 9 16 May Jun Jul Aug Sept Oct Nov Figure 4. Monitoring of leafminer fly population. Comparison between (A) direct sampling observation in the field and (B) field sampling and laboratory analysis, 21 days after sampling. Note that in the laboratory analysis healthy larvae and larvae affected by parasitoids and insecticides can be detected . (1 = cyromazine treatment, a = sampled 1 day after treatment.) 184 Progrom 4 Large-Scale Implementation of IPM for Sweetpotato Weevil in Cuba: A Collaborative Effort J. Alcazar\ F. Cisneros\ and A. Morales2 CI P started investigating the sweetpotato wee- vil , Cy/as formica rius (Fabri c ius), the most injuri ous pest of sweetpotato in the Caribbean area, in 1990. Act ivities were concentrated in the Dominican Republic on the use of sex pheromone traps. In 1993, studies were ex- panded wi th financial support from the Or- ganization of Petroleum Export ing Countri es Fund fo r Intern ation al Deve lopment, and Cuba was included. Cuban sc ientists parti c i- pating in an IPM workshop in Santo Domingo, Dominican Republi c, expressed their inter- est in co ll aborating with CIP in managing thi s pest in their country, where 60,000 ha are planted to sweetpotato each year. Scope of the Problem The As ian sw eetpotato weev i I is the most important sweetpotato pest in the Caribbean, and practica ll y the onl y one in Cuba. Dam- age is so severe that when insecticides from the Soviet Union were no longer ava ilable in Cuba (199 1-92), damage amounted to 40- 50% of production. Previ ously, Cuban farm- ers had sprayed their sweetpotato fie lds 10- 12 times per season, with damage amount- ing to around 10% of producti on. The pest is present in all provinces of the country. Histori ca ll y, reported yields of tradi- ti onal cul t ivars were low (6 t/ha), bu t in areas w here new vari eti es deve loped by th e ln st i tuto N ac io nal de ln vestigac i6n de Viandas Tropicales (INIVIT) were p lanted, yields reached 20-30 t/ ha. It is diffi cult to express yield losses in eco- nomic term s, because of the spec ial charac- 1 CIP, Lima, Peru . 2 Head, IPM team, IN IVIT, Cuba. teri sti cs of the Cuban economy. But if no con- tro l measures are taken, weevil damage to sweetpotato is equiva lent to a total loss of 24,000-30,000 ha annually, about half the area pl anted to the crop. Sweetpotato is one of the staple foods in Cuba. The other crops include banana, cas- sava, yam, and cocoyam. Rationale of IPM Components and Their Integration The urgency created by the sudden lack of insecti cides and the resultant rapid increase in weevil damage obviated the need fo r as- sessment and characterization of the pest, whi ch normall y constitutes the first phase of CIP's strategy. The need to produce enough food w ith limi ted resources during what Cu- bans ca ll the "special period" of the Cuban economy made sweetpotato production even more important. First, efforts concentrated on developing IPM components and filling some knowl edge gaps in the biology and seasonal occurrence of the w eevil. M ax imum efforts were initi ally dedicated to compiling and verify ing informati on about research and experiences for controlling the weevi I worldw ide. Later, knowledge gaps were identified during coordination meetings when new results were discussed and new research acti v it ies were scheduled. A signifi cant amount of new informat ion was generated through research conducted within Cuba. The only new key component foreign to the Cuban experiences was the use of pheromone traps, which CIP had been test- CIP Progmm Report 1995-96 l 85 ing and improv ing in the Domini can Repub- li c for several years . Research led to the identification of a se- ri es of practices to reduce the weev il popula- tion or its damage, and a bas ic IPM program was des igned (Figure 1 ). These pract ices were offered to farmers and agra ri an-cooperati ve technicians as menu options. The in itia l dem- onstrat ion fields were located near I NIVIT's station in Santa Clara. The main practices rec- ommended were biological and cu ltural. Cultural Control Components Healthy planting material Pl anting infested sweetpotato stem cuttings is a primary way to di stribute sweetpotato weevil. More than 95% of the eggs are de- posited in the first 35 cm of stem. By discard- ing the basa l stem portion , the apica l portion of the stem makes a hea lthy cutting. INIV IT estab li shed a system to produce large quantiti es of healthy seed cuttings in Multiplication I cooperat ives . A 1-ha pl ot produces enough hea lthy cuttings to plant 20 ha in 4 mo. Crop rotation About 20% of the area planted w ith sweetpotato is rotated w ith potato. Insecti- c ides and the rel ative ly large amount of wa- ter used in potato f ields destroy all remai ns of the weev il population in one season. With no crop rotation, the number of weev i Is cap- tured by pheromone traps in a 60-d peri od exceeded 8,000 . That was reduced to 41 8 weevi ls after a 1-yr rotat ion w ith crops other than potato, 40 weev il s after 17 mo, and onl y 2 weev il s after 2 yr. Rotat ion is even more effecti ve as more success ive crops are planted within the same rotation period. Neighboring fields In highly in fes ted areas, INIV IT recom- mended separating new sweetpotato plantings from older f ields by at least 1,000 m. lnfesta- Colonization I New sweetpotato field Harvest "?'\ ~ residues V V and I Colonization ~ Sweetpotato field volunteer plants 10 ® = Control measure 1. Early tolerant or resistant cult ivars 2. Healthy or disinfested cuttings (chemically or Beauveria bassiana-treated) 3. Sex pheromone 4. Beauveria bassiana 5. Colonization or predatory ants 6. High hill ing 11 I Mig ration I 7. Irrigation management (avoid soi l cracking) 8. Early harvest 9. Crop rotation 10. Destroy crop residues 11 . Destroy volunteer plants 12. Avoid neighboring, old sweetpotato fields Figure 1. Sweetpotato weevil : population dynamics and management in Cuba. l 86 Program 4 tion risk increases with proximity to older sweetpotato fields. Timely harvest Farmers may delay harvest after physi- ological maturity to increase yields or to get higher prices at a later date. But in Cuba, postponing harvest 30 days means a fourfold increase in damage. INIVIT recommended harvesting mature crops before the level of infestation reached 3%. Destroying crop residues Field sampling from various parts of Cuba showed that crop residues left on fields aver- age 0.7 t/ha. That is enough material to har- bor as many as one million weevils. Elimi- nating harvest residues increases the effec- tiveness of the other IPM measures. Water management In Cuba, 70% of the sweetpotato crop (42,000 ha) is planted in the rainy season and the rest during the dry season. Soil moisture is essential for plant growth and it has a clear effect on weevil infestation levels. Cracking of the soil because of drought or deficient irrigation water faci I itates female wee- vi Is reaching the sweetpotato fleshy roots to deposit their eggs. Well-irrigated fields are com- monly 4-5 times less infested than those suffer- ing from moisture deficit (Table 1 ). Varietal selection Early maturity and deep storage roots help sweetpotato escape weevil damage. Shallow- rooting varieties are four times more infested than varieties that root 8 cm below the soil surface. Early-maturing varieties (90-120 d) are three or four times less infested than late varieties (180 d or more). The recommended varieties in Cuba are INIVIT B-88, CEMSA 85- 48, Cautillo, and CEMSA 78-354. Biological Control Components Disinfestation Under heavy weevil infestations, healthy cuttings (without eggs and larvae) may still harbor adu Its that are attracted by the cut stem. Cubans disinfest cuttings by dipping them in a dilution of the parasitic fungus Table 1. Effects of some selected IPM components for sweetpotato weevil in Cuba expressed as damage (% infested storage roots) and yields in spring-planted experimental fields, INIVIT, Villa Clara, Cuba, 1994. IPM component Irrigated Non irrigated Damage Yield Damage Yield0 (%) (MT/ha) (%) (MT/ha) Pheromone 2.3 c 37.3 abc 21.4 be 12.5 b Beauveria bassiana 5.8 b 29.9 be 27.0 ab 9.0 ( Pheromone + B. bassiana 0.6 de 34.9 a 18.4 ( 12.9 ob Pheido/e megacepha/a 2.7 ( 31 .8 obc 25.9ob 10.2 ( Tetromorium guineense 4.4 b 31.5 abc 25.8 ob 9.9 c Pheromon. + Pheidole 1.6 cd 34.0 o 18.9 c 11.9 b Beauveria + Pheidole 2.0 c 33.5 ab 21.8 be 10.2 ( Pher. + Beau. + Pheid. 0.1 de 35.4 o 17.8 c 12.7 b Insecticides 0.4 c 28.7 c 17.4 c 14.5 a Check 8.2 o 22.3 d 29.4 o 7.1 d a. Plot size: 1,250 m2 (partial influence of neighboring plots not discarded). Duncan's multiple range test was used. Insufficient water affected yields in addition to weevil damage. CIP Program Report 1995-96 l 8 7 Beauveria bassiana, which kills infected wee- vils within 2 or 3 d after treatment. Fields planted with disinfested cuttings showed 3-4 times lower weevil populations than those with nondisinfested cuttings. Predators Two species of predatory ants, Pheidole megacephala and Tetramorium guineense, are common inhabitants of banana planta- tions. INIVIT has devised a simple system using rolled banana leaves as artificial inter- mediary nests to transport the ants from their natural reservoir to the sweetpotato fields where they prey upon weevils and other in- sects. Colonizing fields 30 d after planting with 60-110 nests/ ha can keep weevil infes- tation at low levels (3-5%). Pheromone trapping Mass trapping of male weevils using sex pheromone traps efficiently controls the sweetpotato weevil. Farmers consider phero- mone traps as an essential IPM component (Figure 2). Number of traps 140,000 120,000 100,000 80,000 60,000 40,000 20,000 0 19,200 1993 1994 With 16 traps/ha, the weevil population is significantly reduced. Captured weevils are killed by insecticides or by spraying the para- sitic fungus Beauveria bassiana. The Collaborative Approach: The Role of INIVIT Successful implementation of IPM requires collaborative commitments from institutions devoted to helping farmers in their crop pro- duction. In the Cuban case, an ideal situa- tion was presented with the interest of INIVIT in collaborating with CIP to develop a pro- gram for management of the sweetpotato weevil. INIVIT is the Cuban research institu- tion in charge of tropical staple food crops in that country. INIVIT's activities to improve sweetpotato production include developing new, more productive, short-season varieties, improving agronomic and plant protection practices, and training farmers on management of the sweetpotato crop during the rainy and dry seasons. 128,000 72,000 1995 1996 Year Figure 2. Number of sex pheromone traps used in Cuba to manage the sweetpotato weevil, Cy/as formicarius, INIVIT-CIP, 1996. 188 Progrom 4 The strategy for implementing IPM for the sweetpotato weevil in Cuba was based on CIP1s experience in managing potato pests such as the Andean potato weevil Premnotrypes spp. and th e potato tuber moths Phthorimaea operculella and Symmetrischema tangolias in the Andean region. Some INIVIT researchers visited CIP's headquarters to become aware of CIP's work on IPM. Other INIVIT staff partici- pated in training courses organized by CIP in the Dominican Republic and Cuba. INIVIT took the responsibil ity for all the work on research and implementation in Cuba. CIP's staff parti cipated during the planning phases of the research and reviews of the program, and supported the diffusion activities. One or two visits to Cuba a year to jo intly analyze progress and suggest adjustments to the prog ram completed the coll aborat ion. After 3 years, Cuba's experience w as used as the background for a 1995 workshop on man- agement of the sweetpotato weevil , w ith the parti cipation of spec ialists from Cuba and the Domini can Republi c. The Pilot Units The first pilot unit, w here most of the IPM components were integrated, was establi shed in Arimao, Cienfuegos Province, with govern- mental enterprises, cooperatives, and pri vate farmers w ith small landholdings. The pil ot area increased from 250 ha in 1993 to 908 ha in 1995. A second pilot uni t w as established in Santa Clara, Vill a Clara Province, w ith 120 ha in 1993 and 620 ha in 1995 . Other pilot units were in La Habana and Sant iago de Cuba, w ith a total of 400 ha in 1993 and 78 1 ha in 1995. Training Training rece ived a high priority for develop- ing the large-sca le phase of sweetpotato IPM in Cuba. In 1995 1 INIVIT trained 4,965 people through 4 courses, 78 talks, 23 fi eld days, and numer- ous fi eld and mo nito ~in g visits. Fo l low-up trainin g was co nducted in 1996 and an IPM Intern ati onal Works hop was held w ith parti c ipants from the Dominican Republic, Peru , Ve nezuela, and all the prov inces of Cuba. Large-Scale Implementation The Cuban exper ience of successfully man- aging sweetpotato weevil in a fairl y large area (10,000 ha in 1996) created a demand for large-scale implementat ion in the rest of the country. Figure 3 presents the di stri bution of the implemented areas. As a result of the IPM program, weevil dam- age decreased from an initi al level of 40-50% to 4-8%, after 3 yr. The number of insecticide sprays was reduced from 10-12 per season in 199 1-92 to zero in 1996, the only exception being limited sprays surrounding pheromone traps (16 traps/ha). The area sprayed is around 480 m2 compared w ith 10,000 m2 convention- ally sprayed over an entire hectare. The fungus B. bassiana has since replaced the use of insec- ti cides. Finally, yields have increased from the reported 6 t/ha to 1 5-30 t/ha. The INIVIT IPM team, our counterpart, re- ce ived the Cuban Government 1996 Relevant Award fo r the Cuban IPM project. A ll the components of the IPM program are ava il able loca lly except for the sex phero- mones. That limitation is the only foreseeable constraint to reaching national coverage of 60,000 ha in the near term. Selected Reading Cisneros, F., J. Al cazar, M. Palac ios, and 0. O rtiz. 1995. A strategy for developing and implementing integrated pest management. CIP Circular 21 (3) :2-7. Cisneros, F. and P. Gregory. 1994. Potato pest management. Aspects Appl. Bio l. 39: 11 3- 124. CIP Prog ra m Report 1995-96 1 89 1,562 1,340 /.~~ I .-!Sanct1 :iCiego -:· ~ '- Spiri tusL~e Av~~r·' 684 601 Figure 3. Management of the sweetpotato weevil in Cuba : area (ha) under IPM implementation in 1996, INIVIT-CIP. 190 P1ogrom 4 Integrated Pest Management for Sweetpotato in East Africa N. Smit1 and B. Odongo2 In East Africa, sweetpotato is mostl y grown as a subsistence crop by resource-poor female fa rmers, who do not use inputs. Often, a few roots are dug up and so ld to generate cash for househo ld necessiti es. Production plots are small , rarely larger than 0.5 ha, although some larger-sca le produ ct ion exists. The crop is grown in two types of food systems. One is cerea l-based, where sweetpotato is a food- secu rity c rop, as in Kenya. Th e other is nongrain, starchy staple-based. In this system, sweetpotato is one of severa l staple foods, and pl ays an important ro le in Uganda, Rwanda, Burundi, and adjacent parts of Zaire and Tan- zani a. Uganda alone produces 2.0 milli on t annually, mak ing it the world 's fourth-largest sweetpotato producer. Agroecolog ica l conditi ons under w hich sweetpotato is grown range from semiarid to high-altitude, temperate climates. There is almost no fresh storage of roots . Farmers prac- ti ce in-ground storage and piecemea l harvest- ing. This means th at cro ps are left in th e ground for 7 mo to more th an a year, and roots are removed as necessa ry for family mea ls. Thi s practice guarantees that fresh roots are ava il able for consumption during a large part of the year. But it also means that sweetpotato crops are in the field throughout the year in many areas where they are sus- ceptible to pest infestation . Since 1990, CIP researchers have co ll abo- rated with the national root crops programs of Kenya, Uganda, and Tanzani a; the Crop Science Department of Makerere University; the Natural Resources Institute (NRI), United Kingdom; and the Internation al Institute of 1 CIP-Uganda. 2 Nationa l Agri cultural Research Organ ization (NARO ), Uganda. Biologica l Control (llBC) to deve lop integrated pest management (IPM) for sweetpotato in East Africa. The major pest in the reg ion is the sweetpotato weevi I Cy/as spp. IPM development can be di v ided into f ive distin ct and consecuti ve phases, some of which may overl ap. They are (1) pest prob- lem assessment and characteri zation, (2) de- ve lopment of management components, (3) integration of key components, (4) implemen- tation of IPM in pilot units, and (5) implemen- tation of IPM on a large sca le. Thi s CIP strategy has proven to be very successful in managing potato pests in the Andean region and North Afri ca, and man- aging sweetpotato weevi I Cy/as fo rmicarius (Fabri c ius) in the Caribbean. However, these successes were obtained with commerc ial crops grown by relatively resource-r ich farm- ers w ho commonly use insecti cides to pro- tect their crops. The situation of sweetpotato produ ct ion in East Afri ca is very different and requires adjustments in strategy and expec- tati ons. Pest Problem Assessment Soc ioeconomi c resea rch into sweetpotato production and use pointed toward some important considerations relating to pest con- trol. First is farmers' perceptions of pest prob- lems. It makes I ittle sense to help fa rmers solve problems that they do not consider important. During surveys throughout East Africa, fa rm- ers in areas with long dry seasons indicated that root-damaging sweetpotato weev il s (Cy/as spp. ) were a major constraint to thei r sweetpotato production. An insect pest of CIP Program Report 1995-96 1 91 regional importance is the sweetpotato but- terfly Acraea acerata in the high-altitude re- gions of Uganda, Rwanda, and Burundi. Throughout these regions, most sweetpotato roots are destined for home consumption. Consequently, quality demand is low and what entomologists consider high pest levels are tolerated. Biological research is one part of pest prob- lem assessment. The sweetpotato weevi I spe- cies found in East Africa, Cy/as puncticollis (Boheman) and C. brunneus (Fabricius), are unique to the continent. Their counterpart C. formicarius (Fabricius) is a pest on sweetpotato in Asia, the United States, and the Caribbean. It has been widely studied, but little or no published information is available on the biology and ecology of the African sweetpotato weevi I species or on the sweetpotato butterfly. Basic studies revealed differences in the biology of the two weevil species (Table 1) and the presence of species- specific pheromones for the weevils. Knowledge of sweetpotato weevil ecology, biology, and behavior may indicate how to manage the pest. For instance, weevils can- not dig down through soil; the only roots fe- males can lay their eggs in are those they lo- cate through soil cracks or those exposed aboveground. Soil cracks are more common during the dry season, thus exposing roots for attack. Possibi I ities for control are the use of varieties that escape weevi I damage by pro- ducing roots deep in the soil, covering ex- posed roots with soil and filling soil cracks to protect roots from weevil attack, and adjust- ing planting and harvesting times so that roots are not present in the dry season. Development of Management Components The technological options for an IPM program include host-plant resistance; biological con- trol; cultural control; chemical control; and behavior-influencing techniques such as pheromones, sticky traps, and repellents. Methods described in the literature for con- trolling C. formicarius were validated under ambient conditions for the local species. Prac- tices that farmers traditionally use to control pests were considered as a good basis for re- search and verification trials. Host-plant resistance So far, research has not identified germplasm immune to C. formicarius. But some studies suggest that sweetpotato clones differ in their levels of resistance. These lev- els are low and do not stand up under high weevil pressure. CIP and its collaborators have conducted several field and laboratory experiments with varieties from local germplasm collections to detect resistance to the African sweetpotato weevil species. Although some varieties were clearly less susceptible than others, no reliable source of resistance has been identified. Conventional breeding techniques appear to have limited potential to incorporate weevil resistance into sweetpotato. An alternative approach under study is the development of transgenic sweet- potato with proteinase inhibitors for Cy/as spp. Table 1. Life cycle data on two Cy/as species taken in Kenya at 27±1°( and 45±5% relative humidity, 1992. Cy/as Developmental species period (d) puncticollis 20-28 brunneus 32-41 Longevity• (d) 140±100 92±12 b Number of eggs per female 103±16 a 100±18 a a. Means ;;;ithi~~~ ~oluion followed bydifferent letters o ;~~ig~ificantly diff~rent by th~t~test (P 1 /16" and with A type embryo, TPS-13 was selected as a new poll inato r clone, which proved to be as good a general combiner as the parenta l clone TPS-113 , with- Table 2. Simultaneous selection for yield (t/ha) and stability in the first set of hybrid TPS lamillies in performance trials. Family Yield Yield rank 1. Atzimba x Rl 28.6 20.3 16 2. Atzimba x 7XY. l 17.8 11 3. Atzimba x 104.12 LB 21.3 19 4. CEW-69 .l x 104.12 LB 14.7 3 5. CFK-69. 1 x l 04.12 LB 22.0 21 6. Chiquita x 104.12 LB 18.5 12 7. Chiquita x Rl 28.6 17.5 10 8. 1-1035xR128.6 17.4 9 9. l-1035 x 104.12 LB 17.l 8 10. LT-8 x 104.12 LB 22.5 22 11. LT-8 x4 .1 DI 18.8 14 12. LT-8 x Rl 28.6 18.5 13 13. MF- II x 104.12 LB 19.8 15. 14. Serrano x l 04.12 LB 20.9 18 15. TPS-2 x 104.12 LB 14.0 2 16. TPS-7x 104.12 LB 16.4 7 l 7. TS-5 x l 04 .12 LB 15.7 5 18. TS-6 x l 04 .12 LB 21.8 20 19. TS-9x 104.12 LB 16.0 6 20. TS- 11 x104.12 LB 15.1 4 21. TS-13 x l 04 .12 LB 13.5 l 22. TS-1 4x 104.12 LB 20.6 17 Mean 18.2 LSD (P=0.05) 1.6 a. + = selected families. out showing segregat ion for tuber skin co lor. TPS-1 3 a lso produces a hi ghe r amount of pollen and has pollen ferti lity above 80%. In the second set, the TPS fa mily Chi quita with TPS-11 3 was se lected, whereas in the first set Chiquita as a fema le with two of the male parents was not selected. Adjustment Ad justed Stability Stabi lity YS{i) to rank variance rating 2 18 178.4 - 8 10+ -1 10 56.7 - 8 2 2 21 31.3 - 8 13 + - 3 0 154.3 - 8 - 8 3 24 9.1 0 24 + 1 13 43.9 - 8 5 - 1 9 41.1 - 8 - 1 8 5.7 0 8+ -1 7 14.1 - 2 5 3 25 10.4 0 25+ 15 22.6 - 4 11 + 14 0.9 0 14+ l 16 28.7 - 8 8+ 2 20 19.5 - 4 16+ - 3 - 1 17.9 - 4 - 5 - 2 5 9.6 0 5 - 2 3 22.8 - 8 - 5 3 23 38.5 - 8 15+ - 2 4 12.8 0 4 - 2 2 46.4 - 8 - 6 - 3 - 2 42.0 - 8 - 10 2 19 32.4 - 8 11 + 6.5 Sim ilarly, in the first set, 1-1 035 as a fe- male with R-128.6 got selected. In the sec- ond set, however, in comb ination with e ithe r of the two po llinators, TPS-13 and TPS-67, the TPS famili es were not se lected. The clone LT-8 with both the male clones prod uced hybrid famili es that were se lected, CIP Progrom Reporl 1995-96 21 7 Table 3. Performance (yield int/ha) of the second set of hybrid TPS families as tra nsplants for global distribution in trials at La Molina and Huoncayo (1995 and 1996). Family La Molina Huancayo Spring Winter Spring Summer 1. HPS-7 / 67 8.4 9.5 6.4 24.3 2. HPS-7/ 13 7.7 9.3 5.7 25.0 3. HPS-25/ 13 5.9 11 .1 4.9 28.6 4. HPS-1/67 7.0 15.7 4.1 24.5 5. HPS-1/13 4.7 13.2 5.9 29.5 6. HPS-11/13 9.3 12.6 4.7 24.8 7. Arney x TPS-113 8.7 10.4 7.1 24.1 8. CEW-69.1 x TPS-113 5.3 10.5 4.5 26.9 9. Chiquito x TPS-11 3 7.1 12.1 6.8 24.0 10. C320LM86B x TPS-67 10.7 10.l 3.1 18.9 11. C320LM86B x TPS-13 9.0 8.0 2.9 18.2 12. C914LM86B x TPS-13 14.3 9.7 4.8 24.3 13. C914LM86B x TPS-67 12.6 11.l 7.9 26.3 14. 1-1035 x TPS-67 9.8 8.3 4.1 18.7 15. l-1035xTS-13 5.7 3.7 0.9 11 0 16. Kotohdin x TPS-13 9.2 6.5 3.9 19.7 17. Kotahdin x TS-67 11.3 9.4 4.2 18.8 18. LT-8 x TPS-13 9.6 12.4 5.9 23.8 19. LT-8 x TPS-67 9.0 10.3 8.2 28.3 20. Serrano x TPS-13 12.3 11.7 7.1 28.0 21. Serrano x TPS-67 9.8 12.4 5.8 23.3 22. TS-5 x TPS-6 7 7.8 10.2 7.4 24.l 23. TS-6 x TPS-13 4.3 12.2 10.0 28.3 24. TS-8 x TPS-1 13 6.1 l 0.7 3.5 19.6 25. TS-9 x TPS-113 9.4 13. l 4.9 29.7 26. TS-9 x TPS-67 13.9 13.0 8.8 24.3 27. TS-10 x TPS-13 6.1 10.4 5.8 25.l 28. TS-10 x TPS-67 6.1 9.9 8.6 24.3 29. TS-12 x TPS-113 9.1 10.3 3.6 19.3 30. TS-13 x TPS-67 8.7 10.3 2.6 18.0 31. TS-13 x TPS-13 8.7 10.3 3.1 25.5 32. TS-14 x TPS-67 11.l 5.3 3.3 27.4 33. TS-14 x TPS-13 10.2 10.0 2.0 15.4 34. Atzimba x TPS-13 6.1 11.0 10.3 26.9 35. Atzimbo x TPS-67 6.3 10.5 8.6 24.5 36. Desiree (check) 23.7 20.7 26.8 17.5 OMS 0.05 = 3.66, CV% = 185 218 Program 5 Table 4 . Simultaneous selection for yield and stability in first set of hybrid TPS families in performance trials. Family Yield Yield Adjustment Adjusted Stability Stability YS(i) ' rank to rank variance rating 1. HPS-7/67 12 .1 17 -1 16 2.1 0 16 + 2. HPS-7 /13 11. 9 16 -1 15 5.6 0 15 + 3. HPS-25/13 12.6 22 23 39.6 - 8 15 + 4. HPS-1/67 12 .8 23 24 32.9 - 8 16 + 5. H PS-1/13 13.3 28 29 60.5 - 8 21 + 6. HPS-11/13 12.8 25 26 5.5 0 26 + 7. Aracy x TPS-113 12.6 21 22 0.8 0 22 + 8. CEW-6 9.1 x TPS-113 11.8 12 - 1 11 28.2 - 8 3 9. Chiquita x TPS-113 12.5 20 21 5.9 0 21 + 10. C320LM86B x TPS-67 10.7 9 - 2 7 21.8 - 8 - 1 11 . C320LM86B x TPS-13 9.5 3 - 2 12.9 0 12. C914LM86B x TPS-13 13.3 27 1 28 27.9 - 8 20 + 13. C914LM86B x TPS-67 14.5 33 2 35 5.3 0 35 + 14. 1-1035 x TPS-67 10.2 7 - 2 5 15.0 - 4 1 15. l-1035 xTS-13 5.3 - 3 - 2 47.6 - 8 - 10 16. Katahdin x TPS-13 9.8 4 - 2 2 11 .4 0 2 17. Katahdin x TS-67 11.0 10 - 1 9 24.0 - 8 18. LT-8 x TPS-13 12.9 26 27 0.8 0 27 + 19. LT-8 x TPS-67 14.0 31 2 33 18.3 - 4 29 + 20. Serrano x TPS-13 14.8 34 2 36 8.8 0 36 + 21. Serrano x TPS-67 12.8 24 25 1.3 0 25 + 22. TS-5 x TPS-6 7 12.4 19 20 3.6 0 20 + 23. TS-6 x TPS-13 13.7 30 31 60.0 - 8 23 + 24. TS-8 x TPS-113 10.0 6 - 2 4 7.5 0 4 25. TS-9 x TPS-113 14.3 32 2 34 32.4 - 8 26 + 26. TS-9 x TPS-67 15.0 35 2 37 7.5 0 37 + 27. TS-1 0 x T PS-1 3 11 .8 14 - 1 13 11.0 0 13 28. TS-10 x TPS-67 12.8 18 - 1 17 17.0 - 4 13 29. TS-12 x TPS-113 10.6 8 - 2 6 10.6 0 6 30. TS-13 x TPS-67 9.9 5 - 2 3 17.5 - 4 - 1 31. TS-13 x TPS-13 11. 9 15 -1 14 12.4 0 14 32. TS-14 x TPS-67 11 .8 13 - 1 12 58.7 - 8 4 33. TS-14 x TPS-13 9.4 2 - 2 0 50.5 - 8 - 8 34. Atzimba x TPS-13 13.6 29 1 30 34.1 - 8 22 + 35. Atzimba x TPS-67 11.8 11 - 1 10 7.3 0 10 36. Desiree (check) 22.2 36 3 39 411.7 -8 31 + Mean 12.3 14.9 LSD (P = 0.05) 1.5 a. + = selected families. CIP Program Report 1995-96 2 1 9 as was the case in the first set, thus further proving the value of this clone. Serrana as a female clone was also a good combiner with the two male clones in the second set (Table 4), as it was in the first set (Table 2). Although Atzimba with TPS-13 was se- lected, the family with TPS-67 as the male did not perform well. Among the TS clones, TS-6 with TPS-13 was also selected, as it was in the first set with 104.12 LB as the male parent. However, TS- 14 in combination with TPS-13 and TPS-67 did not produce stable-yielding families, whereas in the first set it was selected with 104.12 LB as the pol I inator clone. The clone TS-9 in combination with TPS-113 and TPS- 67 produced hybrid families that were se- lected, but as in the first set, the family TS-9 x 104.12 LB was not selected. Likewise, the hybrid families with Katahdin as the female parent did not perform well. The clone CEW- 69 .1 did not perform wel I in either of the sets. Conclusions These results will allow us to send fewer hy- brid TPS families to our collaborators in dif- ferent countries who are experimenting with 220 Programs TPS technology. In addition, the performance va lues of female and male clones used as parents will be used to further improve the parental clones, and the selected clones will be used as testers for the newly selected geno- types. Use of the STABLE program to evaluate the performance of hybrid TPS families has proved its worth both as a tool to determine the parental values of the clones used as par- ents for the hybrid TPS fam i I ies and for the selection of high-yielding stable families based on short-term performance trials. Com- parison by NARS of the data from trials un- der different agroecologies will provide the final proof of the usefulness of the STABLE program as a tool for the selection of TPS fami- lies in short-term trials. Selected Reading Kang, M.S. 1993. Simultaneous selection for yield and stability in crop perfor- mance trials: Consequences for growers. Agron. J. 85:754-757. Kang, M.S. and R. Magari. 1995. STABLE: A basic program for calculating stability and yield-stability statistics. Agron. J. 87:276-277. Temperature and Moisture Affect Dormancy and Deterioration of True Potato Seed During Storage N. Pallais and R. Falc6n1 Tru e potato seed (TPS) lots with 90% germi- nat ion or better, w hen tested in the labora- tory, often fai I to emerge in the fi eld. The prob- lem is that newly harvested TPS are dorm ant, and the nature and environmental control of TPS dormancy is not well understood . TPS dormancy ceased to be a problem in clonal breeding after 1961 when it was found that gibberelli c acid (GA) could easil y be ap- pli ed to promote germination. But in spite of treatment with GA, emergence of dormant seeds decreases considerably as the tempera- ture approaches 25°C. M oreover, seedlings of dormant seeds that are induced to germi- nate with GA produce significantly less fo- li age dry matter (OM) th an seedlings pro- duced by dormant seeds. Before the discovery of GA as a "so lution" for TPS dormancy, it had long been known th at new TPS did not germinate as well as older seeds. This study examines the relation- ship between storage conditions and dor- mancy ofTPS with respect to its sowing value. Our work showed that dormant TPS of variety Serrana (Solanum tuberosum) germi- nated readily at 20°C or below, w hereas non- dormant seeds germinated readily at 27°C or above. High storage temperature coupl ed with low seed moisture content (SMC) re- duced the period of storage required for los- ing TPS dormancy from 18 to 4 mo. Research to determine the applicability of these results to open-pollinated TPS of the ancient Peruvian potato cultivar Ccompis (5. 1 CIP, Lima, Peru. tuberosum subsp. andigena) was deemed necessary because of the increas ing demand for its use by small farmers in the Andes of Peru. The objecti ve of the present study was to determine the effects of increasing tempera- ture, SMC, and time in storage of freshly har- vested TPS on germination and emergence from soil under different temperatures. Materials and Methods Experiments were conducted at CIP with open- po 11 i nated seeds provided by Semill a e lnvestigacion en Papa (SEINPA). The TPS were produced according to the recommended prac- tice for high-qual ity seed, in a large plot of cul- tivar Ccompis grown for basic seed tuber pro- duction in Cusco (3,600 m above sea level), Peru. After harvesting the mature berries, the TPS were washed and dried in •fie shade until reaching 7.3% SMC (dry weight basis). About 12.5 kg of TPS were separated into large (51 %) and smal I (49%) lots using a round- hole screen at 1.46 mm over a seed shaker for 4 min . Large TPS were separated with a seed blower into high- (63%) and low-den- sity sublots. Only high-density TPS were sampled at random for thi s study. The treatments were prepared by dry ing TPS sea led with fresh si I ica gel at 22°C and monitoring SMC until the desired moisture levels were achieved. TPS were immediately hermeti ca lly sea led in aluminum packages at 3.4, 4.2, 5. 1, 6. 1, and 7.3% SMC and stored at 15, 30, and 45°C. Each treatment was com- posed of five random samples of 1 g each and packaged separately for opening during each evaluation period. CIP Progrom Report 1995-96 221 Eva luat ion s w ere based on monthl y labo- env ironment was he ld constant at 27°(, un- ratory germin at ion and seed ling emergence t i l the 8th day when it was lowered to a con - (g reen house) tests c o nd ucted in tandem dur- sta nt of l 7°C. W ater was added as needed . ing 6 mo of storage . Before testing, TPS were The germ inating T PS received laboratory light moi stened to a uniform SMC of 13% by plac- and higher temperatu re (22°() onl y fo r about ing them in a sealed env iro nment above wa- 20 min du ring eva lu ations. Germin ati o n w as ter at 22°C for 24 h. G ermin at ion tests con- c hecked dail y for 16 dafter sowing (DAS) and sisted of four rep li cat ions of 100 seeds each germi nated seeds were removed at f i rst sigh t placed even ly over dry f ilter paper in 9 -cm of radi c le em ergence . Percentages of germi- Petri d ishes . TPS were hydrated w ith 5 ml of nation at 8 and 1 6 d are prese nted to sum- deio ni zed water before p lac ing the di shes in mari ze t he resul ts (Tab les 1 and 2). an inc ubator. Temperature in the germin ation Table 1. Effects of increasing storage temperature, seed moisture content (SMC) , and time in storage on per- centage of germination at 27°C. Temp SMC Storage (mo.) (o() (%) 2 3 4 5 6 L• Q 3.4 l 2 0 l 0 ns ns 4.2 0 2 2 0 ns ns 15 5.1 2 4 4 5 6 20 ns 6.1 l 6 8 13 19 46 ns 7.3 3 3 7 13 17 52 *** ns L ** ns Q 3.4 2 3 5 6 12 38 *** ns 4.2 2 8 24 37 44 65 ns 30 5.1 4 21 55 63 72 81 ns 6.1 7 30 48 63 67 78 ns 7.3 9 21 50 59 67 87 ns L *** *** *** ** * Q *** *** ... *** 3.4 2 45 74 79 78 92 ns 4.2 5 75 81 75 79 86 ns 45 5.1 14 69 61 63 34 50 ns ns 6.1 21 21 0 0 0 0 ... 7.3 22 15 0 0 0 0 L Q *** a. Li near (L) and quadra tic (Q) regression;*,**,***, and ns denote P < 0.5, < 0.01, < 0.001, and > 0.05, respectively. 222 Program 5 Table 2. Effects of increasing stora ge temperature, seed moisture content (SMC), and time in storage on percentage of germination at 17°C. Temp SMC Storage (mo.) (o() (%) 2 3 4 5 6 L" Q 3.4 79 93 56 58 87 60 ns ns 4.2 87 96 53 70 87 80 ns ns 15 5. 1 89 91 83 84 96 95 6.1 95 94 93 91 97 84 ns 7.3 97 95 93 95 99 89 ns L ns *** *** Q *** *** *** 3.4 94 94 83 75 98 99 ns 4.2 94 95 100 78 100 95 ns 30 5.1 96 96 100 89 100 98 *** ns 6.1 85 99 100 92 92 95 ns 7.3 87 88 99 94 94 97 ns L ns ns ns ns Q ns ns *** ns 3.4 96 91 99 99 99 99 ns 4.2 98 99 99 93 97 93 *** ns 45 5.1 97 89 99 93 93 95 ** ns 6.1 99 85 0 0 0 0 7.3 83 77 0 0 0 0 *** L ns Q ns *** a. Linear (L) and quadratic (Q) regression; *,* *,***, and ns denote P < 0.5, < 0.01, < 0.001 , and > 0.05, respectively. Seed ling tests cons isted of 5 replications Th e experimenta l des ign was a compl ete of 20 seeds each sown in a steam-ster ili zed randomized facto ria l arrangement. Data were so il mixture of equa l parts m oss and sand. subjected to li near and quadrati c regress io n Emergence was c hec ked daily and cou nted ana lysis w ith SMC as the independent va ri- at f irst sight of the hypocotyl hook during the ab le fo r each eva luation, and with t ime in sto r- first 1 7 DAS, but only the percentage of emer- age for each SMC. G enera l corre lati o n coef- gence at 10 d is presented fo r simplicity of fici ents of germinat ion and seed l ing emer- data interp retat io n (Table 3). Temperature in gence were also ana lyzed (Tab le 4) . th e green h o us e was monitored with a Results and Discussion hygrothermograph and average max imum and minimum d aily temperatures were re- Thi s study showed that TPS lots that germi- co rded (Tab le 3). nate on ly at 17°( are dormant and shou ld not CIP Progiom Report 1995-96 223 Table 3. Effects of increasing storage temperature, seed moisture content (SMC), and time in storage on percentage of emergence under variable temperature conditions'. Temp (o() 15 30 45 SMC (%) 3.4 4.2 5.1 6.1 7.3 L Q 3.4 4.2 5.1 6.1 7.3 L Q 3.4 4.2 5.1 6.1 7.3 L Q 4 2 3 17 9 ns 4 lO 14 19 11 ns ns 7 6 40 41 45 *** 2 19 14 33 49 42 *** 38 64 85 86 86 *** 91 98 93 50 11 3 0 0 4 8 7 *** ns 3 21 52 47 58 59 61 55 0 0 *** Storage {mo.) 4 0 0 3 4 l ns *** 9 33 34 35 32 49 39 0 0 5 0 4 15 25 23 *** 17 63 74 67 76 73 87 74 0 0 *** *** 6 2 3 37 40 54 *** 50 78 84 78 78 91 94 68 0 0 ** *** Q ns ns ns ns ns ns ns ns *** a. Maximum/minimum temperatures in the greenhouse at each evaluation were 32/20°(, 34/21°(, 36/24°(, 38/25°(, 41/25°(, and 37 /23°( for tests after l, 2, 3, 4, 5, and 6 months, respectively. b. Linear (L) and quadratic (Q) regression,*,**,***, and ns denote P < 0.5, < 0.01, < 0.001, and > 0.05, respectively. be used for sowing a crop of potatoes under field conditions (Tables 1, 2, and 3). The re- su Its also demonstrated that a high tempera- ture level close to the maximum allowable for germination can be used as a measure of degree of dormancy. Germination at 27°C accurately predicted relative seedling emer- gence ofTPS from soil under various tempera- ture conditions. The combination of germi- 224 Programs nation results at 27 and 17°C might be used as a practical criterion for estimating present and future sowing value of TPS. The low percentage of germination at 27°C followed by high germination at 17°C can be explained as evidence of dormancy. Seed deterioration can be assumed when seeds do not germi- nate at either temperature. Table 4. General correlations for each evaluation among the various germination and seedling criteria tested . Laboratory Screenhouse emergence(%) germination (mo. in storage) (%) 2 3 4 5 6 At 27°( 0.73° 0.72 0.94 0.88 0.91 0.92 ... ... ... . .. At 17°C 0.07 0.21 0.59 0.48 0.50 0.72 ns ns ... ... a. Co rrelation coefficients {r); ** * and ns denote P < 0.001 and > 0.05, respectively. TPS dorm ancy was effectively rel eased w ith progressive increases in storage tempera- ture, SMC, and time in sto rage. TPS dor- mancy was more strongly preserved at 3.4% SMC at all temperatures. At:?. 5. 1 % SMC ger- mination was suboptimal or decreased w ithin the f irst month of storage at 45°C. The results indi cated th at the c ritical mo isture point (CMP) above w hich seed deterioration in stor- age proceeds at a logarithmi c rate is about 5% SMC fo r TPS. Thi s limit var ies w ithin or- thodox species (i.e., seeds res istant to des ic- cation) from 2% in Arachis hypogaea to 6.2% in Pisum sativum. It is safe to recommend that TPS should always be stored below 5% SMC. Drying much beyond the CMP is believed to serve no additional purpose fo r preserv ing seed quality in storage. Overdrying TPS, how- ever, shou ld be avo ided because germination and seed ling emergence were increas ingly inhib ited as SMC decreased below the pro- posed 5% CMP lower limit. Since this inhibi - tion was also more evident as storage tem- perature decreased, it might be a co nse- quence of enhanced TPS dormancy, whi ch is best preserved under low temperature and low moisture. The nature of TPS dormancy was shown to be similar to that of many other spec ies in w hi ch the true seeds have not been domesti- cated for sexual propagation, such as seeds of noncultivated grasses and frui t trees in the family Rosaceae, and buds and bulbs of un- derground perennials. These "wi ld" seeds are characteri zed by a gradual w idening of the all owab le temperature for germination dur- ing dormancy loss in storage. Storing freshly harvested TPS at high tem- pera tu re and low moisture unt il suffi c ient seeds have lost dormancy, as determined by germ ination at 27°C, however, should onl y be co nsidered a compromise so lution for overcom ing TPS dormancy. For example, the resu lts of this study suggest that the safest and most effective treatment fo r rapidly decreas- ing TPS dormancy could be 2 mo storage at 45°C and 4.2% SMC, even though only 75% of the progeny was apparently nondormant at thi s t ime, according to germin ation at 27°C (Table 1 ). Prote in and li p id alterations result from heat injury to biological organi sms; therefore, further storage at 45°C would be detrimental to seed quality. Dormant seeds are protected from heat injury, but high tem- perature will unavo idab ly damage less dor- mant seeds of a given TPS progeny while the more dormant seeds are still losing their dor- mancy. Consequently, we prefe r to after-ripen freshly harvested TPS at 30°C and 4.5% SMC, even if longer periods of storage are required; 40°C is used only when dormant TPS are ur- CI P Program Reporl 1995-96 225 gently needed for planting. Germination tests at 27°C are regularly conducted and when germination after 8 d at 27°C has increased to >50%, then seeds are stored at lower tem- peratures (1 0-20°C). If parti all y dormant TPS need to be used immediately, sowing rates mu st be increased acco rdingly. Otherwise, co ntinued storage at 10-20°C is recom - mended unti I most seeds become nondor- mant. Conclusions This study showed that TPS that germinate only at 17°C should be considered dormant and inappropriate for sow ing a potato crop. The percentage of germinat ion after 8 d at 27°C, on the other hand, was shown to be an accurate c riterion for estimatin g potential sowing va lue of TPS. W hen thi s measu re is used in conju nction with germination at l 7°C, relat ive dormancy leve ls can also be esti- mated. The ev idence presented also suggests that TPS shou ld always be stored at below 5% SMC to preserve seed viability, regardless of storage temperature. Much lower SMC val- ues inhibited germ ination, and higher values increased the rate of seed deterioration. This work also clear ly demonstrated that a period of dry after-ripening in storage at high 226 Program 5 temperature might be used to accelerate the rate ofTPS dormancy loss. This strategy, how- ever, should be used with caut ion because seeds w ill begin to deteriorate after losing dor- mancy under high storage tempe rature. Th erefore, breeders should seek a genet ic so luti on to the dormancy prob lem so that TPS ca n realize its full potential as an alternative method for propagat ing potatoes. Selected Reading Karssen, C.M. 1982. Seasonal patterns of dormancy in weeds. In: The physiology and biochemistry of seed development and fermentation. A.A. Khan (ed. ). Elsevier Biomedica l Press, New York. p. 243-298. Pallais, N. 1995. High temperature and low moisture reduce the storage requ irement of freshly harvested true potato seeds. J. Am. Soc. Hort. Sci. 120:699-702. Pa l lais, N. , N. Fong, R. Garcia, and J. Santos-Rojas. 1990. Factors affecting seed ling vigor in potatoes. II. Genotype, after-ripening and pre-sowing treatments. Am. Potato J. 67:109-119. Roos, E.E. 1980. Physio log ica l, biochem i- ca l, and genetic changes in seed quality during storage. HortScience 15:19-22. Frost-Tolerant Potato Varieties for the Andean Highlands E. Carrasco1, A. Devaux 2, W. Garcia1, and R . Esprella1 Sensit iv ity to co ld stress is one of the major constraints to potato product ion in the Andes. It is est imated th at around 400,000 ha, rep- resenting 70% of the area under potato cu lti- vat ion in the A ndea n highl ands of South Ameri ca, are affected by frost. In Bo li via, most of the 140,000 ha planted to potato are above 3,000 m, w here frost is one of the major con- stra ints to product ion. Frost affects production in three ways: (1) frost itse lf reduces product ion from 1 0% to 70% for nonto lerant va ri eties; (2) the threat of frost resul ts in extensive, low- input man- agement pract ices because farmers do not wa nt to invest in ri sky operat ions; (3) fa rmers in the most affected areas reduce frost ri sk by using frost-to lerant b itter potatoes se lected over centuries from natural hybri ds between w i Id and cul t ivated spec ies. Because of their high glycoa lka lo id leve ls, the tubers of these b itter varieti es need to be processed before they can be used for consumption. Nonbitter potato varieties cu rrently grown by fa rmers are susceptib le to frost or, in the case of some nati ve ones, to lerant to 0 to -1 °C. It is est imated that if the leve l of frost to lerance of se lected nonbitter potatoes cou ld be ma inta ined at -3 to -4°C fo r as li tt le as 2 h, y ields could be increased by as much as 20% to 30%. PROINPA, the Bo liv ian National Potato Resea rch Program, was estab lished under an agreement between the Bo liv ian Min istry of Ag ri cul ture, CIP, and Swi ss Deve lopment Cooperation (SOC) . One of the main objec- 1 Bo livian National Potato Research Program (PRO INPA). 2 CIP and the Boliv ian Nat ional Potato Research Program (PROINPA), Cochabamba, Boli v ia. t ives of PROINPA's resea rch program is to make potato production systems in the high- lands more sustainable by developing and adapting techno logies that respond to fa rm- ers' ma in constra ints. As pa rt of this work, PROIN PA began developing frost-to lerant po- tatoes, bu il d ing on breed ing and se lect ion work. Materials and Methods New cu lt ivars were obtained using two strat- egies: (1) the introduct ion of genet ic materi al from CIP and the Co lombian Agricultural In- st itute (ICA) for eva luation and se lect ion, and (2) th e generat ion of new c lones by crosses between nat ive Bo li via n cu lt ivars, introduced cul t ivars, and wi ld potato species. The clonal eva luat ion of the materi al was also carried out on-farm, thus invo lv ing farmers' fam ili es in the selection of the best-perform ing clones. Introduction, evaluation, and selection of genetic material Since 1989, about 500 clones introduced from CIP and ICA were tested in areas w here frosts are frequent during the growing season. These areas are located at around 3,500 m, w here the average temperature is around 9°C (the average min im um 1°C, and max imum 15°C) and the yearly pre- cipitat ion is 500 mm. Eva luations of the fo llowing factors were carri ed out during seven success ive grow ing seasons: •Percentage of fo li age damage ca used by frost, and capac ity of recuperat ion after frost damage. • Agronomic cha racter isti cs of p lant and tubers . •Tuber y ield and culi na ry quali ty. CIP Program Reporl 1995-96 2 2 7 • Resistance to other stresses such as nematodes (Nacobbus aberrans) and wart (5ynchytrium endobioticum) . The material was tested in comparison with cu ltivated varieties, including a suscep- tible one (5olanum tuberosum), a tolerant one (5. andigena), and a resistant bitter variety (5. juzepczukit). Generation of new clones We initiated crosses between cu ltivated and wild spec ies in 1990. Table 1 presents the material used in this breeding program. The F, hybrid population obtained from these crosses was exposed to frost; the most v igorous and frost-tolerant clones were se- lected. Many F1 lines genera ll y show resis- tance, but their yield is low and their cook- ing quality is poor. However, one or more backcrosses to cultivated varieties were made to improve yield and tuber organoleptic qual- ity. The material obtained from these back- crosses was evaluated in the growth cham- ber to eliminate the susceptib le c lones. Se- lected clones were field-tested following the sa me clonal evaluation process. Frost data In the Andean high lands, frosts occur only at night, always after sunny days with tern- peratures of 15 to 20°C. The critica l point usu- all y occurs between 0400 and 0630 h, w hen the temperature fa lls below 0°C. For eva luations in the growth chamber, a temperature curve of a frosty night was simu- lated w ith minimum temperatures of-5°C. In the field, plants were exposed to natural frosts. Temperatures were recorded w i th a thermohygrograph or with a maximum-mini- mum thermometer located in the field at plant level. Participatory evaluation with farmers Frost tolerance alone will not ensu re the acceptance of new varieties. Other charac- teristics such as tuber shape and co lor, and cu linary properties, must also be acceptable to farmers, market intermed iaries, and con- sumers. During the last three cropping seasons (1993-96), farmers have eva luated the most advanced material in on-farm trials. One of the techniques used was matri x scoring. Af- ter harvest the new cultivars were placed in piles at the end of each row where they had been grown. A group of farmers ca refu ll y ex- amined the new cultivars. They identified the most important cr iteria they looked for in adopting new varieties (y ield, size of tubers, and level of insect infestation). Taking one Table 1. Some potato species used by PROINPA for frost tolerance. Species Classification Ploidy Chromosomes {no.) 5. tuberosum Cultivated Tetraploid 48 5. andigena Cultivated Tetraploid 48 5. iuzepczukii Cultivated Triploid 36 5. curtilobum Cultivated Pentaploid 60 5. phureia Cultivated Diploid 24 5. aianhuiri Wild Diploid 24 5. acaule Wild Tetraploid 48 5. commersonii Wild Diploid 24 5. toralapanum Wild Diploid 24 5. megistacrolobum Wild Diploid 24 5. sancta-rosae Wild Diploid 24 228 Prog ram 5 criterion at a time, farmers placed a number of grains of maize by each pile of potatoes to indicate how they had scored each one. The potatoes were then boiled and farmers evalu- ated culinary characteristics using the same method. Nutritional and culinary evaluation Nutritional analysis of the selected mate- rial was carried out in collaboration with the Food and Natural Products Laboratory of San Simon University in Cochabamba, Bolivia. Fresh tubers stored for a month were used for these analyses. To determine culinary char- acteristics, the potatoes were evaluated as fried and boiled. In both cases, a tasting panel of six people assessed the clones. The crite- ria used for fried potatoes were internal and external color, internal and external texture, taste, and general appearance. The maximum score of these criteria was 100. For boiled potatoes, the criteria were consistency, flouriness, moistness, color, taste, and flesh discoloration, with an evaluation score of very good, good, average, and bad. Results Four potential varieties were selected from the material evaluated during the last seven Frost damage to foliage (%) years. One came from CIP (389349.1 ), the others from PROINPA. Figures 1 and 2 present the difference in yields and frost damage be- tween potential varieties and control cultivars for field evaluations from 1993 to 1996. These data show a high phenotype stability of the clones compared with control varieties. There is some variability between years, however, due to genotype-environment interaction considering frequency and severity of frost, development stage of the crop when affected by frost, irregular precipitation, and the ef- fects of hai I and drought. During the 1993-94 growing season, three frosts occurred with temperatures from -2 to -5°C for 2 h. During the 1994-95 growing season, only one frost (-3°C for 2 h) and a hailstorm were registered. During the 1995- 96 growing season, two frosts of -2°C for 2 h, a 2-mo drought at the start of the growing season, and one hailstorm during the midgrowth period occurred. The yield and frost tolerance of the four selected clones were stable during the 5 yr of evaluation and were consistently better than those of the control (Figure 3). From field observations, it appears that the four clones can tolerate frost of -3°C for 2 h and of -4°C so ~----------------------~ 70 60 50 D 93-94 . 94-95 • 95-96 C.V. = 33.5% C.V. = 37.8% C.V. = 35.8% LSD .05 = 0.44 LSD .05 = 0.54 LSD .05 = 0.48 40 +-------111--------r->----r-r-----------i 30 +-----rl----111111--f >-------< 20 +-< t-----; 10 0 90-123-22 90-101-3 90-123-18 389349.1 S. lmilla Luk'y Alpha PROINPA PROINPA PROINPA CIP andigena juzepczukii tuberosum Clones/varieties Figure 1. Foliage damage caused by frost in selected clones and control varieties in three growing seasons, from 1993 to 1996. CIP Program Report 1995-96 229 Yield (kg/plant) 1.6 ~-------------------------~ 1.4 +-- - --- --------; 1.2 +--- --- --- -----; 1 .0 0.8 0.6 0.4 0.2 1993-94 1994-95 • 1995-96 CV= 27.8% LSD .05 = 0.30 CV= 23.6% LSD .05 = 0.33 CV= 29.4% LSD .05 = 0.27 0 . . 90 -1 23-22. 90-1 01-3 90-123-18 389349.1 S. lmi lla Luk'y Alpha PROINPA PROINPA PROINPA CIP andigena juzepczukii tuberosum Clones/varieties Figure 2. Yield of frost-tolerant selected clones and contro l varieties in three growing seasons from 1993 to 1996. Yield (t/ha) 40 --- 35 30 25 20 15 10 5 0 I lllimani 90-123-22 Tunari 90-1 01-3 r l r Totorei'ia Condori 90-123·1 8 GIP 389349.1 Damaged foliage (%) n • Yield Damaged fo liage ~-+ S. lmilla Luk'y Alpha andigena juzepczukii tuberosum ro 60 50 40 30 20 1 ~o Cultivar Figure 3. Yield and frost damage to foliage in potential and control varities. Average of five growing seasons, from 1993 to 1996. somewhat less. Moreover, th ey have a lso shown a good capac ity for recuperation and regrowth after frost. Some of the frost-tolerant c lones, such as 389349.1 and 90-1 23 -22, scored as high as or higher than contro l va ri eti es in agronomic and culinary qualities. 230 Progrom 5 There is sti 11 a need to test the se lected clones on a larger sca le in different produ c- t ion areas to confirm their market acceptance. For that reason , these c lones have already been named (Figure 3) and have been in- c luded in PROI NPA's seed product ion pro- gram to initi ate the process of c lean seed pro- duct ion. Prebas ic seed of th ese va ri eti es should be ava il ab le in 1998. Tab le 2 shows react ions of the se lected varieties to biotic and abiot ic stresses, and the earliness eva luation. Bes ides the qua l ity eva lu at ion at the fa rmer's level, the four potential variet ies were also tested for nutri t ional and cu linary char- acteri stics in compari son w ith one of the most popular varieties in Bolivia, Waych'a (Table 3). Conclusions PROINPA's experi ence in Andean conditions indicates that the use of w il d spec ies w ith cu lt ivated potatoes in a breeding program is not as complex as is often argued . In many cases, on ly one or two backcrosses to cu lt i- vated spec ies are required to obtai n potatoes w ith acceptab le agronomic characteri stics . It appears that ecolog ica l and cli matic conditions, genetic variabi li ty, and po llen fe r- t i I ity of the spec ies and cu ltivars in the Andes, Table 2. Earliness and reactions' of selected clones to some biotic and abiotic stresses, CIP-PROINPA, 1994-1996. Clone Earliness Wort Nematode Hail Drought 90-123-22 Late (130-160 days) MR T T T 90-101 -3 Late (l 30-160 days) R T T T 90-123-18 Late (130-160 days) MR T T T 389349. l Late (1 30-160 days) MR T T T Soni lmilla (5. andigena) Late (1 30-160 days) s s T s Luk'y (5. iuzepczuki1) Very late (1 60-180 days) s s T R Alpha ( 5. tuberosum) Early (120 days) s s s s a. R = resistant, MR= moderately resistant, S = susceptible, T = tolerant. Table 3. Nutritional analysis and culinary tests of potential frost-tolerant varieties, CIP-PROINPA, 1996. Nome of cultivor Clone number Nutritional analysis(%} Cul inary quality' Dry molter Storch Protein CQFP CQCP lllimani 90-123-22 24.0 18.6 4.1 69.0 VG Totorefia 90-123-18 26.0 19.0 3. 1 58.5 VG Tun ari 90-101-3 19.8 20.0 3.8 54.0 G Condorib 389149. l 21.0 20.3 0.8 60.0 G Waycha (contro l) (5. andigena) 22.3 16.0 1.9 60.0 VG a. CQFP = Culinary quali ty in fried potatoes (sca le: l 00), CQCP = culinary qua lity in cooked potatoes, VG = very good, G = good. b. CIP selection. CIP P10910m Report 1995-96 2 31 especially in Bolivia, give these potatoes more flexibi lity for genetic manipulation th an po- tatoes in other regions, since potato fe rtility and reproduction are very sensitive to envi- ronmental changes. The yield capacity of the materi al tested was not directly related to percentage of fo- li age damaged by frost. Some va riet ies show good frost tolerance in their fo li age but do not y ield well , fo r example, the bitter variety used as a control. The capac ity of the plant to recuperate after a frost is an important fac- tor in frost tolerance and is an important se- lection cr iterion. There is, however, a need to better understand the relationship between different leve ls of frost intensity and y ield to estimate the potentia l impact of frost-to lerant varieti es under different frost conditions of Bolivian highlands. Research on modeling the response of potato to frost intensity w ill be ca rri ed out in co ll aborati on w ith CIP and CON DESAN. Although these potenti al va rieti es have been se lected through participatory research with farmers, there is st ill the important test of market acceptance. The strategy wi ll be to have the varieti es eva luated in different ar- eas of the country w ith farmers' communi- ti es, development projects, and seed produc- 232 Program 5 tion inst ituti ons invo lving local traders. In the futu re, we plan to invo lve, earli er in these- lect ion process, these other potential users of new vari eties. The two strategies fo llowed by PROINPA's breeding program have led to the se lection of potentia l frost-tolerant varieti es w ith a high potential impact in the Boli v ian highlands w here frost is a major constraint to produc- tion. Some of these vari eties cou ld also be of interest in other cou ntr ies w here frost is a prob lem . The regional networks and the cata- lyt ic ro le played by CIP in the interact ion be- tween nati onal agri cultural resea rch systems of different countri es shou Id contribute to and ensure the exchange of var ieti es and tech- no logies of common interest. Selected Reading Estrada, N. 1982. Breed ing wi ld and primitive potato spec ies to obtai n frost-resistant cultivated var iet ies. In: Pl ant co ld hardiness and freez ing stress. P.H. Li and A. Sakai (eds.). Academic Press Inc. p. 615-633. Estrada, N., W. Garcia, J. Gabri el, and E. Carrasco. 1993. Breeding potatoes fo r to lerance to frost and res istance to late bl ight. Am. Potato J. 70:809 . (Abstr. ) Sweetpotato Seed Units for the Dissemination of Planting Materials of Improved Varieties E.E. Carey', S.T. Gichuki1, 0. Hidalgo2 , and D.P. Zhang2 Relatively little attention has been given to sweetpotato improvement in many develop- ing countri es where the crop is important. In these countries, resea rch programs usually have not selected improved varieties, and, even if they have, formal systems for the dis- semination of planting materi als rarely exist. Instead, sweetpotato farmers in many deve l- oping countries grow large numbers of var i- eties, many of which are low yielding or late matu ring. These varieties are disseminated through informal systems based on farmer- to- fa rmer exchange. The lack of ea rly maturing, high-yielding varieties acceptable to farmers and the lack of timely availability of planting materials are two important constraints to increased sweetpotato production in most places where sweetpotato has received I ittle attent ion from researchers. The problems of timely availabil- ity of planting materials are particularly se- vere in drier upland areas, where resource- poor farmers usually do not have access to irrigation and may have to wait for remnants of previous crops to resprout at the beginning of the rains, or for small multiplication plots to produce adequate quantities of materi als for planting. By the time planting materi als are ava ilable, the rains may be taper ing off, thus limiting sweetpotato yields. When plant- ing materi als are scarce at the beginning of the season, farmers may accept any avail ab le ones, thus contributing to varieta l mixtures, but not necessa rily increasing the frequency of preferred varieties. 1 CIP, Nairobi, Kenya. 2 CIP, Lima, Peru. Research Objectives The basic objective of the CIP sweetpotato seed unit is to conduct research and devel- opment activities to assist with the dissemi - nation of superior varieties, and to improve the effectiveness of sweetpotato seed systems, so that farmers in developing countries will have a better availability of high-quality plant- ing materials of superior varieties. To date, seed unit activities have primarily concen- trated on Peru and Kenya. In Peru, efforts have been made to multiply and distribute substan- ti al quantities of high-qua lity pl anting mate- rials of recently released or promising variet- ies, and to identify varieties with a high ca- pacity to produce cuttings from sprouted stor- age roots. In Kenya, efforts have concentrated on the international di stribution of promising varieties for national testing and dissemina- tion. Results of these efforts are reported here. This report also discusses areas of possibl e future emphas is for the seed unit. Multiplication of Planting Materials In Peru, pathogen-tested in vitro cu ltures were used as the starting material for multiplica- tion of planting materi als of selected supe- rior varieties for distribution to various sites in 1995 and 1996. Multiplications were ca r- ried out using 2-node cutt ings in beds, at a density of 100 plants per square meter. Tabl e 1 shows the quantities of planting materi als produced by variety. Apical cuttings of se- lected varieties harvested from the multipli- cation beds were distributed to various loca- tions to serve as nuclear stocks for subsequent multiplication and distribution to farmers. Varying numbers of varieties were distributed CIP Pmgmm Reporl 1995-96 233 Table 1. Quantities of planting materials of selected sweetpotato varieties multiplied and distributed to Peruvian production zones by the seed unit in Peru. Nome (CIP number) 1995 second season 1996 first season Nacional (187003. l) Costanero (187016.2) Salyboro (187017. l) Yarada (187018.l) Tacna (187019.l} Atacama (l 87020. l) Canetano-INIA (l 88006. l) Jonathan ( 420014) Helena (420068) Jewel (440031) YM89.232 (189008.5) YM89.052 (l 89013.2) Limeiio Morado (420096) Trujillano (420097) Tanzania (440166) No. 29 (440168) OLP 3548 ((89.213 Total 4,930 l l,430 5,430 3,230 3,430 1,430 20,630 19,300 14,430 14,000 98,240 to each location, depending on local adapta- tion and demand for va ri eties under multipli- cat ion . During the first season, planting ma- terials were distributed to 14 sweetpotato-pro- du cing zones in Peru , and during the second season, planting mate ri a ls were distributed to 1 5 locations. Sprouting Ability A study was unde rtaken to examine differ- ences in sprouting ab ility among CIP-bred clones. Fourteen clones from a hi gh dry mat- ter breeding popul at ion were used. Storage roots were placed 15 cm apa rt in beds on Nov. 22, 1995, and the number of roots sprouted per clone was monitored weekly sta rting 26 days after planting and continu- ing until 29 days after planting, at which time 234 P1og1om 5 14,750 3,550 10,250 4,050 2,050 2,050 13,650 11,050 11,250 16,750 5,250 11 ,250 7,050 7,050 4,750 4,750 l 0,750 9,250 149,500 the number of sprouts per plot was counted. Nine roots of each clone were planted per plot, and the tri a l was laid out as a random- ized compl ete block design with three rep li- cations. Data analysis was done using the Proc Anova procedure of SAS, and mea n separation was done using the Waller group- ing method. Table 2 presents results of the tri al. Significant culti va r va riation was found for sprouting ability. In addition, those clones that produced the highest number of sprouts sprouted the ea rliest. The two clones with the best sprouting ability were selected for inclu- sion in cross ing blocks as parental clones. Seed Unit Activities In Kenya, seed unit activities are mai nly con- ducted under the auspices of the KARI Plant Table 2. Variation among sweetpatata clones far ability ta produce sprouts. Clone YM89. ll 8 SR92. 130 CC89.212 YM89. 133 USSC.500 ST87.070 SR90.307 LM89.128 YM89.2 15 SR89.519 YM89. 162 SR90.4ll SR90.323 YM89.099 Sprouts (no.) 149.3 120.3 100.0 86.7 71.3 49.0 48.7 46.0 43.7 43.0 38.3 31.0 19.0 0.0 Waller grouping 0 ob be bed cde def def defg defg defg efg efg fg g Quarantine Stat ion (PQS) at Muguga. Th is work involves the regional d istri bution of rela- tively small quantities of planting materi als of promising va ri eties to be tested and dis- seminated by var ious partners in the reg ion. Thi s activity is d ist inct from the reg ional breedi ng activities that invo lve the distribu- tion and test ing of much larger numbers of genotypes at se lected reg ional testing sites to identify limited numbers of promising geno- types for w ide d istr ibution. Tabl e 3 li sts the names, CIP numbers, and orig ins of six vari- eti es that are cu rrently bei ng w idely dissemi- nated by the seed unit in response to requests Table 3. Promising sweetpatata varieties distributed widely by the seed unit in Nairobi. Name Origin (CIP number) Cemso 74-228 (400004) Cuba Zopo llo ( 420027) Peru Helena (420068) Peru Yon Shu l (440024) Chino Novato (440131) Papua New Gu inea Tanzania (440166) Uganda Exceptional characteristics Early, high yield in g, good taste, good foliage vigor Early, high yie lding, droug ht tolerant, orange fl eshed, low dry matter content Hi gh foliage yield Early, high yielding, fair taste, proven acceptability for refugee relief food in Zaire Early, high yielding, good taste, broad adaptation, good foliage vigor Early, high yielding, brood adaptation, excellent taste, probably most widely grown variety in sub-Saharan Africa CIP Program Repmt 1995-96 235 for improved varieties by national programs and NGOs. The internat ional distribution of planting mate rials of sweetpotato, a vegetat ively propagated crop, requires strict adherence to phytdsanitary regu lations accepted by na- tional plant quarantine services. This has usu- all y been considered to require the in vitro shipment of pathogen-tested planting mate- ri als. In v itro multipli cation and distribution are problematic, however, as in vitro plant- lets are quite perishable, and clones are fre- quently lost during the process of shipment and subsequent transfer of p lantl ets to soi l. To overcome these prob lems, we have be- gun to distribute c lones internationally as cuttings taken from pathogen-tested mother p lants maintain ed in the quarantine screenhouse at PQS. These mother plants ori ginate from pathogen-tested in vitro plants obtained from CIP headquarters. At PQS they are transferred to pots an d retested for sweetpotato v iruses using sero logy and graft- ing to indicator plants. Peri odic re-testing ensures that pathogen-free pl anting materi- als are distributed. At the time of distribution, smal l, 2- to 3-node cuttings are taken from the mother plants, the ir leaves are removed, and they recei ve a fung icide treatment prior to packagi ng and sh ipping, usually by a cou- ri er service. By expanding the production of mother plants in the screenhouse, large numbers of pathogen-tested cuttings can be produced per clone. However, given the rapidity with wh ich sweetpotato can be multipli ed under fie ld conditions, there is usually li tt le need for ship- ping anything more than nuclear stocks (such as 20 to 40 cutt ings per clone) of planting materials from PQS. With the implementat ion of germplasm shi pments as pathogen-tested cuttings, losses during shipping and establishment have dropped markedly in comparison w ith in vitro sh ipments. Furthermore, improvements have been made in packag ing of cuttings, w hich should lead to sti ll further reductions in losses. Init ial shipments of cuttings were wrapped in 236 Progrnm 5 moi st paper towels packed in paper bags. Although few losses of shipments were re- ported, some problems arose with desicca- tion . We have subsequent ly found th at this can be greatly reduced by packing cuttings in plasti c bags . Tab le 4 presents a compari- son of weights of 1 0 2- to 3-node cuttings Table 4 . Weights of 2-node sweetpotato cuttings wrapped in moist paper towels or in plastic bags after 1 wk of storage. Treatment Mean weight of cuttings (g) Moist paper towel 2.2 Plastic bag 8.5 either wrapped in moist paper toweling or packed in plastic bags, fol lowing 10-d sto r- age in the dark at PQS. At the end of the stor- age period, cuttings w rapped in moist paper towe ls weighed rough ly one-fourth as much as cuttings packed in plastic bags. Future Considerations and Challenges There is considerable potenti al fo r further expansion of sweetpotato seed unit activ iti es, particu larl y w ith respect to finding ways of improving seed systems to ensure the timely avai lab ility of pl anting materi als of improved varieti es in developing countries. Parti cularl y noteworthy is the general lack of information on the potenti al benefits of foundation seed program s in developing co untri es. Sweetpotato is a crop that is wel l known for the tendency of varieties to degenerate (i.e., to lose yield potential and varietal character- isti cs over subsequent vegetative generat ions) . This degeneration has been attri buted to the accumulation of deleter ious mutations and pathogens. In the United States, several states operate foundat ion seed programs that pro- v ide farmers with seed stocks of va ri et ies maintained true to type through positive hill se lection for high y ield and through the elimi- nation of off-types . Recommendat ions have also been deve loped to enab le farmers to maintain their own high-quality seed stocks. These programs have clearly demonstrated the benefits of positive hill se lecti on for yield and tru eness to type. Sweetpotato foundation seed programs do not ex ist in most developing countries, nor do recommendat ions for farmers to maintain their planting stocks. In tradition al systems where farmers grow a number of va riet ies, there is a turnover of va riet ies with time, with yields of older varieties often declining, and w ith the introduction of new, better perform- ing varieti es obtained from variou s sources. Farmers also usually pay close attention to the se lect ion of di sease-free planting mate- rial, particularly in areas where virus diseases can severely affect yields. O ne thing that is rarely practiced in the tropics, however, even by many resea rchers, is the se lection of pl ant- ing material on the bas is of per-plant perfor- mance (hi 11 se lecti on). Thi s is because pl ant- ing materi als in the tropics are usually taken as vine cuttings from ex isting pl ants; storage roots are not used as a source of seed. Recent studies reported by ASPRAD re- searchers in Sri Lanka and Thailand have shown the rap id benefits that hill se lection can have in improving yields of sweetpotato in those countries. There is a need for fur- ther verifi cat ion and broad dissemination of the findings from th ese programs in other developing countries. Demonstration of the benefits of hill select ion could lead to the widespread practice of hill se lection by both researchers and farmers, and might lead to increases in sweetpotato yields whil e halting the common ly observed phenomena of vari- etal decl ine and turnover. Another cha I lenging area for resea rch by the seed unit is the search for so lutions to the prob lem of timely availability of pl anting material s fo r farmers in drier areas. Although technologica l innovations may help solve the prob lem, approaches developed in a com- munity-based, participatory fashion wi I I likely be more successful. In thi s area, NGO part- ners may be sought, or work aimed at st imu- lating v iab le community-based seed systems may be conducted as part of multidisciplinary efforts to boost sweetpotato production by diversifying forms of utili zation . Selected Reading Dangler, J.M. (ed.). 1994. Sweetpotato foundat ion programs. HortTechnology 4:223-238. De Silva, K.P.U., and H.D. Jayawickrema. 1996. Selection of true-to-type vari ety Wary ipo la-red . In: E.T. Rasco and V.dR. Amante (eds.). Selected Research Papers, July 1995-June 1996. Volume 2: Sweetpotato. ASPRAD, Manil a. p. 1-3. De Sil va, K.P.U., and A. Premathilake. 1996. Effect of variety and number of generations of vegetative propagati on on vegetat ive and yield traits of sweetpotato . In: E.T. Rascc and V.dR. Amante (eds.). Selected Research Papers, July 1995-June 1996. Volume 2: Sweetpotato. ASPRAD, Manila . p. 4-9. CIP Progra m Repo rt 1995-96 2 3 7 PROGRAM R E PORT Postharvest Management, Marketing Gregory J. Scott1 Postharvest issues are being pushed to the top of the research agenda in many countries that produce pota to es, sweetpot atoes, and Andean root and tuber crops (ARTC). The driving force to overcome postharvest con- straints and capitalize on market opportuni - ties is the growing commerc ialization of ag- riculture combined with urbanization, popu- lati on growth, ri si ng incomes, and increas- ing emp loyment of women. During 1995-96, CIP's postharvest program co ll aborated with industriali zed- and developing-country insti- tutions to deve lop technologies, provide train- in g, and supp ly inform atio n to m ake postharvest research on roots and tubers con- tribu te to reduc ing poverty, improving nutr i- tion, and providing opportu ni t ies for women in Asi a, Afri ca, and Lat in America. 1 Program Leader, CIP, Lima, Peru. 2 Acronyms c ited in this section can be found w ritten out in the sect ion Acronyms, p. 320. Pro Potato Noteworthy results were achieved in • globa l projections of product ion, use, and trade; • storage; and • marketing and process ing. Projections. In antic ipat ion of the Wor ld Food Summit in Rome in November 1996, CIP's Postharvest Management, Marketing Program co ll aborated w ith the Bas ic Food- stuffs Servi ce of FA02 on a study of the world potato economy. We found (1) a tremendous shift under way in the relocation of potato production from developed to develop ing countri es, and (2) an emerging importance of processing and trade in potatoes and potato CIP Program Report 1995-96 239 products, which could lead to a 50% increase in potato production in developing countries by 2000. These findings portend a greater recognition of the importance of potatoes general ly, and the needs of developing coun- tri es specifically, in th e global resea rch agenda. Storage. Seasonal potato production pat- terns in India generate large fluctuati ons in pr ices paid to fa rmers for thei r crop. With improved rust ic stores, small fa rmers would have another opt ion besides se lling thei r crop at harvest when prices are usually the lowest or putting it in co ld storage. CIP worked w ith the Central Potato Research Inst itute of India to document trad iti onal sto rage practices, analyze the experi ence of small fa rmers with co ld storage, and deve lop low -cost rusti c stores. The rustic storage improvements were a technical success, but more cost ly. And the demand for such stores was lower than pre- viously ca lculated. A ware storage manual synthes izing the co ll aborati ve research in In- dia and experiences elsewhere is being de- ve loped because storage remain s high on the li st of major production constraints in Asia. Marketing and processing. Baseline case studies in collaboration w ith the CAAS in China, CORPOICA in Colombia, and COSUDE in Bolivia show that potato fa rmers are increasingly market-oriented. Po lici es aimed at improved commerci alization, such as fac ilitating loca l processing and exp loit- ing export opportunities, merit greater atten- tion. Process ing is a growth sector in China and Colombia in particular. Rapid market apprai s- als of the process ing sector in Kenya w ith KARI , in Mexico and Central America in col- laboration w ith PRECODEPA, in Indonesia with UPWARD, and thesis resea rch in Peru, w ith support from the University of Nymegen in the Netherl ands, all point to a rapidly ex- panding market for processed products. Imports of such products, while currentl y limited, could balloon if loca l supply does not expand to meet loca l demand. This tendency has already emerged in parts of Latin Ameri ca, 240 Progrorn 6 where tariffs co ntinue to fa ll so countri es can ad here to GATT. Field tri als in Peru to se lect processing materi als from advanced clones w ith des ired agronomic traits continue to gen- erate a small number of promising culti vars. All indi cation s are that the demand for such materi als is likely to increase. Sweetpotato Advances in sweetpotato postharvest research have consisted of • macrostudi es of processing potential; • microanalys is of starch, flour, and feed; • development of the commodity database and projections; and • completion of methods mater ials. Macrolevel studies. Studies of national and provincial statisti cs in China, in co ll aborat ion w ith CAAS, document th e sprea d of sweetpotato processing primarily for feed and sta rch. Processing has income- and employ- ment-generating potential for smal I farmers in the poorer parts of the country. The poten- ti al for the expanded use of sweetpotato as a substitute for imported maize in pig feed is a heretofore overlooked way for China to avoid mass ive feed imports predicted by some ob- servers. Regional surveys conducted in Vietn am indi cate an upswing in the use of sweetpotato fo r animal feed and sweetpotato starch fo r making noodles, w ith prospects for even more expansion. The surveys and analysis were done by the National Institute of Agricultural Sciences, the Post Harvest Technology Inst i- tute, and the Animal Husbandry Research Institute, with ass istance from CIP, UPWARD, and CIAT staff. Similar research conducted by economi sts at the University of the Philippines at Los Banos shows neg ligible current use of, but co nsiderabl e potential for, sweetpotato to substitute for imported feed and flour. Analy- sis of secondary data in Indonesia found that current processing accounts for some 25% of total produ ction--far more than previously believed. But, as in the Philippines, lower raw material costs in the form of higher yields and higher conversion rates are needed fo r pro- cess ing to reach its fu ll potential. These macrostudies document the impor- tance of and potential for sweetpotato pro- cess ing, lay out a set of co ll aborat ively forged resea rch priorit ies and polic ies, and provide a common framework for the var ious institu- tions engaged in sweetpotato postharvest re- search. Microanalysis. Detail ed eva luat ions of ex isting techno logy foc used on sta rch, flour, and feed in Chin.a, Kenya, Peru, and Uganda. Work with SAAS on sma ll-sca le starch pro- duction in Si chuan Province, China, was the most promi sing of all. It identifi ed severa l ar- eas for immed iate improvement. Operat iona l analysis in Peru quantified the raw materi al cos ts and conversion rates needed to achieve profitabi li ty at an ex isting starch pl ant. The necessary higher yi elds w ith more extractable starch seem well within reach. Sweetpotato fl ou r appears more problemati c in Peru for a va riety of reasons. However, in Uganda-and to a lesser extent in Kenya-prospects are brighter for flour. Markets are emerging in the wake of declining supp li es of cassava flour. Economic analysis of farm-leve l use of roots for pig feed in China is extremely encourag- ing. Modest changes can improve the pros- pects for even more w idespread use . In Peru, dual-purpose variet ies w ith bal- anced production of vines fo r fodder and roots for human consumpt ion show considerable potent ial to improve the lot of sma ll-sca le dairy farmers. Databases and projections. The pocket- sized compendium of sweetpotato statisti cs for 33 major sweetpotato-producing coun- tri es, issued in 1996, aims at broadening pub- 1 i c awareness abo ut the po tenti a l for sweetpotato in developing countri es. Methods materials . In addition to research results and poli cy recommendations, a ma- jor effort has been made to improve loca l re- search capac ity by preparing posth arvest methodologies appropri ate to conditions in Africa, Asia, and Lat in America. Adding value to root and tuber crops, the manual on prod- uct development, co-pub I ished with CIAT and llTA, prov ides operational guidelines and a co mm on framework fo r sweetpotato postharvest researchers wor ldw ide. Prices, products, and people, a compendium of methods for analyz ing agr icultu ra l marketing in develop ing countries, prepared with sev- eral IARC socia l sc ientists, is intended to serve a similar purpose. Andean Root and Tuber Crops In our work on ARTC, considerable progress was made in each of three spec ial projects: (1) biodiversity, (2) commodity systems, and (3) starch process ing. Work in the COSUDE biod ivers ity project has emphasized eva luat- ing traditional postharvest practices for crops such as ulluco, oca, and native potatoes with a view to their eventua l improvement. IDRC commodity systems resea rch has focused on deve lop ing commercial products such as protei n-enri ched snacks or new uses for tra- ditional flours. BMZ-supported posth arvest work involves documenting process ing tech- niques and consumption patterns for starches derived from ARTC, for example, from achira, and investigat ing the biochemica l properti es of the products . As the consortium of CON DESAN partners conso lidates, progress has been made in forg- ing a common conceptual framework and resea rch agenda to fo llow up on findings to date. Closer links with PRODAR, the Lat in Amer ica n network for small-sca le agroen- terpri se, and CIAT's rural agroenterprise ini - tiative have also been established. CI P Program Report 1995-96 24 1 Making Sense of Agricultural Marketing in Asia, Africa, and Latin America G.J. Scott1 Recent trends in domest ic, as opposed to fo r- eign, agri cultural marketin g are the focus of a 1995 book, Prices, products, and people: Analyzing agricultural markets in developing countries. Thi s co llection of papers by soci al sc ienti sts at va ri ous in ternati onal agri cul tural research centers and their co ll aborato rs in national resea rch systems covers both data co ll ection and data analys is methods. Marketing and Economic Development Agri culture has become more market-ori- ented in v irtu ally all develop ing co untri es over the past 30 years. Subsi stence produc- ti on has declined in re lati ve importance in part because techno log ical improvements mean that producers have more output to sell. Mushrooming urban areas- pa rti cularl y in sub-Saharan Afri ca-mean th at mo re and more consumers depend on agri cultural mar- ket ing fo r their da ily food requi rements. Even env ironmentali sts have show n a growing in- terest in the benefits and costs associ ated w ith agri cultural marketing. 1 CIP, Lima, Peru. Most important, inc reased produ cti o n, popul ation grow th , and improvements in in- frastructure have meant that the sheer vo lume of agri cultural goods traded in domesti c mar- kets has expanded eno rm ously (Tabl e 1 ). Hence, the potenti al rewards to society and individual s from improvements in domestic agri cultural marketing have mult ip li ed in cor- responding fa sh ion. Limitations of Previous Research Much of the literature on agri cultural market- ing in developing countri es can be catego- rized into one of three types: (1) compilations of bas ic principles or concepts, (2) rev iews, syntheses, o r concept papers, and (3) case studi es of particul ar marketing system s or agri cultural marketing enterprises. In publi cations such as these, the end re- sults of the appli cation of economic analys is to the study of agri cultu ra l marketing often abound . But often lacking are step-by-step expl anations of the procedures by w hich such locati o n- , t ime-, and product-spec ifi c re- search might be extrapol ated to another pl ace Table 1. Trends in food production and foreign trade in developing countries, 1961-91. Country Commodity Phil ippines Rice India Wheat Kenya Maize Colombia Potato a. Exports plus imports divided by production. Source: FAO, PC-Agrostat, unpublished, 1993. 242 Program 6 Production {000 t) Trade (%)0 1961 1991 1961 1991 3,910 9,673 4.8 0.1 10, 997 55, 134 28.l 1.2 940 2,340 11 .2 0.8 551 ~225 0.1 1.4 or another commodity. It is thi s gap between the conceptual and case study I iterature on domestic agricultural marketing in develop- ing countries that Prices, products, and people attempts to breach. Because of the frequent need for prompt attention to particular policy-related market- ing issues, this publi cation emphasizes prac- ti ca l and rap id research procedures. Al so included are simplified substitutes for more complicated approaches. In setting out a co l- lect ion of methods to help practitioners ad- dress the types of marketing quest ions th at analysts are most frequentl y confronted with in developing countries, the volume also aims at providing researchers with a cross section of techniques most appropriate for the task at hand. Many contributions have appendices that spell out guidelines on survey procedures or analytical techniques in greater detail. Near ly al I chapters provide examples of the appl ica- ti on of these techniques to specifi c commodi - ti es based on research in different develop- ing countries. The intent here is to enrich the general exposition of a seri es of methods with an array of commodity-spec ific experiences invo lving their use . Selected Reading Fleming, E.M. 1990. Proposa l for a market- ing systems research approach in agricul- tural development pl anning. Agric. Sys- tems 32:97-111 . Kinsey, J. 1988. Marketing in developing countries. Macmill an Educati on Limited, London . Scott, G. (ed .). 1995. Prices, products, and people: Analyz ing agri cultural markets in developing countries. Lynne Ri enner Pub- lishers, Boulder, Colorado, USA. Sellen, D. , W. Howard, and E. Goddard. 1993 . Production to consumption systems resea rch : A rev iew of methods and ap- proaches. Report prepared for the Interna- ti onal Deve lopment Research Centre, De- partment of Agricultural Economics and Business, Uni versity of Guelph, Guelph, Canada. (IP Prog mm Report 1995-96 2 4 3 Perceptions Versus Projections for Potatoes: New Estimates Point to a Changing Global Research Agenda G.J. Scott1 and A. Coccia2 In preparat ion for the Wor ld Food Summit in Rome in late 1996, FAO carri ed out a seri es of studi es to prov ide an overv iew of past trends and future outlook for the world ' s ma- jor agri cultural commoditi es . CIP and FAO staff jo intly prepared the study (begun in 1994 and written up in 1995) on potatoes--the first of its kind for FAO fo r thi s commodity. Thi s paper summari zes the principal f indings from th at co ll aborati ve effort, incl ud ing the analy- sis of past trends in product ion, area planted, and y ields; the evoluti on of marketing and use patterns ; and proj ect io ns to th e yea r 2000--what those trends predicted and how they compare with actu al output in recent years. These findings served as the catalyst for initi ating a collaborati ve effort w ith econo- mists from the International Food Po li cy Re- search Institute (IFPRI) to generate projections fo r potatoes, sweetpotatoes, and cassava in the year 2020. Preliminary results from that work are also included here to lend greater w eight to the data presented. Materials and Methods We analyzed histori ca l growth rates in pro- ducti on, area, and y ield by reg ion and by country for the peri od 1961 -93 . Point-in-time compari sons for consumption, use, and trade fo r 196 1-63 vs. 1991-92 were similarl y ana- lyzed. The stati st ics are al I from FAO databases w ith the exception of use data fo r the United States from the U.S. Department of Agricu 1- ture. The latest FAO fi gures on these va ri abl es 1 CIP, Lima, Peru . 2 FAO, Rome, Ita ly. 244 Program 6 are cited se lecti ve ly to substant iate the analy- sis w hen necessary. The FAO/CIP projecti ons are fo r a si ngle commodi ty and fo r indi v idual countries in major economic and geographi ca l reg ions such as Europe, the area of the fo rmer USSR, Afri ca, As ia, and Latin Ameri ca and the Ca r- ibbean. The base period is 1987-89 and the project ions go to the year 2000. O n the de- mand side, the model is basica ll y dr iven by changes in popu lat ion and per capita incomes inasmuch as relati ve pr ices are assumed to remain constant throughout the period (1987- 89 to 2000). The resul ts cover two iterati ons: (1) the base period to 2000, and (2) incorpo- rating the effects of the Uruguay Round of the General Agreement on Tari ffs and Trade (GATT) fo r the same time frame. Prel iminary results from the IFPRl/CIP co l- laborat ion are onl y fo r potatoes. But those projecti ons fo r the year 2020 are made using a model that simultaneously estim ates sup- pl y, demand, and trade for al I the major food commod it ies. The model also expli cit ly al- lows fo r changes in relati ve pri ces and gen- erates results fo r se lected countries, subre- gions (e.g., East Afri ca), and regions. Results and Discussion Misperception 1. As deve loping countries grow economica ll y, potato producti on and consumpti on w ill foll ow the pattern of sharp dec line in Europe. Recent reports on potatoes in Europe are replete w ith statisti cs on declines in area pl anted and production . Data on use also show sharp drops in per capita consumption of fresh potatoes and in the use of potatoes for animal feed. Based on these fi gures, Eu- ropean observers often assume that, as the economies of other regions of the world ma- ture, sim ilar patterns mu st already ex ist or will certainl y preva il in the not-too-d istant future. But empiri ca l evidence indicates just the op- posite. Potato production in developing countries rose from 29 to 85 million t between 1961- 63 and 1994-96, more than offsetting the drop in product ion in industri ali zed countri es . Hence, world potato output rose slightly, from 265 to 275 million t, even though output in Europe dropped by more th an 50 million t during the same peri od. China is now the world ' s largest potato produ cer since th e breakup of the USSR; India ranks sixth. In fact, area planted to potatoes grew faster in devel- oping countries than fo r any other major food crop over the 30-yr period studied (F igure 1 ). Production increases were second only to those of w heat (F igure 2) . Production in- creases have been particularly strong in Asia. The annual growth rate in production aver- aged 4.0% over the past 30 yr as area planted expanded 2.6% yearl y and yields increased by 1.4% annually. From a base peri od fi gure of 75.6 million t in 1987-89, the FAO/CIP projections devel- Potato Cassava Wheat Roots and tubers Rice Pulses Coarse grains 0 20 40 oped in 1994 ca lled for potato output in de- ve lop ing countries to reach 105.4 milli on t in 2000. By 1994-96, FAO product ion figures showed output at nearly 100 million t, we ll on target to reach the proj ected tota l. Further- more, pre liminary est imates from the IFPRI/ CIP projections indicate that potato produc- tion in developing countri es will continue to expand at an average annual rate of 2.0% over the next 25 yr. Misperception 2. The bulkiness and per- ishability of potatoes severely limit intern a- tional trade. In econom ists ' jargon, potatoes are th e archetypica l " nontradable," that is, the g0od th at is class ified as one fo r which on ly an in- ternal (or domestic) market ex ists. Reasons for this class ifi cation abound. Potatoes indeed are bulky and peri shabl e, and therefore diffi cult to transport over considerable di stances. Their low unit va lue-to-weight rat io and co rre- sponding transport costs also discourage trade. Loca l (consumer) preferences for skin co lor, fl esh co lor, and tuber size, as well as for packaging and grading, constitute add i- tional constraints to international commerce. Nevertheless, trade in table potatoes and seed rose from 3 million t in 196 1-63 to 7.5 million t in 1991-93. If trade in processed potato products is added, recent est imates put 60 (%) 80 100 120 Figure 1. Percentage increase in area of selected food crops in developing countries: 1961-63 and 1994-96 averages. CIP Program Report 1 995~96 245 Cassava Roots and tubers Pulses 0 50 100 150 200 250 300 350 (%) Figure 2. Percentage increase in production of selected food crops in developing countries: 1961 -63 and 1994-96 overages. world potato trade at 10 million t year ly, or about 4% of annua l globa l production. By way of comparison, wor ld trade in rice is roughly 3% of annual output. Trade patterns for potatoes have also di- versified over the past 30 yr. Exports of pota- toes by developing countr ies have risen from 400,000 t to near ly 1.3 million tin this pe- riod. Shipments of seed and processed po- tato products from industrialized to develop- ing countries have also expanded sharply. By the late 1980s, for examp le, more than one quarter of al I Dutch seed exports were to de- ve lop ing countries in North Africa. A world market for potatoes is clearly emerging, with countries tending to specialize in different types of potatoes as well as developing re- giona l trading partners. Part of the misperception about trade is the result of inadequate statistics. FAO includes in its trade data only trade in fresh potatoes and seed. Until governments supply better figures, data ava il ab le to FAO w ill continue to underestimate international trade in pota- toes and potato products. More important, many developing and deve loped countries have onl y recent ly begun to apprec iate the implications for their potato sectors of liber- alized trade po li cies and reg iona l trade agree- ments. In a number of Latin American cou n- 246 Program 6 tri es, imports have risen sharply even though tariffs are st ill comparative ly high. As these tariffs fa ll in the years ahead to comply w ith CATT, the pressure to improve competitive- ness or be forced out of production w ill be considerable. The window of opportunity to effectively add ress such developments is clos- ing rapidly, given that tariffs for many cou n- tries are due to reach the agreed-upon mini- mums in 6 or 7 yr. That is roughly the mini - mum lead time required to fully promote de- ve lopment and diffusion of improve d germplasm. Misperception 3. Potatoes produced in developing countries are cultivated mainly by subsistence farmers for on-farm consumption. Potatoes are produced in nearly 100 coun- tries in Asi a, Africa, and Latin Ameri ca, typi- ca lly by sma ll -scale farmers w ith holdings usually less than 5 ha. A "big" potato farmer in Bangladesh, for examp le, might have 2 ha planted to the crop . These growers usuall y depend on family as we ll as hired labor to plant and harvest. Potatoes are nearly always consumed by those farm households that pro- duce the crop. In fact, potatoes in develop- ing countries are often harvested w hen the basic stap les of rice, wheat, or maize are in seasona l short supp ly. Hence, potatoes are often referred to as a subsistence crop. The bul k of potatoes produced in As ia, Lat in Ameri ca, and North Afri ca--even on small farms--is so ld . Potato sa les are an im- porta nt source of cash income, even in those parts of sub-Saharan Afri ca and in more iso- lated pockets in Lat in Ameri ca w here on-fa rm co nsumption preva il s. Commerc ial potato production dominates in developing countries because y ields-part icul arl y in Asia-are high enough fo r fa rm families to be ab le to eat some of w hat they harvest and still se ll most of the crop. Returns for potato, a crop w ith a vegetative period of 100-1 30 days in many parts of the trop ics and subtropics, are extremely lucrati ve because of strong off-farm demand and re lati ve ly high pri ces. Misperception 4. As per cap ita incomes ri se in deve loping countri es, per capi ta po- tato consumpt ion w i 11 fa l I. Per cap ita consumpt ion of potatoes in Western Europe fell by 22% between 1961- 63 and 1991 -93; even more in Eastern Eu- rope and the former Soviet Union. As incomes generally increased du ring the past 30 yr, European consumers have preferred to eat fine gra ins, fru its, and other vegetab les over potatoes. Hence, per capi ta co nsumpti on shrank in Western Europe from 102.3 to 79.3 kg/yr, in Eastern Europe from 11 7.4 to 80.3 kg/yr, and in the fo rmer Soviet Union from 13 1 .8 to 78.0 kg. As inco mes in c rease elsewh ere-it is often infe rred-per capi ta potato consumpti on w ill fa ll accordingly. Per cap ita potato consumpt ion in devel- oping countri es rose by 43% during the past 30 yr. It is much lower in As ia (11.7 kg), Af- ri ca (8 .0 kg), and Latin Ameri ca (20.6) than in Europe (78-80 kg) . Incomes are also lower. Potato is typ ica lly a comp lementa ry vegetable or seasonal stapl e in deve lop ing countries rather th an a princ ipa l source of carbohy- drates. As income goes up, consumers often prefer to eat more potatoes to d iversify their cerea l-based d iets. Detail ed info rmati on on consumer expen- d itures by commodity by income group in a given developing country is often hard to fi nd. Such data are typi ca ll y fo r "a ll roots and tu- bers," or "a ll vegetables," and for "a ll income groups." Where such d isaggregated stat istics are ava il abl e, the resul ts typi ca ll y show that consumpti on increases w ith income, even among low- income consumers. In addit ion, potatoes play an importa nt ro le in deve lop- ing-country diets as a source of vita min C and essential amino ac ids li ke lys ine that are lim- it ing in ri ce. Misperception 5. As the area planted to cerea l stapl es increases in As ia, the area pl anted to less importa nt crops like potato mu st decrease. Increased food product ion in deve lop ing countri es--part icularl y in Asia----over the past 30 yr is most commonly assoc iated w ith the adopti on of new, high-y ielding varieti es of w heat and ri ce. This t rend is particularl y true in South As ia, w here dwarf w heat and high- yield ing ri ce varieti es were gradual ly taken up by large, medium, and even sma ll fa rm- ers. Thousands of ha are now under cu lt iva - tion w ith these varieties, w hich in large part exp lains the increases in area planted and production (Figures 1 and 2). The spread of the improved w heat and ri ce variet ies greatly fac ilitated the expansion of area planted to potatoes in several respects. The new, hi gh-yie ld ing cerea ls have a much shorter durati on than the traditional va rieties . Hence, the t ime required to grow a ri ce crop in Bangladesh was shortened from fi ve or more months to four or less. This left a per iod during the agri cul tural ca lendar w hen an add itional crop could be grown. Further- more, to take fu ll advantage of the improved varieties, fa rmers need irr igation. But w ith the much higher y ields of the improved va ri e- t ies, investments in irrigat ion were now full y just if ied and spread rap id ly in the cerea l- growing areas. A lthough there is not enough irri gation water in the off-season(s) to grow ri ce, there is suffic ient water to grow a potato crop. As higher y ields from improved cerea ls meant higher incomes fo r small fa rmers in many areas, the increased purchas ing power often translated into more demand fo r pota- toes. That meant addi t ional increases in po- CIP Program Report 1995-96 24 7 tato production and, consequently, further growth in area planted. The most recent (1991-95) national statis- tics for area planted to potato in China show continued expansion in the area under po- tato cultivation. Similar trends are evident elsewhere in Asia. Furthermore, preliminary estimates from the IFPRl/CIP projections to the year 2020 also indicate continued expansion in area planted in the decades ahead, albeit at a more modest rate than before. Conclusions Recent analyses of past trends and future pro- jections for potato clearly indi cate the crop's increasing importance as a world food crop. Statements that wou Id suggest otherwise are frequently based on mispercept ions about the performance and potential of the crop in de- ve loping countries. As such, potatoes merit c loser attention by poli cymakers and re- searchers concerned with global food prob- lems in both developing and industrialized countries in the years ahead. The concentration of potato production is shifting from Europe to Asia and other parts of the developing world at an astounding pace. In 1961 , potatoes produced in devel- oping countries accounted for about 11 % of global output. Potato production reached some 30% in the early 1990s and is projected to be around 40% in 2020. No other major food crop has experienced such a shift in the location of production over this period (1961- 93). In addition, developing countries are in- creas ingly the target of industrialized coun- tries' potato exports and, to a lesser extent, vice versa. Therefore, the needs of produc- ers and consumers in Asia, Afri ca, and Latin 248 Program 6 America increas ingly will be the focus of po- tato researchers worldwide. Postharvest issues such as trade and processing are entering a critical phase in this regard inasmuch as tar- iff barriers are set to come down and the de- mand for processed products is certain to in- crease for many reasons. Process ing and trade are the two fastest growing sectors in the global potato economy. The internat ional database for process ing and trade--p arti cularly for developing coun- tries--needs to be improved. Otherwi se, fore- casters will simply continue to underestimate their growing importance. Building on past successes in the collaborative work on pro- jections, this might well be an area for future joint initiatives between CIP and FAO. Selected Reading Crosnier, J.C. 1994. L'evo lution de la culture de la pomme de terre en Europe et en France. In : Proceedings of the 12th Trien- nial Conference of the European Associa- tion for Potato Research . Held in Paris, France, 18-23 July 1994. lnstitut de Recherche Agricole (INRA), Paris, France. FAO/C IP (Food and Agriculture Organization of the United Nations/ International Po- tato Center). 1995. Potatoes in the 1990s: Situation and prospects of the world po- tato economy. Rome, Italy. 39 p. Rosegrant, M. , M. Agcaoili-Sombilla, and N. Perez. 1995 . Global food project ions to 2020: Implications for investment. Food , Agriculture, and the Environment Discus- sion Paper 5. IFPRI, Washington, D.C. , USA. Scott, G. 1994. The emerging world market for potatoes and potato products. Economie et Gestion Agro-alimentaire (France) 30: 19-27. Sustainability of Potato Consumption Developing Countries: The Case of Bangladesh . 1n G.J. Scott1 and H.E. Bouis2 Since the ea rl y 1960s, a number of horticul - tural crops have experi enced very rap id in- creases in product ion in developing countri es. Typi ca ll y, they are high-va lue, short-durat ion, labor-in tensive crops grown primaril y fo r sa le rather than fo r on-fa rm consumption. As a resul t, several countri es, part icularl y in As ia, look increasingly to hort icul tural crops as a source of increased output, consumption, and income. Potato (So lanum tuberosum) is particularl y important, partly because product ion growth has been so noteworthy. The percentage in- crease in product ion fo r potatoes in deve lop- ing countries from 196 1 to 1996 was greater th an that fo r any other major food crop ex- cept wheat. For example, potato product ion in India increased by 525% during the pe- ri od to near ly 17 milli on t. In Pak istan, po- tato production increased from 28,400 t in 1947-48 to over 1 milli on t by 1994-96. Tur- key, Iran, North Korea, and Bangladesh have also had substantial increases in potato out- put. The domesti c market has absorbed vir- tually all of thi s increased production. These trends have raised a seri es of ques- tions about the potential fo r expanded potato consumption and product ion. • Has the observed increase in consumption been due to changes in relative pri ces or changes in income? • To w hat extent would even lower pri ces, through increased producti on, stimul ate even higher demand? 1 CJP, Lima, Peru. 2 International Food Policy Research Institu te, Washington, D.C., USA. We attempt to answer these quest ions by examin ing historica l demand elastic ity esti- mates fo r potatoes in the case of Bangladesh and then compar ing them w ith observed trends in use. Materials and Methods Resul ts presented in this paper draw upon three different methods and sets of data. Be- cause of space limi tat ions, onl y the results themse lves and their interpretat ion are in- cluded here. The f irst method invo lves est i- mating the demand parameters for a number of key food groups based on househo ld ex- penditure data and a simpli f ied analyt ica l technique des igned espec iall y to est imate a demand matr ix fo r a highl y di saggregated group of foods . The second method consists of a more tra- diti onal procedure used to estimate the same parameters, but w ith househo ld expenditure data co llected some 15 yr later. The third method includes the resul ts of a rapid market appraisa l and descript ive analy- sis of the latest secondary data on produc- tion and consumption gathered in fie ld work and stat istica l moni tor ing since 1982 . The cross-checking of resul ts is intended to test the pred ictab i I ity of the estimated pa rameters as we l I as the rel iab i I ity of such est imates fo r predict ing longer-term trends in consumer behavior. Results and Discussion Baseline hi stori ca l data for Bangladesh are taken from the Household Expenditure Sur- CIP Program Report 1995-96 2 4 9 vey for 1973-74, conducted by the Bureau of Statistics. Selected food consumption patterns using these data are shown for urban and ru- ral , high- and low-income consumers (Table 1 ). Relative price data per 1,000 calories are expressed as a ratio of the price paid for the least expensive grain (in this case, for wheat) by the low-income, urban quartile. For ex- ample, in 1973-74, the lowest income urban consumer in Bangladesh paid 1 .63 times the price of wheat for 1 kg of potatoes (Table 1 ). However, since potatoes contain roughly one- fifth the calories of wheat on a price per calo- rie basis, potatoes were 7 .39 times more ex- pensive than wheat, due partly to their higher price, but more importantly to the lower quan- tity of calories (Table 1 ). Rice and wheat are the two main staple foods consumed in Bangladesh. In calorie equivalents, rice was more than twice as ex- pensive as wheat in urban areas at the time of the 1973-74 survey. Based on these his- torical data, apparently both urban and rural populations would buy higher quality wheat and rice as incomes rose, although the ten- dency is more marked for rice. Wheat was more expensive in rural areas than in urban areas; the price of rice was about the same between urban and rural areas. Per capita annual potato consumption in Bangladesh in 1973-74 was higher in urban areas than in rural areas (Table 1 ). More than 85% of the total population resided in rural areas at the time. Still, in urban areas, pota- toes were more than seven times as expen- sive a source of calories as wheat for low- income groups and nearly nine times as ex- pensive for high-income groups, which pur- chased higher quality potatoes. Also in urban areas, potatoes were more than three times as expensive a source of calories as rice. The price differential between wheat and potatoes and between rice and potatoes was not nearly so great in rural areas. Historical information about food imports and real wages helps to put these figures in perspective. In 1973-74, Bangladesh had just recently achieved nationhood after a war of independence. The devastation of the war was aggravated by the famine of 1974. Hence, in 1973-74 Bangladesh was heavily dependent on food aid and commercial ce- real imports. Table 1. Consumption per annum and calorie prices for potato, wheat, and rice in Bangladesh, 1973-74. Income• Per capita Calorie priceb Market price group consumption (kg) (per 1,000 calories) (USS) Potato Wheat Rice Potato Wheat Rice Potato Wheat Rice Urban 3.6 68.6 83.2 7.39 l.00 2.18 l.63 1.00 2.24 2 5.2 64.0 98.8 7.98 1.01 2.39 l.76 l.01 2.46 3 7.3 58.2 111.8 8.03 l.04 2.49 l.77 1.04 2.56 4 8.8 54.l 125.3 8.98 l.06 2.67 l.98 l.06 2.75 All 6.8 Rural l 1.0 41. l 75.4 6.62 l.43 2.28 l.46 1.43 2.34 2 2.6 41.1 103.0 6.39 l.50 2.36 l.41 1.50 2.43 3 3.1 34.8 130.5 7.44 l.46 2.41 l.64 l.46 2.48 4 4.7 26.0 162.8 7.94 l.64 2.66 l.75 l.64 2.74 All 3.5 a. Income groups refer ta total expenditure quartiles with l designating the lowest expenditure quartile. b. Relative to price of cheapest groin calorie source. 250 Program 6 Total wheat supp ly was in the neighbor- hood of 1.7 million t. According to FAO, nea rly 95% was imported. Domest ic wheat production at the time was sli ghtly over 100,000 t. Domesti c potato supp ly consisted of some 725,000 t, virtu ally al l of which w as produced loca ll y. Loca l wheat prices were depressed by aid, imports, and food subsi- di es, particul arly in urban areas. Hence, the large differences in relative pri ces per calorie for potatoes vs. wheat that preva iled at the time the household survey was carri ed out were at least partly due to these unusual cir- cumstances. Rea l wages in Bangladesh w ere also at a relative ly low po int in 1973-74. Reasons cited for this include the political unrest leading up to independence, destruction ca used by the war for independence, and the famine of 1974. The 1973-74 survey data revea l large per- centage increases in per cap ita consumption of potatoes ac ross income quartil es in both urban and rural areas (Table 1 ). Consump- ti on was low, however, even for high-income groups when compared w ith industriali zed count ri es. Simulated demand changes Using hi storica l demand character istics for potatoes in Bangladesh, we ran four sets of simu lations to examine the effect changes in the pr ices of part icu lar foods and income might have on potato demand. We appli ed the consumption level in 1973-74 to the food demand elasticities given in Table 2 to yie ld the simulations presented in Table 3. The f irst set of simulations assumed a 25% increase in income for each income quartile. Not unexpectedly, consumption increases are large given the low ini tia l levels of potato consumpt ion and the high income elast ic i- ties. The second set of simu lations assumed a 25% increase in the prices of all foods. That is somewhat equivalent to a dec rease in in- come (except that nonfood expend itures be- come re lative ly more attractive). Under this assumption, consumption decreases preci pi- tously. Table 2. Summary of selected income, own price, and cross-price elasticity estimates for potatoes by income quartile and urban and rural populations for Bangladesh, 1973-74. U rba n-ru rol/i ncome Income Own-price Cross-price elasticity in quartile elasticity elasticity demand for potatoes Food# l Food# 2 (Rice) (Wheat) Urban l (law income) 1.17 -1.02 -0.20 -0.20 2 1.09 -1.02 -0.19 -0.15 3 1.02 -1.02 -0.18 -0.11 4 0.87 -1.04 -0.l l -0.05 Rural l (low income) l.67 -1.01 -0.58 -0.37 2 l.44 -1.02 -0.50 -0.22 3 l.32 -1.01 -0.43 -0.12 4 1.02 -0.99 -0.27 -0.04 (IP Program Report 1995-96 2 51 Table 3. Simulated changes in demand for potatoes by income quartile and urban and rural populations for Bangladesh, 1973-74. Urban- Initial rural/income per capita 25% increase quartile consumption in per capita per year income Urban 1 (low-income) 3.6 l.06 2 5.2 1.41 3 7.3 l.86 4 8.8 l.93 Rural 1 (low-income) 1.0 0.43 2 2.6 0.93 3 3.1 1.03 4 4.7 l.19 A th ird set of simu lations assumed a 25% decrease in the price of rice, which effect ively increases income. Even thou gh rice becomes an even cheaper source of calories than po- tatoes, potato consumption increases because of the increased income. The fourth simulation assumed a 25% de- cline in the price of potatoes. The increase in per cap ita consumption of potatoes based on that assumpti on is on the order of 25%. Actual demand changes Given the rise in per capita consumption of potato and its growing importance in the lean season in Bangladesh since 1973-74, potato has evolved from a minor vegetab le to the most important vegetab le in the diet and an occas ional partial substitute for rice. That is consistent with the hi stor ica l demand parameter estimates reported in the previous section, and w ith an observed increase in the price of rice relative to potatoes. The price ratio rose 45% from 1973 to 1983 in Dhaka. The decrease in the relative cost of potatoes great ly contributed to the increase in rural 252 Program 6 Change in per capita consumption per yeor 25% increase 25% decrease in 25% decrease in price of price of primary in price of all foods staple foods potatoes (kg/year) -1.10 0.18 0.93 -1.51 0.25 1.32 -2.04 0.33 l.86 -2.30 0.25 2.22 -0.42 0.15 0.26 -0.93 0.32 0.66 -1.06 0.33 0.78 -1.36 0.31 l.16 consumption from around 5 kg per capita in 1973-74 to around 16 kg per cap ita in 1981-82. More recent consumption patterns Patterns in 1988-89. Several factors led to changes in potato consumption patterns in Bangladesh during the remainder of the 1980s. First, potato production increased markedly due to increases in yield and area planted. Yield increased an average 2.7% an- nually from 1961 to 1993, w hereas area planted increased at an annual rate of 1 .8%. By 1993, annual potato production was 1.3 million t. Second, wheat production increases were also impress ive during the period, although they have tended to leve l off in recent years. Wheat production was 1.08 million tin 199 1- 93 , onl y sl ight ly higher than the 1.03 million t produced in 1981-82. Area planted to w heat grew by about 7% in the 1980s, from 563,000 to 604,000 ha. Not only did loca l production of w heat increase, but imports in 1988-89 were higher in abso lute terms than in 1973- 74. However, population growth in Bangladesh virtually eliminated any increases in per capita avail ab ility of wheat. Third, total rice supplies increased by more than 40%, from 12.9 to 18.8 million tons. The spread of high-yielding variet ies was a prin- cipal factor. Rice import increases were mi- nor compared with increases in domestic pro- duction. Fourth, rea l income increased sharply dur- ing the 1980s (Figure 1 ). Among the possible reasons are the spread of improved rice vari- eties capab le of high yields in the dry sea- son. Fifth, resu lts from the Bangladesh house- hold expenditure survey for 1988-89 show that per capita potato consumption rose sharply (Table 4). That appears consistent with the shift in relative prices for potatoes vs. wheat-part icu larly in urban areas where relative pri ces fell by 50%. The relative price for potatoes vs. rice, however, remained virtually unchanged in 1949 = 100 urban areas and actually rose in the country- side (Table 4). As rice consumption increased, particularly among low-income consumers, this added sense of wealth engendered in- creased potato consumption as we ll. In- creased potato consumption is cons istent with the rise in real wages. Potatoes are more than just an alternative source of ca lories, or a tasty alternati ve to a strictly cerea l-based diet. In Bangladesh, po- tatoes are an important source of vitamin C and also provide essential amino acids that are low in rice. Patterns in 1995-96. Potato production in Bangladesh rose by nearly 250,000 t between 1988-89 and 1995-96 to 1.45 million t. De- tailed consumption figures by place of resi- dence and in come quartile are harder to come by. But FAO Food Balance Sheet data indicate that average per capita consumption in 1992-94 was 9.8 kg/yr, more than double the 4.8 kg/yr reported for 1961-63. The Inter- national Food Policy Research Institute and CIP estimate continued increases in potato 115 110 - - 115 -- 100 -• - Nominal wage/ 95 -- 90 85 -- 80 - 75 - 70 65 - 60 55 50 1948 1953 1958 1963 computed RCPI -x- Trend 1949-64 -¢- Trend 1965-80 -c- Trend 1981-91 1968 1973 1978 Year Figure 1. Real wages of agricultural laborers, Bangladesh, 1949-93. 1983 1988 1993 CIP Progra m Report 1995-96 253 Table 4. Consumption of rice, wheat, and potatoes, and computed relative prices per 1,000 calories in Bangladesh, 1973-74 vs. 1988-89. 1973-74 quartiles 1988-89 quartiles High Low High Low Consumption (kg/cap/year) Urban Rice Wheat Potato Rural Rice Wheat Potato Relative prices (per 1,000 calories) Urban Potato vs. wheat Potato vs. rice Rural Potato vs. wheat Potato vs. rice 125.3 54.1 8.8 162.8 26.0 4.7 8.47 3.36 4.84 2.98 production and consumpti on wel l into the next centu ry. Conclusions Potatoes a re of increasing inte rest to policymakers and planners in developing countries because of their proven potential to raise farm incomes, rural emp loyment, and food consumpt ion. This last consideration is particularly important in South Asia where large segments of the rural population are low-income consumers. To better understand past increases in po- tato consu mpti on and to more accurately es- timate the potential for further growth , basic information is required on the demand char- acteristics of the commodity. In North America and Europe, potatoes are regarded as a starchy stapl e. But in Asia they are val- ued for their dietary variety, taste, and the 254 P1ogram 6 83.2 68.6 3.6 75.4 41.1 1.0 7.39 3.39 4.63 2.90 149.8 19.7 29.7 196.3 17.0 23.7 4.18 3.16 4.14 3.28 135.3 20.1 14.4 132.3 25.1 8.5 3.93 3.19 3.84 3.18 essential vitam ins and amino ac ids they pro- v ide. Potatoes are st ill an expensi ve sou rce of calories relative to wheat and rice . But as we have shown in Bangladesh, the relative price vs. wheat has fallen substanti all y. Esti- mates for Bangladesh indi cate that the pros- pects for increased potato consumption are favorab le, yet modest in per cap ita terms, if incomes increase and potato prices continue to decline relative to staples. Government policy can greatly influence the future demand fo r potatoes. Various mea- sures might be adopted to help further lower production costs and the retail price. Among them are lower ing the cost and improving the ava il ability of planting mater ial and storage faci lities, and cutti ng subsidies on imported wheat. Given the est imated demand param- ete rs, increased potato supp li es cou ld be readily consumed in the loca l market. Selected Reading Bou is, H .E. and G.J. Scott. 1996. Demand for high-value secondary crops in deve lop- ing countries: The case of potatoes in Bangladesh and Paki stan. Food Consumption and Nutrition Division Discussion Paper No. 14. Internat ional Food Policy Research Institute (IFPRI ), Washington, D.C. 42 p. Coletti , F. 1993. Analysis of demand fo r potatoes in Bangladesh. IFPRI, Washington , D.C. 35 p. (In mimeograph.) Scott, G. 1988. M arketing Bangladesh's potatoes: Present patterns and future prospects. Intern ational Potato Center (CIP)/Australi an Development Assistance Bureau (ADAB), Dhaka, Bangladesh. 107 p. CIP P1ogrom Repo1f 1995-96 255 Improved Rustic Storage in South Asia S.G. Ilangantileke1, V.S. Khatana1, J.P.Singh2, and D. Kumar India ranks second only to China among the Asian countries in potato production. The country produced 20.3 million t of potatoes from 1.26 million ha area during 1994-95. If the trend of the 1990s continues, tota l pro- duction could reach 28 million t by the year 2000. Thus the country should be ready to handle an additional 8 million t of potatoes within 5 yr. This may be a difficul t task, par- ti cularly when postharvest facilities to hand le existing potato production are limited. Potato production is mainl y concentrated in the northern plains c:omprising the states of Uttar Pradesh , Bihar, and West Bengal. About 81% of the total potatoes produced in the country are harvested from January to April and are avail ab le for market, but onl y 28% of those potatoes are demanded as ware potatoes during the period. This creates a sur- plus of 53% of ware potatoes, w hich have to be used or stored over the next 6 months from May to November. In addition, about 10% of the total potato produce harvested in the same season needs to be stored as seed to be used for the next produ ction yea r. A major portion of the harvested potatoes are stored in co ld stores for long-term stor- age. Potato pri ces are lowest at the peak har- vest season, generall y from earl y February to ear ly March w hen the major crop of potatoes is harvested in Uttar Pradesh. The usual trend in price variati on results in about a 50% in- crease in pri ces within 2-3 mo of the peak harvest ing period . Th e ava il ab le co ld-storage capac ity in many of the potato-producing states in India falls short due to the high demand experi- enced during the harvest. The installed ca- 1 CIP, South and West As ia Reg ional Office, New Delhi , Ind ia. 2 Central Potato Research Station, Modipuram, India. 256 Progrom 6 pac ity of cold stores is 8.7 t. Estimates indi- cate that 92.5% of total co ld-store space is used to store potatoes. Assuming that the entire exist ing installed co ld-storage capacity for potato operated at full effic iency, on ly 40% of the potatoes pro- duced in the country could be co ld-stored. Because of the shortfal I of co ld-storage ca- pacity and low pri ces experienced at harvest, about 4 mi 11 ion t of harvested potatoes are still traditi onall y stored, since thi s is the only cho ice left with fa rmers . Trad itional storage practices vary by dis- tricts within a state. The end use of the har- vested potato governs the storage practice. It could vary from domestic storage in house- ho lds, mainly for family consumption, to com- mercial storage in traditional structures, heaps or clumps under trees in orchards, and un- derground pit-storage structures found in the state of Madhya Pradesh. The main force be- hind pit storage, w here potatoes are kept fo r over 2 mo under ambient conditions, is the demand for potatoes for processing. In hu- mid West Benga l, potatoes are stored in ordi- nary rooms, generall y on raised bamboo pl at- fo rms. Storage in Uttar Pradesh begins in late wi nter and the temperature increases rapidly from late February onward. The increas ing temperatures resu lt in heavy storage losses. Co nsid erin g FAQ est im ates of 17% postharvest potato losses in India, then the monetary losses fo r a season could be ap- proximately US$55 million. Farmers who are unable to ava il themselves of co ld-storage space w ill benefit from improved traditional storage methods that decrease losses and in- crease their income. Earlier studies indicated that rust ic stores w ith evaporative coo ling to lower tempera- tures and increase humidity provide a better storage atmosphere than the traditi onal stor- age systems. But research is limited and the effecti veness of the storage system on-farm has not been studi ed. We therefore undertook ru sti c-storage tri als in Uttar Pradesh to test such storage systems in fa rmers' fi elds fo r ware potato storage. We also conducted stor- age surveys to determine the status of trad i- t ional and co ld stores in potato-growing d is- tri cts . Materials and Methods Over the past three years, CIP constructed and eva luated evaporative coo led ru sti c stores (EVS) in fa rmers' fi elds. The storage experi - ments were done in co ll aboration w ith the Indi an Cou nc i I fo r Ag ri cu ltura I Research (I CAR) in the Central Potato Research Insti - tute (CPRI ). Sectional Plan All dimensions are in meters Figure 1 . Drawing of an evaporative cool store. The actual experimentation was done in co ll aborati on w ith sc ient ists from the Central Potato Resea rch Stati on (CPRS), in Meerut Di stri ct, and in fa rmers' fi elds in Meerut, Muzaffarnagar, and Farrukhabad d istri cts of Uttar Pradesh. Storage surveys were co n- ducted in co ll aboration w ith sc ienti sts from the CPRI. Eighteen improved EVS were con- structed in different fa rmers ' fi elds using lo- ca ll y avail ab le materi al such as brick, mud, wood, and thatched roofs (Figure 1 ). Cement and bricks were used to build the evapora- t ive chamber at the bottom of each store. Cement troughs were constructed w ith an ar- rangement of bri ck channels, to increase the movement of air under the store. Sand used as the evaporating surface w as placed in the troughs to a depth of 6 cm between the brick channels and was kept wet during potato stor- age. The water level w as kept at about 0.5 cm Rotary exhaust tu rbine CIP Program Reporl 1995-96 25 7 above the sand to ensure adequate moisture for evaporation. A loca lly constru cted w ind turbine was fit- ted to the roof of each store. The suction of the turbine was set to minimize des iccation of the stored potatoes. Westerly w inds pre- va il ed du ring most of the storage season, so the stores were constructed fac ing the north- south direction. The easterl y li p of the water trough was higher than the westerly lip to re- strict the hot, dry air moving through the trough to the outsi de. This enab led a longer res idence time for the hot, dry air inside the water trough, th us provid ing better evapora- tion. The potatoes were stored on bamboo mats placed on the brick channels inside the store. Walls were plastered w ith about 3 cm of mud and straw mi xture to give better insulat ion . Holes (2 cm diam) were made on the side wa lls of the eastern and western side of each store at a height of 1 .5 m to measure tem- perature in side th e sto re usin g a lo ng- stemmed thermometer. The holes were sea led with mud between temperature readings. In itial temperature and relati ve humidity (RH ) data were collected at the time of filling stores and regularly thereafter. At the CPRS store, dial thermometers were permanently insta lled w ith the probes in direct contact w ith the stored potatoes to give temperature of potatoes during storage. A Stevenson's screen was fi xed for ambient temperature and RH measurements. Continuous monitorin g of store temperature and RH at the CPRS store was done using a co ntinu o us-reco rdin g Thermohydrograph insta lled at a height of 1.6 m. Observat ions on potato quality were taken at 2-wk interva ls by opening the stores earl y in the morning w hen the di ffe rence be- tween the inside and outside temperature was a minimum. Similar observat ions were made in fa rmers' fields during regular visi ts to moni- tor the progress of the experiment. Potatoes for storage were so rted and weighed; damaged tubers were discarded. Stores we re filled in M arch. The var ieti es 258 Prog ram 6 stored were Kufri Bahar, Kufri Badshah, Kufri Chandramukhi, Jl-5857, and TPS families. Potatoes were piled to a max imum height of 1 .5 m. The effect ive storage capacity of the stores was 10 t, although in some instances the capaciti es were increased at fa rmers' re- quests and at their cost. The quantity stored ranged from 2.5 t to as much as 14.6 t. After the sto res we re loaded, their doo rs w ere sea led w ith mud to give better insul ation. Moisture or weight loss of potato in the im- proved rusti c stores as we ll as in the heap/ ord inary room was eva luated by placing 5-9 10-kg bags of potatoes in different places (bot- tom, middle, and above the stored potatoes). Potatoes were stored under these condi- tions unti I fa rmers decided to se l I, based on the prevailing market price. Observations taken on the day of se l Ii ng the stored pota- toes were (1) number of sprouted tubers, (2) number and weight of rotted tubers, and (3) the final weight of good tubers. The prevail- ing price of potato in markets in close prox- imity to the stores under study was noted at the time of both storage and sa le. Results and Discussion Cooling efficiency Daily and weekly temperature data taken from the beginning of storage in March to 1 June (the time of se lling) indi cated that there was a uniform reduction in temperatures be- tween ambient and the inside of the store. The morning (0800 h) ambient temperatures were not signifi ca ntly different from the tem- perature inside the store (Figure 2). The out- side temperatures in the morning began to rise signif icant ly from March to the end of May. The difference between the temperature from 0800 h to 1400 h averaged around 16°C. Temperatures increased from about 35°( in March to more than 40° C at the end of May for readings taken at 1400 h; inside the EVS the temperature recorded at 1400 h was 12-19°( lower than the amb ient. The turbine influenced uni fo rm upwa rd movement of evaporating moist, coo l air through the po- tato pile and reduced fluctu ations in tempera- Degrees C 45 ~····················································· ··················································································· ...................................................................... ............................. ' 40 35 30 25 20 15 10 5 -+- Inside min. - Inside max. __.._ Ambient min. -- Ambient max. 0 +-+--+-+-+-t ....... ..._........., ....... +-+-+-+-++++-+-++-+-+-+ ____ .........,t-+-+-+-+-+-+-+-+-1 0) Days 0 N Ci <( "' N (!) N m N 0) Figure 2. Comparison of minimum and maximum inside and outside air temperatures in EVS, March to April 1996. tu re inside the store. The temperatures inside the store observed at 0800 h and 1400 h were not significantly different. Therefore, ea rly morning hours are the best times to inspect the stores. The airflow provided by the turbine re- duced the desiccation of potatoes in storage, and maintained a uniform temperatu re inside the store depending on outside temperature and humidity. Reduction in losses In 1995, average total losses were 11 .3% in Muzaffarnagar and M eerut, and 20.4% in Farrukhabad. The losses were substanti ally less (10%) at the resea rch station, which may be because of better maintenance and man- agement of the store at the research farm. The minimum losses of 3.5% were observed at a fa rm in Bopada village in Muzaffa rnagar. This fa rmer was very progressive and worked on the construction of the EVS, and took a keen interest in monitoring temperature and water leve ls in the trough. During 1996, farmers of Farrukhabad suf- fe red losses of 2.8-1 0.4%, over a storage period of 24-61 d. The losses in potatoes stored for 61 d in the ordinary room were 13.9% compared with 10.4% in EVS during the same per iod (Fi gure 3). The reason fo r short-duration storage in some stores w as the sale of potatoes by farmers when the pri ces were comparat ive ly higher than at harvest. In 1996, losses of the EVS in M eerut var- ied from 6% to 11 % over a storage period of 43-53 d. The farmers who stored in M arch, when temperatures w ere still low, had only 8% losses in an 85-d storage period. Rotted tubers varied from 0.7% to 3% and sprouted tubers varied from 5.5% to 98.7%. Maximum sprouting was noticed in the case of the farmer who stored the longest (85 d). Farmers in the study area without storage fac iliti es heaped their produce. The losses in heaps were 4. 3% and 7.5% in storage peri- ods of 33 and 35 d, respectively. The 7.5% losses in heaps during 35 d was higher than a 6.9% loss in EVS during 47-d storage. Farm- ers w ho stored their potatoes in heaps were paid lower prices because of the low qua I ity. There was a signifi cant reduct ion in tem- perature and increase in humidi ty inside the EVS during the hot, arid months from April to June, thu s maintaining a higher quality of stored potatoes. Weight losses in potato were CIP P1ogra m Repoit l 995-96 2 5 9 Losses(%) 14 12 - 10 8 ~ - - 6 """ 4 2 0 2 3 4 - w M r 5 6 Farms """ 7 8 - 9 10 0 Loss/EVS I Loss/heap Figure 3. Comparison of storage losses for potatoes stored in EVS and those stored in heaps after 61 d, 1996. influ enced by variety, preharves t and postharvest management practi ces, store con- struct ion, and store management. This study was the first major attempt to take the technology of rustic storage to farm- ers ' fields. The results indicate that potatoes stored in EVS had lower weight loss and higher quality at sa le time. An added advan- tage is that the EVS provided a convenient means of keep ing potatoes in bad weather, whereas those stored in heaps were exposed to rain and desiccation . Losses in the Farrukhabad stores were hi gher than those in Meerut and Muzaffarnagar. There was no climatic differ- ence in these districts and the losses may be attributed mainly to leakages and poor man- agement of water in the troughs. Although stores were located in close proximity to tube wel ls, the nonavailability of an assured source of power and the lethargy of farmers to keep the sand surface moist resulted in lower cool- ing efficiencies. Lower losses in 1996 at all sites compared w ith 1995 strengthens the line of thinking that store management plays a key role in potato storage. Farmers were more knowledgeab le about potato storage during the second year of the trial and thu s reduced their storage losses. 260 Progra m 6 Profitability Profitability of an improved techno logy is the main factor that influences adoption. The EVS techno logy was more profitable than the farmers' practice of storing potatoes in houses and in heaps. In Farrukhabad, the price of potato was Rs 2.1 / kg ($0.06/kg) when stores were fi I led. It had increased to approximately Rs 3.75/kg ($0.11 /kg) after a period of 24-61d. The gross retu rn was 11 % higher per kg of potato stored in EVS than in ord inary rooms and in heaps. In Meerut, overa ll weight loss in potatoes stored in EVS varied from 6% to 11 % over a storage period of 43-85 d. The gross returns in heaps ranged from Rs 189/t ($5.30/t) to Rs 494/t ($14/t) compared with Rs 269/t ($ 7 .62/ t) to Rs 1,030/t ($29/t) for potatoes in the EVS. Th e highest gross return was earned by a farmer who stored for the longest period (85 d) in the EVS, because he received the high- est sale price in June. But when net profits are considered, heaping of potatoes is more profitab le because of the high cost of con- structing EVS. Storage in heaps, however, has a risk of high rottage due to rain. Neither can potatoes be stored as long in heaps as in EVS. More than 20% of the potatoes grown in the Uttar Pradesh plai ns are used for seed and stored in co ld stores after harvest. Uttar Pradesh is traditi onall y a seed supplier to the early months of storage, fo r those who do not rest of the potato-growing areas and there- want to be at th e mercy of co ld-store owners, fo re depends heavi ly on co ld stores. This has and for those unab le to ava il themselves of resulted in a concentration of co ld stores in co ld-store space at harvest. If the storage cost Utta r Pradesh. Any new techno logy mu st of the EVS cou ld be sign ificant ly red uced to have signi f icant advantages over co ld stores, below the cost of co ld storage, then EVS could traditi onal heaps, and other structures. The be just ified fo r short-term storage. If not, Uttar cost of EVS is stil I higher than that of the heaps Pradesh fa rmers w ill cont inue to use co ld or other structures, and not significantly lower stores, which guarantee better produce qua I- than the cost of co ld storage per kg of potato. ity even though their ava il ab ili ty fl uctuates. Table 1 compares costs and return s for the The EVS has a defini te advantage for fa rm- EVS, heaps, and pits w ith the cost of co ld stor- ers wanting to dispose of the ir potatoes in the age. Table 1. Comparison of storage cost of EVS, heaps, and pits with the cost of cold storage'. Value EVS Rs USS Value Heap Rs USS Value Pits USS Fixed cost Cost of constructing store storage: 20,100 574 l,000 29 3,000 86 Sa lvage cost 804 23 40 l 150 4 Life span lO 10 Depreciation l,930 55 240 285 8 Interest on investment(%) 14 2,814 140 4 14 420 12 Total fixed cost 4,744 136 380 ll 705 20 Variable costs of storage Labor for loading and unloading and water use/yr 360 lO 500 14 600 17 Repairs/yr 480 14 100 3 200 6 Weight of potato in storage (kg) 10,000 10,000 10,000 Price of potato at storage/kg 2.50 2.50 2.50 Cost of pota to in storage 25,000 714 25,000 714 25,000 714 Total variable cost (including potatoes) 25,840 738 25,600 731 25,800 737 Total cost (incl udi ng potatoes) 30,584 874 25, 980 742 26,505 757 Total cost (without potato)/kg/yr 0.56 0.016 0.1 0 0.003 0.15 0.004 Cold· storage costs/kg 0.60 0.02 0.60 0.02 0.60 0.02 Returns from storage Losses in storage(%) 6.9 7.5 7.5 Pr ice at time of so le (Rs/kg ) 4.47 0.13 4.13 0.12 5.00 0.14 Risk loss in kg 75 Weight at sole (kg) 9,310 9,175 9,250 Gross returns from sole 41,615 l,189 37,893 l,083 46,250 l,321 Net profit 11 ,032 315 ll ,913 340 19,745 564 Net profit/ton l.10 0.03 l.l 9 0.03 l.97 0.03 a. This calcu lation is mode fo r traditional heaps in the fie ld for a similar duration as the EVS of 75 doys for the same former. CIP Program Report 1995-96 26 1 Adoption domains The cooling efficiency of the EVS during the hot, arid months from April to June gave a higher-quality potato to the market. But farmers have yet to understand the need for overall management of both produce and stores during the storage period, if the tech- nology is to be successfu I. The Uttar Pradesh plains are semiarid. Temperature rises to about 44 °( in summer and drops to about 3-4 °( during winter. The states of Madhya Pradesh and Gujarat have temperatures similar to Uttar Pradesh, but the humidity is even lower. The scope for EVS adoption is higher in these three states than in all the other potato-growing areas in In- dia. Increasing interest and demand for pro- cessed potato, mainly for chips, is creating a significantly increased demand for indig- enously stored potatoes . Large processing companies pay premium prices for good- quality potatoes that have not been in cold storage in Madhya Pradesh. These potatoes have lower sugar contents and optimum pro- cessing quality. In Madhya Pradesh, Gujarat, and Karn ataka, cold-store space fal Is much short of the optimum level of 55% of the total pro- duce (Table 2). Cold-store charges in these states are much higher than in Uttar Pradesh . Farmers in these states may find the EVS tech- nology quite profitable for ware and for pro- cessing. These states cater to the needs of potato processors by storing potatoes in in- digenous structures, built with high invest- ments. The investment in EVS is significantly lower than the investment in traditional stores, which do not have the cooling and lower weight loss advantage of the EVS. Conclusions Rapidly increasing potato production in In- dia is placing a strong demand on the cur- rently available, limited postharvest facilities, especially those to store ware potatoes. By the year 2000, the country needs to be ready to handle an additional 8 million t of pota- toes, although cold-store space is limited. Store losses in the different districts varied from 4% to 11 % . Loss depended mainly on the initial condition of potatoes going into storage and the duration of storage. Other factors such as store management and water control in the EVS troughs also contributed to losses. Losses in heaps were lower than expected, but still higher than losses in EVS for a similar duration. Table 2. Potato production and cold-storage capacity in major potato-producing states of India, 1992-93. State Production Cold-storage Cold-store Cold-storage (000 t) capacity space being capacity as {000 t) used for other percentage of purposes(%) potato production Uttar Pradesh 5907.6 3585.3 0 60.7 Gujarat 444.3 203.9 30 32.l Madhya Pradesh 367.5 158.3 20 34.5 Kornotoko 452 .4 19. l 55 1.9 Maharashtra 64.5 129.l 73 54.0 Source: Compiled from the information given by National Horticultural Boord, Jan. 1995, 262 Program 6 Profits (based on prevai I ing pri ces) of fa rm- ers were significantl y higher fo r those who stored their produce effi ciently fo r 2 mo or more. Postharvest management of potatoes be- fo re storage and the management of stores after fi II ing affect EVS effic iency. Proper man- agement of the total storage system provides a better quality of stored produce. CIP Progrom Report 1995·96 263 Recent Advances in CIP's Strategy for Collaborative Postharvest Research on Sweetpotato G.J. Scott and C. Wheatley1 Postharvest issues are parti cular ly relevant to the deve lopment of root and tuber crops . Unlike the cerea ls, these crops are peri shable. And given their seasonal production patterns, their effect ive use, either as pl anting material or for consumption, depends on proper pres- ervati on through storage or process ing. In fresh fo rm, root and tuber crops are bulky and low va lue in relati on to transport costs. They are often produced in rural areas at consider- ab le di stances from urban co nsumption centers-a problem particularly acute in sub- Saharan Afri ca. Greater ava il ability of sweetpotato in the market and at reduced cost to consumers is contingent upon improved process ing or mar- keting or both . Roots and tubers often conta in either toxic (g lycoa lkalo ids) or unattracti ve (fl esh co lor) traits that can be eliminated in process ing. Postharvest act iv iti es enabl e roots and tubers to achieve their full potential as sources of food in other th an fresh fo rm (fl our, meal , starch), as raw materi als fo r secondary pro- cessed products (modi fied starch, noodles, glucose, candy), or in the fo rm of feed (fo- li age, roots, by-products) . In the case of roots and tubers, th e combinati on of traditional knowledge and modern sc ience offers par- ticu larl y good promise fo r impact. Thi s paper presents a synthes is of the strat- egy and recent advances from CI P's co ll abo- rati ve pos th arves t resea rch prog ram for sweetpotato . Th e secti on on materi als and methods summarizes the obj ect ives and gives an overview of the four phases (Table 1) of 1 CIP, Lima, Peru , and CIP, Bogor, Indones ia, respect ivel y. 264 P1og1om 6 CIP's postharvest research strategy. The pa- per then examines the resul ts of the first two phases. Emphasis is on phase 2 acti v ities dur- ing 1995-96. Materials and Methods Postharvest research at CIP has as its overa ll objecti ves those of the Center itse lf, namely, to increase product ion and use of sweetpo- tato. Postharvest research concentrates on va- ri eta l tra its, raw materi al characteri stics, pro- cess ing techniques, and soc ioeconomic is- sues. The goal is to make processed products made from sweetpotato more affordable and nutrit ional and their producti on more profit- abl e. As such, posth arvest research aims not onl y to increase smal I farmers' and proces- sors' incomes from ex isting production, bu t to provide the incentive fo r the adopti on of y ield-increas ing, cost-reducing techno logies provided by other areas of research at CIP. As the pl anning of postharvest research on sweetpotato at CIP evolved in the 1990s, a globa l strategy w ith guiding principles gradu - all y emerged (Tabl e 1 ). The strategy has four ph ases, each w ith a spec if ied but fl ex ible t imeframe, set of activiti es, and expected out- puts. These phases are: • Ph ase 1. Identi f ication of prob lems and opportuniti es (1989-92). • Phase 2. Eva luati on of ex isting technologies and ex ante assessment (1993-96). • Phase 3. Co ll aborati ve technology deve lopment, market resea rch, and training (1997-2000). • Ph ase 4. Enterpri se development and ex post assessment (2000-2003) . Table 1. Elements of CIP collaborative sweetpotato research strategy. Objectives • Increase and diversify crop use • Increase incomes/welfare • Increase nutrition levels Means • Make better products at more affordable prices • Make processing more efficient and profitable • Make sweetpotato more nutritious Target beneficiaries • Small formers (producers/consumers) • Smal l-scale processors • Undernourished consumers Phases • Phase 1. Identification of problems and opportunities (1989-92) • Phase 2. Eva luation of existing technologies and ex ante assessment (1993-96) • Phase 3. Collaborative technology development, market research, and tra in ing (1997-2000) • Phase 4. Enterprise development and ex post assessment (2000-2 003) Guiding principles • Build on rural small-enterprise innovations • Selective and prioritized locations and issues • Participatory diagnostic and action research • Fl exibili ty Th e pha ses in CIP 's strategy for sweetpotato postharvest research close ly re- flect at a globa l, or macro, level, the steps invo lved in posth arvest technology and mar- ket ing fo r root and tuber crops at the loca l, or micro, level. Microlevel steps are: (1) prob- lem and opportun ity identification, (2) re- sea rch on products and processes, (3) pilot project, and (4) commerc ial-sca le expansion . Phase 3 involves interdiscip linary and insti - tutiona l co ll aboration, and direct invo lvement of users' groups (processors, consumers, and animal produ cers) as active research partners. To ensu re w idespread effect, phase 4 brings the postharvest research resu Its together with other aspects of enterprise development. In the fo llowing sect ion we describe the out- comes in ph ase 1 to provide the context for the more deta iled revi ew of recent results in phase 2. Results and Discussion Phase 1 Phase 1 ended with the conso lidat ion of a globa l agenda for sweetpotato posth arvest research that had been deve loped in co ll abo- ration with developing- and developed-cou n- try partners. Th at agenda was now to focus on 7 of some 100 sweetpotato-produc ing cou ntries w ith the greatest process ing poten- ti al, at parti cular sites w ith in those countries, and on three types of priority products (Table 2) . The products were starch, fl ou r, and feed (i.e., roots and v ines for pigs in Asia and vines for cattl e in Afri ca and Latin America). CIP Progrom Report 1995-96 265 Phase 2 Phase 2 of CIP's sweetpotato postharvest st rategy calls for a dichotomous approach . Macrolevel or regional stud ies look at th e more long-term outlook for sweetpotato pro- cess ing by examining ex ist ing or potential practices in light of production and use trends, prices, and trade patterns. This ex ante re- search seeks not only to assess the econom- ics of exist ing processing but also to set tar- gets for competitiveness and eva lu ate the prospects for achieving them. In keeping with the guid ing principle of buildi ng on local rural momentum (Table 1 ), phase 2 also calls for eva luating ex is t in g technologies to complement the mac rostud ies . Key factors here are drying and extraction rates, costs, and identification of the postharvest attributes of potential processing cu lti vars . Asia China. China produces about 85% of the developing world's sweetpotatoes. Since the early 1960s, the percentage of production used for feed or processed products has risen steadily to over 50% of annual output. Rela- tively little information was ava il ab le for pro- jecting future sweetpotato uses. A more de- tailed assessment of the potential for sweetpotato process ing in China invo lved four research activities: (1) a macroreview of Chinese secondary data and associated lit- eratu re on production and use trends, (2) case studies in a se lect number of provinces to analyze county data on production and use, (3) a micro level analysis of sweetpotato feed use in Sichuan Province, and (4) the eva lua- tion and improvement of small-scale process- ing of sweetpotato roots into starch. The macroreview and provincial case stud- ies detected a recent reversa l in the down- ward trend in area planted to sweetpotato in south, southwest, and central China (F igu re 1 ). Sichuan Province in the southwest, w hich had become the largest pig-produ cing prov- ince in China, was suffering from a chro ni c shortage of feed maize. Farmers substituted sweetpotato for maize in their hog rations as a resu It. Small-scale processing of roots into starch for noodl es also had expanded tremendously, espec iall y in Shandong and Sichuan Prov- inces. In Anyue County, Sichuan, for example, sweetpotato noodle production increased from 3,000 to 26,000 t between 1990 and 1996. Anyue nood les are marketed from the fa r west of China to coastal cities. Urbaniza- tion and income increases in China appear y = 1.593E10 - 1.605E7x + 440.9542 1650000 .....------------------------..--~ 1600000 • 1550000 • 1500000 • Southwest 1450000 1400000 • • 1350000 • 1300000 -+----~--~---~--~---~--~----! 1980 1982 1984 1986 1988 1990 1992 1994 Year Figure 1. Sweetpotato sown area {ha) trends in southwest China, 1982-93. 266 Progrom6 to be dri ving demand for better-quality starch noodles, thus offering opportuniti es for small - sca le processors to produce goods w ith high added va lue. Collaborative research w ith the Sichuan Academy of Agri cultural Sciences (SAAS) focuses on quality improvement for both sta rch and noodles, and invol ves links w ith food sc ientists at Hong Kong University and processing enterpri ses themselves. Feed and processing together account for wel I over 50% of use in most of the prov- inces studied and are clearly on the increase. Process ing includes dried chips for feed, as well as sta rch. Research in China ca rried out by th e Australian Center for Intern at ional Agricultural Research (ACIAR) estim ated that the expected benefits from sweetpotato were higher than for any other crops except ri ce and wheat. Ninety-fi ve percent of the pigs produced in China are raised on small fa rm s. The initi al Center for Integrated Agri cultural Deve lopment (C IAD)-SAAS fa rm-leve l ana ly- sis of pig production in Sichuan Province showed that the traditional sweetpotato feed system used by smal I farmers is ineffi cient due to nutri ent-poor rations, which lead to low pig growth rates and prolonged production times . Adding appropri ate feed supplements to sweetpotato offers the potential to increase efficiency and profitability on a small sca le. The compos ite pi cture of sweetpotato pro- cessing emerging from these studi es is one of a great dea l of local innovation with tremen- dous potential for growth to the benefit of the poorest households in rural China. Priority research topics include effo rts to improve technical effic iency in feed operati ons, and commercialization studi es to target the most dynamic markets for starch and related prod- ucts. Philippines. Ex ante impact assessment in the Philippines focused on the potential for sweetpotato use as animal feed and fl our. The study notes th at the limited demand for fresh roots offers littl e incent ive for sweetpotato producers to increase production. Instead, the report identified a site in Cen- tral Lu zon as the most promising site fo r sweetpotato process ing enterpri ses. Success will depend on government poli cies to (1) fa- cilitate the adoption of high-yielding va rie- ti es w ith high dry matter content, and (2) im- prove process ing efficiency. Since the study was completed, the private sector has estab- 1 ished medium-sca le sweetpotato sta rch pl ants in Central Luzon, targeting high-va lue export markets for nood le production in Ko- rea. The performance of these enterpri ses, especially in relati on to the supply of fresh roots, w ill be relevant for assessing the feasi- bility of producing starch or other products for domestic markets. Latin America Rece nt res ea rch on sweetpotato postharvest issues in Peru resulted from the collapse in pri ces paid producers in earl y 1990 and th e ever- increas ing imports of wheat flour. The private sector responded by parti ally substituting raw, grated sweetpotato for wheat flour in a new type of sweetpotato bread. Subsequently, exploratory research projects al I provided positive indicators of the potenti al for sweetpotato bread and oth er products. Beginning in 1995, CIP coordinated a Dutch Ministry of Foreign Affairs (DGIS)- funded project on sweetpotato product de- ve lopment for flour, starch, and foliage. Results from the work on sweetpotato as a wheat fl ou r substitute have c~en mixed. The raw, grated method turned out to be hard to sustain on other than a novelty basis because of seasonal and year-to-year price fluctua- tions, and the logisti cs of securing and grat- ing regular supplies of fresh roots. Trade lib- erali zat ion and the lifti ng of subsidies drove the price of wheat fl our down, not up as ex- pected. Th at made the substitution of sweetpotato flour even more problemat ic. On the other hand, CIP-supported research found the use of sweetpotato in fl our-I ike form in a ch ildren' s weaning food highly attrac- ti ve. This product is intended for mass distri- buti on in the government-sponsored ca m- paign to help eradicate malnutriti on in the most vul nerabl e segments of Peru's popula- tion . The weaning food formula has attracted interest by large-sca le food processors. The CIP Program Report 1995-96 2 6 7 cost of sweetpotato fl our has resurfaced as a key issue, especially as the price of imported wheat flour has experienced some sharp price hikes in recent months. The government- sponsored weaning food program also calls for some 70-80% of the ingredients to be pro- duced domestically. Sub-Saharan Africa Sweetpotato postharvest research in east- ern Afri ca focused initially on poor, female farmers and women's groups engaged in sweetpotato processing in the dry, western part of Kenya. Earli er consumer su rveys and product testing had shown there was an in- terest in sweetpotato as a flour substitute. A CIP multidisciplinary team and sc ientists from the Kenyan Agricultural Resea rch Institute (KARI) at the outset faced the challenge of doing postharvest research with a crop that people were remarkably unfamili ar with when it came to uses other than simply boil- ing it for direct human consumption. The col- laboration had two principal thrusts. One involved using sweetpotato to make a variety of loca lly popular baked goods (chapatis, mandaz is). Different ways were tried, but all were within the capabilities of the intended, limited-resource beneficiaries. These new uses for sweetpotato and the as- sociated simple processing techniques proved quite attracti ve . Some procedures were tried for a while and dropped; others were quickly adopted as in co me-generating activities, sometimes in modifi ed form. Another thrust was to improve nutrition leve ls through a package of rel ated initi ati ves. Amon g th em was encourag in g women sweetpotato fa rmers to plant varieties shown to do well under loca l growing conditions and to contain high amounts of beta carotene. This would help overcome vitamin A deficiency, which is high in western Kenya. Proposed work includes (1) improving qua I ity control, (2) incorporating the high beta carotene-containing va rieties into processed products, (3) increasing the widespread di f- fusion of improved varieties and simple pro- 268 Program 6 cess ing practices, and (4) conducting an ex post appraisal of the effect of these initiati ves. In add ition , the technical and economic fea- sibility of using sweetpotato as an ingredient in composite flours needs to be assessed if growers in western Kenya are to penetrate rap idl y growing urban markets in that part of the country and elsewhere. Conclusions Recent results from phase 2 encompass the eva luation of existing technologies and as- sessment of the potential for future commer- c ial expansion of processing acti v ities. This work centered primarily on the seven target countries and the three priority products iden- ti f ied in ph ase 1 (Table 2) . We achieved the following results: • Documented the importance and potential of postharvest activities. • Generated initi al impact through the ready transfer of ex isting ski I ls and technology. • Developed in stitutional l inkages necessary for add ressing the tasks in phase 3. • Sharpened the geographi c and product-specific focus of future research on technology development. • Empowered national scientists to achieve their research objectives more quickl y through training and workshops. Based on these results, CIP's collaborative postharvest research now enters phase 3. Greater priority w ill be given to work on sta rch and roots for feed in Asia, and flour and fodder in Latin America and Africa. Selected Reading Cabanilla, L.S. 1996. Sweetpotato in the Philippines: Production , processing, and future prospects. International Potato Center (CIP) and the University of the Philippines at Los Banos, Lima, Peru. 100 p. Gitomer, C.S. 1996. Potato and sweetpotato in China: Systems, constraints, and potential. International Potato Center (CIP), Lima, Peru. 183 p. Table 2. Sweetpotato research sites and products. Target Site Ecology country Chino Sichuan Subtropica l Indonesia Mo long Tropical post-rice Kenya Houma Boy Semiarid Peru Central Coast Desert lowland Condebombo Tropical highland Philippines Central Luzon Tro pical lowland Uganda Lira Semiarid Vietnam Than h Hoo Subtropica l Wheatley, C. 1996. Sma ll -sca le enterprises: What might the future hold fo r rural househo lds and root crops? In : Into act ion resea rch: Partne rships in Asi an root crop Priority products Animal feed Starch Flour Roots Foliage + + + + + + + + + + + + + research and development. UPWARD, Los Banos, Laguna, Ph ili ppines. 1996. p.171-190. CIP P1og1om Repo1f 1995-96 2 6 9 Enhancing the Role of Small-Scale Sweetpotato Starch Enterprises in Sichuan, China C. Wheatley1, Lin Liping2, and Song Bofu3 Sweetpotato, lpomoea batatas (L.) Lam., is a majo r component of sma ll-scale Chinese farming systems, which produce more than 100 million t of fresh roots annuall y. Since the 1960s, use of the crop has shifted from direct human co nsumption toward animal feed and indu strial uses. Shandong and Sichuan provinces, w ith annual sweetpotato root production of over 20 milli on t each, have developed important starch industries based on small-sca le extraction of starch and its use in transparent nood les, wh ich also are produced large ly on a small to medium sca le. In Shandong, some 30-40% of sweetpotato production is used for starch extraction; in Sichuan, about 1 5%. In Sichuan Prov in ce, w here CIP has worked w ith the Sichuan Academy of Agri- cultural Sciences (SAAS) for over 7 years, sig- ni ficant progress in upgrading the technology leve l of small-sca le enterpri ses has been achieved. Manual processing of the roots has been largely rep laced by small- and med ium- sca le mechanized equipment (root washers, grinders/ raspers, horizontal and drum sepa- rators, and inclined channel sed imentation). The introduction of small-scale, single-screw extruders has allowed increases in produc- ti vity and output of sweetpotato sta rch nood les. Several items of equipment also used for potato starch processing are now manu- factured commerc iall y and distributed be- yond Sichuan Province . Sichuan is an inland province, distant from the booming coastal regions of China. The 1 CIP, Bogor, Indones ia. 2 Crops Research Institute, Sichuan Academy of Agricultural Sciences (SAAS), Sichuan, China. 3 Ch inese Academy of Agricultura l Sc iences (CAAS) and CIP. 270 Program 6 province is densely populated (110 mil li on people). Incomes are be low the national av- erage: rural per cap ita annual income in Sichuan was US$85.00 in 1993 (75% of na- tional average rural income, and only 29% of nat ional average urban income). Conse- quentl y, Sichuan is now a major source of econom ic migrants. Increased use of sweetpotatoes in Sichuan, w hether for feed or starch, offers rea l poten- tia l for rural income generat ion in those ar- eas w here sweetpotatoes are grown, and w here the need fo r poverty all ev iati on is acute. SAAS and CIP research has, since 1994, focused on eva luat ing the status of sweetpotato starch enterpri ses and identi fy- ing opportunities for their further develop- ment. Thi s has entailed integrating market and technical resea rch . Materials and Methods Starch enterprise appraisals Three starch enterpri ses in Santai County were se lected as case study units on the ba- sis of ownership type (pr ivate or co llecti ve), sca le of operation (0.2 -0.8 t fresh roots per h), w hether manual or mechani zed, and the starch sedi mentation method used (natural sed imentat ion in tanks, sour liquid , or in- clined channels). Three batches of fresh roots were fo ll owed through the normal process operat ions for each enterpri se. During November 1994, a part1c1patory appraisa l of the smal I-scale starch extraction enterprises was carried out. Data on the tech- ni ca l operat ion of the process, through direct measurement and observation of normal op- erat ions, were complemented by relevant in- formation on the food system in which each enterpri se operated. Fresh roots of Xushu 18 from one field plot were used as raw materi al fo r all appraisa ls. To determine the effects of differing raw ma- terial quality on technical performance and process economics, at Guanqiao enterpri se fresh roots of Mi anfen 1 were also processed. Mianfen 1 has a high fresh-root starch con- tent, but lower fresh-root yield , than Xushu 18. Technica l performance of the enterpri se was eva luated on a batch basis (per kg fresh roots processed), and then reca lculated on a product basis (per ton dri ed starch produced). Most (60-80%) of the starch was so ld or used immed iate ly as wet starch cakes (moi sture content 45%) fo r noodle production. For the purposes of analysis and comparison, how- ever, resu Its are presented as dry starch equiva lents. On a dry matter basis, prices for wet and dry starch are nearly th e same. Drying is worthwhil e only if the starch is to be mar- keted or stored. Economic performance was determined on a per metri c ton dried sta rch basi s plus addi- tional revenues from the sale of moist resi- dues. In many cases the res idues are used direct ly for pig feed by the same household, but for the purposes of enterprise analysis the 1994 market price ($2.40/ton wet res idue) has been used. Market research A 1995 market survey, conducted jo intly by the Center for Integrated Agri cu ltural De- ve lopment (CIA D) of Bei jing Agri cultural University and SAAS assessed the potential demand for sweetpotato starch in both food and nonfood industr ies. We used secondary information to develop an inventory of Si chuan enterpri ses that use starch as a raw mater ial for the manufacture of food and other products, for se lecting a representative sample of enterprises for inter- views, and for estimating starch demand in Sichuan. The fi eld study included sweetpotato starch and wholesale markets, sample starch enterpri ses, and key informant interviews with sc ientists and industry leaders in Sichuan. Process research for quality improvement During 1996, SAAS process research sc i- entists stud ied two ways in which small-sca le enterpri ses could improve the quality of the starch they produce through (1) process mod i- fi cations appropriate to their sca le, and (2) use of different sweetpotato vari et ies. Using Nanshu 88 in an orthogonal experi - mental design, the effects of varying the ratio of water to we ight of fresh roots (3: 1, 6:1 , and 8: 1 ), mesh size during separat ion of sta rch suspension from fibrous res idue (80, 100, and 120 mesh), and sedimentation time (8, 16, and 24 h) on sta rch yield and quality were studied. Starch yie ld was calcu lated at stan- dard 14% moisture content. Analyses of sta rch (Layn e and Eynon method), lipid (Soxhelet extract ion), ash (in- c inerat ion at 525°C for 1 h), and protein (macro-Kjeldahl ) of the dried starch samples were carried out by SAAS Central An alytical Laborator ies. Whiteness was determined by fluorescence analys is, and by a sensory panel. Similar analyses were carried out in a sec- ond exper iment to determine the vari ation in starch purity across 10 sweetpotato var ieti es or promising sel ections. Results and Discussion Starch enterprise appraisals The appraisa l data showed that all enter- prises were profitable w ith high returns on investment, despite low use of installed ca- pacity (Table 1 ), caused by problems in raw materi al suppl y. In 1994, drought lowered sweetpotato production in Si chuan, and re- duced the length of the processing season. The dry sta rch yield (14% moisture con- tent) ranged from 10% to 18%, depending on both process and va riety (Table 2). Improve- ments to process efficiency, espec ially at the separation stage, wou Id increase profitabi I- CIP Progrom Report 1995-96 2 71 Table 1. Principal performance and profitability data from the three sweetpotato starch extraction enterprises from Santai County, Sichuan, China , 1994. Enterprise Changping Gaoyan Guanqiao (household, manua l) (collective) (family) Sweetpototo variety processed Xushu 18 Xushu 18 Xushu 18 Mianfen 1 Yield of dried products (14% moisture content) as % initial fresh-root weight Dry starch 10.0 13.4 13.l 18.1 Dry residue 12.0 8.4 12.4 14.4 Production costs (US$/ton dry starch) Variable 389 256 264 232 Fixed 22 37 15 15 Total 411 293 279 247 Variable costs (%) due to: Fresh roots 85.8 97.4 96.2 96.2 Labor 12.5 1.2 2.4 2.4 Profit Net profit/ton dry starch (USS) 48 100 116 146 Total profit 1994 (USS) 315 662 l,525 1,752 Profit as% soles income 10.5 25.5 32.1 37. l Return on investment 71. l 23.9 140.0 160.8 Table 2. Starch quality (purity) specifications of Sichuan industries compared to the chem ical composition of sweetpotato starch produced by small-scale enterprises in Santai County, Sichuan, China , 1994. Composition Sichuan industrial specifications for starch purity• Sweetpotato starchh Super grade First grade Second grade Moisture content(%) 14 14 14 15-18 Ash(%) 0.10 0.20 0.20 0.38-1.07 Protein(%) 0.35 0.50 0.80 0.21-0.57 lipids(%) 0.10 0.15 0.30 0.02-0.10 a. Source: Survey of industrial starch users in Sichuan, Center for Integrated Agricultural Development (CIAD} and Sichuan Academy of Agricu ltura l Sciences (SAAS}, 1995. b. SAAS analyses of starch samples from enterprise appraisals. 2 72 P1ogrom 6 ity si nce more starch would be produced per unit of roots processed. At Guanqiao enterprise, where two vari- eties were compared, Mianfen 1 y ielded more dry starch and greater profits, allowing the enterpri se to pay farmers a higher unit price for Mianfen 1 roots. The low labor costs, as a proportion of total costs, in the two mecha- nized enterpri ses are encouraging. Any future increases in labor costs should have a small effect on overall enterpri se profitability. Use of higher starch var ieties (such as Mianfen 1) by farmers should be encou raged by the pro- cess ing industry, and adequate price incen- ti ves should be given, as justified by process economics . Demand fo r starch process ing equipment is high. O ne of four starch processing equip- ment manufacturers in Santai Co unty, Sichuan, reported that from 1992 to 1996 sales of sweetpotato process ing equipment more than doubled (240%) for starch extrud- ers, and near ly tripled (288%) for root wash- ers. The sa les va lue of root washers, starch separators, and extrud ers totaled over US$180,000 in 1996, 85% of total agri cul - tural equipment sales. Between 1992 and 1996, sweetpotato processing equipment sales were valued at $630,000. Although neither sweetpotato production area nor vo lume changed appreciably be- tween 1989 and 1995 (production in Santai County va ri ed between 82,000 and 93,000 t/yr), the proportion of sweetpotatoes th at were processed increased from 36% to 76%. Of 91,000 t produced in 1995, 69,000 t were processed. That was accompanied by an in- crease of 70% in the number of pigs produced in the area during the same period (110,000 in 1995). Pi g production is closely linked to starch processing through use of residues for feed. Market research The results of the market survey were sur- prising. Companies comp lained of severe shortages of maize starch (the major starch used) because of (1) insuffici ent maize pro- duction in Sichuan itse lf, w ith demand for maize starch exceed ing loca l supply by over 100,000 t in 1995, (2) com pet ition from the feed industry for raw materials, and (3) diffi- culties in importing starch or maize from other provinces. Thi s last problem was due to con- gested transport infrastructure and the un- avail ability of maize elsewhere in China be- cause the ri cher coastal provinces bought up ava i I ab le supplies for themse lves . This situation represents an opportunity for sweetpotato sta rch to fill a supp ly gap. Sweetpotato starch in Sichuan has consis- tent ly been priced below maize starch (F ig- ure 1 ). But sweetpotato sta rch extract ion on a smal l sca le is more profitable than maize extract ion on a large sca le. At 1994 prices, the net profit on maize sta rch extract ion was $35 / to n compa red w ith $60/ ton fo r sweetpotato, even after including the income from maize sta rch by-products. Sweetpotato starch product qua I ity, how- ever, was rated poorly by industry. The starch contains impuriti es such as excess ive mois- ture and ash (Table 1 ), and is not sufficiently whi te. Impuri t ies can be reduced through pro- cess improvements. Shandong Province pro- duces refined sweetpotato starch using equ ip- ment si milar to that in Sichuan by add itional purification and drying steps. Research for quality improvement of sweetpotato starch Mesh size and volume of water used were fou nd to significantly (P=0.0001) affect starch extract ion rate and ash content of the ex- tracted starch. A l I treatments produced starch that met maize starch standards for lipid and protein impuri t ies. Starch w hiteness increased with increased volume of water used and finer mesh sizes, bu t decreased w ith increased sed imentation t ime. The ash content of even the best starch sampl es, however, exceeded acceptab le limits (lowes t sample va lu e: 0.38%) (Table 2). For both sta rch yield and product quality, the optimum processing conditions were found to be 1 :6 ratio of root weight to pro- CIP Prog rom Report 1995-96 273 100 OWholesale 90 •Retail 80 70 60 50 40 30 20 10 0 1990 1991 1992 1993 1994 1995 Figure 1. Wholesale and retail price of sweetpotato starch as a percentage of maize starch price in Sichuan Province, China, 1990-95. cess water, 120 mesh (f inest size), and 8 h precipitat ion. More efficient root washing to ensure more effi c ient so il remova l before pro- cess in g coul d redu ce ash co ntent of th e starch. Starch was also extracted, unde r standard conditi ons, from 1 0 di ffe re nt sweetpotato va rieti es/promi sing se lections, from the SAAS Breeding Section. Significant differences were fo und betwee n th e 10 va ri eti es tested fo r starch content percentage, and lipid and ash content of the extracted sta rch (Tab le 3). This suggests a potenti al to improve starch purity through a va rietal approach if the fresh-root and starch yie lds remain attract ive to fa rm- ers. Table 3. Effect of variety on purity of extracted starch (14% moisture content) . The varietal effect was significant for all three parameters (starch, lipid, and ash contents) . Letters denote mean separation at a =0.05. Sweetpotato Starch content Lipid content Ash content variety/clone {%) (%) (%) Chuonshu 27 81.3 a 0.11 e 0.653 b 9014-3 74.9 e 0.32 abc 0.646 a Chuon 778 73.7 f 0.13 de 0.594 c 53-5 78.9 b 0.13 de 0.492 f Sheng nan 78.9 b 0.43 ab 0.482 d 303 7 6.6 cd 0.45 a 0.460 a 9102-1 79.5 b 0.30 be 0.408 g 92-113-70 77.0 c 0.27 cd 0.392 h Nonshu 88 75.7 de 0.26 cd 0.385 g 89-1524 82 .1 a 0.43 ab 0.358 e 274 P1ogrom 6 Conclusions Any diversification of markets for sweetpotato starch in Sichuan will depend upon produc- ing a product with greater purity than that cu rrently available in the market. Such a high- grade sta rch would also permit the produc- tion of nood les of higher quality and va lue. Recent private-sector investment in Shandong Province points to improved starch quality as a prerequisite for producing export-quality nood les. The starch and noodl e agro indu stry in Sichuan is predominantly sma ll sca le. These rural entrep reneurs have made sign ifi ca nt investments in equipment during the last few years, despite the strong seasona l nature of the business. The processing season may last only a few weeks from late October to mid- November each year. Such seasona li ty is un- favorab le for the establishment of large-sca le starch extraction plants, which would rema in idle for most of the year unless they were in- vo lved in processing other raw materials. That situation provides an opportunity for smaller- sca le enterprises to benefit from the expan- sion of market demand for sta rch in Sichuan, and fo r the development of nood le products of higher quality and value. The technical and market resea rch act iv i- ti es between CIP, the Chinese Academy of Agricultural Sciences (CAAS), and SAAS wi ll feed into these and other efforts. Finall y, close links have been estab li shed with reg ional centers of excellence in starch research (Hong Kong University and Nat ional University of Singapore), which will ass ist in relevant tech, ni ca l research and professional development of project co ll aborators in China. Selected Reading CIAD and SAAS. 1995. Potential for market diversifi cat ion of sweetpotato sta rch and flour in Sichuan Province. Center for Integrated Agricultural Development (CIAO) and Sichuan Academy of Agricultural Sciences (SAAS), Beijing Agricultural University, Beijing, China. 28 p. Collado, L.S. and H . Corke. 1997. Properties of sta rch nood les as affected by sweet potato genotype. Cerea l Chem. 74:182-187. Marter, A.D. and W.H . Timmins. 1992. Smal I-sca le processing of sweetpotato in Si chu a:n Province, China. J. Trap . Sci. 32:241-250. CIP Progrom Report 1995-96 2 7 5 Feasibility, Acceptability, and Production Costs of Sweetpotato-Based Products in Uganda V. H ageniman a1 and C. Owori2 Sweetpotato (/pomoea batatas) is one of three stap le crops in Uganda, but its posth arvest use is remarkab ly narrow and limi ted to hu- man consumpt ion in a fresh, boil ed fo rm. The small number of ways in w hich sweetpotato is used, and the limi ted process ing technolo- gies ava il able and adapted to the area, do not permit the potential benefits of the crop to reach farmers and consumers. Research on the identifi cati on of potenti al markets, the assessment of consumer acceptability for new sweetpotato products, and the feas ibility and adaptabi I ity of new rec ipes and processes are important in the w idening of sweetpotato use. Sweetpotato is grown everywhere in Uganda. It prov ides a good part of the dietary starch throughout the year. Fresh sweetpotato is consumed w ith sauces containing beans, cowpeas, or vegetabl es. Boil ed or steamed fresh sweetpotato is the preva iling form of consumption. To a limited extent, sweetpotato is also chipped or cru shed and dried for stor- age from November through January. The dried sweetpotato is then bo iled w ith sauces or tea, or milled into flour and mi xed with millet to process the local porri dge called atapa. M y ri ad prod ucts ca n be made usin g sweetpotato as a major ingredient, but some attitudes and cultural habits related to accept- ability of a new food product need to be over- come. Sweetpotato, either fresh, grated , cooked and mashed, or made into flour, could, w ith high potential for success, repl ace the expensive w heat flour in making bread, c hapati s (Indi an- t y pe fl at brea d ), and 1 CIP, Sub-Saharan Afr ica region, Na irobi , Kenya . 2 Nati onal Agricu ltural Research O rgan isation, Kampa la, Uganda. 27 6 Progrnm 6 mandazi s (doughnuts). Sweetpotato-based products are of hi gh quality and could com- pete w ith existing produ cts on the market. Thi s paper reports the results of a study undertaken in Lira Di stri ct, Uganda, to ex- amine the feasibil ity and profitability of par- t ia ll y subs t i t utin g sw eetpotato coo ked , mashed, or as fl our fo r w heat fl our w hen m akin g bread o r bun s, c hapati s, and m andazi s; and th e acce ptabili ty o f sweetpotato products by consumers. The w ork was conducted prim aril y w ithin the community together w ith women's groups and indiv iduals at Lira's main trading mar- ket. Methods The method used to co ll ect info rmation in this report w as a combination of analys is of avail- able secondary data, informal interv iews, fo r- mal questionnaires, and new product process- ing and taste testing by users. The process ing steps of the products had to be adapted to the equ ipment and utensil s ava il abl e. Members of the community were asked to evaluate sampl es of a product con- taining sweetpotato and compare it to the type of product just purchased or usuall y bought. Most people interviewed were from the Lango ethni c group, and sweetpotato is their stap le food. Results and Discussion Food production and consumption in Lira Agri culture is the main economic acti vity in Lira. Cotton once dominated as a cash crop, but poor producer incentives and inadequate suppl ies of cotton seeds have ca used a big slump in cotton production. Consequently, farmers are shifting their attention to other, new cash crop opportunities. As a stap le food, finger mill et ranks first fo ll owed by cassava , sorghum , and sweetpotato. Cassava, which was an impor- tant cash crop and fam ine reserve in the past, has been devastated recentl y by the high in- cidence of the African cassava mosa ic viru s. The pri ce of cassava on the markets of Lira Di strict is now 7-8 times higher than in 1989. In contrast, the price of other major staples has o nly doubled or tripl ed . H ence, sweetpotato is becoming an important and cheap stapl e food in the district because of cassava virus problems and the poor yield and other uses of finger millet and sorghum. Varieties of sweetpotato in Lira Municipality Neither sweetpotato retail ers nor buyers pay any attention to variety in Lira. During our survey, there were heaps of sweetpotatoes in the market with red-skinned ones mi xed with wh ite-skinned ones. The var iet ies were the loca l variet ies Tedo-Olokeren (white sk in, wh ite fl esh), Luacer or Edopolap (red skin, white fl esh), Mbale or Tanzania (white skin, yel low fl esh), Anamuyito (red sk in , wh ite fl esh), and Odyek-Aw ili (pink skin , white fl esh). Tedo-Olokeren was by far the most common variety. In Lira it is be lieved that Tedo-Olokeren is hard and res istant to wee- vi I attack. It is also w hi te, starchy, and fl oury when cooked as are most other varieties gen- erall y preferred in Uganda. The price of fresh sweetpotato in Lira's main market was from US$0. 01 to $0.02 per 1 kg (US$1 .00 = 1,000 Uganda shil lings). Consumer food preferences As mentioned, finger millet along with sor- ghum, cassava, and sweetpotato are the staple foods of the area . Breakfast food is usuall y a light porr idge made from a flour mixture of finger mill et, sorghum, or cassava. Bread is not a breakfast item, but a rare and expen- sive snack food. It is usually eaten with tea in the evening. Its consumption was fou nd to be as important as th at of mandaz is. Chapati was not considered a snack; it was eaten for lunch or dinner as a main food of the mea l. From the ranking exercise for preference of availab le foods on the market, sweetpotato and bread ranked first, whereas chapatis and mandazis were fourth. About 25% of consum- ers interviewed ranked cassava second, but most others ranked it sixth or seventh, mainly because of its current low availability. Iri sh potato chips were less known in the area, because potatoes were not grown in areas near Lira. These chips were ranked most ly last I ike kabalaga la (a sort of cassava-banana pancake). Availability of baked and fried products in Lira Table 1 lists the snack products and bread sold daily in d ifferent markets of Lira Munic i- pality. Th e act iv ity of tradin g mandaz is, chapatis, kaba laga la, ca ke, and bread in - volves an exchange of about US$700/day, and th e acti v ity employs more than 1 00 people. Half of trade is in bread and a third in mandaz is. During the survey, we noted that many households baked buns. We counted 11 sma l I and loca ll y made ovens for bun bak ing. The bread so ld in Lira is mainly local ly processed, but there is also bread from Kampala, Jinja, and Ki sumu (in Kenya) . The price of bread was US$0.60-0.80 for a 500-g loaf. But few people in Lira can af- ford to buy a whole loaf. Instead, bread was being so ld by th e sli ce for $0.05 each; mandaz is and kaba laga la were also sold for $0.05 each and were popular in the munici- pal markets . Women were the major players in selli ng fri ed and baked products. Feasibility and acceptability trials for baked sweetpotato products Fresh sweetpotato roots used for the tri als were purchased from Lira's main market. Th ey we re a mi xture of white- and red- skinn ed var iet ies w ith white fl es h . Sweetpotato fl our was processed from dried sweetpotato sli ces purchased from a farmer CIP Prngram Repo11 l 995-96 2 7 7 Table 1. Fried and baked products sold in different markets of Lira Municipality, March 1995'. Market Mandazis Sellers Quantity Price/unif' Sellers (no.) sold (no.) (USS) (no.) Cuk-Alok 3 1,000 0.05 0 Junior Quarter 3 250 0.05 0 Main Market 7 1,200 0.05 3 Obong Apewony 4 350 0.05 0 Kokoge 2 100 0.05 0 Soroti Rood 4 350 0.05 0 Borogole 6 500 0.05 l Kitgum Rood 3 550 0.05 0 Aduku Rood 4 500 0.05 2 Total 36 4,800 6 Total (US$) 240 Chapatis Quantity sold 0 0 600 0 0 0 60 0 50 710 71 Price/unit {USS) 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 Sellers (no.) 0 0 2 4 4 2 3 0 16 Kabalagala Quantity sold 0 0 500 350 550 250 700 0 250 2,600 26 Price/unit (USS) 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 a. Bread was sold only at one place near the bus station of Lira Municipality. It was traded as a 500-g loaf or as a 12-bun pocket. The cost of a loaf of bread varied between USS0.60 and $0.80 while that of a 12-bun pocket was $0.60. About l 00 loaves of bread and 450 packets of buns were sold daily. Other baked products at the market were cokes worth $20.00. Each was sold at $0.10. b. USSl .00 = 1,000 Uganda shilli ng s. in Lira Municipality, then ground in the usual hammer mill. The wheat flour was that usu- all y used for bun- or bread-making in Lira Municipality. Bread-baki ng, chapati, and mandazi pro- cessi ng trial s were conducted under local commercial conditions using a mixture con- taining various proportions of wheat flour, sweetpotato flour, or cooked and mashed sweetpotato roots. The recipes used have been reported elsewhere. Because chapatis and mandazis are deep fat-fr ied, the amount of fat absorbed by the products is nutritionally and aesthetica ll y important and influences production costs. In the experiment to determine w hether differ- ences in fat absorption var ied by the form of sweetpotato (bo il ed and mashed, raw and grated, or dried flour) , we found that substi- tuting 50% of w heat flour with boiled and 278 Progrom 6 mashed sweetpotato sign ifi cantly reduced o il uptake in mandazis (F igure 1 ). Acceptab ility was evaluated by compar- ing sweetpotato bread, chapat is, and mandazis (using sweetpotato flour or cooked and mashed sweetpotato to subst itute for wheat flour) with similar products the con- sumer has just bought from the market or had ever tasted. Attributes such as overall taste, texture, freshness, appearan ce, sweetness, and co lor were judged subjectively by bread, chapati , and mandazi consumers. Consum- ers were informed that our products contained sweetpotato as an ingredient. The results (Figure 2) show that bread, chapatis, and mandazis contain ing cooked and mashed sweetpotato were preferred by consumers for taste, texture, freshness, ap- pearance, sweetness, and co lor. Consumers expressed a willingness to pay the same price Oil (g/100 g sample) 20 15 B 10 5 0 -+-~-- A D Chapati D Mandazi 0% substitution 50% substitution 50% substitution 30% substitution (sweetpotato flour) (boiled and mashed) (raw and grated) Level of substitution and sweetpotato Figure 1. Oil content in processed sweetpotato products. (Means followed by common letters for the same product are not significantly different at the 5% level by LSD.) for sweetpotato products as for similar prod- ucts they had been buying. Figure 3 shows the sa les trend for sweetpotato buns at a women's group kiosk in Lira over a 7-mo period. Production costs The data co llected during the study indi- cated that i t was c heape r to produce sweetpotato bread or buns, chapat is, and mandazis than to produce similar products using 100% wheat fl our. Thi s was mainly due to the reduct ion in the amount of wheat fl our, sugar, and o il required to make mandaz is compa red with the cost of ingred ients of tra- ditional mandazi rec ipes th at incorporate sweetpotato. Table 2 compares net revenue for chapatis, mandazis, buns, and bread made using only wheat flour and sweetpotato, and shows a signifi cant profit increase when sweetpotato is used. Conclusions Baked and fried products hav ing sweetpotato as an ingred ient are highly acceptable to the community. Cooked and mashed sweetpotato as an ingred ient improves the taste, texture, Table 2. Comparative gross margins of sweetpotato products and wheat flour products, Lira , Uganda . Product Wheat floor Cha pa tis (l piece) 0.032 Mandazis (l piece) 0.014 Buns {l piece) 0.007 Bread (l loaf of 500 g) 0.058 Net revenue per product (US$) Sweetpototo cooked & mashed 0.045 0.023 0.014 0.065 Sweetpotato flour 0.047 0.023 0.015 0.069 CIP Program Report 1995-96 279 Answers(%) 100 80 60 40 20 0 Taste Answers(%) 80 60 40 20 0 Taste Answers(%) 80 60 40 - 20 0 Taste Freshness • Bread usually bought o Cooked and mashed sweetpotato bread • Sweetpotato flour bread Appearance Texture Sweetness First choice • Chapati usually bought Color o Cooked and mashed sweetpotato chapati • Sweetpotato f lour chapati n I I ! ! 111 Freshness Appearance Texture Sweetness Color First choice • Mandazi usually bought o Cooked and mashed sweetpotato mandazi • Sweetpotato flour mandazi Freshness Appearance Textu re Sweetness Color First choice Oiliness Oiliness Figure 2. Consumer acceptability of baked and fried sweetpotato products. 280 Progrom6 Unsold buns (%) 60 50 40 30 20 10 Sept 95 Oct95 Nov95 Feb96 Mar 96 ___._._ Pure wheat buns --o- Sweetpotato buns Apr96 May96 Figure 3. Percentage of unsold buns for the ls! day on the market at women's group kiosk, Lira . Sweetpotato buns accounted for one-third of total buns. freshness, appearance, sweetnesss, and co lor of bread or bun s, chapatis, and mandaz is. It also signi f ica ntl y redu ces o il uptake in mandaz i and chapati processing. Sweetpotato flour is easy to store and process, and is highly profitab le. St ill , much more fundamenta l re- search is requi red to upgrade the quality of sweetpotato-based fr ied and baked products. Selected Reading Bashaasha,B., R.O.M. Mwanga, C. Ocitti p'O bwoya, and P.T. Ewell. 1995 . Sweetpotato in the fa rming and food sys- tems of Uganda: A fa rm survey report. CIP and NARO, Uganda. 63 p. Hagenimana, V. and C. Owori . 1996. Feasibi I ity, acceptabi I ity, and product ion costs of sweetpotato baked products in Li ra Municipal ity, Uganda: A study report. CIP/ NRI and NARO, Na irob i, Kenya. Hagen imana, V., E.G. Karuri, M.A. Oyunga. n.d. O il content in fr ied sweetpotato processed products. J. Food Proc. Pres. (In press.) W heatley, C., G.J. Scott, R. Best, and S. Wi ersema. 1995. Add ing value to root and tuber crops : A manua l on product deve lopment. Centro Intern ac ional de Agri cu ltura Trop ica l (CIAT), Ca li , Col ombia. 1 66 p. CIP Progrom Report 1995-96 28 1 The Potential of Root Crop Processing for Rural Development in Vietnam G. Prain1, C. Wheatley2, and Nguyen Doy Duc3 Vietnam is a leading world producer of sweetpotato (/pomoea batatas) and cassava (Manihot spp.) , and is probably the global leader in area planted to edib le canna (Canna edulis) . Historically, these crops were impor- tant seasonal , supplementa l, and emergency food sou rces, but in the past 15 years, they have begun to assume greater importance as raw materials for va lue-added processing. Feed and starch are the dominant intermedi- ate processed products. Pork, transparent noodles, candies, and baked goods are com- mon end products. The expanded process- ing activity has been almost exclus ive ly based on household initi ative. 1 CIP, Los Banos, Philippines. 2 CIP, Bogor, Indonesia. 3 Post H arvest Technology Institute, Hano i, Vietnam. Legend Bae Thai = Main research site §o_n .!,.a_= Secondary site Since the mid-1980s, however, public-sec- tor research agencies have been working to introduce the processing innovations that were developed in vi llages around Hanoi to other parts of the cou ntry. The mixed results from those efforts led CIP and several inter- national and national research partners to begin an assessment, in 1995-96, of current use of root crops in four key sites in north and central Vietnam (Figure 1 ). The sites w ere in Bae Thai , Hatay, and Thanh Hoa Provinces in the north , and Quang Nam Da Nang Prov- ince in central Vietnam. Sweetpotato and cassava are the two main root crops grown in farming systems in north and central Vietnam. Edible canna is less im- portant, but was included in the study be- Quang Nam Da Nang gu~n.\! ~h~ Ho Chi Minh City Figure 1. Provinces in north and central Vietnam where root crop utilization research was conducted. 282 Prog ram 6 cause of its Andean root crop ancestry and its importance in nood le mak ing. The major ity of fa rm households in the prov inces under study depend on crop-live- stock systems to secure a ri ce-based food sup- ply for the household as well as cash income. Both food and cash depend on an adequate supply of manure for the crops and an ad- equate supply of feed for the animals. When process ing is undertaken fo r add itional in- come, it provides a further supply of feed in the form of by-product, but it increases de- mand for raw materi al. In Thanh Hoa Province, fresh consump- tion of sweetpotato by producer households decl ined from around two-thirds to one-third of total production from 1980 to 1994. A simi- lar decline is also ev ident in cassava, but the change of function of these two crops has been different. Wh ereas the primary function of sweetpotato has shifted from household food to household an imal feed, cassava has become a cash crop. It is grown primarily as raw material for the growing process ing in- dustry, but the process ing by-product finds a ready market as animal feed. Sown area and production trends for cas- sava over the past ten years show a gradual downward trend throughout the period in all four provinces (Figure 2) . The abrupt decline in cassava in Hatay, w hich is close to Hanoi, can be attributed to a switch to sweetpotato by some farmers and a move from culti va- tion to more lucrative food processi ng and nonfa rm enterprises. A strong upward trend in sweetpotato began around 1990, especially in Hatay and Thanh Hoa, where it is increas- ingly culti vated fo r pig feed. Role of Root Crops in Pig Raising From 1985 to 1995, pi g liveweight in the north rose from 299,000 t to 580,000 t; per capita pork consumption more than doubled. The massive increase in sweetpotato area in Hatay, from 5,000 ha to more than 20,000 ha between 1990 and 1993, is directly attrib- utable to the use of sweetpotato for feed. Pig-raising systems Two household-based pig-rai sing systems prevail w ith different impli cations for pig feed resea rch. In one system, piglets are fattened fo r slaughter. M any households comb ine these systems (Tab le 1 ). In Da Nang, households commonly con- centrate on ly on breeding pi glets, part ly be- cause of better access to the piglet markets of southern Vietnam and Laos. Fattening is the more common pract ice, espec iall y in the more remote, poorer areas of the north . In Bae Thai , for example, there is an average of three fattening pigs per household, whereas only one in four famili es has a breeding sow. In high mountain areas of Bae Thai, Thanh Hoa, and Da Nang, catt le are more common domest ic animals, mainly because of the ava il ab ility of open pasture. There is cons iderable variati on in the num- bers of pigs managed by different households (F igure 3), which is an important consider- ation in technica l intervention. Famili es fat- tening only one pig are least likely to be in- terested in innovat ion s. With more pi gs, im- proved performance becomes commerciall y attract ive. Da Nang agai n stands out as different from the other sites . Not only do many hou seholds have two or more breeding sows, many more households are fattening only one piglet. That suggests the low priority and noncommercial character of fattening in Da Nang. Bae Th ai and Hatay have similar numbers of households engaged in fattening. Both ap- pea r to have greater commercia l pig-raising than Thanh Hoa. In Hatay and in Thanh Hoa, of households fattening more than four pigs, only five have more th an 10 head, and of those the largest operation has 40 head. These are the on ly large-sca le pig-rais ing enterprises encountered, emphas iz ing the backyard char- acter of most pig-raising. Feed regimes The two most notable findings about swine feed reg im es are (1) the dominance of CIP Program Report 1995-96 283 Sweet potato 300 1,000 tons 250 200 150 100 50 0 --+-~--.-~--.-~--.-~--.-~--.-~--.-~--.-~--.-~--.-~~ 1985 Cassava 250 200 150 100 50 1990 1995 -0-Da Nang --{}---Thanh Hoa --!::s- Bae Thai ~Hatay 0 -+-~---..~~r-~......-~--r~~.----~,...-~-.-~---.~~~~--. 1985 1990 1995 Year Figure 2. Sweetpotato and cassava production in lour provinces of Vietnam, 1985-95. 284 Progrom 6 Table 1. Pig-raising strategies in case households of four provinces, Vietnam, 1995. Province Househalds (no.) Fattening Breeding Bath Total Bae Thai 21 Hatay 15 Thanh Hoa 8 Do Nang 9 Total households (%) 70 60 50 40 30 20 10 0 0 9 1 14 2 7 5 6 • F 1 head &'!!! F 2-4 head rm F > 4 head CJ B 1 head CJ B 2-4 head 30 30 17 20 Bae Thai Hatay Thanh Hoa Da Nang Figure 3. Pig-raising systems among case households in four provinces of Vietnam, 1995. (F = fattening, B = breeding.) sweetpotato v ines over roots, and (2) the lim- ited use of protein-ri ch supplements, even in coastal areas w here fi sh by-products are rela- tive ly cheap. Vines are the most important component of pi g diets in 7 out of 12 p ig-rearing house- holds in Bae Th ai . In one case, v ines ac- counted for 45% of the approx imate ly 2.4 t of feed given to a single animal over a 12-mo growing period. Cassava is the most important feed com- ponent in mountain hou seho lds. In Hatay, the by-product from starch-processing v i 1- lages accounts fo r up to 80% of total feed, in combinati on with root crops. A large part of the root crop component, however, is vines rather than roots. O n th e coas ta l pl ain of Th anh Hoa , sweetpotato roots, both fresh and dri ed, are much more important as feed than at other sites . Thi s is an area w ith as many as three harvests of sweetpotato per year and w here storage of dried sweetpotato chips is relatively common. Sweetpotato vines are also impor- tant as feed. CI P Program Report 1995-96 285 The feed regime also vari es in Da Nang. In th e p la in s, 75% of fee d co nsists of sweetpotato v ines and small roots, w hereas at hi gher elevati o ns cassava by-produ ct dominates, although sweetpotato vines still account fo r about 30% of total feed. Vines dominate over roots in Da Nang because of the concentrati on of househo lds mainta ining sows to produce pig lets . Fresh or dried v ines, cooked w ith oth er vegetati ve by-products, prov ide a better diet for piglets th an roots. Socioeconomic importance of pig-raising In the plains of Bae Thai and in the Red Ri ver Delta, pig-raising contributed on aver- age about 40% of income in case households in 1995, ranging from as little as 15% to as high as 64% . However, the purpose of pig- raising in Bae Th ai is not prim aril y to gener- ate d irect income. Pi g-fa ttening offers a w ay to convert nonfood fa rm output into manure and m arketa bl e po rk usin g surplu s labor-that of the elderl y, the young, and the limited spare time of w omen. Househo lds th at keep breeding sows fo l- low a different strategy. For instance, in Da Nang fa rmers seek to reduce feed costs and use of nonfood fa rm output, and to reduce the ri sk of di sease, by selling off p iglets earl y. Thi s system tends to generate less manure for the crop operat ion. Thi s profit-ori ented but risk-averse approach works well in the spe- cial c ircumstances of Da Nang, w ith its good links to piglet markets. The highest profi ts were made by families that had combined breeding and fattening pig operati ons, and by those w ith the most ani - mals. Constraints and opportunities Pig fa ttening and piglet breeding pose dif- ferent constraints and offer d iffe rent opportu - nities for intervention in root crop-based feed. Househo lds invo lved in mul t ip le pig-fat- tening are interested in increasing the vo lume of feedstuffs produ ced on -farm, feed conver- sion characteri st ics of their animals, and the rate of weight ga in. In Thanh Hoa, the far- row -to-f ini sh time ranges from 6 to 12 mo. 286 P1og1om6 The diffe rence is largely due to d iet, both the vo lume of feed and the incl usion of prote in- ri ch supp lements. There are cl earl y opportu- niti es here for exp loring improved v ine and root producti v ity and fo r eva luating alte rna- t ive feed regimes. Starch and Noodles: Prospects for Sweetpotato Starch production in north and central Vietnam Major sources. The two major sources of starch in north and central Vietnam are cas- sava and ca nna; sweetpotato is onl y a minor starch source at present. Cassava is by fa r the largest and cheapest source and small-sca le processing occurs in most provinces (Table 2). Canna starch process ing occurs o n a small er sca le, primaril y fo r the specialty pro- duct ion of transparent nood les. Canna starch enterpr ises were found in a few vi ll ages in Bae Thai and Thanh Hoa close to areas w here the root was grown; canna process ing is not done in Da Nang. In northern Vietnam, most cassava and canna starch producti on is con- centrated in fi ve v ill ages in Hoai Due Dis- tri ct, Hatay Prov ince, near Hano i. A tota l of 5,679 househo lds were processing cassava starch in 1995, consuming between 100,000 and 200,000 t of roots to produce around 50,000 t of starch. In the sa me area, 276 househo lds processed about 34 ,000 t of canna into 7,500 t of starch in 1995. Processing efficiency. The procedu res to extract starch from cassava , can na, and sweetpotato are simple and cheap. Roots are washed (cassava roots are pee led fi rst to pre- vent starch disco lorati on), grated, and fi lte red through c loth in to a concrete container of water to make a slurry. The slurry is filtered four times and then allowed to sett le fo r 6 h (longer in w inter). The water is drained, leav- ing wet starch of 35 -45% moisture content. From 1 00 kg of cassava roots, about 42 kg wet sta rch can be extracted, yielding about 28 kg dry starch. Canna roots convert to 28- 30 kg wet sta rch from 100 kg fresh wt, yield- ing 21-23 kg dry sta rch. Sweetpotato has the lowest extraction rate, yielding 25 -30 kg wet Table 2. Costs and benefits' of processing l 00 kg of cassava roots into starch in four provinces of north and central Vietnam, 1995. Province Casts (USS) Fresh Labor Other Total roots Hatay 4.36 0.75 0.32 5.43 Bae Thai 3.64 0.91 0.53 5.08 Thanh Hoo 4.18 0.91 0.09 5.18 Do Nang 2.21 l.14 013 3.54 a. The 1995 exchange rate was Vietnam D 11 ,000 = USS 1.00. starch or 1 7-20 kg dry starch from 100 kg fresh wt, depending on var iety and time of year. Cassava processing benefits. Producers of fresh cassava and can na and a vari ety of as- sembler and transport businesses supply the roots over long d istances for sa le to starch processors in Hoa i Due. Th e co ncentration of processin g house- holds in one area in Hatay increases the effi- ciency of transportat ion, making it feas ible to truck in fresh roots from up to 320 km away. It also fac ilitates the specializat ion of process- ing act iv ities among hou seholds and the ex- change of products and serv ices. The large vo lume of f inal products generated enters a compl ex marketing system of processed prod- ucts throughout north and central Vietn am . Cons ignment arrangements are common in th is system between processors and traders. Costs and benefits va ry greatly among the cassava processors stud ied. Hatay processors have high t ransportat ion costs, w hi ch in- creases the pri ce of the roots. But that is partly compensated for w ith higher processing effi- ciency. The profitability in Bae Th ai is due to a hi gher se llin g price for wet starc h of U5$0.16/kg, compared with between $0.12 and $0 .1 3/kg at the other sites, and a much higher va lued by-prod uct. Income (USS) Net profit (USS) Storch Block By- Total starch product 6.11 0.32 0.27 6.70 l.27 6.36 0.91 7.27 2.19 5.27 0.29 0.73 6.29 l.11 3.78 0.34 0.91 5.03 1.49 Though profitability is qui te low for cas- sava starch processors, it represents about 60% of annual income for some households. In Th anh Hoa, the process ing season runs from October to March. Average process ing during the season fo r the case households is about 30 t of fresh roots, w hich y ields just under $400 income. Sweetpotato could be an alternative source of bu lk starch in the north. At present there is almost doub le the quantity of raw material ava il ab le (1.8 million t compared with 1 mil- lion t of cassava). However, the low starch extraction rate of sweetpotato makes cassava ultimately a cheaper raw mater ial. Transparent noodle supply and demand Per cap ita noodle consumption in Vietnam in the ear ly 1990s was 12-20 kg; transparent noodle may represent 10-15% of that. The most common raw material for transparent nood le over the past 30 yr has been canna root. Canna sta rch and nood le processing is undertaken on a sma ll sca le in Bae Thai and Th anh Hoa, but the major production area is the same intens ive process ing di strict of Hoai Due, Hatay. Un li ke cassava-based process- ing, w ith its cl ea r division of labor between starch producers and maltose processo rs, CIP Program Report 1995-96 2 8 7 canna starch and nood le processing is often combined. Of the 276 starch-produc ing households in Hatay, 230 also prod uce noodles using their own sta rch. The long growth duration of canna, usu- all y from March to November, means that the processing season is quite short, gener- all y from November to March or April. Al- though thi s covers the period of the Vietnam- ese New Year when nood les are most in de- mand, it limi ts the opportunities for opt imiz- ing investment at other times of the year. In Bae Thai, where up to a ton of fresh roots can be processed in a day, average net prof- its of case househo lds were around $16/d (Table 3). Thi s compares to around $1/d for agri cultural laborers. The bulk of starch and noodle demand is met from Hatay, thus in- fl ating the costs of the fi nal product in areas beyond Hano i. Th e lower profitability of starch from Hatay reflects both the more ex- pensive raw materia l, trucked in over large distances, and a lower sta rch price. Nood le process ing clearly offers the high- est returns of the processing activities stud- ied (Tab le 4). The net benefit recorded in Thanh Hoa may be overestimated. Un like the other ca lculations, the Thanh Hoa ca lcul a- tion fai led to include deprec iation and un- derestimated costs fo r fue l and other inputs. The rel ative ly low profitabil ity of noodle pro- duction in Bae Th ai is ca used by a much lower conversion rate from starch to nood le, w hich is due to a different and less effic ient process ing techno logy that uses manual ex- trus ion rather th an steaming. The potential of sweetpotato starch in noodle enterprises The relatively narrow adaptation of ca nna at high altitudes, its remoteness from process- ing areas, and the long growi ng season mean th at the supply of raw material is limi ted to a few months. But the period of raw materi al suppl y co incides w ith the per iod of highest demand fo r the sta rch, lead ing up to the New Year celebration s. Therefore, pri ces remain high, even during thi s glut per iod. These raw materi al supply prob lems, how- ever, have resulted in very little exp loration of sweetpotato as an alternat ive. Thi s is de- sp ite the prox im ity and stab le supp ly of the crop-there are three harvests a year in some areas- and the limi ted fresh root market for sweetpotato, w hich frequent ly resu lts in de- pressed pr ices. Table 3. Costs and benefits0 of processing 100 kg of canna roots for starch in three provinces of north Vietnam, 1995. Province Costs (USS) Income (USS) Net profit (USS) Fresh Lobor Other Total Starchb By Total roots -product' Boe Thai 3.76 0.42 1.52 5.70 7.32 7.32 1.62 Ho toy 4.09 0.91 0.31 5.31 6.70 0.04 6.74 1.43 Thanh Hoo 3.18 1.09 l.03 5.30 7.00 7.00 l.70 o. In 1995 the exchange rote was Vietnam D 11,DOO = USSl .00. b. Based on extraction of 22 kg starch at USS0.30/kg in Hotoy, extraction of 22.4 kg starch at USS0.30/kg in Boe Thai, ond extraction of 22 kg starch at USS0.32/kg in Thanh Hoo. c. Conno starch by- prod uct is less valuable than cassava by-product as animal feed because of its fibrousness. It is used mostly as compost. 288 Progrom6 Table 4. Costs and benefits' of processing 100 kg of canna starch into transparent noodleb in three provinces, Vietnam, 1995. Province Costs {USS) Income Net profit {USS) {USS) Canna starch Labor Other Total noodlesb Bae Thai 30.00 3.02 2.21 35.24 40.02 4.78 Hatay 30.45 4.09 18.86 53.40 60.36 6.96 Than h Hoa 31.81 3.64 3.91 39.36 69.36 30.00 a. The 1995 exchange rate was Vietnam D 11 ,000 = US$1 .00. b. Based on conversion rates from sta rch to noodles of l :0.59 in Bae Tha i, l :0.83 in Hotay, and l :0.88 in Thanh Hoo. The reluctance has been due to techn ica l and cultural factors. Sweetpotato has a lower starch content than canna and the extracted starch has a grayish co lor. The nood les made from the starch also tend to be more br itt le. Sweetpotato is also perce ived as a food and feed crop, not an industr ial raw mater ial. Technology development and participatory pilot studies in Hatay and Th anh Hoa Prov- inces have shown that it is technica lly fea- sible and profitab le to use sweetpotato starch in transparent noodles. Using l ime during sta rch filt rat ion improves separat ion and yields a wh iter starch. And combining canna starch w ith sweetpotato starch overcomes the britt leness of 100% sweetpotato. Using sweetpotato starch, however, in- vo lves more work and processors require a lower pri ce for sweetpotato starch to justify the extra effort. In add it ion, the supp ly of sweetpotato starch remains a probl em and is not yet stab le. In 1996, pilot work on starch production , begun earlier in a southern d istri ct of Thanh Hoa Province, t ransferred to Hoang Hoa, nearer the provincial capita l. Twe lve house- hold s have enthusi as ti ca ll y adopted sweetpotato and can na starch production, first to supply noodle processors in Thanh Hoa City, and more recentl y to produce nood les themselves. Almost all of the sweetpotato/canna nood le produced so far has been marketed and priced as canna nood le. In focus group dis- cuss ions w ith a stratif ied sample of consum- ers in Hanoi during 1995, high-income con- sumers reacted negatively to sweetpotato as a raw material. But medium- and low- income consumers, w ho were more concerned with nood le quality than sweetpotato's image, gave a positive eva luation . Conclusions and Opportunities The enormous opportun iti es fo r econom ic development in Vietnam are most ly generated at the base, among rural households. The role of agricu lture is increased product iv ity, of course, in add ition to va lue-added products made ava il ab le through processing. The pro- cess ing of root crops for pi g feed, starch, and nood les is an important part of that process- ing potential. As a resu lt of the appraisal of the use of root crops for pig feed, starch, and noodles, two projects have been established. The fi rst project involves: • eva luat ing geneti c material fo r high foliage production and high protein content for use in pig-fattening enterprises, • eva luating geneti c material fo r high dry matter, • assess ing the use and costs of ava ilable CIP Progrnm Report 1995-96 2 8 9 protein supp lements to energy-based sweetpotato and other feed, and • exp lor ing alternative ways of preparing feed to improve digestibility and quality. The second project aims to improve the efficiency of sweetpotato and canna for starch and noodle processing. It involves: • supporting pi lot units for sweetpotato starch and noodle production, 290 Prog ram 6 • evaluati ng new sweetpotato cultivars for potential in starch processing, • evaluating new canna var ieties for use in starch and noodle production , and • supporti ng improved marketing linkages between starch producers and noodle producers . Perspectives on Sweetpotato: Dual-Purpose Varieties C. Le6n-Velarde1, J. Roca1, J. Arteaga2, L. Quispe2, and A. P arraga2 Sweetpotato (/pomoea batatas) is one of 12 main plant species used as hum an food throughout the world. The roots are used for human and animal consumption, and the vines are generall y used for animal feed along with crop res idue and unmarketabl e roots . In the trop ics, one of the main limi tations to small fa rmers in mi xed crop-li vestock pro- duction is a reliable, year-round supply of feed and energy. Sweetpotato might help to over- come this limitation with minimal environ- mental damage. However, small farm size and low income from the crop limit the area grown to sweetpotato for use in livestock pro- duction only. The germpl asm collection held at CIP in- cludes a large group of sweetpotato clones and var ieties. Breeding efforts concentrate on root product ion, with emph as is on dry mat- ter content, flour, and starch. One group of clones, however, produces mostly forage with low root production, which makes it idea l for animal feed. Although in recent years demand fo r these c lones has increased, root produc- tion is also necessary fo r human food and as an income component of small crop-livestock fa rm s. On sma ll crop-livestock operati ons, dual- purpose sweetpotato c lones hold a compara- tive advantage over single-purpose varieties grown fo r roots or for forage. In an opt imal integrated livestock management system, farmers could take advantage of sweetpotato's regrowth ability by continually or period ica lly harvesting the vines for animal feed through- out the growing season before finally harvest- 1 CONDESAN/CIP, Lima, Peru. 2 Universidad Nac ional Alcides Carri on, Oxaparnpa, Peru. ing the roots for human food. Because there is littl e information on the management of dual-purpose clones, this work focused on de- termining a preliminary roots-forage classifi- cation , and on deve loping management strat- egies fo r dual-purpose sweetpotato variet ies. The work was done jo intly by CONDESAN/ CIP and the Universidad Nacional Alcides Carrion, Oxapampa, Peru . Materials and Methods We first established a preliminary classifica- tion for dual-purpose varieties based on the relati ve contribution of foliage and roots to dry matter (DM) content. Then we looked at how crop management affected the partition- ing of DM in foliage and roots in f ive clones identified as having dual-purpose potential. Relation of roots and foliage A data set of 1, 1 68 accessions from the germplasm co ll ection held at CIP was ana- lyzed in 1995 in La Molina, Peru , to estab- lish a preliminary classification fo r dual-pur- pose vari et ies. We identi fied potential dual- purpose varieti es based on the rat io of root to foliage (R/ F) contribut ions in the produc- tion of total DM. Ca lcu lated va lues of R/F ranged from 0 to 4, but in some cases actual va lues exceeded the va lues in the range con- sidered . As a result, we deleted 11.8% of the inform ation. Th e extreme valu es resulted when in some cases a group of two or three pl ants was considered as one pl ant. When the germplasm co llection was being eva lu- ated , some varieti es produced roots more quickly, whereas others did not. As a consequence, we need uni form data collection over time from germpl asm plots to CIP Progrom Report 1995-96 2 91 accurately classify potential dual-purpose varieties based on R/F. We also need to complement that information with other bioeconomic characteristics not included in this preliminary classification . Systematized information of that type will be valuable for breeding to produce primarily roots, prima- rily forage, or roots and forage. Table 1 shows a classification scheme of fi ve categories based on R/F, ranging from 0, indicating no roots present, to 3-4, indicat- ing high root production. There is some over- lapping of categories on root and foliage pro- duction. Categories of low dual-purpose and high dual-purpose showed no significant dif- ference for root production. Similarly, catego- ries of high dual-purpose and low root pro- duction showed no significant difference for foliage production. This work indicates the potential of crop varieties from low root production or high dual-purpose categories to perform well as a dual-purpose crop. A multivariate analysis (analysis of principal components) showed that the group of accessions classified within the range established (Figure 1 ). The forage category shows a wide dispersion, whereas categories 3 and 4 are grouped together, showing the potential of the accessions for dual-purpose use as wel I as for root produc- tion. Management effects on R/F The frequency of cutting forage from the growing crop is the main management fac- tor. For livestock production, several cuttings of foliage are required throughout the crop- growth period, even though root production is sacrificed. When crop, rather than forage, production is most important, root develop- ment is paramount. However, in the crop-livestock operation, both foliage and roots are desirable. There- fore, we evaluated potential dual-purpose clones to determine the effect of foliage cut- ting frequency on forage and root production. Five clones were evaluated: DLP-3548, R/F 1 .43; ARB-265 I R/F 0.51; ARB-142, R/F 0.42; RCBIN-5, R/F 0.20; and ARB-UNAP55 (cv. Helena), R/F 0. The evaluation was car- ried out in Oxapampa, Peru, in 1996 during the dry and wet season, at 1,850 m above sea leve l. The zone presents an average tem- perature of 16°( and annual rainfall of 600 mm. Soils are neutral to slightly alkaline. A split-plot design was used with the five accessions in the main plot, and foliage cut- tings at 45, 90, and 1 3 5 d as treatments in subplots. Six cuttings were made at 45-d in- tervals, three at 90-d intervals, and two 135 d apart. DM and protein content were ana- lyzed. Table 1. Classification of sweetpotato dual-purpose varieties based on the relation of roots and foliage production of dry matter' , CIP, La Molina, 1995. Classification Range R/F Observations Roots Foliage R/F (no.) (g/plant)0 (g/plant)0 1. Forage 0-1 .0 632 136.51 ± 3.21 d 140.72 ± 3.70 a 0.24 ± 0.01 2. Low dual-purpose > 1.0-1.5 113 164.31 ::': 7.60 c 133.58 ± 8.77 a l .24 ± 0.03 3. High dual-purpose > 1.5-2.0 82 180.97 ± 8.92c 104.79 ± 10.29 b 1.74 ± 0.03 4. Low root production > 2.0-3.0 127 212.04 ± 7.17b 87.26 ± 8.27 b 2.45 ± 0.03 5. High root production > 3.0 76 235.89 ± 9.27 a 68.l9±l0.68c 3.47 ± 0.03 a. Least mean squares and standard error. In a column, means followed by a common letter are not significantly different atP<0.05. 292 Program 6 PRIN2 5 - 4 - .I 3 - . I 2 - I - II I.II. I I II . I I. I . II II I. I . I . I I . I 111 111 111 I 11. I I I 111111 .I I . I 11111 I . I .. I 2 II Ill I. Ill . Ill 2 111 11 111 1111 11.1 I I 2 .. 22. 2. 1111111111 . 1 I Ill 11.1 . 2 2 2. 1111111111 111.1 .1 112 2 2 2 . 2. 2 2. 2 0 - .11111111111.11111122222 2.22 2 .. 22 1111111111 I 1 222222222 222.22 2 3 3 3 3 3 . 111111111122222222222 3223333 333 33 3. 11111112322323323333333333 333 3 3 . . 112225333344444434443334444.34 4 . .44 4 4 4. 4 -I - 5445555455544444444 44 44.44 .4 4 .4 4 55555 55555.4554454 .4.4 4 4 .4 555 5 555.5 5 5 5 . 4 5 5555 5 .. 5. 4 -2 - 5 -.r.t.t.f.f.t.t.trJJJJJJ.f.fr.f.t.f.f.f.f.HrJ.f.f.f.fJJ.tFJ.f.f.f.fJ.f.IFJ.HJ.fJJ.tr.t.tJJJ.f.f.tFJ.f.fJ.tr -2 -I 0 I 2 3 4 5 PRIN1 Figure 1. Graph of l, 168 sweetpotato accessions class ified within five dual-purpose categories and plotted considering two principal components. The five categories are l , forage ; 2, low dual-purpose; 3, high dual -purpose; 4, low root production; 5, high root production. Results Tab le 2 shows total DM forage production at cutt ing frequencies of 45 , 90, and 135 d. For- age production was significantly hi gher (P<0.01) in cuttings at 90-d intervals than in cuttings at 45- and 135-d frequencies. Cut- tings at 45-d intervals give good forage pro- duction, but root production is nil. A cutting frequency of 90 d appears to strike a reason- ab le balance between forage and root pro- duction. A cutting frequency of 135 d in- creases root production but decreases forage production. R/F values increase as the inter- vals between cuttings increase, thus indi cat- ing better root formation. R/F values are within the range described in Tabl e 1. The choice of a cutting frequency depends on whether for- age or root production is the object ive of the mixed crop-livestock farm. The clone ARB-UNAP55 requires special mention . Thi s clone is considered to have an R/ F of 0, being stri ct ly considered a forage variety. However, in certain zones such as Oxapampa and in Africa, it has been found to produce some roots, which may be ev i- dence of a genotype x environment interac- tion. In thi s study, the R/F of ARB-UNAP55 is less than that of the other clones eva luated, and close to 0. Table 3 shows the management effects of cutting frequency and ferti I ization on clone ARB-U NAP55 . A lthough there is a positive effect of fertilization at rates of 60, 120, and 180 kg/ha N, DM production decreases with increased cutting intervals. A similar reduc- tion in DM production was observed when the clone was eva luated for dual-purpose use. Although fertilization increases DM produc- CIP P1ogrom Repo111995-96 293 Table 2. Average production of dry matter int/ha of forage and roots in five sweetpotato dual-purpose clones, and the relation of roots/forage (R/F), Oxapampa, 1996. Clones Frequency (days) 45 90 135 Forage• Roots0 R/Fb Forage" Roots• R/Fb Forage• Roots0 R/Fb ARB142 7.1 0.0 0.0 12.38 1.4 0.11 8.1 6.6 0.82 ARB265 10.4 0.0 0.0 14.25 3.7 0.25 9.7 8.3 0.85 DLP3548 9.5 0.0 0.0 14.51 2.4 0.16 10.5 6.7 0.63 RCBIN5 4.5 0.0 0.0 8.40 2.0 0.23 5.9 7.9 l.35 ARB-UNAP55 6.4 0.0 0.0 13.23 2.0 0.15 10.4 3.9 0.37 a. Dry matter production. b. Relation of DM content of roots and foroge; percentage of DM of forage production ranged from 13. l % to 16.1 %; percentage of DM of roots ranged from 19.8% to 24. l %. Table 3. Total production of dry matter (t/ha) of clone ARB-UNAP55 (cv. Helena) at different levels of N fertilization and different cutting frequencies, San Ramon, 1996. Cutting frequency (days) 45 (6 cuts) 90 (3 cuts) 135 (2 cuts} 0 6.7 6.1 5.4 tion, the cost of this input must be consid- ered in any technological alternative for sweetpotato forage production. Conclusions The analysis done on s 1 elected sweetpotato dual-purpose clones indicates that a cutting frequency of 90 d during crop growth gives the best balance between forage and root pro- duction. Harvesting foliage is hard work. Whether farmers are willing to invest in the labor required to gain maximum forage pro- duction depends on the crop-livestock mix. Studies on grazing cattle are necessary. Con- sidering the wide number of clones available, future selections for consideration as dual- 294 Program 6 Nitrogen (kg/ha) 60 120 180 7.7 6.8 5.5 8.3 9.1 7.8 8.0 6.0 6.2 purpose varieties should include those char- acteristics that favor livestock production. Selected Reading SAS (Statistical Analysis Systems). 1986. SAS system for linear models. 1986 edition. SAS Institute Inc. , Cary, North Car. 211 p. Scott, G.J . 1991 . Sweet potatoes as animal feed in developing countries: Present patterns and future prospects. In: Roots, tubers, plantains and bananas in animal feeding. Animal Production and Health Papers95. FAO, Rome. p. 183-189. Woolfe, J. 1992. Sweet potato: An untapped food resource. Cambridge University Press, Cambridge, U.K. 214 p. Crop Growth and Starch Productivity of Edible Canna M. Hermann\ R. Uptmoor2, I. Freire\ and J.L. Montalvo3 Edible canna (Canna edulis Ker-Gawler) is a starchy root crop that is grown sporadi ca lly in the tropical highlands for food security. Tax- onomists now consider edible ca nna con spe- cific with ornamental or feral Canna indica . But edible ca nna's much larger rhizomes, inconspicuous flowers, and high starch con- tent set it apa rt from C. indica. In its nat ive Andean range, as well as in other parts of the developing world, the use of canna for direct consumption is about to disappea r, mainly because of the long cooking time required to soften rhi zome tissue (>3 h). In Vietnam and southern China, however, there has been a new appreciation of ca nna as a source for sta rch in th e manufacture of transparent noodles, a lu xury food widely eaten across Asia. In Vi etnam alone, the canna area is es- timated at 20,000 to 30,000 ha. Canna starch has the largest grains known and it settles quickly out of a suspension of grated tuber ti ssue. Starch recovery in rural factories is therefore high (>80% of total starch content). The starch is hi gh in amylase and functionally similar to mungbean starch, the traditional raw material for transparent noodles. Canna starch is, however, less ex- pensive to produce. Canna is an outstandingly versatile and robust crop. It is typi ca lly not ferti I ized and significant pests or di seases as well as repl ant problems are unknown. Where rice-based cropping systems predominate, as in Taiwan, Vietnam, and Indonesia, canna is grown on unirrigated uplands, in backyard ga rdens, or other areas where it does not compete with other food crops. In open (unshaded) pl ant- ing sites, plant deve lopment ca n be exuber- 1 CIP, Lima, Peru, and Qu ito, Ecuador. 2 Kasse l University, Witzenhausen, Germany. 3 Universidad Central, Quito, Ecuador. ant and rhi zome yields high. Canna is also a promising ca ndidate for agroforestry systems in trop ical mountains where it produces un- der significant shading and on marginal soils. Drawbacks to canna are its long crop du- rati on of 1 yr, which limits the crop's use in ti ght cropping systems. Also, its sta rch pro- ductivity and plant architecture have not been investigated so far. Therefore, thi s study aims to ga in a bas ic understanding of crop growth and sta rch formation of canna. Recommen- dations for crop management and breed ing of this poorly known-species will be derived. Materials and Methods In addition to taking field surveys in several countries, we eva luated plant performance of canna in the greenhouse to determine growth variabi I ity and yield characteristi cs . Greenhouse experiment For the greenhouse experiment, we used 26 accessions from the internati onal canna collection held at CIP. The accessions were selected to represent the full geographic range of ca nna in the Andes and to inc lude diploid and triploid material s. Knowl edge of ran- domly amplified polymorphic DNA in the collection allowed us to identify distinct geno- types for the experiment. The materi al was grown in a quarantine screenhouse in the equatorial Andes nea r Quito, Ecuador, at 2,400 m above sea leve l. Seasonal temperature va riation in Quito is minim al; the diurn al temperature in th e screenhouse during the experiment ranged from 12 to 27 °C. W e planted ro o ted propagules at a density of 2 plants/m2 in natu- ral so i I of a sa ndy texture and moderate fer- CIP Program Report 1995-96 295 tility. The crop was irri gated 2 or 3 times/ week, but was not fert ili zed. After 12 mo, we harvested the mater ial and took 4 plants at random from each accession for analysis. Field study For the f ield study, we chose Patate, a fa rmer community in the highlands of Ecua- dor (nea r the equator at 2,350 m), w here can na is grown commercia lly fo r starch . The highl y drained soi ls in th is area range from sand to loamy sand, w ith neutra l pH and high nutrient content. Pl ant ava il able N was 144 mg/100 g soi l; P, 153 mg/100 g so il ; K, 30 mg; and Mg, 32 mg. Furrow irrigat ion was app li ed week ly, but overall so il moistu re was low and most likely represented a growth-lim- iting factor. Mean monthl y temperatures were 15-1 7°( w ith d iurna l amplitudes of 10-25°C. During the fi rst 5-6 mo after p lanting, ca nna was intercro pped w ith potatoes and veg- etables. We harvested plots ranging in size from 10 to 20 m2 in di fferent fi elds after 6, 8, 10, and 12 mo of crop duration. Four repeti- t ions were used . We determined dry matter (DM) at 105°( and mechani ca ll y ex trac ted sta rch by a method allow ing the recovery of about 90% of total starch. So luble so lids were measured refractometri ca ll y. Leaf area was assessed by a linea r regress ion describing the relat ion of leaf area to the product of leaf length and w idth. Pl ant nutrient contents were measured by standard methods. Results Canna's plant architecture is typical fo r a member of the Zingiberales (F igure 1 ). The plant forms a branching rhi zome, w hich in mature plants consists of several segments or generations th at deve lop consecutive ly. Each segment ca rri es a shoot, w hich concludes its deve lopment w ith the formation of a termi- nal infl orescence and se nesces . Sim ulta- neously, a new shoot w ill give ri se to a new rhi zome segment. Each segment has the ca- pac ity to deve lop sprouts from several buds. The size of the rhizome segments and differ- ent branching modes di stin gui sh rhi zome vari abi I ity between accessions (Figure 2). 296 Progrom 6 So me pec uli arities charac ter ize p lant growth of ca nna and its measurement. The dead aboveground plant matter remains at- tached to the plant and ca n be gathered and determined at harvest. Canna is also d iffe r- ent from most other perennial roots in th at thi ck adventi t ious roots anchor the rhi zomes firm ly in the so il and ca use strong pulling re- sistance at harvest. The youngest rhi zome segment (typi ca ll y we ighing 50-200 g) is mostly used as the propagu le. It ca n be stored in the co ldest months unti l planting. In equator ial climates, however, on ly the uppermost ap ica l tip of the rh izome, w ith the emerging shoot attached, is used. That minimizes the use of sta rchy ti s- sue for repl anting. Although diploid cul tivars fo rm botan ica l seeds, these are never used to propagate the crop . At harvest, after 1 yr of crop duration , to- ta l plant DM in a co ll ect ion of 26 ca nna ac- Figure 1. Ecuadorian accession of edible Canna. Scale indicates 2 m. Figure 2. Variability of canna rhizome between genotypes. Scale = lO cm. cessions averaged 24 t/ha; one access ion achieved total DM accumulation of 54 t/ha or 5.4 kg/m2• Rhizome y ields vari ed consid- erably (17-96 t/ha) with an average of 56 t/ha. Harvest index (HI), the fraction of total DM accounted for by the rhi zomes, was the sec- ond least var iab le character (56 ± 8%), as indicated by the coeffi c ients of vari ation (CV). HI was not correlated with rhi zome y ield (Tab le 1 ). On the other hand, the va ri at ion of total plant DM exp lained 69% of the varia- tion in rhi zome yi eld (ca lcul ated from the corre lation coeffici ents shown in Table 2). Although the content of physica lly extract- able sta rch in the rhizome was low compared w ith other starchy roots--14 ± 4% of rhi zome fresh matter (FM)-ca nna has respectab le starch yields (2.8-14. 3 t/ha). As expected, the starch content of th e rhizome was c lose ly corre lated w ith rhizome DM content. Surpri s- ingly, it was also positively and signifi cant ly corre lated to the content of so luble so lids (Table 2), wh ich varied between 5 and 11 °Brix (Table 1 ). This means that access ions that are high in so luble so l ids also tend to have high starch contents. Starch accounted for on ly 59 ± 14% of rhi zome DM (range 32-88%). The significance of the number of shoots indi cated in Table 1 is that it rough ly equa ls the number of rhizome segments and thus approx imates the number of propagul es that can be derived from a p lant. In other words, it indicates the multipli cation rate, and thi s vari able showed the hi ghest vari ation of the characters (means and standard deviation of 28 ± 16) under study. Interestingly, the num- ber of stems w as inverse ly correlated to HI, meaning th at pl ants w ith a large number of stems (or a higher degree of rhi zome ram ifi - cation) invested re lat ive ly more DM in aboveground plant structures. Spec ifi c leaf weight of canna is compara- tively high; it was the least variabl e character in our eva luat ion (CV=l 3%). Tab le 3 gives data on crop growth, in terms of abso lute and relative DM all ocation, fo r a co mmerc ial CIP Program RepOJ t 1995-96 297 Table 1. Variation of growth and yield characteristics at harvest in a collection of 26 canna accessions, 1996. Variable • Unit Mean Min Max SDb CV{%) Total pla nt DMh (t/ha) 24 8 54 9 39 Rhizome yield (FM) (t/ha) 56 17 96 20 35 Starch yield (t/ha) 7.8 2.8 14.3 3.1 40 Rhizome fraction of total DM (harvest index) % 56 35 74 8 15 Sta rch fraction of total DM % 33 12 54 11 32 Stem fraction of total DM % 19 9 34 7 35 Leaf fraction of total DM % 23 15 46 7 28 Adventitious roots fraction of total DM % 2.5 1.5 5.0 0.9 35 Shoots per plant n 28 13 79 16 57 Specific leaf weight g/dm1 0.44 0.35 0.54 0.06 13 DM content of rhizome % 24 12 31 4 18 Starch content of rhizome (FM basis) % 14 4 22 4 30 Sto rch content of rh izome (DM basis) % 59 32 88 14 24 Soluble solids in rh izome 0Brix 8.1· 5.0 11 .0 1.6 20 a. Starch data refer to physically extractable starch. b. OM = dry matter, FM = fresh matter, SD = standa rd devia tion, OJ = coefficient of variation. Table 2. Pearson's corre lation coefficients between growth characteri st ics of 26 canna accessions, 1996. Variable Total Rhizome Content Starch Stem Harvest plant yield of content fraction index DM0 FM0 soluble of of rhizome solids in rhizome total DM fraction rhizome (FM basis) of total DM Total plant DM -0.32 Number of shoots 0.16 0.68** -0.57** Rhizome yield (FM) 0.831 ** -0.15 -0.26 -0.10 OM content of rhizome 0.66** 0.66** 0.42* Starch content of rhizome (FM basis) 0.55* * * Significant ot P= 0.05, ** significant at P= 0.01. Only meaningful correlations are shown. a. OM = dry matter, FM = fresh matter. 298 Pmgmm6 Table 3. Absolute and relative dry matter allocation and leaf area index of edible canna according to crop age0 • {t/ha) (%) Plant age (months) 6 8 10 12 6 8 10 12 Leaf blades 0.48 b l.76 a 2.16 a 1.86 a 23 16 14 12 Stems, inflo rescences 0.43 ( 2.86 b 4.22 a 3.83 a 21 26 28 24 Dead plant ports 0.16 b 1.1 7 a 1.17 a 1.60 a 8 10 8 10 Rhizomes 0.93 ( 4.77 b 7.02 a bl.83 a 41 43 45 49 Adventitious roots 1.15 b 0.60 a 0.82 a 0.79 a 7 5 5 5 Total plant 2.16 ( 11 .16 b 15.39 a 15.91 a 100 100 100 100 Leaf area index 0.95 b 3.64 a 3.61 a 3.32 a a. Field data token from a commercial plantation in the highlands of Ecuador (2,350 m) . In a row, means followed by the some letter are not statistically different at P < 0.05. b. Corresponding to a rhizome fresh matter yield of 36 t/ho. canna starch crop after 6, 8, 10, and 12 mo of crop duration . DM buildup in canna is slow. Si x months after pl anting, tota l OM ac- cumulation is only 14% of that achieved at commercial harvest 12 mo after pl ant ing. But the proportion of DM allocated to rhi zomes has reached 41 % at 6 mo and increases dur- ing the second half of crop durat ion to only 49%. Likewi se, the fraction of dead pl ant parts vari es little from 6 to 12 mo, although in ab- so lute terms there is a tenfo ld increase. Thi s and the insignifi cant ch ange in leaf area in- dex from month 8 to month 12 illustrate the perennial nature of canna, whi ch does not rea lly become senescent at harvest. The rela- t ive DM all ocation to leaves and stems does not show the dramati c changes most other crops undergo during canopy development. Starch content of the rhi zomes, however, in- creased significantl y from 5% at 6 mo and 7% at 8 mo to 13% at 10 and 12 mo. The nutri ent analysis in rhi zomes showed th at per ton of harvested rhi zomes, or per 120- 130 kg of extracted dry starch, nutrient offtake from the fi eld amounted to 0.75 kg N, 0.74 kg P, 3.24 kg K, 2.48 kg M g, and 0.98 kg Ca. But when extraction residue is returned to the field (the common practi ce in Ecuador), 28% each of N and P, 3.8% of K, 0.6% of M g, and 62.5% of Ca removed by the rhi zomes are returned to the f ield. Most of the K (95%) and Mg (>99%), however, is lost in the washing water during starch extraction. Thus, the net nutri ent offtake from a canna field would be on the order of 0.54 kg N, 0.53 kg P, 3. 11 kg K, 2.47 kg Mg, and 0.37 kg Ca per ton of har- vested rhi zomes or per 1 20-1 30 kg of ex- tracted dry starch. Discussion With rhi zome yi elds averaging 56 t/ha in an international germpl asm collection , edible canna is clea rly a highl y productive root crop that compares well with cassava, sweetpotato, and potato. G iven an average propagul e we ight of 50-200 g, and a pl anting density of 20,000 pl ants/ha, the seed wei ght of canna amounts to only 2.5-10% of a typi ca l y ield (40 t/ ha). For a primitive crop that has never been subjected to systemati c pl ant breeding, ca nna also has a remarkably high HI (56 ±. 8%). Since thi s character shows some geneti c vari ati on, it shou Id be possibl e to breed fo r increased HI. We found only a weakl y nega- CIP Prog mm Repmt 1995-96 299 tive but insignifi ca nt correlation between HI and total plant OM (Table 2); therefore, se- lection for hi gh total OM in breeding does not imply reduced HI as observed in other crops. On the basis of our germplasm eva luation , we suggest that canna cu ltivars for starch ex- ploitation should have large rhi zomes. In- deed, predomin antl y commercial culti vars exp loited for starch have comparatively large rhi zomes. Large rhi zomes are found in plants with a low degree of ramification or a low number of shoots. These, in turn, are inversely related to HI. Canna breeding should also aim at increasing starch content, w hi ch currently accounts for only 59 ± 14% of rhi zome OM. All cultiva rs have significant concentrations of soluble solids (mostly sugars). If genotypes th at polymerize sugars into starch more readily were to be found, the starch produc- tivity of canna could be improved by several percentage points from a current average of 14 ± 4% in rhizome fresh matter. Another concern for breeders wi 11 be to deve lop cultivars with few adventitious roots. These trap soil and compli cate manual har- vest for two reasons. First, soil particles con- tam inate the starch or interfere with its ex- traction. Second, the metabo li ca lly active roots seem to contain high concentrations of oxidases, which are involved in the brown- ing of starch during grating and extraction . In Asia and Lat in America, remova l of these roots accounts for most of the working hours spent during harvest. Simple washing equip- ment, however, such as the revo lving drum or the static paddle washers used in small cassava starch factories in Brazil , is sufficient to remove dirt from canna's adventitious roots. Our crop growth data, from plants of com- mercial cann a plantations harvested after 6, 8, 10, and 12 mo, suggest that the causes for the initial slow pl ant development and result- ing long crop durat ion might be slow leaf area repli cation during the juvenil e phase. Several lines of evidence support this. Leaf area indi- ces remain low well into the second half of crop durat ion. Although the field crop under 300 Progrnm 6 study had accumulated only 14% of its final biomass at 6 mo, OM al located to the rh i- zomes accounted fo r 42% of total pl ant OM. Thu s, even before the crop deve lops signifi- cant leaf area indi ces, it compromises large propo rtions of assimilates in photosyntheti- ca lly unproductive structures (the rhi zomes). High HI early during deve lopment and the peculiar generational development of the rhi- zome, however, provide an opportunity to increase planting density and to harvest ear- li er than is currently done. Further studies are needed to elucidate what effect increased planting densiti es would have on rhizome starch contents, w hi ch seem to remain low during the first 8 mo of crop development in trad itional systems. Our data reveal that canna is a nutrient- effic ient producer of starch, especially w ith regard to the environmentally relevant N and P, of w hich less th an 1 kg is removed per ton of harvested rhi zomes. A significant fraction of these nutrients ca n be returned to f ields by applying either fresh or fermented process res idue as ferti I izer. The data also show that major quantities of K and M g are lost in the water used to wash out starch from grated pulp. It would therefore seem important that the washing water be recycl ed to fields and not discarded in to sewer systems as is currently done. Be- cause the amounts of water used in manu- all y extracting and refining starch are enor- mous (wel I over 1 00 L/ kg dry starch under Andean conditions), locating extraction pl ants above growing sites and distributing wash- ing water by gravity would be a preliminary recommendation . Conclusions In conclusion, yie ld potential, effic ient OM partitioning, nutri ent efficiency, and starch quality make canna an interesting starch crop for resource-poor fa rmers in the tropical high- lands. Howeve r, except for some parts of Southeast Asia, w here the noodle industry has fueled demand for canna starch, this crop 's potential is not being recognized. There is mounting ev idence that canna starch cou ld replace potato starch because of a high de- gree of funct iona l similariti es. Product ex- amples include several types of As ian nood les (for wh ich an increas ing share of potato starch is used), certa in bakery goods, and thermo- pl astics. M any other applications are likely to be found w ith appropr iate resea rch and deve lopment. Canna starch has a potentially vast demand if industri al-grade starch we re to beco me ava il ab le. Current ly, canna starch does not meet the quali ty standards requ ired by most users . Moreover, can na starch producti on is at a disadvantage because of the lack of ap- propriate harvestin g and extract ion equ ip- ment. Even in deve lopi ng economies, the excess ive labo r costs ar ising from manual process ing render canna starch too cost ly to stay competitive versus inexpensi ve sta rch imports. Improvements in the crop's starch productivity must therefore be accompanied by the development of improved extraction technologies . Further resea rch is needed to address spe- cifi c crop constraints, such as the long crop duration; experiments des igned to clarify the effect of planting densities on crop y ields and precoc ity also seem warra nted. Researc h aimed at eva luat ing the feasib ili ty of breed- ing culti va rs w ith improved starch productiv- ity and process ing characteristi cs also de- serves attent ion. Andean ca nna germpl asm has recentl y become ava ilable and superior cul t ivars can likely be identif ied fo r use in areas of narrow genetic va riation of this crop, such as in Asia. CIP hopes that interested na- tiona l programs w ill join this effort. Selected Reading Hermann, M. 1996. Starch nood les from edible canna (Ca nna edulis). In : Jani ck, J. (ed.). Progress in new crops . Am . Soc. Hort. Sci ., Alexandria, VA. p. 507-508. Imai, K. , K. Shimabe, K. Tanaka, and T. Kawana. 1994. Studies on matter product ion of ed ible canna (Can na edulis Ker.). Ill. Changes of production structure with growth. Jap. J. Crop Sci. 63(2) :345-35 1. Ripperton, J.C. 1927. Carbohydrate metabolism and its relat ion to growth in the edible canna. Hawaii Agri c. Exp. Stn. Bu ll . (Wash ington, DC) No. 56. 35 p. CIP Program Report 1995-96 30 l Cross-Program Training Courses 1995-1996 Patricio Malagamba Cl P's Training Department conducted diverse ac ti v iti es in 1995-96, mainl y in vo lv in g courses, workshops, conferences, and sym- pos ia, of genera l coverage of subjects across programs. A symposium on "Current Ap- proaches for Developing Concepts and Prac- tices Re lated to Agri cultural Promot ion and Deve lopment" was organized in co ll abora- tion w ith the Escuela para el Desa rrol lo, an NGO in Peru and member of CONDESAN, fo r other national organizat ions and institu- tions. The symposium's main object ive was to conceptuali ze and define the base line re- quired fo r agricultural development in Peru . Attempts were made to estab li sh the relat ion- ship between promot ion poli c ies and prac- tices w ith princ ipl es related to deve lopment. It also focused on the need to strengthen in- terdi sc iplinary interventions that link differ- ent sectors by proposing directions and de- fining concepts and practi ces invo lved in ag- ri cul tural deve lopment. A course on "Agroecology and Rural De- ve lopment" was organ ized by the Consorcio Latin oa meri ca no de Agroeco log ia y Desa rrollo, CLADES, and directed to sci en- tists and deve lopment agents whose activ i- t ies are related to susta inab le development of the rural sector. CIP sponsored the pa rt ici- pat ion of two nat iona l sc ienti sts from Cusco and Cajamarca . PROINPA spon sored five parti cipants from Bolivi a. The course used di s- tance- lea rning elements and procedures. A " Production Course on Bio log ica l Con- tro l in Potato" was held at CIP for staff from the national organization responsible for san i- tary control of agricultural products, SE NASA. The training program covered biologica l tech- niqu es re lated to th e prod uct io n of entomopathogens (fungus, virus, and bacte- ria) and parasito ids. A two-week "Workshop on Integrated W aters hed Ma nage ment in th e Andea n Ecoreg ion " was organized by loca l deve lop- menta I o rga ni zat ions in Ca j amarca (ASPADERUC and ADEFOR), w ith support from CONDESAN and CIP. Thi s workshop emphasi zed theory and practi ces related to systems analysis, ag roeco logica l zoning, simulation models, and geographic informa- tion systems as too ls for identi fy ing alterna- tives for th e susta in ab le deve lopment of Andean agri culture. Two watersheds were used as case studies, one located on the west side of the Andes (J equetepeque), and the ot her an in te rm o nta ne wate rshed (Cajamarca) . Parti cipants included technica l staff working in different di sc iplines (agricul- turi sts, animal production spec iali sts, ecolo- gists, etc.) as well as poli cymakers (mayors) from both Cajamarca, Peru, and Carchi Prov- ince in Ecuador. A course on f ield des ign, layout, and ad- ministrat ion of fie ld experiments was con- ducted at CIP. It covered five main topics: scientific research methods, experimenta l plot des ign and management, competitive effects, mechanica l errors in conducting experiments, and characterist ics of the major experimen- tal des igns . CIP organi zed a " Pl anning Workshop for Mountain Forum Act ivit ies in Lati n America," sponsored by CONDESAN, to promote the development of po li cy actior.s directed to the equitab le and sustainable deve lopment of mountain regions. Parti cipants were from the Pro-Sierra Nevada Foundation of Colombia, United Nations Development Programme, the Mountain Institu te of Huaraz, ASPADERUC, In te rn at io nal Conse rvat ion , and the Universidad Nacional Agrari a La Molina. CIP P1ogrom Repmt 1995-96 303 A "Workshop on the Appli cation of Crop Simulation Models" was conducted in Cusco, Peru. It was sponsored by CON DESAN in co llaboration w ith the lnstituto de Manejo de Aguas y Medio Ambiente, Centro de Estudios Barto lome de las Casas, and Pl an Meriss lnka. The main objective was to expose participants to crop simulation models as a methodologi- ca l tool fo r use in more effect ive decision making w ith respect to the best use and man- agement of natural resources and to their ap- plication in improving the productivity of those crops included in agricultura l systems prevailing in the Andean ecoreg ion. 304 Training Sponsored by FA O, ll CA, UNEP, CONDESAN, and CIP, a " Regiona l Confer- ence on Sustainab le Agricu ltural Systems in the Central Andes" was conducted. The main object ive of th is confe rence was to eva luate causes and mechanisms related to soc ial and economic deteri orat ion in rural areas as we ll as to assess the poss ible consequences of dif- ferent management strategies. Di fferent strat- egies fo r the sustainable deve lopment of the Andean ecoregion th at could be used to gen- erate technica l ass istance plans in countri es were analyzed and discussed in deta il. Scientific Publications 1995-1996 Abad, Z.G., J.A. Abad, and C. Ochoa. 1995. Histor ica l and sc ientifi c evidence that sup- ports the modern theory of the Peruvian Andes as the ce ntre of orig in of Phytophthora infestans. In: Dowley, L.J ., E. Bannon, L.R . Cooke, T. Keane, and E. O'Sullivan (eds.). Phytophthora infestans 150: Proceedings EAPR Pathology Sec- tion, September 1995, Dublin, Ireland. European Association of Potato Research/ Boole Press, Dublin , Ireland . p. 239-245. Anguiz, R.J. and H.A. Mendoza. 1995. Corre- lation between seed ling and adu lt potato plants for res istance to early blight (A lterna- ria so /ant). Fitopatologia 30(2): 100-106. Arbizu, C. 1995. Agroecologia de la achira en el Peru. Agroenfoque 70:4-5. ASPADERUC/CONDESAN-CIP/Fondo Peru- Canada. 1995. La Encafiada: Caminos hacia la soste nibilidad. Proyecto PIDAE/ Asociaci6n para el Desarro llo Rural de Cajamarca/Consorcio para el Desarroll o Sostenible de la Ecorregi6n Andina-Centro Internaciona l de la Papa/Fondo Contravalor Peru-Canada, Lima, Peru. 11 2 p. Austin , D.F. and Z. Huaman. 1996. A synop- sis of lpomoea (Convol vulaceae) in the Americas. Taxon 45(1 ):3-38. Bamberg, J.B. , Z. Huaman, and R. Hoekstra. 1995. Intern at ional cooperation in potato germplasm. In: R.R. Duncan, D.M. Kral , and M.K. Viney (eds.). Intern atio nal germpl asm transfer: Past and present. Crop. Sci. Soc. of America Special Publi - cat ion No. 23. p. 177-1 82 . Blondet, C. and R.L . Duran. 1996. El impacto potencial de la difusi6n de tecnologia panificadora de camote en panaderias de Villa El Salvador. Documento de Trabajo no. 1996-2, CIP, Lima, Peru. 52 p. Bouis, H.E. and G. Scott. 1996. Demand for high va lue secondary crops in develop- ing countri es: The case of potatoes in Bangladesh and Pakistan. FCND Discus- sion Papers no. 14. International Food Policy Research Inst itute (IFPRI), Food Consumption and Nutrition Di v isio n (FCND), Washington, D.C. 1 USA. 46 p. Bouma, J., A. Kuyvenhoven, B.A.M. Bouman, J.C. Luyten, and H.G. Zandstra (eds.). 1995. Eco-regional approaches for sus- tainable land use and food production. Proceed ings of a symposium held 12-16 Dec. 1994, ISNAR, The Hague, Nether- lands. Kluwer Academic Publishers in cooperation w ith the International Potato Center, Dordrecht, Netherlands. 596 p. Bradshaw, J.E., R.L. Wastie, H.E. Stewart, and G.R. Mackay. 1995. Breeding for res is- tance to late blight in Scotland. In: L.J. Dowley, E. Bannon, L.R. Cooke, T. Keane, and E. O'Sullivan (eds.) . Phytophthora infestans 150. Boole Press, Dublin, Ire- land. p. 246-254. Cabanilla, L.S. 1996. Sweetpotato in the Phil- ippines: Production, processing, and fu- ture prospects. Internat ional Potato Cen- ter, University of the Philippines, Lima, Peru/Manil a, Philippines. 100 p. Canizares, C.A. and G.A. Forbes. 1995. Fo- liage resistance to Phytophthora infestans (Mont.) de Bary in the Ecuadorian national collection of Solanum phureja ssp. phureja Juz. & Buk. Potato Res. 38(1):3-10. Castillo, R. and M. Hermann.1995. Collecting Andean root and tuber crops (excluding CIP Progrom Report 1995·96 305 potatoes) in Ecuador. In: Guarino, L., V. Ramanatha Rao, and R. Reid (eds.) . Col- lecting plant genetic diversity: Technica l guidelines. IPGRl/FAO/IUCN/UNEP. CAB International, Wallingford, UK. p. 639-646. CIP (I nternational Potato Center). 1995. Vines to roots, sweetpotato breeding for impact: Recommendations and abstracts. Interna- tional workshop held 1-3 June 1994, Lima, Peru. CIP, Lima, Peru . 46 p. CIP. 1995 . Manual de producci6n de papa con semi Il a sexual. Fasciculos. Lima, Peru. CIP. 1996. Marchitez bacteriana. Manual de Capacitaci6n. Lima, Peru. CIP. 1996. Bacterial w ilt manual. Lima, Peru. Cl P. 1996 . M anu al de m anej o de germop lasma de batata o ca mote (lpo- moea batatas). Fasciculos. Lima, Peru. CIP. 1996. Accounting fo r seed potato pro- duction costs: A computer spreadsheet- based management tool. Training Manual. Lima, Peru. CIP. 1996. La batata en cifras: Un compendio de info rmaci6n clave y anal isis para 33 importantes paises productores de batata, producc i6n, uso, consumo y alimentac i6n animal. Lima, Peru. v p. CIP. 1996. Late blight: A global initi ative. Lima, Peru . 21 p. CIP. 1996. Major potato diseases, insects, and nematodes. Lima, Peru. 111 p. CIP. 1996. Prin c ip ales enferm eda des, nematodos e insectos de la papa. Lim a, Peru. 111 p. CIP. 1996. Sweetpotato facts: A compendium of key figures and an alys is fo r 33 impor- ta nt sweetpotato-prod ucing countries, production, use, consumption, feed use. Lima, Peru. v p. 306 Publications CIP/COTESU (Centro Internacional de la Papa/ Cooperaci6n Tecnica Suiza) . 1995. Programa Colaborativo Biodiversidad de Raices y Tuberculos Andinos. Memorias 1993-1994. CIP/COTESU, Lima, Peru. 322 p. CON DESAN (Consortium fo r the Sustainable Deve lopment of the Andean Ecoregion) . 1995. A post-UNCED progress report on Andean natural resource research. CIP, Lima, Peru . 16 p. Cottrell , J.E., C.M. Duffus, L. Patterson, and G.R. Mackay. 1995. Properties of potato starch: Effects of genotype and growing conditions. Phytochemistry 40(4):1057-1 064. Cri ssman, C. and A. Hibon. 1996. Establish- ing seed potato prices: Concepts, proce- dures and implications for research and training. Social Sci ence Department. Working Paper Seri es no. 1996-1, CIP, Lima, Peru . 24 p. Chavez, R. , M. Upadhya, H . Mendoza, J. Esp inoza, R. Cabello, P. Siles, G. Bolio, T. Melendez, J. Eyzaguirre, C. Maquera, E. Guevara, K. Monasterio. 1996. Nuevas hi p6tes is y avances de se lecci6n en el mejoramiento genetico de papa y camote para adaptac i6n a sue los aridos y sa linos. Rev. Consejo de lnvestigaci6n (COIN), Ciencia y Desarrollo, Univ. Jorge Basadre Grohmann (Peru). p. 32 -54. Chujoy, E. 1995. Root crops germpl asm re- search in Vietnam. ( IP/ National Institu te of Agricul tural Sciences (I NSA)/ lnterna- tion al Deve lopment Resea rch Centre (IDRC), Manila, Philippines. 62 p. Dayal , T.R., G.J. Scott, G.T. Kurup, and C. Balagopalan (eds. ). 1996. Sweetpotato in South Asia: Postharvest handl ing, pro- cessi ng, storage, and use (Proceed ings of a workshop). Tri vandrum, India, 9-13 Sept. 1991. Central Tuber Crops Research In- st itute (CTCRI ), CIP, Lima, Peru. 137 p. Diaz, J., P. Schmiediche, and D.F. Austin . 1996. Polygon of crossability between eleve n spec ies of lpom oea: Section Batatas (Co nvo lvul aceae). Euphytica 88(3):189-200. Engels, C., J. Schwenke!, R. El Bedewy, and B. Sattelmacher. 1995. Effect of the de- velopment stage of potato seed Ii ngs on recovery after transp lant ing to the field and o n tuber y i e ld . J. Agri c . Sci. 124(2):2 13-2 18. Estrada, R.D. 1996. Analisis ex post de la influencia de algunas var iab les macro- econ6m icas sob re la co mpetitividad agropecuar ia, equ idad y conservac i6n de recursos. In : Rivera, B. and R. Aubad (eds.). IESA America Lat in a 1995: El enfoque de sistemas de producci6n y la inco rporac i6n de criter ios de politica: Memori as 2. Simposio Lat inoameri cano sob re ln vest igac i6 n y Exte nsion en Sistemas Agropecuarios. Santa Fe de Bogo ta, Co lo mbi a, 7-9 N ov. 1995. Corporaci6n Colombiana de lnvest igaci6n Agropecuari a (CORPO ICA), Bogota, Co- lombia. p. 219 -230. Fano, H. , 0 . O rtiz, and T. Walker. 1996. Peru: Inter-inst ituti onal cooperat ion for IPM. In : Thrupp, L.A. (ed.). New partnerships fo r susta inab le agr iculture. World Resources Inst itute (WRI), Washington , D.C., USA. p. 85-98. FAO (Ofici na Regional para America Latina y el Ca ribe). 1995. Manual tecni co de producc i6n de papa a partir de sem ill a sexual: Curso audi ov isua l. FAQ/Centro I ntern ac i ona I de I a Papa/I nst it uto de ln ve.st igac iones Agropecuarias (INIA), Santiago, Chil e. 72 p. FAO/C IP (Food and Agri culture Organization/ International Potato Center). 1995. Potatoes in the 1990s: Situation and prospects of the world potato economy./ La papa en la decada de 1990: Situaci6n y perspectivas de la economia de la papa a nivel mundial./ La pomme de terre dans les annees 90: Situ- ation actuel le et perspectives de I' economie mondiale de la pomme de terre. Rome, Italy. 39 p. , 50 p. , 44 p. French, E.R. and G.R. Mackay (eds .). 1996. Enhancing the globa l late blight network: Report of the project design meeting on the Globa l Initiative on Late Blight. CIP, Lima, Peru. 43 p. Fuentes, S., M.A. Mayo, C.A. Jolly, M. Q uerci, and L.F. Sa lazar. 1996. A novel lu teov irus from sweet potato, sweet potato leaf speck lin g v iru s. Ann. Appl. Bio l. 128(3):491-504. Gibson, R.W. , R.O.M. Mwanga, S. Kasul e, S. Fuentes, and E.E. Carey. 1996. Sweet potato feathery mott le and other viruses in Uganda. In : Proceed ings of the regional sweetpotato workshop. Commission of the European Union, Technical Agricultural Center (TAC), Libreville, Gabon. p. 87-90. Gitomer, C.S. 1 996. Potato and sweetpotato in China: Systems, constraints, and poten- tial. CIP/ Chinese Academy of Agricultural Sciences (CAAS), Lima, Peru . 183 p. Golmirzai e, A.M. and L.F. Salazar. 1996. Ap- plication of biotechnology for conserva- tion and uti I izat ion of Andean root and tuber crops and biosafety guidelines. Af- ri can Crop Sci . J. 3(3):341-344. Gregory, P. 1996. Global program to develop late blight res ista nt potato cul t ivars . CGIAR News 3(3) :1. Gutarra, L. , T. lcochea, and E.R. French. 1995 . Etiologia de pud ri c iones bacteri anas de tuberculos de ulluco (Ullucus tuberosus Caldas). Fitopatologia 30(2):92-95 . Hagenimana, V., R.E . Simard, and L.P. Vez ina. 1996. Method for the hydrolys is of starchy mate ri a Is by sweetpotato endogenous amylases. U.S. Patent no. 5,525, 154. June 11, 1996. 11 p. Hardy, B. and E.R. French (eds. ). 1995. Inte- grated management of bacterial wi It: Pro- ceed ings of an international workshop held in New Delhi , India, 11 -16 Oct. 1993. Internat iona l Potato Center (C IP)/ CIP Program Report 1995·96 307 Indian Council of Agricultural Research, Lima, Peru. 196 p. Hardy, B., P. Malagamba, and C. Martin (eds.). 1995. True potato seed in the Middle East and Africa: Proceedings of an international workshop held in Cairo, Egypt, 9-15 Apr. 1994. International Potato Center (CIP), Lima, Peru. 97 p. Hijmans, R.J. and M.K. van lttersum. 1996. Aggregation of spatial units in linear pro- gramming models to exp lore land use options. Neth. J. Agric. Sci. 44:145-162. Hoekstra, R. and Z. Huaman. 1995 . CIP pro- motes wider use of root and tuber crops. PROPHYTA Annual 1995. p. 102-103. Holle, M. 1996. La biodiversidad es una de las bases del desarrollo sustentable. In: Rivera, B. and R. Aubad (eds.) . p. 77-78.* Huaman, Z. 1995. Tecnicas citologicas para determinar el numero cromosomico y la fertilidad de las papas. Guia de lnvestigacion CIP no. 10. CIP, Lima, Peru. 18 p. Huaman, Z. , F. de la Puente, and C. Arbizu. 1995. Collecting vegetatively propagated crops (especially roots and tubers). In : Guarino, L. , V. Ramanatha Rao, and R. Reid (eds.). Collecting plant genetic diversity: Technical guidelines. IPGRI/ FAO/IUCN/UNEP. CAB International , Wallingford, UK. p. 457-466. lnstituto Boliviano de Tecnologia Agropecuaria (IBTA). 1995. Primera reunion nacional de recursos geneticos de papa, raices y tuberculos andinos: Memorias. Cochabamba, Bolivia, 7-10 Feb. 1994. I PG Rl /C I P/11 CA/PROCISU R/COTE SU / PNUD, Cochabamba, Bolivia. 341 p. Jacobsen, S.-E. , J. Hill , and 0. Stollen. 1996. Sta bi I ity of quantitative traits in qui noa (Chenopodium quinoa). Th eo r. Appl. Genet. 93:110-116. Khatana, V.S. , M.D. Upadhya, A. Chi Iver, and C.C. Crissman. 1996. Economic impact of 308 Publications true potato seed on potato production in eastern and northeastern Indi a. CIP, Lima, Peru. 19 p. Korva, J.T. and G.A. Forbes. 1995. A simple non-destructive technique for healthy and diseased leaf area assessment of potato plants. Potato Res. 38(1):31 -37. Kreul , W. and B. Trognitz .1995. Forschungss- chwerpunkte des lnternat ionalen Kartoffe l Forsc hungszentrums - CIP - in Peru . Entwicklung + Landli c he r Raum (Germany) 29(5):30-32. Landeo, J.A. , M. Gastelo, H. Pinedo, and F. Flores. 1995. Breeding for horizonta l res istance to late blight in potato free of R genes. In: Dawley, L.J ., E. Bannon, L.R . Cooke, T. Keane, and E. O'Sullivan (eds. ). Phytophthora infestans 150: Proceedings EAPR Pathology Section, September 1995, Dublin, Ireland. European Association of Potato Research/ Boole Press, Dublin , Ireland . p. 268-274. Martin , C., P. Demo, D. Njualem, T. Gass, and C. Ntonifor. 1995. Late blight resis- tant va rieties for Cameroon. Am. Potato J. 72 (9):513-522. Martin, C., P. Demo, T. Gass, V. Fondong, and J. Koi. 1995. Development of a seed pro- duction system from in vitro in Cameroon: Experiences from the first two years. Am . Potato J. 72(5):299-302. Martinez, C.A., C. Guerrero, and U. Moreno. 1995. Diurnal fluctuat ions of carbon ex- change rate, praline content, and osmotic potential in two water-stressed potato hy- brids. Rev. Bras. Fisiol. Veg. 7(1 ):27-33. Ma zo n Ortiz , N. , R. Castillo Torres , M . Hermann , and P. Espinosa. 1996. La za nahoria blan ca o arracacha (Arracacia xanthorrhiza Bancroft) en Ecuador. Publicacion Miscelanea no. 67. lnstituto Nacional Autonomo de I nvestigacion Agropecuaria (IN IAP)/ Dept. de Recursos Fitogeneticos y Biotecnologia (DENAREF)/CIP, Quito, Ecuador. 41 p. *See the entry for Estrada, R.D., for detail s on this publication. Mendoza, H.A, E.J. Mihovilovich, and F. Saguma. 1996. Identification of triplex (YYYy) potato virus Y (PVY) immune progenitors derived from Solanum tuberosum ssp. andigena. Am . Potato J. 73(1 ):13-19. Mujica B., E. and J.L. Rueda S. 1996. El desarrollo sostenible de montaf\as en America Latina: Actas de la segunda consulta regional intergubernamental sobre el desarrollo sostenible de montanas. CONDESAN/CIP/FAO, Lima, Peru. 318 p. Pallais, N. 1995. High temperature and low moisture reduce the storage requirement of freshly harvested true potato seed. J. Am. Soc. Hort. Sci. 120(4):699-702. Pallais, N. 1995. Storage factors control ger- mination and seedling establishment of freshly harvested true potato seed. Am. Po- tato J. 2(7):427-436. Pallais, N., J. Santos-Rojas, and R. Falcon. 1996. Storage temperature affects sexual potato seed dormancy. HortScience 31(1):99-101. Prain, G.D., l.G. Mok, T. Sawor, P. Chadikun, E. Atmodjo, and E. Relwaty Sitmorang. 1995. Interdisciplinary collecting of lpo- moea batatas germplasm and associated indigenous knowledge in lrian Jaya. In : Guarino, L., V. Ramanatha Rao, and R. Reid (eds.). Collecting plant genetic diversity: Technical guidelines. IPGRI/ FAO/IUCN/UNEP, CAB International, Wallingford, UK. p. 695-711. Programa Colaborativo Biodiversidad de Raices y Tuberculos Andinos . 1996. Memorias 1994-1995. CIP/Cooperacion Teen ica Su iza (COTESU)/Consorcio para el Desarrollo Sostenible de la Ecorregion Andina (CON DESAN), Lima, Peru . 370 p. Querci, M., D.C. Baulcombe, R.W. Goldbach, and L. F. Salazar. 1995. Analysis of the re- sistance-breaking determinants of potato virus X (PVX) strain HB on different po- tato genotypes expressing extreme resis- tance to PVX. Phytopathology 85(9):1003- 1010. Quiroga, J., G. Watson, H. Cardozo, 0. Navia, R. Casso, and R. Andrew. 1995. Diagnostico interdisciplinario del cultivo de la papa en la zona de lscayachi, Tarija (campana 1991-1992) . Documento de Trabajo no. 1/95. lnstituto Boliviano de Tecnologia Agropecuaria (I BTA)/ Program a de lnvestigacion de la Papa (PROINPA), Cochabamba, Bolivia. 35 p. Quiros, C.F., A. Epperson, J. Hu, and M. Holle. 1996. Physiological studies and determi- nation of chromosome number in maca, Lepidium meyenii (Brassicaceae). Econ. Bot. 50(2):21 6-223. Quiroz, R., C. Leon Velarde, B. Arce, and D. Genin. 1996. Resultados de ensayos en ganaderia con modelos de simulacion. In: Rivera, B. and R. Au bad (eds.). p. 113-123. * Quiroz, R.A., R.D. Estrada, C. Leon Velarde, and H.G. Zandstra. 1995. Facing the chal- lenge of the Andean zone: The role of modeling in developing sustainable man- agement of natural resources. In: Bouma, J., A. Kuyvenhoven, B.A.M. Bouman, J.C. Luyten, and H.G. Zandstra. Eco-regional approaches for sustainable land use and food production: Proceedings of a sym- posium held 12-16 Dec. 1994, ISNAR, The Hague, Netherlands. Kluwer Aca- demic Publishers in cooperation with the International Potato Center, Dordrecht, Netherlands. p. 13-31. Rivera, J.J. and R.D. Estrada. 1996 . Cuantificacion ex ante del intercambio en- tre equidad, productividad y sostenibilidad para el disef\o de alternativas tecnologicas: El caso del cultivo de arracacha en Colom- bia. In: Rivera, B. and R. Aubad (eds.). p. 101-112.* Rodriguez Quijano, P. 1996. La papa y el desarrollo economico en Colombia. CIP/ Corporacion Colombiana de lnvestigaci6n Agropecuaria, Lima, Peru. 114 p. CIP Program Report 1995-96 309 Rousselle-B ou rgeois, F. and S. Pri ou. 1995. Screening tuber-bearing Solanum spp. for resi stance to soft rot caused by Erwinia carotovora ssp. atroseptica (van Hall ) Dye. Potato Res. 38(2): 111-11 8. Salazar, L.F. 1995. Los virus de la papa y su control. CIP, Lima, Peru . 226 p. Sal aza r, L.F. 1996. Aplicaci6n de la microscopia electr6nica en la v iro logia vegetal . In : Castillo de Maruenda, E., N. Rojas Moran, and A. Lopez Mill a (eds.). Microscopia e lectr6ni ca en e l Peru. Consejo Nacional de Ciencia y Tecno logia (CONCYTEC), Lima, Peru . p. 43-48. Salazar, L.F. 1996 . Potato v iruses and their control. Lima, Peru. 214 p. Salazar, L.F., M . Querci, I. Bartolini , and V. Lazarte. 1995 . Aphid transmission of po- tato spindle tuber viroid assisted by potato leafroll virus. Fitopatologia 30(1 ):56-58. Sanchez, I. , M . Milian, A. Morales, and F. de la Puente. 1995 . Prospecci 6n de recursos geneticos de raices y tuberculos tropicales en Cuba en 1991. Cultivos Tropicales 16(1 ):84-87. Sanchez, I. , M. Milian, A. Morales, and F. de la Puente. 1995 . Prospecci6n de recursos geneticos de raices y tubercu los tropicales en Cuba en 1992. Cultivos Tropi ca les 16(1 ):88-95. Scott, G.J. (ed.). 1995. Prices, products, and peopl e: An alyzing agricultural markets in developing countries. Lynne Ri enner Pub- lishers in cooperation with the Interna- t ional Potato Center (CIP), Boulder, Colo- rado, USA. 493 p. Scott, G.J . 1996 . Paradoxes and projections for potatoes in the 1990s. Entwicklung + Landlicher Raum. 30(3):20-22 . Seminario Cunya, J. and A. Seminario Cunya. 1995. Co lecci6n reg io na l de germoplasma de ra ices andinas. Universid ad Nacional de Cajamarca 310 Publ ications (UNC)/Programa Bi od iversidad de Raices y Tuberculos Andinos Convenio CIP- COTESU, Cajamarca, Peru . 32 p. Smit, N.E.J.M. and L.O. Matengo. 1995. Farm- ers' cultural practi ces and their effects on pest control in sweetpotato in South Nyanza, Kenya . Int. J. Pest Manage. 41 (1):2-7. Thiele, G., A. Devaux, and C. Sori a. 1996. lnnovac i6n tecnol6gica en la papa: De la oferta de tecnologia al impacto mac ro- econ6mico. In : Rivera, B. and R. Aubad (eds.). p. 79-88. * Thro , A.M., R.N . Beachy, M . Bonierbale, C. Fauquet, G. Henry, G.G. Henshaw, M.A. Hugh es, K. Kawa no , C.J.J.M. Raemakers, W. Roca, C. Schopke, N. Tay- lor, and R.G.F. Visser. 1996. Internat ional research on biochemistry of cassava (tap i- oca, Manihot escu lenta Crantz) and its rel- evance to Southeast As ian economies: A Cassava Biotechnology Network review. Asian J. Trap. Biol. 2(1 ):1-30. Torres, H., M.A. Pacheco, and E.R. French. 1995. Resistan ce of potato to powdery scab (Spongospora subterranea) under Andean field condit ions. Am. Potato J. 72(6):355-363. Trognitz, B.R. 1995. Analysis of poll en tube growth in situ to invest igate self-incom- patibi I ity in the w ild potato So/anum commersonii. Euphyt ica 86(2): 149-1 56 . Trognitz, B.R. 1995. Female fertility of potato (Solanum tuberosum ssp. tuberosum) dihaploids. Euphyti ca 81 (1):27-33. Upadhya, M. 1996. True potato seed tech- no logy to fight potato blight in develop- ing countries. Diversity 12(3) :6 7-68. Upad hya, M.D. , B. H ardy, P.C. Gaur, and S.G. 1 langanti I eke. 1996. Production and uti lization of true potato seed in Asia. CIP/ Indian Council of Agricultural Research, Lima, Peru. 233 p. Valdivia, R. , E. Huallpa, V. Choquehuanca, and M . Ho lle. 1996. Monitoring potato and Oxalis va riet ies in mixtures grown on farm famil y fields in the Titicaca Lake ba- sin, Peru, 1990-95. In : Eyzaguirre, P. and M. lwanaga (eds.). Participato ry plant breed ing: Proceed ings of a workshop . Wageningen, Neth erl and s, 26-29 Jul y 1995. Intern ati onal Plant Geneti c Re- sources In st itute (IPGRI ), Rom e, Italy. p. 144-1 50. Valkonen, J.P.T. , M . Orrillo , S.A. Sl ac k, R.L. Pl aisted, and K.N. Watanabe. 1995. Res istance to viruses in F1 hybrids pro- duced by direct cross ing between diploid Solanum seri es Tuberosa and diploid 5. brevide ns (series Etuberosa) using 5. phureja for rescue pollination. Plant Breed. 114:42 1-426 . Van den Berg, J.H. , E.E . Ewing, R.L. Plaisted, S. McMurry, and M .W. Bonierbale. 1996. QTL analys is of potato tuber dormancy. Theor. Appl. Genet. 93(2):3 17-324. Van den Berg, J.H. , E.E . Ewing, R.L. Plaisted, S. McMurry, and M.W. Bonierbale. 1996. QTL analys is of potato tuberi zation. Theor. Appl. Genet. 93(3):307-3 16. Velit Tamanaja, E. 1996 . ldentifi cac i6n de virus en papa (So lanum tuberosum). In: Castillo de Maruenda, E. , N. Roj as Moran, and A. Lopez Milla (eds.). Microscopia electr6nica en el Peru. Consejo Naciona l de Ciencia y Tecnologia (CONCYTEC), Lima, Peru . p. 75-80. Walker, T.S. and C.C. Cri ssman. 1996. Case studies of the economic impact of CIP-re- lated technology. Lima, Peru. 157 p. Walker, T.S. and C.C. Cris sman. 1996. Estudios de casos del impacto econ6m ico de la tecnologia relac ionada con el CIP en el Peru. Lima, Peru . 32 p. Watanabe, K., M . Orrillo, and A.M. Golmirzaie. 1995. Potato germplasm enhancement for resistance to biotic stresses at CIP: Conven- tion al and biotechno logy-ass isted ap- proaches using a wide range of Solanum species. Euphytica 85:457-464. Watanabe, K.N., M . Orr illo, A.M. Golmirza ie, S. Perez, J. Crusado, and J.A. Watanabe. 1996 . Generation of pest res istance: Dip- loid potato germplasm w ith short day ad- aptation from diverse genet ic stocks. Breeding Sci. 46: 329-336. Watanabe, K.N ., M. Orrillo, S. Vega, J.P.T. Valkonen, E. Pehu, A. Hurtado, and S.D. Tanksley. 1995. Overcoming cross- ing barri ers between nontuber-bea ri ng and tuber-bear ing Solanum spec ies: To- wards potato germpl asm enhancement with a broad spectrum of so lanaceous geneti c resources. Genome 38:2 7-35. Watan abe, K.N ., M . Orrillo, S. Vega, M . lwanaga, R. Ortiz, R. Freyre, G. Yerk, S.J. Peloquin , and K. lshiki. 1995. Selec- tion of diploid potato c lones from dip- lo id (hap loid X w ild species) F1 hybrid families fo r short day conditi ons. Breed. Sci. 4 5:3 41-347. Watanabe, K.N .. J.P.T. Valkonen, and P. Gre- gory. 1995. Use of plant biotechnology tool s in plant protection, genetic resources management and crop geneti c improve- ment: An interdi sc ip linary approach with potatoes at the Intern ational Potato Cen- ter. In : Altman, D.K. and K. N. Watanabe (eds.) . Plant biotechnology transfer to de- veloping countries. Academic Press Inc., Texas. p. 179-188. Wh eat ley, C., G.J. Scott, R. Best, and S. Wiersema. 1995 . Adding va lue to root and tuber crops: A manual on product deve lopment. Centro Intern ac ional de Agricultura Tropical (CIAT), Ca li , Co lom- bia . 166 p. Yencho, G.C., M.W. Bonierbale, W.M. Tingey, R.L . Plaisted, and S.D. Tanksley. 1996 . Molecular markers locate genes fo r res is- tan ce to th e Colorado potato beetle (Leptinotarsa decemlineata) in hyb rid Solanum tuberosum x S. berthaultii po- tato proge ni es. Entomol. Exp. Appl. 81:141-154 . CIP Prog ram Report 1995-96 3 1 1 Zandstra, H .G . 1996. The CGIAR response to new chall enges in agroenv ironrnental resea rch in deve loping co untri es . In : Rivera, B. and R. Aubad (eds.). p. 335-342. * Zandstra, H .G. 1996. Twenty-five years of in- ternational potato research, a retrospec- 3 12 Publications tive and fo rwa rd look. Potato Res . 39(4):395 -401 . Zhang, D., W.W. Collins, and M. Andrade. 1995. Estimation of genetic variance of sta rc h digestibility in sweet potato. HortSc ience 30(2):348-349. CIP Board Members 1995-1996 Dr. Lindsay lnnes1 (Chairperson) Deputy Director Scott ish Crop Research Institute Dundee, Scotland Mrs. Martha ter Kuile (Chairperson) Ontario, Canada Dr. Hubert Zandstra Director Genera l Internation al Potato Center Lima, Peru Dr. Setijati Sastrapradja Indones ian Institute of Sciences National Centre for Research in Biotechno logy Bogor, Indonesia Dr. K. L. Chadha (Vice Chair) Deputy Direc tor General (Hort.) Indian Council of Agricultural Resea rch New Delhi, India Dr. Klaus Raven Universidad Naciona l Agraria La Molina, Lima, Peru Dr. Durward Bateman Dean, Coll ege of Agricultural and Li fe Sciences North Carolina State University Raleigh, N .C., USA Dr. Toshihiro Kajiwara1 Director, Japan Plant Protection Assoc iation Tokyo, Japan 1 Left in 1995. 2 Joined in 1996. Dr. Moise Mensah Cotonou, Benin Dr. Lieselotte Schilde Institute of Plant Biochem istry University of Tubingen Tubingen, Germany Dr. Koenraad Verhoeff Wageningen, The Netherlands Dr. Chukichi Kaneda2 Assoc iat ion for International Coopera- tion of Agri culture & Forestry Tokyo, Japan Dr. Vo-Tong Xuan 2 Vice Rector & Director Mekong Delta Farming Systems R & D Inst itute University of Cantho Cantho, Vietnam Dr. Adrian Fajardo-Christen Ministerio de la Pres idencia Lima, Peru Dr. Alfonso Cerrate1 Executive Director lnst ituto Nac ional de lnvestigaci6n Agrari a Lima, Peru (IP Progrom Report 1995-96 3 1 3 CIP Staff in 1995-1996 Directors Hubert Zandstra, PhD, Director General Jose Valle-Riestra, PhD, Deputy Director Gen- eral for Finance and Administration Peter Gregory, PhD, Deputy Director Gen- era l for Research Roger Cortbaoui, PhD, Director for Interna- tional Cooperation Edward French, PhD, Associ ate Director for Research (since May 1996) George Mackay, MS, Director of Genetic Re- sources4 Program Leaders Production Systems Thomas S. Walker, PhD Germplasm Management and Enhancement Ali Go lmirzaie, PhD Disease Management Edward French, PhD (unti l April 1996) Luis F. Sa lazar, PhD (s ince May 1996) Integrated Pest Management Fausto Cisneros, PhD Propagation, Crop Management Mahesh Upadhya, PhD Postharvest Management, Marketing Gregory J. Scott, PhD 1 Staff w ho jo ined in 1995 2 Staff w ho joined in 1996 3 Staff w ho left in 199 5 4 Staff w ho left in 1996 5 Sta ff funded by spec ial projects 6 Project leader 7 Died in 1995 314 Staff International Cooperation (country) =post location, but acti vity regional in scope country = post locat ion Latin America and the Caribbean (LAC) Fernando Ezeta, PhD, Regional Representative (Peru) Liaison Office-Ecuador Char les Cr issman, PhD Sub-Saharan Africa (SSA) Peter Ewe I I, PhD, Regional Representative (Kenya) Liaison Office-Nigeria Humberto Mendoza, PhD6 Middle East and North Africa (MENA) Az iz Lagnaoui , PhD, Regional Representative (Tunisia) (from June 1995) Carl os Martin , PhD, Region al Representative (Tunisia) (until June 1995) Liaison Office-Egypt Ramzy El-Bedewy, PhD South and West Asia (SWA) Sarathchandra llangantil eke, PhD, Regional Representative (India) East and Southeast Asia and the Pacific (ESEAP) Peter Schmiediche, PhD, Reg ional Represen- tati ve (Indonesia) Liaison Office-People's Republic of China Song Bo Fu , PhD Liaison Office-Philippines Gordon Prain, PhD Internationally Recruited Staff Departments Breeding and Genetics Juan Landeo, PhD, Acting Head of Depart- ment, Breeder6 Edward Carey, PhD, Breeder (Kenya)6 Enrique Chujoy, PhD, Geneticist (Indonesia) Ha il e M. Kidane-Mariam , PhD, Breeder (Kenya)6 Humberto Mendoza, PhD, Geneticist (Nige- ria)6 II Gin Mok, PhD, Breeder (Indonesia) Genetic Resources Ali Golmirzaie, PhD, Genet ic ist, Head of Department6 Carlos Arbizu, PhD, ARTC Consultants Fermin de la Puente, PhD, Germplasm Col- lector4 Marc Ghislain, PhD, Molecular Biologist6 Michael He rmann, PhD, Andean Crop Spe- cialist, Ecuador" Z6simo Hu aman, PhD, Germplasm Curator6 Koshun lshiki, PhD, Associate Expert, Ecua- dor4·5 Carlos Ochoa, MS, Taxonomist, Consultant Peter Schmiediche, PhD, Breeder (Indonesia) Bodo Trognitz, PhD, Genetic ist6 Kazuo Watanabe, PhD, Cytogeneticist (USA)4 Dapeng Zhang, PhD, Breeder6 Nematology and Entomology Fausto Cisneros, PhD, Entomolog ist, Head of Department6 Ann Braun, PhD, Ecologi st (/ndo nes ia)6 Aziz Lagnaoui, PhD, Entomo logist (Tunisia) Ni co le Smit, MS, Associ ate Expert, Ugandas Jea n Loui s Zeddam, PhD, Entomovirologist Pathology Luis Salazar, PhD, Virologist, Head of Depart- ment6 John Elphinstone, PhD, Adjunct Scientist, UK Gregory A. Forbes, PhD, Pl ant Path o logist (Ecuador) 6 Edward R. French, PhD, Bacteriologist6 Teresa lcochea, PhD, Pathologist, Consultant6 Upa li Jayasinghe, PhD, Virologist (Philip- pines)6 Rebecca Nelson, PhD, Molecular Pathologist2 Pedro Oyarzun, PhD, Mycologist (Ecuador)s Sylvie Pri ou, PhD, Bacteriologist5 Maddalena Querci, PhD, Molecular Virologist6 Lod J. Turkensteen, PhD, Adjunct Scientist, Neth erlands Physiology Mahes h Upadhya, PhD, Breeder, Head of Department6 Catherine Brabet, PhD, Associate Expert2 Ramzy El-Bedewy, PhD, Breeder, Egypt Vital Hage nim a na, PhD , Food Scientist (Kenya)5 Oscar Hidalgo, PhD, Senior Seed Spec ialist6 Sarath c ha ndr a ll a ngantileke, PhD, Postharvest Specialist (/ndia) 6 Jukka Korva, MS, Fe llowship Agron omi st, Ecuador4·s Noel Pa ll a is, PhD, Physiologist6 Marian Van Ha l, MS, Associate Expert1 Christopher Wheatley, PhD, Postharvest Spe- cia li st (/ndo nesia)6 Social Science Thomas Wa lke r, PhD, Economist, Head of Department6 Johan Brons, MS, Assoc iate Expert (Philip- pines)2·s Alwyn Chilver, MS, Associate Expert, lndo - nesias Charles Crissman, PhD, Economist (Ecuador)6 Peter Ewe ll , PhD, Economist (Kenya) 6 Robert Jan Hijmans, MS, Associate Experts Jan Low, PhD, Economist (Kenya) 4 ·s Dai Peters, PhD, Rural Sociologist (/ndone- sia)2 Gordon Prain, PhD, Anthropologist (Philip- pines)6 Gregory Scott, PhD, Economist6 Juli a Wright, MS, Associate Expert1•4 Training Patric io Malagamba, PhD, Head of Depart- ment Communications Edward Sulzberger, MS, Acting Head of Unit (from Sept. 1996) Bill Hardy, PhD, Engli sh and Spanish Write r/ Editor, Acting Head of Unit (until Sept. 1996) CIP Program Report l 995-96 31 5 Michel L.Smith, Journalist, Head of Depart- ment4 Information Technology Anthony Collins, Heads Directors' Offices Office of the Director General Edward Sulzberger, MS, Senior Adviser Office of the Deputy Director General for Research Jose Lui s Rueda, PhD, Coordinator, Andean Natural Resources Office of the Deputy Director General for Finance and Administration William A. Hamann, BS, Assistant to the DDGF&A3 Special Country Projects FORTIPAPA, Ecuador Alberic Hibon , PhD , Econ o mi st, Team Leader4,s Corinne Fankhauser, MS, Associate Expert1 PROINPA, Bolivia Andre Devaux, PhD, Seed Spec ialist, Team Leaders Enrique Fern andez-Northcote, PhD, Virolo- gists Javier Franco, PhD, Nematolog ists,6 Graham P. Thi ele, PhD, Technology Transfer Specialists Uganda Nico le Smit, MS, Associate Experts Consortium CONDESAN Robert Jan Hijmans, Ir, Associate Experts Miguel Holle, PhD, Biodiversity of Andean Cropss,6 Carlos Leon-Velarde, PhD, Animal Produc- tion Systemss Osvaldo Paladin es, PhD, Andean Pastures (Ecuador)s Roberto Quiroz, PhD, Land Use Systems Spe- c iali sts 316 Stoff Joint Appointments with Other Institutions Walter Bowen, PhD, Nutrient Cycling Spe- cialist, IFDC Ruben Dario Estrada, MS, Natural Resources Economi cs (Co lombia) , CIATs Networks ASP RAD Eufemio T. Rasco Jr. , PhD, Coord inator (Phil- ippines)s UPWARD Gordon Prain, PhD, Coordinator (Philippines) Controller's Office Carlos Nino-Neira, CPA, Controller Office of the Executive Officer Cesar Vittorelli , Ing. Agr. , Acting Executi ve Officer Nationally Recruited Staff Departments Breeding and Genetics Walter Amoros, MS, Agronomist Miguel Ato, Food Industries Specialist3 Raul Anguiz, MS, Agronomist4 Luis Calua, MS, Agronomist4 T.R. Daya l, PhD, Assoc iate Expert (India) Luis Diaz, MS, Agronomist Jorge Espinoza, MS, Agronomist Manuel Gastelo, MS, Agronomist Shimon Gichuki, MSc, Breeder, Kenya Hugo Gonza les, Ing. Agr. , Agronomist, Chile Pamela Jean Lopez, MS, Breeder, Philippines Eli sa Mihovilovich, MS, Biologist Daniel Reynoso, MS, Agronomist K.C. Thakur, PhD, Breeder, India Genetic Resources Cesar A. Aguilar, Agronomist, Huancayo Victor H. Asmat, Biologist Milciades A. Baltazar, Agronomist, San Ram6n 3 Jorge Benav ides, Bi o logist Fausto Buitron, Agronomi st4 Patrici a G. Cipriani, Biologists Walberto M. Eslava, Agronomist Rene A. Gomez, Agronomist Maria del Rosa rio Herrera, Biolog ist Ana M . Hurtado, Biolog ists Lui s H. Nopo, Biolog ists Maria Gise ll a Orjeda, PhD, Bi o logist4 Matilde O rrill o, Bio logist Ana Luz Panta, Bio logist Flor de M aria Rodriguez, Biologists Alberto Salas, Agronomist Roxa na Salinas, Agronom ist4 Jorge Tenorio, Bio log ist Judith To ledo, Biologist2 Fanny Vargas, Agronomist Nematology and Entomology Jesus Alcazar, MS, Assoc. Agronomist6 Juan Cabrera, MS, Agronomist2 Manuel Canto, PhD, Nematologists·6 Veroni ca Canedo, Biologist Jav ier Ca rhu amaca, Ing. Agr.s Wilfredo Cata lan, Ing. Agr. 4 Victor Cerna, Ing. Agr. s Roberto Delgado de la Fl or, Agronomist3 Oder Fabian, Ing. Agr.4 Erwin Guevara, Ing. Agr. , Agronomist6 Rocio Haddad, Biologist3 Ange la Matos, Ing. Agr. 6 Norma Mujica, Agronomist Maria Palac ios, Assoc . Biologist6 Alcira Vera, Biologist W ilberto Vill ano, Agronomists Pathology Pedro A ley, MS, Pl ant Pathologist6 Ciro Barrera, MS, Pl ant Pathologist Ida Barto lin i, MS, Biochemist Car los Chuquill anqui , MS, Pl ant Pathologist Christian Delgado, MS, Biochem ist6 Vio leta Flores, Biologist Segundo Fuentes, MS, Pl ant Pathologist6 Lili am Gutarra, MS, Pl ant Pathologist Char lotte Li zarraga, MS, Pl ant Pathologist Hans Pinedo, Agronomi st3 Hebert Torres, MS, Plant Pathologist6 Jose Luis Zapata, MS, Plant Pathologist, Co- lombia4 Physiology Ro lando Cabe llo, MS, Assoc. Agronomist Nelly Espinola de Fong, MS, Nutriti onist Rosar io Falcon, Bio logist M.S. Kadian, PhD, Agronomi st, India Joseph Koi , MS, Agronomist, Cameroon Jose Lui s Marca, Ing. Agr. Jorge Roca, Bio logist Sonia Sa las, MS, Assoc. Food Scientist K.C. Th akur, PhD, Breeder, India Social Science Cherry Banga lanon, MS, Philippines Rosario Basay, BA, Economist Patri cio Espinoza, Economist, Ecuador" Hugo Fano, MS, Economist Cri stina Fonseca, MS, Agronomist6 V.S. Khatana, PhD, Socioeconom ist, India Maria Lozano, Computer Assistant Lui s Maldonado, BA, Economist Margaret Ngunjiri , MS, Socio logist, Kenya Oscar Ort iz, MS, Agronomist4·s M arice! Pini ero, Ecologi st, Philippines Victor Suarez, MS, Statistic ian Inge Verdonk, Ir, Nutriti oni st, Phil ipp ines Training Nelson Espinoza, Training Spec iali st Martha Huanes, Training Logistics America Va ldez, MS, Training Material Spe- ciali st Research Support Victor Otazu, PhD, Superintendent, Support Unit Lombardo Cetraro, Biologist, Field & Green- house Supervisor, San Ramon Roberto Duarte, Ing. Agr. , Field & Greenhouse Supervisor, La M olina Hugo Goyas, Ing. Agr., Fi eld & Greenhouse Supervisor, Huancayo Lauro Gomez, Supervisor, Huancayo3 Ulises Moreno, PhD, Phys io log ist Mario Pozo, Ing. Agr., Supervisor, La Molina Experiment Station4 Miguel Q uevedo, Ing. Agr. , Off-Station Field Supervi sor, Cajamarca3 Statistics Unit Alfredo Garcia, MS, Experimental Statisti cs Felipe de Mendiburu, Statisti cs Eng. CIP P1ogram Repo1t 1995-96 31 7 Consortium CONDESAN Blanca Arce, MS, Assoc. Scientist, Quito, Ec- uador Elfas M uj ica, M S, Anthropo logist, Adjunct Sc ienti sts An a Marfa Ponce, PhD, INFOANDINA Ad- mini strators Jorge Reinoso, MS, Agri c. Economics, Puna, Peru5 Mario Tap ia, PhD, Agroeco logists Roberto Va ld ivia, MS, Agronomi st, Puna, Peru5 Latin America and the Caribbean Regional Office Sven Vill aga rcfa, PhD6 Communications Unit Cec ili a Lafosse, Chief Des igner Godofredo Lagos, Product ion Chief Emma Martinez, M S, Spanish Ass istant Edi- tor/Supervisor Med ia Product ion Gigi Chang, MS, Photographer, Audiovisual Sect ion Coordinator Information Technology Unit Edith Aguilar, Telecommunications Officer4 Jorge Arbul u, Telecommunications Systems Superviso r3 Monica Ari as, BS, Systems Engineer1 Pab lo Bermudez, Systems Analyst1 Lili ana Bravo, BS, Systems Engineer2 Oscar Bravo, Microvax Systems Supervisor3 Roberto Castro, BS, Systems Development Eduardo Manchego, Systems Ana lyst4 Jose Navarrete, Systems Analyst2 Pia Maria O liden, Systems An alyst Eri c Romero, BS, Systems Engineer Edgardo Torres, BS, Systems Engineer A lberto Velez, MS, Systems Engineer Library Cec i I ia Ferreyra, Head Librari an Controller's Office Miguel Saavedra, CPA, General Accountant Edgardo de los Rios, CPA, Senior Accountant Vilma Escudero, Accountant4 Mi lagros Patino, BA, Accounta nt2 318 Staff Accounting Unit Jorge Bautista, Accountant Blanca Joo, CPA, Accountant Rosari o Pastor, CPA, Senior Accoun- tant Eduardo Peralta, Accountant Budget Unit Deni se Giacoma, CPA, Accountant A lberto Monteb lanco, CPA, Senior Accountant Treasury Unit Sonnia So lari , Chief Cashier Luz Correa, CPA, Accountant (Super- visor)3 Office of the Director General Visitors' Office Rosa Rodriguez, Manager7 Mari ell a Corvetto, Supervisor Office of the Deputy Director General for Finance and Administration Haydee Zelaya, Internationally Recruited Staff Officer Office of the Executive Officer Foreign Affairs Liaison Marce la Checa, Li aison Officer Travel Ana Maria Secada, Supervisor Logistics and General Services Aldo Tang, Comdr. (ret .), Manager Front Desk Micheline Moncloa Maintenance Antonio M orill o, Chief Purchasing Supervisors Artu ro Al varez Roxana M orales Berm udez4 Jose Pizarro Security Jorge Locatelli , Capt. (ret. ), Supervisor4 Transportation Hugo Davis Paredes, Vehicle Main- tenance Officer At il io Guerrero, Vehicle Program- mer Jacques Va ndernotte, Pil ot4 Djordje Ve li ckovich, Pilot4 Percy Zuzunaga, Pilot Warehouse Jorge Luque, M BA, Supervisor Human Resources Lu cas Reano, M anager Juan Pab lo Delgado, M anager4 Auxiliary Services M onica Ferreyros, Superviso r Sor Lapouble, Ass istant Compensation Estani slao Perez Aguilar, Supervisor Labor Relations Luis Caycho, Act ing Supervisor2 Medical Office Dav id Halfi n, M D Lu cero Schmidt, Nurse Social Work Martha Pi erola, Supervisor CIP Program Report 1995-96 31 9 Acronyms and Abbreviations ACIAR ACM VD ADEFO R AEZ AFLP AGERI AMOVA ANOVA AP APW ARC ARTC ASPADERUC ASP RAD AUDPC AVA AVB AVRDC BAC BMZ Bt BW CAAS CaM V CGIAR CIAO CIAT CIRNMA CLAD ES CMP CMS CON DESAN CORPOICA COSUDE CPRI CPRS CpTI CRH DAS DGIS DM DNA ELISA 320 Acronyms Australian Center for International Agri cultural Research Afri can cassava mosa ic v irus di sease Asociac i6n para el Desarro llo Forestal, Ecuador agroeco logical zones amplifi ed fragment length polymorphi sm Agri cul ture Geneti c Engineering Research Insti tute, Egypt analysis of mol ecular vari ance analys is of vari ance arracacha potyvirus Andean potato weevil Agri culture Research Center, Egypt Andean root and tuber crops Asociac i6n para el Desarrol lo Rural de Cajamarca, Peru Asian Sweetpotato and Potato Research and Development, network area under the disease progress curve arracacha v irus A arracacha virus B Asian Vegetable Research and Deve lopment Center, Taiwan bacteri al arti f ic ial chromosome German Mini stry for Economic Deve lopment and Cooperati on Bacillus thuringiensis bacterial w ilt Chinese Academy of Agri cultural Sciences cauliflower mosa ic v irus Consultati ve Group on Intern ational Agri cultural Resea rch Center for Integrated Agricultural Deve lopment, China Centro Intern ac ional de Agri cultura Tropica l, Colombia Centro de lnvest igac iones de Recursos Naturales y M edio Ambiente, Peru Consorc io Latinoamericano de Agroeco logfa y Desarro llo criti ca l moisture po int cytopl asmic male sterility Consortium for the Sustainable Deve lopment of the Andean Ecoreg ion Corporac i6n del lnstituto Co lombiano Agropecuario Cooperac i6n Tecnica Suiza Central Potato Research Institute, India Central Potato Research Stati on, India cowpea trypsin inhibitor Horticultural Research Center, Canada days after sow ing Directorate General for Intern ational Cooperation , Netherl ands dry matter deoxyribonucleic ac id enzyme- linked immunosorbent assay ESEAP EVS FAO FFS FORTI PAPA GA GAAS CATT GBF GE GILB GMI GV HEC HI Hi IARC IBTA ICA ICAR I CARDA ICIMOD ICM ICRAF IDIAP IDM IDRC IFPRI llBC llCA llTA ILRI INIA INIA INIAP INI BAP INIFAP INIVIT IPGRI IPM KARI LAC LB LMF MOS MENA MIP MSU East and Southeast Asia and the Pac ific, CIP reg ion evaporative rusti c stores Food and Agriculture Organization of the United Nations farmer fi eld schoo ls Fortalec imiento de la lnvest igaci6n y Producc i6n de Sem illa de Papa en el Ecuador gibberelli c ac id Guandong Academy of Agricultural Sc iences, China General Agreement on Tariffs and Trade Society for Biotechnologica l Research, Germany genotype x environment Globa l Ini tiative on Late Blight Globa l Mountain Initi ati ve granulosi s virus hen egg cystatin harvest index heterozygos ity index international agri cultural research center lnstituto Boliviano de Tecnologfa Agropecuari a lnstituto Colombiano Agropecuari o Indian Council for Agri cultural Research Internationa l Center for Agr icultural Research in the Dry Areas, Syria International Centre for Integrated Mountain Development, Nepal integrated crop management Intern ational Centre for Resea rch in Agrofo restry, Kenya lnstituto de lnvestigac iones Agropecuari as de Panama integrated d isease management International Development Research Centre, Canada International Food Policy Research Institute, USA International Institute of Bio logica l Control , Kenya lnstituto lnteramericano de Cooperac i6n para la Agri cultura, Costa Ri ca International Institute of Tropica l Agriculture, Nigeria Intern ational Livestock Research Institute, Kenya lnstituto Nacional de lnvesti gaciones Agropecuarias, Uruguay lnstituto Nacional de lnvesti gaci6n Agrari a, Peru lnstituto Nac ional Aut6nomo de lnvest igaciones Agropecuari as, Ecuador Internationa l Network for the Improvement of Banana and Plantain, France lnstituto Nac iona l de lnvestigac iones Foresta les y Agropecuarias, Mex ico lnstituto Nacional de lnvesti gaci6n de Viandas Trop ica les, Cuba International Plant Genetic Resources Institute (formerl y International Board for Pl ant Genetic Resources), Italy integrated pest management Kenya Agricu ltural Research Institute Latin Ameri ca and the Caribbean, CIP region late bl ight leafminer fl y multidimensiona l sca ling Middle East and North Afr ica, CIP reg ion Programa de M anejo lntegrado de Pl agas, Domini can Republic Michigan State University, USA CIP Program Report 1995-96 32 1 Mv NAARI NARO NARS NASH NGO NPT NRI OCI ODA PAGE PapMV PBC PBRV PCR PIC PICTIPAPA PLRV PMSF PNG PPRI PQS PRA PRACIPA PRAPACE PRECODEPA PROCIPA PRODAR PROINPA PRONAMACHCS PSTVd PTM PVP PVS PVX PVY QTL RAPD R/F RFLP RH RILET RNA RRD RT-PCR SAAS SADC 322 Acronyms marker value Namulonge Agricultural and Animal Production Research Institute, Uganda National Research Organization, Uganda national agricultural research systems nucleic ac id spot hybridization nongovernmental organ ization neomycin phosphotransferase Natural Resources Institute, UK oryzacystatin I Overseas Development Administration , UK polyacrylamide gel electrophoresis papaya mosaic virus Potato Base Collection at CIP potato black ringspot virus polymerase chain reaction polymorphic index content Programa Intern ac ional Cooperativo del Tizon Tardio de la Papa, Mexico potato leafro ll v irus phenylmethylsulfonyl fluoride Papua New Guinea Plant Protection Research Institute, Egypt Plant Quarantine Station, Kenya participatory rural appraisa l Programa Andino Cooperativo de lnvestigaci6n en Papa, CIP network Programme Regional de !'Amelioration de la Culture de la Pomme de Terre et de la Patate Douce en Afrique Centrale et de l'Est, CIP network Programa Regional Cooperativo de Papa, CIP network in Central America and the Caribbean Programa Cooperativo de lnvestigaciones en Papa, CIP network in Southern Cone Programa de Desarrollo de la Agroindustria Rural para America Latina y el Caribe Proyecto de lnvest igaci6n de la Papa, Bolivia Programa Nacional de Manejo de Cuencas Hidrograficas y Conservaci6n de Suelos, Peru potato spind le tuber vi roid potato tuber moth potato virus P potato virus S potato vi ru s X potato vi ru s Y quantitative trait loci randomly amplified polymorphic DNA root to foliage ratio restriction fragment length polymorphism relative humidity Research Institute for Legumes and Tubers, Indonesia ribonuclei c acid Red River Delta, Vietnam reverse transcription polymerase chain reaction Sichuan Academy of Agricultural Sciences, China Southern African Development Community SAN REM SAPP RAD SARR NET SBTI soc SEIN PA SENA SE NASA SINGER SM SMC SPFMV SPVD SSA SWA TALPUY TAP TCRC TMV TMV TPS UKDW UMATA UMV UNA UNALM UNC UNCP UNEP UNSAAC UN SCH UNT UPWARD URP USAID USEPA uvc VASI VTMoV WARD A WBV WWR XS PRC Sustainable Agricultural and Natural Resource Management Southeast Asian Program for Potato Research and Development Southern Africa Root Crop Research Network soybean trypsin inhibitor Swiss Development Cooperation Semilla e lnvestigaci6n en Papa, Peru Servicio Nac ional de Aprendizaje, Colombia Servicio Nacional de Sanidad Agraria, Peru Systemwide Information Network for Genetic Resources simple matching coefficients seed moisture content sweetpotato feathery mottle virus sweetpotato virus disease Sub-Saharan Africa, CIP region South and West Asia, CIP region Grupo de lnvestigac i6n y Desarrollo de la Ciencia Andina, Peru transmission access period Tropica l Crops Research Center, Bangladesh tobacco mosaic virus total marker variance true potato seed Duta Wacana Christian University, Indones ia Unidad Municipal de Asistencia Tecnica Agropecuaria, Colombia ullucus mosaic virus Universidad Nacional del Altiplano, Peru Universidad Nacional Agraria La Molina, Peru Universidad Nacional de Cajamarca, Peru Universidad Nacional del Centro del Peru United Nations Environment Programme Universidad Nacional de San Antonio Abad del Cusco, Peru Universidad Nacional de San Cristobal de Huamanga, Peru University of Turku, Finland Users' Perspectives With Agricultural Research and Development Universidad Ricardo Palma, Peru United States Agency for International Development United States Environmental Protection Agency ullucus virus C Vietnam Agricultural Science Institute velvet tobacco mottle virus West Africa Rice Development Association, Cote d'Ivoire whitefly-borne closterovirus watery wound rot Xuzhou Sweet Potato Research Center, China CIP Program Report 1995-96 323 !Uplnd opal!IV 'ofi1op1H uo111w '·9 04iuo)-zapu9uia:1 !UIN :1n0Ao1 puo ufi!saa 8SSO}Dl D!l!J8) :JaAO) sofio1 op8ljOpo9 'asSO}Dl 01Jll8) :sJOjDU!PlOO) UO!PnpoJd 411ws 11!8 'ApJOH 11!8 :sJOJ!P3 :(A11oi11aqo4d10) SJ!P8JJ UO!PnpOld