S 589 .7 B58 BIOLOGICAL AND CULTURAL /.r DIVERSITY [he role of indigenous agricultural experimentation in development Edited by GORDON PRAIN, SAM FUJISAKA and ; MICHAEL D WARREN I .,/'-";- ,,,""----.., _'. "_" .-J'" . ... '. .~ '. ~ :. !,,-" .• _.-.<' -,- ~ .. ~ INTERMEDIATE TECHNOlOGY PUBLlCATIONS 1999 ''1 " This book is dedicated to !he memory of Professor Michael Warren, whose untimely death occurred during Ihe preparation of the volume. Mike"s profes­ sional life was committed lo rhe practica! applicatíon of social seicnees lo development. His mos! powerful message was tha! in order to achieve suceessful development, we have to work with loeal people using ¡heir knowledge as Ihe number one resource. In imp!ementing that message, he helped establish indigenous knowledge centres tbroughout the world and demonSlrated very c1early Ihe potency of local knowledge in understanding, domesticating and utilizing biological diversity. His interest in Ihis theme \Vas the driving force behind the present volume. Mike Warren's energy, commilment and cloquence in pursuit of pcople-centred devel­ opment wil! be sadly missed. Intermediare Technology Publications LId, 103-105 Southampton Row, Londan WClB 4HH, UK © the individual authars; tbis collection JT Publications, 1999 A CIP catalogue record for tbis book is available from the British Library ISBN 1 85339 443 2 Typeset by J&L Composition Ltd, Filey, North Yorkshire Printed in the UK by SRP, Exeter Contents Preface v DA VID BROKENSHA Notes on Contributors vi Introduction THE EDITORS 1. Rice cropping practices in Nepal: indigenous adaptation to adverse and dífficult envíronments. 6 JIT BHUKTAN, GLENN DENNING and SAM FUJISAKA 2. Farmer-based experimentation with velvetbean: innovation within tradition. 32 D. BUCKLES and H. PERALES 3. Side-stepped by the Green Revolution: farmers' traditíonal rice cultivars in lhe uplands and rainfed lowlands. 50 SAM FUJISAKA 4. Environmental dynamics. adaptation and experimentation in indigenous Sudanese water harvestíng. 64 DAVID NIEME/JER 5. The indigenization of exotic inputs by small-seale farmers on the Jos Plateau, Nigeria. 80 KEVIN D. PHILLlPS-HOWARD 6. Farmer managernent of rootcrop genetic diversity in Southern Philippines. 92 GORDON PRAIN and MARICEL PlNIERO 7. Farrner experimentation in.a Venezuelan Andean group. 113 CONSUELO QUlROZ 8. Indian farmers opt for ecological profits. 125 VITHAL RAJAN and M. A. QAYUM 9. Indigenous agricultura] experimentation in horne gardens of South India: conserving biologícal diversity and achieving nutritional security. 134 B. RAJASEKARAN 10. Living local knowledge for sustaínable development. 147 NIELS ROLING and JAN BROUWERS 11. Varietal díversity and farmers' knowledge: the case 01' ¡he sweet potato in Irían Jaya. 158 JURG SCHNEIDER iv Contents 12. The indigenous concept of experimentation among Malian farmers. 163 ARTHUR STOLZENBACH 13. Umnotho Wethu Amudobo: the clash between indigenous agricultural knowledge and a Western conservation ethic in Maputaland, South Africa. 172 DAN TAYLOR 14. Local-level experimentation with social organizalion and management of self-reliant agricultural development: lhe case of gender in Ara, Nigeria. 184 D. MICHA EL WARREN and MARY S. WARREN 15. Chinese farmers' initiatives in technology dovelopment and dissemination: a case of a farmer association for rural technology development. 192 Ll XIAOYUN, Ll OU and LI ZHAOHU Notes 199 References 202 PREFACE WE WELCOME THIS important volume as the tenth in our series on Indigenous Knowledge and Development. Whíle much attemion has becn paid to biological diversity, ils c10se association with cultural diversíty has been comparatively neglected: we hope that Ihis volume will help to demonstrale lhe íntimate relationship between ¡he lwo. As ¡he title indicates, our 15 contributors all deal with farmer experimenta­ tion, iIluslraled with case-studies from many par!s of Ihe world. The crops Ihat are used for experimenls inelude rice, maize, beans, and rool-crops; Ihere are also sludies of water harvesting (in lhe Sudan) and of home gardens (in South India). The studies show that: o farmers do experiment. In any society, there may be only a few individuals who are truly experimental, but this is true of all societies, including Western ones o farmers know theír local environments intimately, and lheir experiments are usually very site-specific o farmers also have a c10se and detailed knowledge of local cultivars, and are well aware of lhe need lo promote biological diversity o lhe experiments underline the importance of in sÍ/u conservation. inc\uding the protection of wild plants o any atlempts at conservation of natural resources should involve the local inhabitants Q farmers' experiments are 'constantly being re-ínvented' ~ indígenous know­ ledge is a dynamic topie o while contributors stress the value of farmers' experiments, they are also aware that these are no substitute for conventional on-farm rescarch, but ¡hey do provide a valuable resouree for lhe latter () women farmers are often also actively involved io experiments. This collectioo, wíth ílS detaiJed. descriptions aod aoalyses of specific and varied situatioos, should broaden the debate about biological diversity; it should also emphasize the importance of taking a broad approach, one that indudes ¡he cultural factors. David W. Brokensha January 1998 Notes on Contributors BHUKTAN, JIT Researcher, Internationallnstitute of Rural Reconstruction, Philippines, BROUWERS, JAN Programme OjJicer, Netlzerlands Development Organisarion in CameroOfL BUCKLES, DANIEL Interna/ional Deve/opmen! Research Centre, PO Box 8500, Ottawa, Canada, DENNING, OLE!'!\' Director of Development, International Center Jor Research on Agro~ [oyestr}', Nairobi, Kenya. FUJlSAKA, SAM Agricu/cural Anthropologist, CIAY, Colombia, NíEMEUER, DAVID Department 01 Irriga/ion and Soil aná Warer Conserva/ion, JYageningen Agricultura/ Untversity, fhe Netherlands. Ou, LI Depuly Director, Centrefor lntegrated Agricultura! Developmem (CIAD), Chinese Agricultural University, Beijing, People's Republic 01 China. PERALES, H, EcologíSI, Colegio de Postgraduados, Montecillos, Mexíco. PHlLUf'S-HoWARD, KEVIN ProJes:mr and Head 01 Geograpny, University 01 Transkei. Soutn Africa umil hi. .. untimely death in 1995. P1NIERO, MARICEL Doctoral candidme, Departme!1l of Amhropology ami Linguistics, Uni­ versity 01 Georgia. PRAI:"-l, GORDON D. Regional Director. International Polato Centre, East, Soulkeasl Asia and the Pacific Division, Jakarta, Indonesia. QAYUM, M. A. ProJect co·ordinator. Action far World Solidarity. India. QUlROZ, CONSUELO Dm?ctor, Vtmezula Resource Center Jor lndigellous Knowledge ( V ERC/K), University o/ the Andes, Trujillo. Venezuela. RAJAN, VITHAL Founder and Director, Decean Del'eJopment Society, India. RAJASEKARAI'-i, BHAKTHAVATSALAM Senior Information Scientist for Agriculture, Consortium for lnternational Earth Science Information Nelwork (C1ESIN), Saginaw, Michigan, USA. ROLlNG, NIELS Professor, Department of Communication and lnnMalion Studies, Wageningen Agricultural University, the Netherlands. SCHXEIOER. JURO Anthropologist, Institut fiir Ethnologie, Universitiit Bern, Switzerland. SrOLZENBACH, ARTHUR Department 01 Extensioll Science, Wageningell Agricultural Uni­ versity, lhe Nether/ands. TAYLOR, DAN Director, Centre for Low Input Agricultural Rescare/¡ and Deve!opment, University 01 Zululund, KwaZulu, Natal, South Ajrica, WARREN. D, M1CHAEL Director. Cenler jor Indigenous Knowledgefor Agricullure and Rural Development (CIKARD) and Unfl.'éfsity Professor, Iawa Sta/e University. Ames, 1011'a, USA. unlil his untimely death in 1997. WARREN, MARY S, Consultanl in Rural Development and Gender Issues in Development. XIAOYlJN, LI Director, Centre for lltlegrated Agricultural Development tCJAD), Chinese Agricultura! University, Beijing, Peop!e's Republic 01 China. ZHAOHU, LI Lecturer. Agronom}' Deparlment, Chinese Agricultural University. and Deputy Director, Guioxin Association for Rural Technology Developmenl, Peop/e's Republic al Chifla. Introduction DURING THE PAST decade development professionals have focused theír attention on mechanisms thal can facilitate participatory decision-making, collaboration with clientele groups, and sustainability of development project efforts (National Research Council, 1991). Sin~'C the 1980, lhe role of indigenous or community­ based knowledge systems has been demonstrated as a powerful aUy in lhe efforls lo make development more c1íentele-fríendly and more cost-effe<:tive. Indigenous knowledge is now viewOO by many as lhe basis for indigenous approaches lo decision-making, parlicularly lhrough indigenous organizations where commu­ nity problems can be identified and discussed, frequently leading to indigenous approaches to innovation and experimentation (see p.l58 in Warren and Pinkston, 1997). In the past few years there has been an enormous inerease in the number of publications on lhese topies, especially on indigenous experimen­ tatíon (e.g., Adams and Slikkerveer, 1997; Amanor, 1991; Brush and Stabinsky, 1996; Leakey and Slikkerveer, 1991; Sillitoe, 1998, Thurston, 1992; Thurston, 1997; Thurston and Parker, 1995; Warren, 1994). Their effons should expand with the new availability of user-friendly manuals and guides for recording indigenous knowledge (e.g. Cen!er for Traditional Knowledge, 1997; T1RR, 1996). Global coneern with lhe issuc of biodiversity is al so now bcing matched with a concomitant concero for lhe human knowledge generated in every com­ munity that is relatOO to the bíological realm (Natíonal Research Councíl, )992). Interest in lhe role of indigenous agricultura) experimentation in fostering bíologieal diversity and lhe cultural knowledge of that diversity has likewíse expanded rapidly in recent years. An example of the interest ís reflectOO in lhe Intemational Conference on Creativity and Innovation at the Grassroots, co­ ordinatOO by Anil Gupta and held al the Centre for Management in Agriculture, Indian Instituto of Management, Ahmedabad, India in January 1997 with doz­ ens of papers presented by participants from all over the globe. This volume provides additional case-studies that show how indigenous knowledge. practice, afid experimentation can and do serve to advance agricultural developmcnt. This íntroduclÍon places the current efforts to record examples of indigenous agri­ cultural experímentation into a historical context, with a bibliographical sectlOn that reftects lhe rapidly growing volurne of current publicatjons on Ihis subjec!. That farmers work actively to solve theír problems and do so based on locally developOO knowlOOge are not new ideas. Al the turn of the century King (1911) documented practice, in east Asian rice-growing countries which formed the bases of many 'modero' rice-production technologies. Sauer (l969) encouraged building technical ínnovations on traditional knowledge in Mexico in the early 19405. Cultural ecologísts (or 'ecological anthropologists') documented farmers' practices and knowledge - albeit without much of a concern for experirnentation aud 'development' (Conklin, 1957; Evaus-Pritchard, 1940; Forde, 1949; Leach, 1968; Netting, 1968; Rappaport, 1968; Steward, 1955). A next wave ofresearchers documented cases of active and inventive indigenous agricultural problem-solving. Johnson's (1972) work with smal! farmers in Brazil, Bunch's appeal for farmer-centred agriculture (1985), aud Richard's 2 Biological alld Cultural Díversity work in West African farming systems (1985) were among important carlier contributions. In addition lo defining different types of indigenous experimenta­ lion, Rhoades and Bebbinglon (1995) reviewed some of the key contributions supporting Ihe 'farmer as experimenter' written sinee Johnson's work more than 20 years earher (Biggs and Clay, 1981; Box, 1989; Denevan, 1983; Rhoades, 1987; Thompson, 1973). Richard, (1986) expressed Ihe general problem in reference to African agriculture: Many commentators see the prevalence of 'primitive' cultívation practices as a major factor in present food-production difficullies. Permanent solulions lO Ihe famine problem, it is supposed, will depend on e10sing the 'gap' between scíence and a 'backward' peasantry, henee Ihe slress placed on 'Iop-down' agricultural extension systems for the efficient delivery of 'modern' inpuIs lo small-scale farmers. According to Ihis viewpoinl, the key lO an agricultural revoJulion in Africa is technology Iransfer. This book challenges the assumptions of this approach. It argues that tech­ nology transfer is par! of the problem not the solution. There is indeed a major gap belween science and the peasantry, but Ihis gap is conceplual not historicaL Tropical agricullural scienee is out of touch with ils clientele. The problem is nO! how lO bridge lhe centuries and reacb 'backward' farmers with 'modern' inputs and advice, bul Ibat Ibese inpuls and advice are in many cases totally inappropriate lo African food-crop producers beca use they have beco designed without reference to Ibe problems, priorities, and interes!, of those who are supposed to use them (1986: 1). Richards furlber discussed the need to mobilize local initiatives in solving problems given farmers' interest in, commilment lo_ and suceesses in technolo­ gical change: Careful analysis of peasant innovation and experimentation ... provides clues as lo the nature of local priorities and Ihe scope and lrajeclory of endogenons agricultural change. Ralher Ihan concentrate on lechnology transfer, formal­ seclor R&D should lock on to local initiatives. Scientific research should be carried out within design parameters set by Ihe farmíng community. .. Formal­ seclor research agencies look at land and land resources from an analytical perspective. .. By contrast, farmers view the landscape in synthetíc terms (1986: 156). Agricultural research began lo incorporate índigenous knowledge and farmer parlicipalory researcb in the 1980s. Studies of indigenous knowledge syslems and development were presented in a landrnark collcelion (Brokensha el al.. 1980) and in a follow-up (Warreo el al., 1995). Ioternalional agricultural rese>lrch increasiugly built researcb upon indigenous knowledge tbroughou! Ihe same period, with findings published in collections aud reviews (Dvorak, 1993; McCorkle, 1994; Moock aud Rhoades, 1992; Warren, 1991). In off-farm case-sludies of innovations and experimentation (e.g. Adcgboye and Akinwnmí, 1990; Oamser el al., 1990) demonstrate Ihat 'Ihe poor are experimenting con­ slantly, innovaling in a struggle to survive' (ibid.: ix) and Ihal people's teehool­ ogies 'develop in a way tha! retains and builds upon local ,kills and dosely rellec!s Ihe priorities of local people' (ibid.: x). lmfoduclion 3 Many of (hese recenl ,ludies have described farmer experimentation and innovation. In a USAID-funded project in Niger: , ... farmers were observed to manipulate such research·related concepts and techniques as the functional equivalent of a (necessarily oral) literature revíew before mountíng field tríals. .. , split plots and trial replications, experimental versus non·experimental variables, costlbenefil and risk assessments; team-like analyses of resulls, and hypothesis generatíon for fulure research' (McCorkle and McClure, 1995: 327). Olher sludies have had more oC an elhnographíc basís, focusing on indigenous concepts and (erms. Work with lhe Ara of Nigeria indícaled Ihal 'Management and development planning concepts often exist in locallanguage ... Localleaders and citizens have a comprehensive understanding of weaknesses and strengths in Iheir own indigenous organization. aud are very open lo experimenting with new management approaches' (Warren el al., 1996: 48). In a critique of farming systems research, den Biggelaar (1991) nOled Iha! 'liUle bas been done to develop indigenous technology generariog and dilTusing capaeities already present in the rural arcas.' He proposed a model: ... based on eo·operalion and collaboration between the exogenous and indi­ genous knowledge syslems leading to a synlhesis of the two. The underlying principIe of lhe model is Ihal the ultimate soJution for rural development is nOl Ihe dumping of more .dentisls upon rural people (of whatever discipline) lo make exogenously-generated tecbnologies more adaptahle and in-line with people'. problems, but lo st,englhen, empower, and legitimize indigenous capacíties for identifying problems and developíng solulions for Ihese pro­ blems (1991: 25). Indigenous capabilities need not be Iimiled to farmers in developing countries. Jules Pretty (1991) noted that for the UK: Over a period of two centuríes crop and livestock production increased 3~ fold as innovalÍve technologies and techniques developed by farmers were extended to olher Carmers Ihrough tours, farmer groups, open days, and publications, and then adapted to local conditions by rigorous experimenta. liún. These technologies maximized the use of on-farm resourees al a time when there was no government ministry of agriculture, no research slations, and no exlension instítutions (132). Farmer participatory research which builds on farmer experimentation and índigenons knowledge - has become the norm, generating a huge amount of Iiterature (Ashby el al., 1989, 1996; Berg, 1996: Biggs, 1980; Brouwers, 1993; Bunch, 1989; Bunch, 1990; Chambers, 1997; Chambers el al., 1989; den Biggelaar and Hart, 1996; Hiemstra el al., 1992; Innis, 1997; Joshi and Wítcombe. 1996; McCorkle and Bazalar, 1996; Mallon, 1988; Page and Richards, 1977; Rhoades and Booth, 1982; Richards, 1985, 1986; Scarborongh el al., 1997; Selcner el al .. 1997; Sthapit el al., 1996: Systemwide. 1997; Tripp, 1989; Van Veldhuizen el al., 1997; 1998: Warren, 1989; Willcocks and Gichuki, 1996; Witcombe, 1996; Witcombe el al., 1996), reviews (Amanor, 1989; Bemley, 1994; Farrington and Martin, 1987), and catchy phrases sueh as 'farmer-baek· to-farmer' (Rhoades and Bootb, 1982), 'putting Ihe fírst las!' (Chambers, 1983), 'farmer first' (Chambers el al., 1989), 'joining [armers' experíments' (Haverkort el al., 1991), 'farming for the future' and 'Iow-external-input and sustainable 4 Biological and Cultural Diversüy agriculture' (Reijntjes el al, 1992), ando more reccnlly, 'beyond farmer !lrst' (Seoones and Thompson. 1994). These contributions have led to the shared vision that, 'If we are to be serious about the deveJopment of a sustainable agriculture, il is critical that local know­ ledge and skills in experimentatíon are brought to bear on the processes of research' (Pretty. 1995: 180). Farmer experimentation and knowledge is íncreasingly acknowledged as a key component of agrobiodiversity conservation and erop genetic-improvement. A recent collection of case-studies addressed 'the need to develop appropriate research ami development strategies whieh build upon both the capacities of farmers to experiment with crops, and the knowledge they have acquíred of diversity". Farmer experíments with crops have been important in promoting diversity and the conservation of species and varieties' (de Boef el al., 1993: 1), A more recent collection of experiences on participatory plant breeding (Eyzaguirre and Iwanaga, 1996) covers a continuum from breeder-controlled to farmer-controlled systems of breeding and selection, and indicates clearly the cost-effectiveness of working with farmers' skills in experimentation and innovation, Sumberg and Okali (1997) have provided Ihe most recent and very compre­ hensíve overview of the literature dealing with indigenous agricultural experi­ mentation and innovation. The authors also present empirical data on farmers' experiments from case-studies in Kenya, Zimbabwe, Ghana, and the UK and eonclude that: ... farmers' experimentation is widespread, an importan! part of everyday farming, and shares many characteristics with formal agronomic experimenta­ tion, Thus, through their experiments, farmers are involved in ongoing pro­ cesses of local knowledge creation through site-specifle leaming, which, in ¡he short term, results primarily in small adaptations to farming practice and, in the long term, contri bu tes to the development of new farming systems. However, we conclude at the same time that although in many situations the arguments for greater participation of farmers in agricultural research are compelHng and relevant, relatively Iiule potential synergy wiU be realized tbrough formal research and farmers' experimentation beíng more closely linked. In addition, beca use of the site-speciflc nature of the knowledge created through farmen' experiments, lhe claim that there is significant unrealized development potential associated with them, which couId some­ how be used to make an impaet on a larger seale, is also called into question (1997: 7. .8 ). Both farmers' experiments and much formal experimentation aim to develop practical solutions to immediate problems or to seok smal! gains within the context of pro ven produetion methods and systems. Both are largely empírical and iterativo, combining experience, observation (both methodical and oppor­ tunistic), intuition, persistence, skill and luck (1997: 149), Bunders eC al. (1997) presented a colleetion of case-studies looking at tbe application of both indigenous biotechnology and science-based biotechnology 10 agricultural devclopment, building on farmers' knowledge while making use of the latest scientific insights, The case-s!udies deal with biotechnologlcal prac­ tices of farmers in animal health, biopesticides, food processing, and crop genetic resources. A section on 'science-based biotechnology' assesses its potential Introducliol1 5 wilhin lhe sodo-political context: while a section on 'building on farmers' practice' calls for an integrated approach to biotechnology development that builds on farmers' knowledge and makes use of Ihe latest scientific insighls. Similar topies had been discussed al a USAID-funded workshop on the role of indigenous agricultural knowledge in biotechnology lhat was conducled al Obafemi Awolowo University in !le-Ife, Nigeria in 1996 (see Warren, 1996). The 15 contributions to Ihis volume span lhe globe, with case .Iudies from Afriea (Nigeria and one each for Sudan, Benin, Mali, and South Afríea), Asia and lhe Pacific (Philippines, Papua New Guinea, Indonesia, India, China, and Nepal), Latin America (Mexico, Peru, Venezuela), Europe (Nelherlands, France), and Australia. The authors come from Canada, the USA, UK, the Nelherlands, India, Venezuela, South Afríea, Nigeria, China, Nepal, and the Philippines. Mosl of lhe studies deal with various Iypes of on-farm experimenta­ lion reJated lo produetion agriculture and biodiversity, such as lhe deveJopment of new erop varieties, manipulation of germplasm, lesting of exo!ie inputs, and pest management. Related sludies focus on lhe dynamie nature of indigenous organizations in China and Nigeria, the clash of systems in South Afriea between small-scale farmers and conservationists, indígenous concepl. of experi­ menlation (shifteli) in Mali, and water resouree managemenl in the Sudan. Of particular interest is lhe role of indigenous approaches to biotechnology including farmer expertise in Ihe conservalíon, evaluation, and exploitation of plant genetic diversily (Prain and Bagalanon 1994), lt is elear from Ihis eollccrion of case-studies lha! farmers world-wide are involved in experimenting, recreating and reinvenling, a capacity Ihat mus! be harnessed in participatory technology development. I01d'Z>lP 1. Rice cropping practices in Nepal: indigenous adaptation to adverse and difficult environments )lT BHUKTAN, GLENN DENNING, and SAM FU}/SAKA Abstract NEPALESE SMALL FARMERS employ a wide range of indigenous practices and correspondíng technical knowledge in organízing rice farming. Practices are well adapted to fragmented land and other diverse and dífficult environments, while techníeal knowledge has been developed over years. Farmers conceptualiz. farm problems and design experimental processe. for whole systems, and super­ impose experiments on regular farming aCtÍvities, to adjust the sy.tems accord­ ing to the changing environs. Knowledge of the indigenous practices can be invaluable in the design and conduct of rice research for increasing system productivity or maintaining su.tainability. Introduction Rice occupies 27 million ha of 54 per cent 01' Nepa!'s total cultivated area (2.65 mmion ha), Mean yields have becn 2,1 MT/ha, Between 1961-62 and 1988-89 rice production increased at the rate of O. 41 MT annually, whereas population grew from 9.4 to 17.13 millions from 1961 to 1986. As a result, per capita rice availability decreased from 224 kg in 1961 to 138 kg in 1986 (DFAMS, 1990). Attempts made 10 improve rice production have not becn based adequately on the traditionally iohented local farmiog practices. Agricultural research in diverse and difficult environments has increasingly relied on ¡he understanding of farmers' traditional practice. and knowledge as a practical starting point fOf research intended to inerease productivity and maintain sustainabi1ity (Brokensha el al" 1980; Rhoades and Booth, 1982; Richards, 1985; Thurston, 1990; Warren and Cashman, 1988). Researchers have e1sewhere incorporated farmer,' perspectives and worked to develop farmer parHcipatory research as mean s to make rice ínnovations farmer­ appropriate and adoptable (Maurya el al" 1988; Osborn, 1990; Prakah-Asante el al., 1984; Richards, 1986, 1987). Sucb an approach is crucial for eountries like Nepal, which is resanrce-poor and where rice is grown in diverse and difficnlt environments. This chapter examines local rice cropping practicos of Nepalese small farmers, íncluding their rice-growing envíronments and indi­ genous experíments. Method Descriptive data were gathered via 26 in-depth household interviews and via interviews, observatíons and group discussions with 190 rice-farming small­ holders from 17 rural communities loeated at varians altitudes of mid-HiII and Tarai rural communities of central NepaL The data were analysed using simple statistics. and findings are presented in simple frequencies, descriptions, farmers' statements and narratives. Rice cropping praclÍces in Nepal 7 Results The results are organized in four sections: rice farming environments, rice crop­ ping practices, farmers' experiments, and farmers' varielal organization strategies. Rice farming environments Rice farming environments in tbe study area were characterized by variable altitude and diverse bio-physical conditions. The altitude in the Hill ranged from 525 to 1710 metres and in Ihe Tarai from 60 to 185 metres. The Hill average monthly temperature ranged from 2°C to 27°C, and that of the Tarai 15°C to 30'C. Minimum temperature went down to -O. 15'C in the Hills and 8"C in the Tarai and the maximum went up to 32°C aud 37°C, respectively. The mean annual raioral! in the Hills and Tarai was 2141 mm and 1786 mm, respectively, with most prevaiHng during June lo Seplember. The mínimum rainfall prevailed from November lo Deeember in Ihe Hills, and from Oclober to April in the Tarai. Rice farming environmenls in Ihe HiIIs and Ihe Tarai were different, while fragmented land was lhe most pervasive environmenl in rice farming. Land fragmentation. The 91 HiII and 99 Tarai rice farmers cultivated 0.48 ha and 0.67 ha which were fragmented into three and five pareels, and scattered al a mean dislanee of 1.02 km and 0.57 km from home, respeclively (Table 1.1). The distance was more of a problem in the HiIIs, where many rugged. steep-slopíng narrow trails were pas.able only on foo!. Hill ricefields were often localed on river terraces, in basins or valleys; while olher parcels and homestead were scattered over ¡he HíIIs. Sorne Tarai parcels were located in other villages, Tabla 1.1. Land fragmentation in the Hills and Tarai of Nepal HíIIs Tara; Total Number 01 sample households 105 99 204 Mean 5ize 01 total landholding (ha) 0.7 0.8 0.8 Mean numbar 01 parcels 5 (1-12) 4 (1-13) 5 percentage households with 1 parcel 3 7 5 2-4 parcels 39 47 42 5-7 parcels 41 25 32 8-10 parcels 16 18 17 10+ parcels 4 3 3 Mean size 01 parcels (ha): 0.14 0.24 0.19 percentage parcels 01 size up lo 0.11 (ha) 0.1 43 20 0.11-0.2 39 23 32 0.21-0.4 15 15 15 0.4+ 3 20 11 Mean dislance: home to parcels (km) 0.9 0.5 0.7 parcentage parcels w/dislance (km) 0.0-0.5 33 45 39 0.51-1.5 55 49 52 1.51-2.5 4 3 3 2.5+ 8 3 6 8 Biological and Cultural Diversity allhough these were usually accessible by oxen-carts and bicycles. Parcel charae­ teristíes infiuenced farming praetices. Land forms. In the Tara; where the relativo elevation of /la! land was of key lmporrance parcels were categorized as lowland (43 per cen! of areas), midland (23 per eent), upperland (17 per cent), and deepland (9 per cent); (Table 1.2). HíIl parcels were charaeterized by elcvatíon, slope and relativo width of the cultivable surface. Hill lands were tirst categorized as upper hill (including summÍt), intermediate hill (pahad¡, foothill (fedi), or river basin (byansi). Each of these was further characterized as level (favourable for transplanted rice (TPR)) or sloping (less favourable good for only upland direct-seeded rice (DSR». Level and sloping terraces were further divided into wide and narrow: an oxen pair could be lurned around easily on a wide terrace, while animal­ drawn implements had to be unmounted for turning animals around to avoid Iheir falling down the terrace wall on narrow terraces. The narrower and sleeper Ihe sJope, lhe harder it was to cultivate, Overall, Híll rice parcel areas were 53 per cenl narrow-Ievelled lerrace, 31 per . Farmer., therefme, preferred Ihose that could easily be threshed bul which did not shaller. Farmers also, at times, examined the shape and size of the graio. The finer the graio with bright white kernel, Ihe more attractive it was for market and the higher the priee it commanded. Women considered a bright brown hull colour as advantageous for cleaning the grains conveniently, as they could easily detect foreign materials mixed with the grains. Rice cropping practices in Nepal 25 Tbe higber milling recovery of grain was important both for food and market purposes. Farmers discarded ríce varieties with poor milling turo-out for no! fetching better priees and having poor cookíng quality, Cooking quali!y of grain being important for bolh domestic consumption and the market, farmers tested Ihese varietal traíls, Poor cooking quality mean! loss of grain, thereby requiring more rice lo foed fewer persons - and ít was also a sticky rice, They preferred a ríce variety for food which increased in volume when eooked and was heavy when eaten, so ¡hat a smaller quantity of l'Íee satisfied the hunger of more persons for longer Farmers also believed that heavier and non­ stieky riee food was more palatable and nutrítious. Plant type. Farmers examined various aspects of plant type of rice for varíous purposes. They required fairly tall plants, mainly lO harvest more slraw for animal feed, mal weaving (for human beds and seals), making tying ropes, roofing and fence materials, Farmers also stored and fed it lo animal, almosl throughoul Ihe year, Farmers tested Ihe lodging hah;t of rice plants. A lodged crop mean! 1055 of grain and more labour cost to harvest il. They therefore preferred shorter varieties lo the lodging taller ones. Farmers carefully examined the tillering vigour of rice varieties, especially Ihe proportion of productive tiller,. Thcy had experienced in the past Iha! crops with a higher proporlion of unproduclive tillers not only yielded less grains bul al so produced straw which was less ,torable and palatable for animals, and less durable for roofing and weaving ma!s, As Ihe unproductive tiller, got spoiled faster, these reduced the storabilíty of the bulk of Ihe straw. Farmers preferred rice varíeties with vigorous tillers, mos! of which were productive, Farmers also examined the tilleríng habit and preferred one Ihat Iillered gradually to one tíllering simultaneously. In the former type, even if the ínitial tillers were damaged by ínseets, the subsequent tillers had a chanee lo survive, or tillering could be induced through subjecting Ihe crop to waler stress eonditions. The erop maturity period was an important consideration, especially when they grew rice in relation to other crups. Farmers preferred a variety that matured at a defini!e time. even if planted lale, One of (he preferred a(lributes of Masuli was Iha! even if its planting was delayed by two months it malured in early November logether with others planted in July. However, preference for Ihe maturity period of a varielY differed according to the environment, and Ihe type of cropping patlern (lf which rice was a compo­ nen!. For Ihe rainfed or water stTess eondition, and where moisture became ,caree at Ihe later par! of the wet scason, farmers preferred short-duration cultivars. Also Ihey preferred early maturing varietie, for the distant pareds, to harvest earller to avoid thef!, and they grew rice as an early crop or in between Iwo erops, especially when delaying the establishment of the suceeeding crop was not environmentally permissible, In the land where water remained for a long time aud other erops could not be grown thereafter, farmers preferred long duralÍon varietíes Ihat matured after excess water problems and aquatic leeches had subsided. Under normal condi­ lions farmers preferred medium-duration varieties beca use it was difficult to manage the too-short duration cultivars wíthin a limited time period, and too­ long duration ones upse! the succeeding crops. Farmer, belonging to different eeologícal zorres and havlng their rice laud parcels scattered over different microecological pockets faeed difl'erent types of 26 Bíologícal and Cultural Díversíly inseet and disease problems, They therefore preferred the ríee cultivars that were resistant to local pests, and often discarded those which were susceptible, AH farmers tested to determine the water requirements of any new or alien variety, They had experienced that lhe less vigorous and shorter rooting of a rice plant had several disadvantages, ineluding higher susceptíbility to moisture stress and lodging, Likewise, plants wíth vigorous rooting required ample and running water, whieh farmers often lacked in mos! parts of their rice land, Farmers preferred a cultivar with relatively fewer and longer roolS, for their light and moisture-stressed fields, and heavy rooled varielies for moisture-rich lowlands, Fanners also did not like rice varieties requiring high levels of chemícal fertilizers (despite theír higher yields), Cash spending, unavaílabílíty of appro­ priate types of fertilizer, and more importantly, lbe adverse effeets of chemicals in .oils, had dissuaded small farmers from adopting such variehes, Farmers' varietal choices retlected the environmental diversity of their pareels. As an example, one Tarai farmor grew lleven varieties on 0,7 ha and nine pareels, An early MV (CH45) was planted in the spring, followed by an MV (Masuli) in the summer on two irrigated, middle-light loam soil parcels, A short,duratÍon TV (Mansara) was planted 00 a distant rainred upper paddy. Mansara did well on the two irrigated pareels with heavy soils, Muluri was selecled because it did nol lodge due to exeess water and high ,oil fertilíty, Muturi was followed by trans­ planted Masuli, A spring crop of Muturi was followed by a non,lodging tall TV (Hakijhukan) on a boggy, fenile lower parceL A late-maturing TV (Katika) thal required less water at later growth stages was planted on a heavy,soil, moderately fertile and partially irrigated pareel, Different varielies were suited to different areas within the hills, Cold and shading were problems in the upper hills, making cold- and ,hade-tolerant TVs such as Thankote desirable, In .haded lands irrigated with wann spring water, shade-tolerant TVs such as Chandramarsi were grown, In sunnier irrigated areas of the mid-hills, .ome widely adaptable TVs such as Chandramarsi were growo, In sunoier irrigated arcas of the rnid-hills, sorne widely adaptahle TVs such as Pokhreli Masino (see Table Ll3) were common, In Ihe lower hills and river Table 1,13. Farmers' In = 48) reasons for growing Pokhreli Masino % Farmers Cold (alr and water) tolerant 100 High straw production 10 0 High grain quality & market price 1 00 Good storage 100 High milling recovery 100 Good disease resistance compared with other TVs 100 Does not require too much water 100 Lower seed rate than other varie!ies 100 Bes! far delicacies and ceremonial purposes 100 Compatible with following wheat crop 95 Does not require too much fertilizar but responds lo moderale dose 88 High yield among LVs 83 Productive tillers 68 Full grains 34 Rice cropping proctices in Nepol 27 basins wilh warmer climales and less shade, lhe MV Masuli was planled exten­ sively. The shon-duration MV, CH45, was adapted lO light soils and drought; and lhe MV K-39 was adapted lo ferlBe soils and drought. In areas with seeure early irrigatíon and possible droughl later in the season, the relatively early­ maturing MV IR-8 was eommon. B090 also grew in lhese arcas; bUl IR-8 was planled beca use il matured earlier and was less alfected by rice bug. The diversily of rice cultivars also reflected farmers' mulliple uses of rice, ineluding consumption, fuel, livestock feed, crop production, a souree of eash, medicines, roofing and handieraft material" and liquor (Table 1.14). Farmers also grow sorne TV s as high-quality eash erops. Tliese included Pokhreli Masino and Sikarimarshi in the HiIIs and Kalauamak, Kasluri, and K anakjeera in Ihe Tarai. Farmers' rice cultivars changed over lime as they roulinely collecled aud lested varieties. HiIl farmers adopted lhe TV, Thankole, in Ihe 19 60s. Thankote replaced several TVs (Le., Ratemarsi, Chioia, Falame, Jungemarsi. and Belí); and now farmer, were lesting MV Himali lo replace Thankote. Farmers ín the low HiIls are adopting a TV, Batlísara, for ils hígh yield, and it was likely to replace sorne curren! TVs. IR-8 aud Masulí previously replaeed several TVs, such as Rajbhog, Manbhog, Biramphool. aud Baglunge. Farmers in Ihe Taraí were testing MVs such as Janaki, Laxmi. Sabítri, MaJ­ lika, aud BindesworÍ. Farmers had recently dropped sorne TVs such as Jogini, Saltari, Aga and Barmabhusi. Farmers liad increasingly sought varieties for early-planled erops. HiII farmers Iríed CH45 and K-39. These performed well with warmer spring water irrigatiou; bul farmers were stíll seeking varielies adapted lo Ihe cold stream water. Table 1.14, Farmers' (n = 37 in the Hills and 73 In fue Taral) reasons for plantlng Masull HiIIs Tara; Adapted to varíous envlronments (except cold, shade) 100 100 Good graín qualíty and market price 100 100 Hígh tílleríng and low seed rate 100 100 Good storage and germínation 100 100 High straw yield and lodder quality 100 100 Does nol shaller bul easily threshed 100 100 Straw good lor mats 100 100 0005 nOI nead running waler 100 100 Volume increases upon cooking and is fairly heavy 100 100 Doas nol upse! following crops 100 100 R05ponds to moderale amounls 01 fertilizer 100 82 High yield 83 95 High milling recovery 69 90 High tlllaring wrth few unproductlva tillers 64 82 Heavy grain and minimum weight loss over time 59 85 Gradual tilierlng means less sl9m-borer damage 16 46 Rools nol vigorous but some are deep, maklng II suitabla for both lighl and heavy soll 3 28 Good lar double Iransplanting O 60 28 Bíological and Cultural Diversi/y Two dominanl varteties. Pokhreli masino covered 32 per cent of the hill farmers' rice area. Introduced in Ihe late 19605, il replaced many TVs. It was popular for a wide range of reasons (Tab\e 1.15). Ma5uli was Ihe only widely adopted MV grown in the lower Hills and in Ihe Tarai. It had replaced TV s since its adoption in Ihe 19705. Although farmers reported many favourable characteristics (Table 1.14), Masuli had reportedly started to deteriorate, as indicated by declining yield, increasing susceplibility lo pests, reduced tillering and increasingly unproductive tillers. Farmers are looking for a replacement. and have rotated it among rice environments, readjusted inputs, and exchanged Masuli seed from household lO household wilhin and between communitlcs, Table 1.15. Farmers' (n = 91 in the HiIIs, 99 in the Taraij uses of rice HiIIs Taraí Consumption Primary stapla 10 95 Secondary stapla 70 5 Livestock feed Straw 92 100 Rice hull, bran 79 100 Green foraga 18 21 Fuel Straw mixed with dung cakes 91 Straw 26 94 Rice hulls 47 Crop produetion Seed 100 100 Slraw converted lo FYM 100 100 Residues incorporated 100 100 Straw used as mulch 27 38 Slraw bumad lo enrieh soll 21 11 Income generation Grain is sold for cash or bartered 79 100 Milled rice Is soldlbartered 24 53 Straw is sold for cash or bartered 12 42 Milling by-products are .old 4 22 Miscellaneous Mats, baskels and ropes from straw 100 100 Foods from rice prepared for lhe sick 100 100 Ceremonial use 100 100 Poullry nesls from straw 91 54 Straw USed In pyre for cremating dead 71 94 Bear, liquar from graln 54 42 Roots usad as animals medicine 37 21 Roafing malerial from straw 35 28 Wall & floor plaster from strawlhulls 35 89 Rice cropping practíces in Nepal 29 Variety protective strategies Sorne of the rice practices of farmers were strategies for sustaíníng the varíetal viahílity. Among all crops, rice varíeties were relatively ,hor! lived. Among olhers, seed selection and varietal rotation were the main strategies. Seed selee/ion. Farmers generally used Iheír own home-grown seed. They selected seed to maintain the hest of varielal qualitíes, and lo preven! seed unavailahility at Ihe time of sowing, cash spendíng in huying sceds and hazards thal often accompanied sceds from oulside. Farmer' seed selection practices included: seleclÍon and collection of heavy panicle-bcaring matured, full and undamaged grains, tillers hearing heavy panicles, and whole plant CUI [rom the best yielding patehes - all from within Ihe erop field; selection and collection of heavy paniele lops, tillers bearing heavy panieles, or harvesl bundles containing heavy panides before slacking Ihe harvesl (seed viability was believed lO deleriorate during curíng in Ihe stack). The tirst type was more common in the HiIIs and lhe olher one in Ihe Tarai. Varietal ro/alion Rice varietal combination pattems were nol permanent arrangemenls of rÍee erop in relation lO olher crops, even wilhin lhe viable age of any rice variely. Farmers believed thal eontinuously growing a rice variely in the same land, year after year, would bring about the early varietal degradation, in terms of declined grain and straw yields, increased succeptibility to inseel and disease pesls, and dislortions in olher Irails. Unlike in Ihe case of other crops, in whieh the crop ilselr was rolated over space and time, the varieties were rolated in rice fanning, and was locally known as dhan ko nal ferne. The smaU landholding size with diverse micro-environment did nol usuaUy permit a planned rice varíetal rotation over space, whích was otherwise a eommon strategy among the medíum-sízed and large holders. Slabilizatíon of rice yicld was, however, more crucial for small farmers, and spalíal and temporal variety rotation was a strategy to attain il. Spatially, farmers changed growing-space variety after Ihree to five years, by growing it in other parcels wíth ncar suitable envíronments for it. Whenever Ihis was nol environmentaIly possible farmers exchanged the variety (seed) for anolher. The change in growing space often reslored sorne of lhe degraded qualities, if not aH. Farmers re-exehanged the variety after growing it for a rew years. When Ihe ínter-farm seed exchange withín the communily ceased to show desírable improvemenls, farmers exehanged the seed wíth those from distanl communilies, and if slill no sign of ímprovement was seen, ít was finally replaced. However, together with, prior to, or after Ihe spatial rotations, farmers al so rotated rice varieties temporally. Farmers often used differenl rice varieties for the spríng and the summer rice, hut Ihey also used sorne varieties grown earlier as spring crop the nexl year as summer crop, and vice versa. Also, farmers dropped sorne varietíes for a rew years and grew them again ¡horeafter by acquiring Ihe seed from olher households. Co,n dusion and implications Nepalese small farmers employ a wide range of Iraditional practices and corre­ sponding technica] knowledge in organizing rice farming. Practice, are well 30 Biological and Cultural Diversily adapted to fragmented land and other diverse and difficult environments, while techníeal knowledge has been developed over genefalions, Household rice land was a cluster of mulliple heterogeneous smaller pareels scat!ered over diverse ocologiea! pockels, The individual pareels were no! only dífferent with respeet to size and location, but also in terms of vital farming environmenlal anributes. Eaeh oflhe individua! parcels had a unique ecosystem, Overal!, farmers managed an aggregate of mulliple agroecosystems. Eco­ syslems characterization procedures need lo consider land fragmenlation and the parcel-specífic nature of farmers' adaptive 'Iralegies, Farmer,' environmen­ tal characlerization systems based on indigenous lechnical knowledge can be combined wilh currently used methods for analysíng rice agroecosystems al Ihe micro leve!. Such un approach can íncrease the effeotiveness of on-fann prior­ itization by dealing direclly with Ihe mullip!e envíronments (on smal! scatlered parcels) aClually co-ordínated aud managed by individual farmers, Farmers mental!y kept farm record s by individual pareels, aud Ihe same was their actual unit of farm operation, implying that individual parcels should be Ihe unit of data,galheríng for farming research, Farmers organized rice farming using diverse indigenous stralegies inheriled traditionally, but they also continuously modified and developed new farmíng techniques through Iheir own kinds of experiments, Farmers undertook experi­ menls for different reasons and had different types of results, The firsl rice experiment tested seedling,bed water-muddíness, seed source, water souree, fer­ tilizer, and, finally a new varíety in order lo find a spring rice cultivar and managemenl practices tbal would nol negatively affeet Ihe summer rice. Another rice experimenl tesled one cultivar in dífferent envíronments lo find it suited to a low, more water,logged environment. More important Ihan Ihe particular problems addressed, however. were sorne of lhe underlying fea tu res of these farmers' experiments. o Farmers no! only Iried lo solve existing problems (e.g, insufficienl food supply), but also tried to create new allernatives (e,g, a spring rice), lesled traditional practices (e,g, seed and fertilizer rates) and tested new questions (e,g, the effeets of water souree or temperalure), o Farmers experimenled with both erop componen!s and cropping patterns, viewed problems holistieally (in terms of on-farm environmental diversity and the 'fit' of crops over time) and, therefore, automatically conducted 'sys, tems researeh'. o Farmers ehanged experimental dcsigns, treatments, and variables in Iheir repeatíng of experiments in succeedíng seasons aud years, based on Iheir previous findings. Bolh farmer. and scientists are agricultural researehers, Achieving sustainable produetion under farmers' conditioos would clearly be easier if partnerships werc fonned belween experimenlÍng farmers and experimental scientists, Attaining sueh parlnerships. however, will require that scÍentists first value, and second undersland, farmers' research capabilities. Rather Ihan Ihe learniog of com­ pletely new practices, farmers may need more assistance in Iheír researeh tech­ niques at the 'adaptive research stago' of lechnology testing and diffusion, Farmers in Nepal planted a wide range of TVs and MVs for a wide range of uses, Farmers soughl. evaluated, and adopted or rejected varíeties; and their selection criteria provide directions for varietal improvement research, Vigorous tillering with fewer unproduclÍve tíllers, fewer but deeper rools, shorler duralion, Rice cropping practices in Nepal 31 good storage, high-quality straw, pesl resistanee, and tolerance lO environmental ftuctuations were farmers' rice-breeding targets, Farmers' indigenous tests of various aspects of rice varieties also reflected their varietal selection critería, which could be categorized into environmental, technical, socio-economic criteria, and others, Consideration of such characteristics means that rice varieta1 improvement need no! oceur at Ibe expense of existing use fuI varietal attributes, 2. Farmer-based experimentation with velvetbean: innovation within tradition D. BUCKLES and H. PERALES1 Abstraet loCAL KNOVlLEDGE AND farmer perspectives were applied to the development of new cover-erop management strategies for maize-based systems in southern Veracruz. Mexico. Farmers' spontaneous innovation with velvetbean (Macuna prurims) was the basis for four cydes of farmer experimentation, resulting in a new management strategy acceptable lO farmers (a mid-season velvetbean inler­ crop wilh Ihe pOlential to amelíorate declining soil fertitity. weed invasion, and drought stress). Farmer perspectives on maizelvelvetbean assodalions are línked to the conceptual framework of shifting cultivation practised for centuries by the indigenous population. Introduction Recent appreciation of farmers' historieal roles in developing and adapting traditional agricultural syslems has focused attention on Ihe potenlial of farmers lo contribule to developing new strategies for sustainable agriculture. Scientists have tried to tap into the innate creativity al' farmers and their intimate knowl­ edge of thoir immediale environment through enhanced [armer participation in problem identification (Harrington el al., 1992; Chambers, 1990; Fujisaka, 1989; Lightfoot el al., 1'}88: Lightfoot, 1987) aud the evaluation of agricultural tech­ nologies (Graf el al., 1991; Ashby, 1990). Bentley (1994) poinls out, however, Ihat examples of offeclive collahoralion between farmers and scientis!s resulting in the development of new technology options are stin rare. Ideas and methods for solving farmers' problems are typically sought in the realm of science. and subsequently introduced to farmers for adaptation lo local conditions. This paper examines the application of loeal knowledge and farmer perspec­ tives to Ihe development of new cover-crop management slrategies for maize­ based systems in the Sierra de Santa Marta in southern Veracruz) Mexico. Spontaneous farmer innovatian with velvetbean (Mucllna pruriens), a vigorous, annual climbing legume, ís described and new velvetbean management aption, developed through collaborative experiments are discussed. Farmer perspectives on maize-velvetbean associations are línked lo the conceptual framework of shifting cultivation practised for eenluries hy the índigenous population. bring­ ing lo líght lhe creative tension between tradition and innovation characteristíc of local knowledge systems. The research indicates tha! local knowledge can be a fruitful paint of departure for collaboration with rescarchers, and farmer parti­ cipation an effective 1001 in developing new technology options acceptable to farmers. It also shows. however, that potential complementarities with science­ based knowledge and experimeotal methods should not he overestimated. lnsights gained through farmer particípation in key researeh decisions are made al considerable cost to lhe reliabilíty of agronomic data and should no! be seen as a substitute for more conventional on-farm experimentation. but rather as an early phase of research on new technologies. Experimentation with velvetbean 33 Farmer innovation with velvetbean Tbe Sierra de Santa MarIa rises sleep!y from sea level al Ihe Gulf of Mexico lO more Ihan 1700 m al ils highest peak (Map 1). The soH, of the Sierra (Andosols aud Alfisols) are moderately fertile bOl highly susceptible lO erosion. More Ihan 3000 mm of rain faH annually on Ihe northern, seaward stope of Ihe range while on Ihe southwestern s!ope annua! rainfall average, sorne 150ú mm. The rains begín in late May or June, peak in October, and gradually drop off through January and February. A shor!, sharp dry scason extends from April through May, interrupting mosl agricultura! aClÍvities. Climatíc conditiol1s facilitate two growing scasons per year, lhe temporal or UNITED SfATES G O Gull 01 ( MOluco '" O l2' ~ l2' ..¡- / C' O t N I I: ÑV'noI--:-'" ;;;;;;¡¡¡¡¡OO"""·~·~I.·. ! ! <1 - '. ... _.~ __ ........-J Map 1 Sierra Sama Marta research area 34 Biological and Cultural Diversity summer scasan and the lapachole or winter scason. A wide range of annual and permancnl erops cao be growo in the Sierra, bul maize aod beans are Ihe mosl important to the indigenous population. Nahua and Popoluca speakers havo inhabíted Ihe Sierra sioee pre-Hispaníc times, using techniques of shifting cuhi­ vation to produce basic foodstuffs. Fallow land, known as an acaual, is slashed and burned prior to Ihe summer scason in preparation for planting maize in lune. Only local varieties are used. Crops are weeded manually or in combina­ tion with herbicides and doubled after reaching physiological maturity 10 facili­ tate drying in Ihe fleld. Fertilizer use is rare, mainly due to cash constraints. Periodic rainfall from November through February permits a winter maize erop lo be grown in most of the Sierra. Winter maize is planted between the rows of doubled summer maize. Weed and erop residues from the summer season are nol burned prior lo planting wioter maíze, bUI rather are chopped and left on the field as mulch to conserve soil mobture. Because of the relatively dry winter conditions, maize plants do not need to be doubled prior to harvest and only one weeding is required, with the advantages 01' lower labonr costs compared with the summer season. Nevertheless, maize can be flattened by strong winds during the early part of the scason, and the risk of crop failure from drought is high during the later part of the season. Yields of winter maize average sorne 800 kgl ha, whereas summer maize yields average 1.4 kglha. However, intercrops such as beans aud eassava, semí-permanonl erops sueh as plantains, volunteer plants used for food (qui//ies), and tree species used for flrewood may a1so he harvesled [rom the same plOI, mainly duriog Ihe summer season. While these basic techniques have varied li!tle over cemuries ofuse iu lhe region (Stnart, 1978), lhe broader environmenl has. Since the 19 50s, Ihe expanding caltle industry aud populatinn growth have increased pressure on land resources, resulting in rapid deforeslation, more intensive cropping patterns and land degradation. Although, tradilíonally. fallow periods of eigbt years or more were used to recover the agricultural potentíal of cultivated laod, fallows have been reduced lo two or three years in many parts of the Sierra. Fallow frelds are no longer dominated by the tree species characteristic of the traditional fallow, bul by grassy weeds. Slashing and buruing grassy fallow eaDllot support erop production without drawing heavily on the limited resourees of the soil and hefty investments in weed controL Declining soil fertilily and weed invasion are currently Ihe mosl important constraints 00 maize productivity in the region (Chevalier and Buckles, 1995; Pare el al. . 1993; Perales Rivera, 1992). Farmers in the Sierra de Santa Marta have not been passive in the faee of these problcms. Interviews during 1991 deteeted a number of regional farmers growing velvetbean in their maize fields to improve soil fertility and eradicate weeds. Velvethean is originally from India or China. where it was at one time widely cultivated as a grecn vegetable (Wilmot-Dear, 1987, 1984; Burkill, 1935). It was probably introduced into tbe Carribean by East Indian indentured workers during the late 19th century and later adapted by orange growers and maizc farmers in parts of the southern U nhed States as a soil-improving crop aud ca!tlc and pig forage (Scott, 1919; Tracy and Coe (918). Transnational banana companíes may have introduced velvetbean into Central Ameríea early in the twentieth century as a forage erop for mules (Buckles, 1995). Jt was subsequently adapted to maize-based farming systems among the Ketchi of Guatemala (Carter, (969), the Chontales 01' Tabasco (Granados Alvarez. 1989), the Mames of south-western Chiapas, and Chinantecos (Arevalo Ramirez and limenes Osornio, 1988) aud Mixes (Narvaez Carvajal and Paredes Hernandez, 1994) in Experimentation with velvetbean 35 lbe Isthmus area of north-western Oaxaca. It also found its way into the Sierra de Santa Marta in soulhern Veracruz, where it drew the attention of Nahua and Popoluca farmers. Two distinct management practice, were idenlified during informal field surveys. Farmers in San Pedro Soteapan, Ihe cultural cenlre of the Sierra Popoluca, indicaled that they had encountered velvetbean growing wild in tbeír ficlds and noted it, ability lo smother wceds and improve maize yields. They collected seed and broadcast it over a larger area, giving rise to a praclice known as 'making a fallow ficld' (Sp., hacer acauaf). Farmers use velvetbean seed on a maize field they intend to fallow beca use of declining yields and weed invasion, typieally at the end of the summer maize cycle. The crop develops fm several months, sets sced, and naturally re-establishes itgelf during subsequcnt scasons. According 10 experienced farmers, weeds are eliminated by the aggressive velvetbean crop and soH fertility is regained: maize yields on land 'improved' "'th velvetbean for two years rival yields on land fallowed for tive years with native trees and shrubs. Fallow enhancemenl with velvetbean .ecms lo be an endogenous innovation in direct response lO declining land and labour productivilY resulling from reduced fallow periods. Farmers can increase Ihe frequency of cultivation wilhout recourse to external inputs and yet mainlain acceptable levels of production. The practi. .' e of 'making a fallow ficld' wilh velvetbean is recognized locally as an improvemenl broughl lO lhe ficld by Ihe farmer, and even helps establish customary rigbts to particular pareels of land, In 1991, an estimaled three-quarters ofthe ejidatarios' ofSan Pedro SOleapan reported the presenee of velvetbean in theír ficlds (Perales Rivera. 1992). It had appeared spontaneously in the fields of half the growers, while lhe rest had origina11y planled lhe crop. Nevertheless, Ihe management of velvetbean to improve soil fertility and control weeds is somewhat hapbazard. Farmers do not plant velvetbean systematically on a11 fields abandoned to secondary vegela­ tion or plant the crop every year, bul rather rely on natural propagation. As a result. the velvethean-stands observed in San Pedro Soteapan were uneven and scaltered in small pockets within farms. Few farmers had velvelbean fields larger Ihan 0.5 ba. The casual qualities of improved fallow management using velvetbean observed in San Pedro Soteapan contrasl sharply with a more systemalic strategy noted among Nahua [armers in the neíghbouring ~iido of Mecayapan. Farmers reported Ihat velvctbean is stick-planted at the end of Ihe dry season in winter maíze fields and allowed to develop as asole crop throughout the gummer season. The abundan! velvetbean growth, known locally as a picapica, is slashed in Novcmber, and winter maize is stíck-planted into the mal of decomposing Icaves and vines, wbere it develops relatively free of competition from weeds. In keeping witb tradítional winler maize cropping practices, farmers do not burn velvetbean residues or incorporate Ihem into the soil bul ¡eave Ihem on the surface as mulch, Velvetbean seed that has matured in Ihe field germinates during the winter, naturally reseeding tbe picapica without significant additional ¡abour inputs. Sorne farmers reported more than 10 years of continuous crop­ ping of winter maize following velvetbean, Yelvetbean summer rotations used by farmers io Mecayapan resemble strate­ gies employed by farmers in Tabasco (Granados, 1989}, Guatemala (Carter. 1969), and Honduras, suggesting that lhe practico may have beeo borrowed and adapled from olher areas. Farmers from Meeayapan tirs! planted velvetbean 36 Biologícal and Cultural Diversíty in Ihe late 1940s in an unsettled area known as Pozo Blanco in the current ejído of Reforma Agraria'- The practice was later adapled to sleeper hillsides when land reforms foreed farmers to Iimit agricultural activities to their home village. Farmers' use ofvelvetbean rotations with winter maize in Mtx:ayapan seems to have been stimulaled by significant land and labour productivity advantages compare0.50 2 (dry) 3 650 (117) 566 (44) >0.50 3 (wet) 6 1.245 (168) 1,345 (233) 0.30 4(dry) 3 499 (96) 180 (SO) 0.01 Note: Pis the sígnificance level tor a two-tailed paired 't' test. In contrast to the positive effeets noted during the winter cycle, no significant differences hetween the control and velvetbean trcalments were found during lhe summer eycle (Table 2.4). High variability between farmers may partly explain the non-significance of these resuhs; differenees between the control and velvet­ bean t¡calment ranged from -420 to +760 kglha during the tirst summer season and remaíned high during the seeond summer season. Despite high variability, however, a slatistieally significant response to the applieation of fertilizer was encountered; maize yields were more than 700 kgiha higher on plots receiving fertilizer compared with other treatment, (P probability <0.0001). This response indica tes Ihat even though the variation between farmees was high throughout the experiment, a dear treatment effeet could he determined under Ihe experi­ mental conditíons. Thus the absence of a significan! effeet was more Iikely due to the modest development of velvetbean 00 maoy fields aod very high rates of mineralízatíon of velvetbean resídues duríng the winter cycle. Even on fields where veJvethean growth had heen good in the previous year, very Iittle velvet­ hean cover remained on the ground at lhe beginning of the following summer cycle. While incondusíve, Ihese observations suggest Ihal the benefits of velvel­ heanlsummer maize associations may not carry over ioto the following Summer cycle, or are too weak to measure under farmer-bascd experimental condítioos. Velvethean effeels on weed populations followed a similar pattern. Weed measurements were taken 20 to 25 days alter maize sowing in 10 randomly ehosen, 1m squares delimited by four hills, In fields with vigorous velvetbean growth in the summer cycle, weed suppression was apparent during the following winter cycle; maize wíth velvethean had less weed cover (20 per cent or less) compared with the conlrol field without velvelbean (60 per cenl weed cover). This sizable effect possibly resulted from weed suppression eaused by velvethean competition with weeds during the growíng season. The effeet of the velvethean intererop was apparent during the following summer season. Again, the limited development of the velvethean crop during the tirst cyeJe and Ihe smal! amount of residue left on the tield at the heginning of the following summer cycle may have accounted for the absence of observable weed control effeets in summer maize. These observatiolls raised concem over Ihe potential impact of velvetbean intercrops on summer maíze. The agronomie data aad researchers' aS"eSSments, although not conclusivo, suggestcd that lhe benefils of velvetbean intererops in summer maize were limÍted to cropping patterns involving winter maize, at leasl during the first year or two after establishment. Summer maize díd not ,eem to henefil ímmediately from management of velvelhean in the previous summer scason, This scemed to be lrue both in fields where velvetbean was left to develop 44 BiolQgical and Cultural Diversil}' as a wínter fallow and in fields where winter maize was planted. While conven­ tional agronomíc trials with adequate repelítions and control of experimental error may have detected measurable differences ín summer maize attributable to velvetbean associations, these effects would probably have becn limited and would certainly havc becn continued to demonstrate great variabílity. Farmers' objectives and evaluation criteria Farmers' evaluation ofvelvetbean associations with summermaize, while general1y consistent with researchers~ observations, was much more favourable. Farmers confirmed tha! a velvetbean association in summer maize, followed immediately by wínter maíze, resulted in a notable yield and weed-contml benefil during Ibe winler cyele. These e!Toet. were attributed by farmers to ímproved soil fertility, soil moistnre conservation, and wecd suppression by !he velvetbean cmp, a perspective compatible with rescarchers' observatíons, Farmers al50 c1aimed, however, Iha! velvetbean planted in summer maize improved field condítions for subsequent summer maize crops - effeets not captured in agronomic data collected by the researchers. Differences between science-based and farmer cvaluation criteria may partly explaín discrepandes betwecn researchers' and farmers' observations. Farmers' evaluation of the trials was more comprehensive and holístic than the partial, fragmenled evaluation realized by tbe rescarchers, a perspective on tcchnology now generally appreciated by rescarchers (Bentley, 1994; Byerlee, 1993; Asbby, 1990; Norman el al., 1989). While the rescarchers foeuscd on short-term yield and weed control effeets, farmers were equally concemed with the multiple and incremental effeets of velvetbean associations on field condítions. Dudng inter­ víews in farmers' tields at the end of each cyc1e, farmers reporled a wide range of positive effects of velvetbean associations with summer maize, including improved soi1 fertility, reduced weed populations, belter soil structure ('softer soil'), less soíl erosion, beller moisture conservation duríng the winter scason, and reduced damage to maíze from soil pests. When ranked in order of impor­ tance by farmers, these factors highlight the importance of soil fertility benetits and labour savings among farmers' priorities (Tablc 2.5). Thís perspeclive reHects the multiple objectives offarmers and their perception thal the technology has the pOlential to respond to a nnmber of pmductíon constraints .ímultaneously. During Ihe trial, farmers managed velvetbean in difTerenl ways to respond lo di!Terent problems, sometimes even within the same field. Farmers concemed Table 2.5. Beneflts of velvetbean soociations with summer malze identified by farmers (no, of fanners) Beneflls Flrst seleetion Second selec!ion Improved soi! fertilíty (abono) 12 5 Weed control (aplasta malezas) 4 12 Moísture conservation (conserva humedad) 2 2 Erosion control (no se lava la tierra) 1 O No response --3 -----3- Total 22 22 Note: Farmers ranked vefvetbean benefits using cards depicting fue characteristics. Experimentation wilh velvetbean 45 Table 2.6. Main reason glven fo, selecting a velvelbean management option. tirst season (no. of farmeNl' Option se/ecfad Main reason Summer rotation Summer intercrop Eradicate weeds 6 3 Improve soil fertilíty 1 14 No reason given 1 3 Total 8 20 Note: Slx farmers selected more than one option. primarily with eradicating persistent weeds from their fields typically opted for velvetbean summer rotations, while farmers concemed mainly about declining soil productívity int.reropped velve!bean in summer maize (Tab1e 2.6). Si. farmers diversified velvetbean management strategies wi!hin !heir fields, planting summer ro!ations or improved fallow in areas invaded by weeds and intercrops in maize fields. Several farmers managed parts of their fields with velvetbean plan!ed 50 days 01 so after summer maize lo improve soil conditions and control weeds during the win!er cyele, while other pans of !heir fields were managed with a midseason velvetbean intercrop followed immediately by winter maíze. Farrner adaptation of velvetbean managemen! slra!egíes lo particular field condifions and production constraints suggests !hat one strength ofthetechnology ís ílS flexibílíty in Ihe race of diverse production environments. Velvetbean can be 'applied' !o fields in different ways as nceded, akin to a 'component' tcchnology such as fertilizer or herbícides. An implícatíon for research is that adoplion of legume associations may be favoured by the devclopment of a wide range of flexible management options rather !han lhe refinement of a single, 'ideal' practice. During lhe evaluatíons, farmer. also expressed concerns about possible pro­ blems arising from the use of velvelbean in lheir fields. Many índica!ed Ihat rals could become a problem as !he pest ís attracted to areas wilh protective ground cover. Sorne Doled Ihat rats had c1imbed up the vines of the velvetbean íntercrop to reed on ripening maíze plants. A few farmers rel! tha! the amoun! of time required to harvest summer maize íncreased owing lo the ahundant growth of velvetbean vines coveríng the doubled maize. Thís was particularly problematíc in cropping patterns involving winter maíze, because land preparations were required long before the velvetbean erop had stopped growing. Perhaps Ihe mos! important limÍlalion on velve!bean assocíations wíth summer maize noled by farmers was íncompatibilily with traditional intercrop­ ping and following practices. Velvetbean in summer maize competes dírectly with volunteer planls used for food (quilites), intercrops such as beans and cassava, semi-permanent crops such as plantains, and teee specíes used foe fire­ wood. While velve!bean associatíons provide a mean S of intensífying maíze production, farmers noted that tbe strategy is appropeiate only for pans of the field sole-cropped with maize. De.pi!e Ihese constraints, virtually all farmers participatíng in lhe trial contínued to plant velvetbean in summer maize field. independently of !he researchers, and more than half of the farmers increased tbe arca under the technology rrom one year to the next compelling evidence of farmers' in!erest 46 Di%gical and Cultural Diversity in lhe practice. Farmers noted that direet costs of lhe technology were limited to collecting sce 58 Biological and Cultural Diversify Tabla 3.4. Characteristics of upland rice cultivars deslre '" g Table 3.5, Farmers' (n = 32) rainfed lowland rice cultivars and evaluations of cultivars, Solana, Philippines, 1992 (continued) Characteristics Cultivars Improved Wagwag Fino Ramsey Java Raminad Malagkit Others ~ 15' ¡¡- Negativa characteristics: late maturing O 44 O 11 O O O "[" insect and disease problems 33 22 O O O 33 O low tillering O 22 11 O O O O ;'":, lodging 50 22 O 11 O O O "'- less drought tolerant 17 11 O O O O O ¡¿ shatlers O 11 11 O O O O :::- requires high input levels 83 O O O 33 O 20 "~ low grain weight O O O 6 O O O difficult to thresh O O O 11 O O O o ~. early maturing O O O O O O 20 ¡¡ poor eating quality 50 O O O O O O ~. none 17 O 44 O 33 O O Side-stepped by the Grem Revolution 61 Table 3.6. Rice cultivar characterístlcs desired by farmers, Solana, Philipplnes, 1992 -Cha_racteristic (%) .. _----- Farmers Good qualíly eatíng/aromallc 41 Hlgh yleld 28 Flood tolerance 22 Low Inpul levels requlred 19 Drought tolerance 19 Long, heavy panicles 16 Early maturing 16 for hígh input Jevels and good water control. Traditional eultivars were desired for good eating quality (although no! al! tradítionals), taH stalure and fiood toleranoe, late maturation, and low input requirements (Table 3.8). As in Solana, Tarlac farmers planted the MVs ín areas having good water control or supplemental irrigation, and matched theír traditional cultivars to lower areas where suhmer­ gence was cornmon, Discussion and conclusions The 'Green Revolution', in terms of developíng, íntroducing, and adopting suitable new varieties, bypassed the less-favourable rice enviroomento - or less­ favourable areas within particular regions, Traditíonal cultivars dominate the uplands of Bukidnon; and traditional and improved or modern varieties dom­ inate in Claveria (upland) and Solana (rainfed lowland), Although Tarlac farmers planted MVs on portions of their rainfed lowland areas having good water control, they had lO continue to rely upon traditíonal cultivars in Ihe remaining areas. Farmer adoption of suceessful tradítiooal varieties cao lead to 1055 of local agrobiodiversity. Examples in which one or two traditional rice cultivan came to dominate (in terms of numbers of farmers plantíng and area planled) are Dinorado in Bukidoon, Speaker and UPLRi5 in Clavería, and Wagwag Fino and Ramsey in Solana, The maintaioing of a range of tradítional rices, but with emphasis 00 a rew has becn eocountered in other upland rice arcas: in Laos (Fujisaka, 1991), Myanmar (Fujisaka el al., 1992), and Indonesia (Fujisaka el al., 1991b), Similarly, in the rainfed lowlands, a few dominant cultivars among many often correspond to [armers' malching of cultivars to lower flood-prone, and míddle and upper drought-prone lerraces (Fujisaka. 1990, Fujisaka el al" 199Ia), Cultivars are oceded that match the local diversity of unfavourable environ­ menls. AIJ farmers do not grow Dinorado in the areas surveyed in Bukidnon, apparently because of slight dilTerences in elevation and temperature. Many farmers lested Dinorado in Claveria (many Claveria selllers came from Bukidoon), but Ihe cultivar was less suíted to lhe area beca use of high losses to birds. Farmers in Solana and Tarlac plant MVs where possible, but continue to match traditionals with dilTerent hydrological conditions on less favourable lan,ds. Farmers' and rice brceders' main se1ection criteria are more similar than different, but are not eompletely congruent. Combined seIs of crÍleria are ideally Table 3.7. Farmers' (n = 33) rainfed lowland rice cultivars and evaluations of cultivars, Tarlac, Phllippines, 1992 el' Characteristícs CuitivafS '" JmpfOved a Benser Okinawa Pagay lioko Hamog b Farmers growlng cultivar (%) 78 30 21 18 12 Mean area (ha) 1.3 1.5 0.8 1.2 3.0 Mean yield (t ha-l) 4.2 3.0 2.1 1.6 1.8 Positive characteristícs: high yield 88 60 14 O 25 earty maturing 50 O O O O soft when cooked 42 O a 60 a t:. good taste/eating quality 31 100 28 100 100 ,,' expands when cooked O 70 28 O 75 ¡;- fOf delicades O 30 56 o 25 ";:-;. white graln o 20 56 83 50 '" tall stature O 100 100 100 100 -§ !Iood toleran! O 100 86 100 25 "'- late maturity O 100 100 100 100 low inputs required O 50 100 100 100 ~ pest and disease resistance 7 90 14 83 25 ," weed toleran! O 30 71 83 75 drought toleran! O 50 14 O 50 -"\:;, ~. long grain O 40 O O 25 ~ high market price O O O O 25 ¡;¡ ~ . Negative characteristics: .;;¡ need. hlgh input levels 100 O O 16 O needs goOO water control 57 O O O O hard when cooked 15 50 100 O O lodging 15 30 71 O 50 poor eatlng quality 12 O 56 50 O susceptible lO drought 7 O O 33 O difficull to Ihresh O 50 O 50 O low yield O O 71 50 75 low millíng recovery O O O 83 O a IR4, IR20. IRaS, IR42, IR48. ¡RS9, IRBO, IR64, IRB6, IR70, BPIR10, and Super. b. A few tarmers also plantad Pínarya. Lamia, Dinaloson, Magsapa, Numero Doce. Azuceña, and Los Baños. Side-Slepped by Ihe Green Revolulion 63 Table 3_8. Farmers' In : 33) desired traditional rice cultivar characteristics, Tar­ lae, Phlllppines, 1992 Characferisfic Farmers choice (%) Good eating qualitylaromatic 100 Late maturing 54 TaJl stature 48 High yield 36 Expands when cooked 18 White 12 Flood tolerance 9 Low input levels required 9 suíled lO newer breeding strategies wherc internalional agricultural research centres provido donor lines featuring major resistance or adaptability, and nalional programmes make more specifie eros.es and selections to develop loeally appropriate varieties. Both upland rice farmers and breeders .eek high yields, adaptability 10 local soils, medium duration (not so 5hort that the erap is unstabic in the face of stresses), lodging resistance, high milling recovery, and pest and disease resistance. Farmers. however, al so consider Iraits such as taste, maintenance of taste in sto red graín. hardness of cooked left-overs, attractivc­ ness to birds and ehickens, and the difficulty of handling awned rices. Farmers in the rainfed lowlands, who face long periods of fiooding, chose cultivars that are tall and purposely of long duration (i.e., photoperiod sensitive) - characteristics quite unlike the short-statured and short-duration, Grecn Revolution rices. Farmers' knowledge and selection criteria related lO fice cultivars are sub­ stantially 'scientifically' sound. For example, farmers are correet in Ihat plant architecture, grain ,ize, and relative tíming of graín development infiuenee bird losses; Ihat plan! architecture infiuences weed competitiveness; thal aromatic plants (e.g., Speaker) can be relatívely more attraetive to sorne inseet pests; thal sorne cultivars are more tolerant of acidic or poor soils; and tha! sharp Bwns are difficult to handle. Providing a 'Green Revolution' in the uplands and rainfed lowlands may be difficult precisely because farmers make sound choices, practice sound manage­ ment, and are successful to Ihe degree thal risks inherent in lhe systemallow.ln the 1988 wet scason, a good year in Claveria in terms of rainfall, farmers' traditionai varíeties on their own fields yielded a mean 1.9 tlha; UPLRi5 without fertilizers yielded a mean 2.8 tlha; and UPLRi5 wíth small amounts of inorganic ferlilizor (25kg N, 6kg P205, and 5kg P20) yielded 3.2 Ilha (Fujisaka, 1991). Conversely, in 1991. a drought year in Clavería, all rice fields of upland humers yielded a mean 0.8 t/ha (Fujisaka, 1993). The uplands and rainfed lowlands pose substantial challengcs and olfer sig­ nificant opportunities for problem-solving research. 'Modern' varietíes, which address the diverse. risky condítions inherent in Ihese environments, and which are clearly superior to traditional cultivars, have nol yet been developed. The opportunities lie in the diversity of traditional rices that farmers have selected over time for suitability lo local conditions, and in underslanding and applying lhe knowledge underlying lhese selections. 4. Environmental dynamics, adaptation, and experimentation in indigenous Sudanese water harvesting DAVID N/EMElJER Abstract $OME 2000 YEARS ago the Beja of North·eastern Sudan first engaged in runoff farming as a side·line lo lheir mainly pastoral life. At present il has become a vital souree of subsistenee for several of the Beja ethnie groups. The best·watered soils were losl lo large·seale irrigatíon schemes and the traditional desert trade was supplanted by modern means of transportatioo. 00 the arid piedmont plains to lhe east and north of Kassala town lhe Hadendoa and Beni Amer peoples, two of Ihe Beja ethnic groups, engage in three forms of indigenous runoff farming: wild Hooding, waler spreading, and teras water harvesting. The highly dynamie environment with low aod irregular summer raiofall and conlinually shifting wadi courseS is hard lo master. Nevertheless, the local agro·pastoralists have succeeded in developing highly adapled [orms of Ihese indigenous runOff farming syslems. The key lo their long·term survival is the degree lo which they have becn successful in coping with environmental change and dynamics. As wadi courses shifl and rainfall varies, Ihe Hadendoa and Beni Amer are able to a!tain a reasonable and sustainable harvest only througb conlinued experimentation. Espccially among the Hadendoa Ihis has Jed lo eXlremely varied shapes and dimensions in their water·harvesting structures, as eaeh farmer altempts to cope successfully wilh Ihe constraints and opportu­ níties tha! a spedfie sile offers in Ihe eourse of time. For the Beja, as for many other agriculturists, experímentation ís lhe key to adaptation - a fael lhal, once recognízed, has importan! implications for attempts al lechnology transfer. Introduction Long before lhe onsel of agriculture the survival of humankind depended on adaplalion and experimentation: adaptation to Ihe dangers and opportunítíes of lbe environment and experimentation with tools and techniques ín order lo further master lhat environment.! Not only did adaptatíon and experimentalíon eventually Icad to lhe development of agriculture, bul these eharacterístic, continue to be among the vital forces that support agricultural livelihoods in our ever.changing world. Leakey (1936) was among !he tirs! lo recognize lhe well adapted nature of Afríean agriculture as he encounlered it among lhe Kíkuyu. In his víew. local methods should not be replaced by (less adapted) European ones. Still, he suggested, they could he improved through scientifie research, for he belícved scicnlific research to be more powerful than what he caHed lhe trial and error rcsearch of lhe Kikuyu. In mentíoníng that 'the methods of researeh by trial and error have led the Kikuyu lo important conclusions' he nevertheless acknow· ledged lhe value of farmer experimentation (Leakey, 1936)" In the eurly 1970s Johnson (1972) drew attt'l1lion to índigenous experimenta· tion again, and maintaíned that índividual diflerences in agricultura1 practice Indigenous Sudanese Water Harvesling 65 Southem and Centra! BorderArea Jebel Yodrot lIatAyot • <> Tibiai Afea .. • Jebel Hourl! Eritrea / , / f f Khartoum • ...- f~/r -------. /' Sudan ", t'" I - - - Natiooa! border I Ral!way I --- Tarmac road I ~ ViUage I ~Wad¡ - '-----' (col.khor) I ¡ ~ Inselberg I : ... (col-¡ebef) I Figure 4.1 Maln features of the southern and central Border Area; adapted from Van Dam and Houtkamp (1gg2) and Westhoff (1985) 66 Biological alld Cultural Diversity and systematic experimentation were probably pervasive in Iraditional agricul­ tural communities and 3n essentÍal component of their adaptive processes; a view that has since gaincd increasing power especially among anthropologists (see for example Richards, 1985), and has led lO a series of new publications starting in the early 1980s. Much of the recent material on farmer experimenta­ lion (for example Rhoades and Bebbington, 1988; Brouwers, 1994; Dangbégnon 1994) [ocuses on agrouomic measures, new crops, aud cultivar seleclion aud breeding (especially with the current interest in biodíversity). In contrast, this chapter will foeus on experimentation in anather field, namely that of ethno­ enginccring. Ethno-engineering is a term that is sometimes used to denote physical structures that local farmers build lO serve tield crop production (Reij. 1991). [rrigation canals. or soil-conserving stone lines are good examples of such ethno-engineering practices, as are the brushwood and earth bunds of the Sudanese Beja peop1es lhal we will discuss here. The material presenled here is based on researeh among the agro-pastoral Hadendoa and Beni Amer, two of Ihe Beja cthnie groups, that live in the Border Area of Sudan's Kassala Provinee (an 8600 square kilometre pie-sliee bordering Eritrea and located to the north and east of Kas,ala town, see Figure 4.1). These agro-pastol"dlists employ a range of indígenous runoff farming techniques for sorghum (Sorghum bicolor. S. vulgare) and millet (Pennisetum typhoideum) eultí­ vation. Through eontinued experimentation and adaptatÍon they have deve10ped an arable farming system that is remarkably viable and sustainable eonsidering the harsh and dynamic environment. Fíeldwork methods The present report is based on field research conducted in Eastern Sudan in Ihe period Oetober 1990 to January 1991.2 Al the time there was a majar drought (with 76 mm rainfal!, only 25 per cen! of lhe 10ng-term average, it was one of the driest years of lhe century) and this was one reason to foeus on lhe engineering structures rather than on agronomic praetices. It also 1imited the numbor of farmer interviews, because many farmers left the village in seareh of alternative sources of live1ihood. The material presented here is based on detailed tíeld observations in several farming areas of the southern and central part of the Border Area,3 informal interviews with key informants and farmers at field sites, leve1 surveys of three runoff farming fields (rerus), air photo interpretation of farming areas, soil survey work, and soil sample analysis. In total, over 30 scattered runoff farming fields were visited. Hístorical background4 Somewhere betwccn 4000 BC and 2500 BC lhe Beja erossed the Red Sea from Arabia and settled in Sudan's Eastern Desert (Paul, 1954). From what is eurrently known, livestoek-keeping has beeo the e\dtural!y preferred mode of subsistence for the Beja ever sinee. Nevertheless, they have a1ways engaged in various ather activities. A1ready befare the Christian era they developed exchange contaels, and by the eud of lhe tirst millennium the Beja had intense trade relationshíps and conducted earavans through the desert, taking care of the local supplies and controlling the hinterland (Hjort af Omas and Dahl, 1991; Pa1misano, 1991). Some 2000 years ago runofffarming gradually replaced wild grain eollection5 and probably took place mainly in the form ofwadi cultivation Indígenous Sudanese Water Harl'esting 67 (Hjort af Ornas and Dahl, 1991; Ausenda, 1987), A notable exception was ¡he large inland delta arca of Ihe seasonal Gash river Ihat allowed Hood-recession agriculture, Ibn Hawqal, an Arabic lraveller from lhe tenth century already noted sorghum cultivation in the Gash delta (Van Díjk, 1995: Hjort af Ornas and Dahl, 1991), and as early as Ihe fífteenth century Gash sorghum was a much famed export prodllct ofnorth-eastern Sudan (Hjor! afOmas and Dahl, 1991), Frorn the literature (Newbold 1935; Paul, 1954; Hassan Moharnmed Salih, 1980; Hjor! af Omas and Dahl, 1991; Serb0, 1991) il becomes clear Ihar lhe Hadendoa and Beni Amer peoples, the two Beja ethnic groups that O~'Cllpy the Border Arca, the territory on whích this paper focuses, traditíonally lived dispersed in small family groups, These unirs lived a more or less sedentary life at a number of seasonal sites. Year after year they would retum to the same dry and wet soasan sites, This scattered way of life was a successful means of eoping wirh the searee environmental resources that a harsh climate allowed, It guaranteed a wel!-balanced population distribulÍon over the whole area, Largo­ seale scasonal migrations were nol necessary, thus sparing the environmen!. Only under certain conditions (lilee very low rainfall) larger scale migrations took place lo lhe Gash and Tokar deltas or to the winter rainfal! arca near the Red Sea. Places like the Gash delta, lhe Red Sea hills and the Eritrean highlands offered dry season's grazing, whereas lhe wet season allowed a wide dispersal of man and beas!. The Boja had achieved considerable flexibility in lbeír exploita­ líon of a natural environment characterized by great fluctuations both in lbe amount and the nature of available resourees, through lheir socio-cultural syslem that supported dispersed residenee, and through the availability of drought-time refuges, As happened everywhere in Afríea, the way of life of the Boja was seriously affected by the establishment of colonial power in the early nineteenth century, Taxation, though certainly at first very ineffective (Hjor! af Omas and Dahl, 1991), influenced agricultural and settlement patterns. Tax gathering expeditions had to be avoided, leading to additional movements of stock and conRicts in the tribal-hierarehy abou! the contributions to taxation, Conflíets over land and land use could no longer be dealt with by traditional law (and arrns), wbich impelJed overexploitation by mulliple users. According to Baker (1867) taxalÍon of cultivated land discouraged agricultural (surplus) produclion, During lbe early years of Egyptian and Anglo-Egyptian occupation environ­ mental effeets must still have been límited, The situation changed, however, when lhe colonial hold on the arca increased around the beginning of the twentielh century and 'development' activities were initiated, Public works and irrigation iuitiatives along the Gash required tabour.rs who were at fírst aUraeted among rhe local Beja, In years of high rainfall or fioods the Beja, however, planted their own erops - leading to a labour ,hortage which was counteracted by lhe govern­ ment by the attraction of immigrant labour in the forrn of Eritreans, Somali" Abyssinians and rhe so-called Westerners (West Afriean Muslim pilgrims), Wilh the creation of Ihe Gash Delta Scheme in lhe early 1920s the labour req uire­ ments grew, and because of lhe unwillingness of the local Beja 10 ehange their lífestyle and seUle as colton farmers in the seheme, still more immigrants were attracled (MeLoughlín, 1966), The Hadendoa did not like these developments and the inlrusions of olher ethnicities on their grounds, Even though the govern­ menl decided to give more thought to the rights of Ihe Hadendoa, the seheme remaioed at odds with tbe ioterests of the Hadendoa, They had to aceept 'foreigners' on theír land, they lost control over grazing and cultivation rights 68 Biological and Cultural Diversity and they were supposed lo gel rid of Iheir sloek and seule for eotton farming. What happened was Ihal they le1 olhers cultivate their land and eventual!y los! control over much of the bes! land, They conld no longer utilize the land in their own way and lO Iheir own accord (Paul, 1954; McLoughlín, 1966; Hassan Mohammed Salih, 1980). In effect, they were lo a large exten! driven into Ihe comparalively marginal areas like lhe Border Area, This increased Ihe popula­ líon pressure in these areas, also during Ihe dry season when they could no longer graze their slock throughout Ihe Gash delta, and has led lO deforestation and land degradatíon. A furlher set-back for the local Beja was the introduction of modem means of transportation and when in 1924 the railway that connected Kassala to the Red Sea coast was completed, the desert trade, which had been a virtual monopoly of the Hadendoa, dwindled (Newbold, 1935). Al! in al!, the Hadendoa and Beni Amer became increasingly constrained in their traditional livelihood activities, as the best watered soil. were occupied by large-seale irrigatíon schemes, and Ihe transhumance routes were curtailed by mechanized farming in the southern pastures. lt is within this context that, in the eariy decades ofthe present century, runoff farming beca me an increasingly important contribulÍon to the livclihood of the population of the marginal Border Area (Niemeijer, 1993). The runoff farming systems and their position within the landscape Geomorphologically the Border Arca may be described as piedmonl plains consisting of pediments largely covered by coalesdng alluvial fans and visually dominated by occasional inselbergs (rock oulcrops), The average altitude of the plains is sorne 500 m aboye mean sea level, with a gentle slope (averaging less Ihan I per cenl) from Ihe east to the west. The plains are dissected by numerous eaSI~west ftowing smaller and larger wadis (ephemeral streams) thal drain the foothills ofthe Eritrean highlands, as well as the local inselbergs (sce Figure 4.1). These slreams 'pread and contrael several limes along Iheir eourse, al points where Ihe gradient respeclively diminishes and sleepens, resultíng in a relatively regular east~west oriented drainage pattem (sce Figure 4.2), This complex drainage pattern and the variabilily of discharge and ,edimcnt load has resulted in soils wíth an extremely high spalial variabililY, often leadíng 10 completely diflerent soil profiles (in terms of toxture and horizon depth) at dislanees of as little as 10 m. The followíng soils may be recognized: Vertic, Haplic, and Calcic Luvisols; Vertic Cambisols; Eulric Fluvisols; and Eutrie Regosols' Long-term average (summer) rainfall decreases from sorne 300 mm in lhe soulh to 150 mm in the northern part of the Border Arca, but in reeent decades the average has decreased by sorne 20 per cent (Niemeijer, 1993). In many ways the Border Area offers an ideal environment for runoff farming, Rainfall is too low for regular rainfed cultivation, while the gentle ,Iope and the relalively low population density allow for largo, sparsely vegetated eatchments. Initially (that is sorne 2000 years ago wben the Beja lirs! engaged in the cultiva­ tion of grains), wildllooding, a type offtoodwater harvesting without artifacIs on Ihe alluvial Hals, was probably the main form of runoff farming. lt required relatívely Iíttle efforl in lhe form of land preparation, while returns could be considerable in mediocre to wet years. During colonial times, the Hadendoa and Beni Amer populations bccame increasingly dependent on Ihe Border Area for their grain production, as a resull of lhe earlíer described colonial policies, and the need arose to expand the cultivated area. The arcas Ihat received water from lndigenous Sudanese Water Harvesting 69 Figure 4.2 Spatial posltlon of the Border Ares runoff farmlng systems lIIustrated schematlcally naturally spreading wadis were Iimi!ed and il is likely Ihat, in order lo expand Ihe cultivable area, Ihe people al sorne pojnt turned lO water spreading, a lype of !loodwater harvesling lhal uses (mainly) brushwood spreaders Oil the alluvial fials. Bu!. as suitable space on Ihe alluvial !lals ran oul, the populalion soon had lo lurn lo ye! anolher technique. Probably it was in Ihe 1940s Ihal lhe local population developed Iheir own form of Ihe leras syslem, a rainwaler harvesting syslem !hat unlike wild!lood­ ing and waler spreadíng (both floodwater harvesting syslems) - did nol depend on Ihe occurrenee of stream-flow. 7 The teras syslem probably has jls rools in Ihe area of Sennar where during Ihe Funj Kingdom (1504--1820) elemenlary forms of Ihis lechnique were developed (Craig, 1991). Laler Ihe lechnique sprcad lo large pans of (semi-)arid Sudan. Jefferson (1949) estimates Ihat during Ihe 1940s sorne 5 per ceot to 15 per cenl of Sudan's sorghum crop was produeed on teras fields.' Presently, Ihe leras syslem is Ihe mosl widely used runofffarming system in Ihe Border Area and covers sorne 40 per cenl of Ihe cullivaled area (Van Dam and Houlkamp, 1992). A teras consísts of a field surrounded on Ihree sides by low earth bunds whjle lhe upslream side functíons as an inlel for surface runoff (see Table 4.1 for some basic stalistlcs of Ihe teras syslem). Figure 4.3 shows a 3D-skeleh (nol lO scale) wilh Ihe basie teras features. AH terus (plural of leras) have a main bund consistíog of a base conlour buod (1) as well as Iwo ouler colleelion arms (2). The numerous small parallel rjdges inside Ihe leras are weeding ridges (3) Ihat result from Ihe weedjng pattern (and also occur on wildflooding and water spreading fields). The inner colleelían arm (4) and Ihe inner diversion arm (5) are optional features Ihal do nol occur on all terus. The Ihreshing circle (6) is 70 Bíologícal and Cultural Díversity Table 4.1. Basle statisties of \he teras system of \he Border Area Dímensions Bund height usually around 0.5 m, bu! old, manually conslrucled bund. may be over 2 m high; base bunds are generally 50--300 m in length; anns usually 20 lo 100 m in lenglh; sizes vary from 0.2 lO 3 ha Labour inves!ment for 7 to 19 person-days per heclare (depending on soil type conslruction and season) Agricultural cycle Land preparation (May/June); sowing (June/July); weeding (June-August, up lo three rounds); harvesling (Oclober/November) Seasonallabour 66 lO 80 person-days per heclare Oand preparalionl requírements maintenance 12-14; sowing 2-4; weeding 14--24; harvesling, threshing, bundling and transport 38) Tools used Sowing wilh a planting Slick (generally on untilledl unploughed soU); weeding with a hoe; harvesling with a hoe, ax, or knife; threshing with a heavy stick Crops Mainly sorghum (Sorghum bicolor, S. vulgare) and millel (Pennisetum typhoideum); occasionally water melons (Citrullas Vulgaris), okra (Hibiscus eseulentus), rosella (Hibiscus sabdariffa), and sesame (Sesamum indicum) on Ihe wetter parts Plan! densities Plan! spacing is variable, as IS the number 01 plan!s per seeding hale (5 lo 10); average plant densities around 58000 lo 62000 planlslha Yields 300 lo 600 kg/ha in mediocre to good rainfall yearss Source: Niemeijer (1993, 1998); Van Dam and Houtkamp (1992); Van Oijk and Mohamed Hassan Ahmed (1993). r------------------~-- .... -----, (5) Inner diversion arm (1) Base contour (el Thr.shing circie bund (3) Weeding ridges Rgure 4.3 3D-sketch of a teras (not to scale) Jndigenous Sudanese Waler Harl'csling 71 only found on a few lerus and is used lo reduce grain losses during lhreshing. A part lha! has no! becn drawn is the shallow channel that is sometimes found on !he insíde of !he main bund (1 & 2). This ehannel is Ihe souree of earth needed for heíghlening or repairing tbe main bund. In sorne cases it funclions as an additional water reservoir for planting fruils or vegetables, usually water melons (Citrullus Vulgaris) or okra (Hibiscus esculenlus). Though the leras teehnique shows local variations in size. shape and height of lhe bunds, lhe same name is used throughoUI the Sudan. The colloquial term leras is also widely used in the Iiteralure (see Tothíll, 1948; Lebon, 1965; Reij el al., 1988; Reij, 1991; Craig, 1991; Critchley el 01.,1992; Van Dijk and Mohamed Hassan Ahmed. 1993; Van Dijk, 1995) ~ for which mason il will also be used here. Runoff farming in Ihe Border Area ls remarkable in Ihe sense Iha! in an area of only 8600 square kilometres lhree major runofffarming syslems are in use. What is more, within the broad calegories of wildftooding, water spreading and leras water harvesling, there is al so plenty of differenliation. The wildflooding fields may be divided iuto those thal are localed wilhin the actual streambed (wadi cultivation), and those located in areas thal are naturally flooded by water that has eilher overftowed Ihe river banks or flows as dispersed overland flow. Of the brushwood water-spreading system two variations may be reeognlzed. The fírst uses communally buíl! spreading slruetures in Ihe main water course al the so­ called projecI-level: fields of several farmers are receiving water from a single large strueture. The second variation uses fíeld·level spreading struetures localed on Ihe individual ficlds. Finally, two types of leras systems may be distinguished. The more eommon system receives only overland flow (raínfed), and the other, speeific lo the Border Arca, makes use of concentrated stream ftow [rom one or more smal! ephcmeral streams (wadi-fed) and is ¡hus not a rainwater harvesling system (pur-sang), bUI a combined flood- and rainwater-harvesting 5ystem (Niemeijer. 1993). This variety of systems is indicative 01' lhe creativily wilh which the local agro­ pastoralists approach runoff farming. Land is not yet in shorl supply (calcula­ lions of Van Dijk and Mohamed Hassan Ahmcd (1993) show that only some 5 per cent of the area is eovered by agricultural field.) and most fields are Ihus located in the hydrologically favourable areas. Farmers usually have fields in several of lhese areas and adapt the type of runofr farming system Ihey use lo lhe local hydrological conditions. As Figure 4.2 shows, wildHooding is found on the alluvial flals, mainly in arcas thal show natural spreading of coneentraled stream flow lo dispersed overland flow. Water spreading is also found on these alluvial Hats, but in those areas where the water flow is still too concentrated to allow wildflooding. The leras system oceurs in intermedia!e positions and on lhe interfluves (only a few metres hígher than the alluvial flats). Those loeated in the intermediare positions make use of bOlh overland and stream flow, whereas those on lhe interfluves depend exc\usively on overland tlow. The spreading struetures and bunds used in Ihe waler spreading and leras systems are by no means uniform in size, material, shape and orientalion. They could not be. for in lhe harsh and dynamic Border Area environment farmers need to adapl theír systems optimally lo the site-specific hydrology and soils. In faet, adaptation alone is nol enough, for wadis and overland flow pattems ,hift from year to year, as do the amount of minfall and discharges. With a coefficient of varialion of 30 per cent there are large inter-annnal fluctuations in the amount and liming of rainrall and, as is cornmon in sueh arid environments, rainfall spalial variabilily is high as well. Farmers need to experiment continualIy to 72 Biological and Cultural Diversity adapt ¡heir stroetures and planting patterns to ¡hese environmental dynamics and seeure an adequate bul non-destructive supply of water to their crops. How ¡his occurs ís bes! secn in the leras system. Adaptation and experimentation in the teras system In the mosl soulhern parl of the Border Arca, that is inhabíted by the Beni Amer, most of the variation in the terus occurs in terms of the size of the fields and the height and orientation of Ihe bunds. Here most of the terus are rectan­ gular in shape, unlike in Ihe central part of the Border Arca occupied by the Hadendoa, where, in addilion, shape is a major variable. That was one of Ihe reasons why the farmlands of the village of Hat Ayot, which is located in this central area (see Figure 4.1), were studied in mosl detail and are tbe source 01' most of the examples below (tbough a lot also pertains to teru", in the other par!s of the Border Arca). Regulating ¡he water supply la ¡he leras Regulation of the water supply is essenlial to successful cropping, The farmers bave lo balance the water supply in such a way as to obtain tbe bighest chance of a good harvest in most years, while eliminating the nsk of destruction of Iheir terus in years wilh higb rainfal! and large wadi discharges. In tbe Border Arca environmenl witb low irregular rainfall and dynamically changing wadi courses choice of location is vilal, since it will determine tbe chances of suecess in the years lo come (Jabour inveslments ror the initial construction of a teras are higb; see Table 4.1). As can be seen in Figure 4.4 (a delail of a tracing of a 1986 aerial photograph of an nat Ayot fanning arca), lerus are nol localed in lhe immediate stream course of a wadi. A frontal assaul! of the water during one of the few discharge events would certainly wreak bavoc on a teras, and damage the bunds severely. So, instead, Ibe terus are carefully constructed in those places that do receive additional wadi water, bu! lhat do no! receive Ihe full force of the discharge, Broadly speaking, two kinds of situations can be observed, Sorne IéfUS (a and e in Figure 4.4) are positioned in those areas where Ihe water of the wadis spread. out, leaves the defined water course and continues to Ilow dispersed as low kinetic energy overland Ilow. Other lerus (d in Figure 4.4) are construcled on the banks of tbe wadis and, on discharge events, receive the low kinetic energy overflow from the only slightly incised wadis. Certain terus (b and e in Figure 4.4), make use of a combination of these kinds of additíonal ephemeral stream water. Once the location has been fixed a lot depends on the adaptation of size, orientation, and sbape to Ibe bydrology and soHs of thal particular site. Tbrougb continuous experimentation Ihe farmer tríes to cope wíth the environmental dynamics. Shifts in a wadi's course or fluctuations in discharge frequently war­ rant modifications of the leras shape. During the yearly maintenance, and some­ times even wilhin a single sea son, the farmer may modify tbe height, ,hape and extent of the various bunds, based on his 'o experiences from tbe previous years. Tbe amount of harvested water may be regulated by extending the outer collec­ tion anns or changing their ,hape. Sorne lerus are very broad, with a long base contour bund, but have very shor! outer collection arms; olhers are very deep and have a short base contour bund and long outer collection arms. The bunds are usually irregularly sbaped and sometimes the size of tbe outer collection Indigellous Sudanese Water Harvesting 73 Approxlmate north r .. ( Remains of , ald bund Teros D o 50 100 150m , , 100 200m Figure 4.4 Environmental pasltlon of Figure 4.5 Extremely long oular col­ some tenIS iection arm of a teras (traclng of 1986 aerial photograph) (traclng of 1986 aerial photograph arms of a single leras may differ considerably. The size of terus in ¡his particular farming area ranges from 0.30 to 1.85 hectares. One of the lIat Ayot farmers really went to extremes to control the water supply of his teras. Figure 4.5 shows what now remains of an extension of one of the outer collec(ion arrns ofthis particular teras. Al one time the outer colleclion arrn had a totallength of sorne 430 metres, of which at preson! sorne 140 me tres ,till remain in good condilion. With Ihis long onter colleelion arm the farmer tried to harvest additional stream fiow from an ephemeral stream. He misjudged lhe power of this stream and was unable lo maintain this extensive onter collec­ lion aTm Da the long run, Regulating ¡he water distribution within the teras The shape and location of the teras d.etermínes the approximate amount of water that is collected in that teras, as well as lhe proportions contributed by local overland fiow and concentrated stream fiow (wadis). FurtheT management 74 Biological and Cultural Dil'ersily teehniques determine the water dis!ribution within the leras and are vital to guarantee a good harvesl in all parls of the field, bu! also lO preven! damage from an excess of water ín a single par!. ' , The most common Iype of interna! struetilre is Ihe inner collection arm (Figure 4.3 (4)). It ís used to lake care of a good dístríbutíon of waler wíthin the teras. In sorne cases it prevent. a very smal! amount of TUnoff bcíng spread over the whole leras, whefeby none of the parts receives enough water lO grow anything. In olher cases it guarantees that some waler remaíns behind in slightly elevaled pans of Ihe teras, hecause Ihe inner colleclion afm prevenl. il from all flowing to the lowesl comer of the teras. Another, less common, Iype of waler regulating bund is Ihe iuner díversíon afm (Figure 4.3 (5». Its maín function is probab!y lo prolecl against damage the part of the teras that receíves the híghest discharge from Ihe incoming wadi, as part of lhe ínftow is diverted 10 the olher side 01' lhe leras and Ihe force of Ihe water is reduced. The mosl extreme regulating structure is the 'molher' and 'chíld' slruetu,• . Insíde one large leras (mother), one or more smaller lerllS (chíldren) are oon­ slructed (sce Fígure 4.7). This technique is nol so common in the Border Area (in contrast to lhe larger lerus on Ihe Clay Plains to lhe west of Kassala). lIs function is símilar lO that of lhe inner colleetioo armo In wet years it al10ws lhe plantcd area to be larger, sínce runofT water is dívided into separate sectors. In dry years the 'child' leras, beca use of íls the smaller size, requiTes less runofT waler lo grow a crop, than Ihe large 'molher' leras Ihal may not receive enough water in proportion to lts aTea. If lhe inner arms and the mOlher--<:hild struetures can be regarded as lhe maero-slruetures tha! regulate the water dístribution inside lhe teras, lbe direc­ lion of the weeding ridge, can be regarded as the micro-slrueture, used to di,lribute Ihe life-giving líq uid. AU lertiS in lhe Ilat Ayo! area are ,ti!! hand ti!!ed. Sowing ís done with a planting stick and if ít were not fm the weedíng activities, the terus fields would havo a ,mooth surface. Weeding is done wilh a hoe, and togelher with weeds some eartb is scraped aside, resulling in smaU eontinuous ridges bel ween the planting rows. The orienlatíon of these rídges is cbanged by changing the direc­ tion followed duríng weedíng. From year to year weeding directions are usuaUy maintained (except when changed discharges roquire alleration) and even afler several years of fallow the weeding ridges can still be recognized and form a consislent water regulatíng struclure. The orienlalíon of the weeding ridges is adjusled to distribute waler oplimaUy over Ihe field. Usually lhe weeding rídges are kept parallel lO Ihe flow direction to guide waler into lhe teras. In sorne cases, however, the current of incoming water is too strong and weeding ridges are created at a right angle to the inflow direction to reduce the kinetíc energy of the water by increasing fríction. Al the same time Ihese ridges prevent water fram leaving the teras again. Another silualÍon in which the farmers change Ihe weeding direction is where water is spread unequally over the land due to height differences or wadi characteristics. Tlús is lhe case in Figure 4.6, whích shows a leras and a part of its catchment. Only one part of thís teras receives wadi water, resulting in a local excess of water. The farmer has therefore oríented Ihe weeding ridges in such a way that excess waler is guided lowards the otber par! of lhe teras. Thís other parl of lhe leras has weedíng ridges ín Ihe normal direction, paralle! lO Ihe inflow of local surface runoff, to enhance harvesting. Figure 4.7 shows another exarnple of a farmer applying multíple weeding lndígenous Sudanese Water Harvesting 75 A..p proximate north Area with an excess of water Weeding ridges guide water to other parts Teras ~-( bunds ~'\'\ Normal weeding direction o 50 100m '---'----'--....... ~ Figure 4.6 Regulating the water supply by adaptation of the weedlng ridges (tracing of 1986 aerial photograph; weeding ridges not lo scale) directions lO regulate Ihe water distribution wilhin his teras, In addition, this teras iIlustrales how the inner colleetion arm may be used. In lhe mothor leras waler of IWO wadis is used separalely, one wadi supplies one side of the leras and lhe olher supplies lhe olher side, In lhis way lhe farmer en han ces lhe chanee of a suecessful erop in at least a part of the ficld if one of the wadis fails, IndividuaJiIy, adaptation and experimenta/ion From the above it is elear that lhe terus are highly adapted lO the Border Arca environment in general, but also lo lhe specific dynamics of particular sites. The farmen cope with lhese dynamics by employing a number of techníques that regulate the water supply as weJl as the dístributÍon wíthin the leras, There is a high degree of individuality as each farmer makes his choice from the common set of techniq ues and adapts them to his requirements and Ihe conditíons of his particular locatÍon. This is an al! bul easy task in an area tha! recently received less Ihan 250 mm of rainfaJl in eight out of ten years and where wadi courses shift from year !o yeaL Under such circumstances experimentation is no! a risk (as it is often regarded in the Iiterature), but a necessity. Through 76 Di%gital and Cu/tural Diversity Breach , Inflowof I 4.-~-- wadiwater ......,...::- '!nflow of ,: ra.lnwater I 111 Bund ¡¡¡;¡¡ Old bund "'-J ~ r ~ Ch~e in weeding " directlOn or end of /~ cultivation Inflow of k:'b -' Inftow of o Direction of 10 20 30 4Qm wadiwater L rainwater /' j ¡ weeding ridges ! I Figure 4.7 A teras with a mother and child structure. Weeding ridges are identical by double sided arrows (grid survey in 1990; 20 x 20m gridcetls) experímentatíon the farmer gaíns the necessary understandíng of the environ­ ment, as well as of lhe hydraulícs of his engíneeríng slrllctllres, Iha! ís prereqllísíte to a successful (re-)adaptation of his teras to lhe ever-changíng círcumstances, Sorne of these experíments are successful and are picked up by olher farmers, Olhers, like the long outer collection arm shown in Figure 4,5, are less successful and remain a very individual, one-off experiment, There are more examples of farmers who go heyond lhe regular experimenls and try sometbing really new, One nat Ayot farmer buill a reras within lhe ,tream course of a major wadi, According lo a key informanl lhe seeds on lhis teras are not washed away by the fiood, bUI lhe leras bunds must be rebui!l every year: 'This requires more work than the olhor tenIS, but yields are higher, because there is more water from the wadi, and a!so from deep in tbe soil: Another farmer designed curved weeding ridges lha! guide water from a central water souree to lhe outer sides of the teras, Sorne farmers use hranches and in a few cases (where lhey are available) stones lO strengthen the teras earth bunds. Apart from these cngineedng cxperiments farmers also expedmen! with different sorghum aud millet varieties and non­ standard crops, Iike okra or water melon, in part of lheír teras, There are about eight different sorghum varieties in the area, each with different characteristic.: e,g" drought-tolerance, taste, growing scason, market vaJue, Sorne farmcrs plant up to lbree differ.n! varieties on a single field lO optimally henefil from soil moisture differences and to minimize risk due to variable rainfalL Plan! spacing is varied from year to year (Van Dijk and Mohamed Hassan Ahmed, 1993) and Ihere are even indications thal sorne farmers adapl plan! densities relative to the expectele, operative. Extension services Frequently Not available Frequenlly Rarely available. available, available. Credil support No eradil No eradi! No eredit No eredil support. support, support, support, Home Gardens o[ South India 143 Space utility experiments Rooftop gardeners conduct experiments with the objective of utilizing efficiently the limited space. Depending on the availability of space, rooftop gardeners alter their sowing techniques. A few rooftop gardeners have sown different varieties of gourd vegetables in a single circular pit, whereas others dug separate pits for each variety of gourd. The size of circular pits al so varies from 12" to 18" in diamcter. In another case, a rooftop gardener lifted the bottle gourd vines on a pongamia tree, sponge gourd vines on a neem tree, and pumpkin vines on the rooftops of his house. The gardener claims that using different types of suppor­ ters to lift various gourd vines leads to lessening the competition among the gourds, and reduces the crowding of vines and thus results in easy harvest and increased yields, with minimal pest and disease inciden ce. Market testing Young farmers of upland gardens are profit-oriented and are interested in analysing the market demand for garden-grown vegetables. They constantly watch the demand for chillies in the market and adjust the picking of vegetables accordingly. The young farmers have joined together as an informal marketing society wherein they brought the merchants to their door, thus eliminating the middle-men. Conventionally, the farmers have to take their produce to the markets or shandies in towns. Constant market analysis and co-operative skills have also tumed the upland gardens into an economic enterprise. erap combination testing Backyard gardeners constantly experiment with crop rotation and crop combi­ nation techniques. Through experimentation, they found various combinations of herbaceous and tree crops. They al so experiment with shade-Ioving and shade-hesitant plants. Elderly women display a remarkable array of knowledge of crop combination techniques in the pump house gardens. When compared with other types of home gardens, the pump houses are prone to more shade. Hence, repeated informal testing and experimentation made the elderly women familiar with shade-Ioving leafy vegetable crops. For instance, karisilankanni, a green leafy vegetable crop, performs well under shady conditions due to the rough leaf surface. Pump house gardens are also sites of poorly drained soils. Women tested different types of banana and coconut in poorly drained soils. One of the women farmers claims that poovan, a banana variety, performs well under the waterlogged conditions of the pump house gardens. Varietal testing The young farmers conduct experiments by constantIy changing the varieties of chillies. Informally, they visit each other's chilli gardens and select seeds. They make identification marks by tying a thread near the selected fruil. According to sorne farmers, changing the seed varieties season after season increases the yield. In a pump house garden, about six different varieties of banana were tested. Such varietal testing was al so observed elsewhere. In one river basin in West Java, the performance of 34 banana varieties were recorded (Soemarwoto and Soemarwoto, 1984). The fruit of sorne bananas are eaten as dessert, or steamed for snacks, and others are supplementary staples. Home gardeners also clearly recognize the long-term importance of gene tic diversity. When asked why an unused tree is found in a garden, they typically respond by saying that they might need it sometime in the future (Soemarwoto and Conway, 1991). 144 Biological and Cultural Díversity Informal IPM experimenlJ; Home gardencrs are meticulous IPM (Integrated Pest Management) experimen­ ters, Farmers experiment with difIerent types of pesticides, especially in the backyard gardens. They found that dimethoate, an insecticide, is highlyeffective in controlling the insect borers that infest Ihe fruil. of brinjal and okra. If the borer incidence is severe, they normally skip the planting of lhat vegetable crop for the successive season. To control minor incidence of pests and disease., home gardeners apply cow dung a.h. During sowing, elderly people are involved in reftecting the sunligbt using mirror. to scare away crows to prevent the birds from carrying away the seeds. Sinee the small seeds of brinjal, okra, and other leafy vegetables are sown on the soil surface, protecting the seeds from birds and ants is crucial lo maintain the required plant population, Two or three days before planting, they check for ant burrows in the garden and apply kerosene to the burrows in order to drive the anls away from Ihe garden. Jt is elear Ihat, through experimentation, farmers demonslrate abílities to cope wíth and adapl to change, to confront contemporary and future situations, and to reassess existíng adaptive strategies lo deal with environmental aud socio­ economic forces (Dei, 1988; 1989). Farmers, with their constant interest in experimentatíon, have innovaled a wide speetrum of home garden lechnologies lhat are ecologically sound and socioeconomically benigno Nutritional security in home gardens Mos! ofthe value ofthe home garden lo the rural household lies no! so much in nel ineome generated but in the range of production and its contribution to the overalllivelihood of the household (Socmarwoto and Conway, 1991). Mainte­ nanee of continuous rood production througbout the year is one of the impor­ tant roles of the home gardens. In general, there is always something to harvest from lhe home gardens, Thougb the iew York, 1982. An execllent book that gíves a chatty bul scíenlific descríption 01' the struggles within the life sciences to bring a more rational and ecologicaJ approach to pest managew ment. Of particular note is the work of pathbreaking scienlists who introduced plants to attract pests n}, whích ín turn wouid attract controHing populaHons of predators. 7. Rcfer to 'Patticipatory Research witb Women Farmers·. a scientific film produced by Development Perspectives., for Dr, Michel Pimbert. Internationa1 Crop Research Institute for the Semi-Arid Trapíes, Potancheru, A.P., India. 8. Jan Seva Mandal recorded great success in mobihzing local chndren to coUeet and de'troy the clearly visible yellow egg masse. of Ihe pest. 9. Pal Mooney was given the Righl Livelihood Award, also known as the Alternative Nobel Prize, in 1986 for his work in alertíng the Third World lO the danger of germ­ plasm concentration in the hands of multinationals and Westem research institutes. lO, Many power groups struggle foc ascendency in a developing country, parücuJarly in a rural situation. We may broadly ídentify four tendencies: one is led by the 'moder­ nists', who indude scientists. government officials and businessmen, who believe in the liberating forces of science and modem organization. Another i5 composoo of Marxisls who also believe in scicnce but want people's control. A third force is that of 'Iradilionalists' who decry modernisms. bul otherwise join hands with the powerfuL A fourth is made up of 'romantics' who walk part of the way with the others, but support none. Under these conditíons, the search for lpurity: with its special lndian meanings, blocks attempts at practical work, .Notes 201 1L Lighl Iraps used by entomologists employ fumiog ehemica!. to klll insect& Water can be used lo trap the molhs, with a soapy film to lessen surface teosioo. Design, size and safety measures need lo be worked out for rugged rural conditions, Also, power failure constantly interrupts efficiency, when most needed. Can gas lighls be used jusI as effeetively? Further, il has becn noted Ihat molhs are more attracted to lhe blue frequeneie. of light, and that the normal florescent tube-lights, used as stand-by. in India during power break-downs, eould very easUy serve lhe purpose. The lights could al.o be useó for night schools, villago elinies, meetings, and henee be more affordabJe to the village communities. 12. M. A. Qayum and N. K. Sanghi (eds) Red Hairy Caterpillar Management through Group Actioo and Non-Pesticidal Methods, ASWand Oxfam (India) Trust. 13. See: M. K. Mukundan~s exceUent study on tradítíonal víllage tank systems. Patriotic and Prople-Based Seienee & Technology Foundation (PPST), P.O. Box 2085, Adyar, Madras, India. 14. Personal communication from Dr. T. Hanumantha Rao. Ex-Director General, Water and Land Management Research Institute, Hyderabad. 15. Dr. N. K. Sanghi, Central Research Institute for Dryland Agriculture (CRIDA), Hyderabad, India, reporto tha! sorne indigenous peoples in Adilabad Districl, A.P., are innovating in field bunding practices. Women sangams coUectiveJy leasing Ín tand in Medak District also innovate for soil control see Deccan Development Socíety's report&. 16. ICRISAT: Annual RepoN 1986. 17. Irfan Habib: An Atlas ofthe Mughal Empire, Oxford University Press, Oxford 1982. 18. Permacu)ture, a concept ror designíng organic farmíng in harmony with nature. was developed by BilI Mollison of Australia, who also won the Alternative Nobel Príze. Permaculture has struck firm roots among Indian NGOs. A Permaculture Associa­ tion has been formed shal'Íng an offiee wíth Ihe Deecan Development Sociely, al Hyderabad. 19. J. D. Bernal, seience historian, used Ihe term 'people-based seienee' bul perhaps nO! too appropriately considering the divergence of interests between common peQP)e and the scientific commurtity. especiaHy physícists. Hving off defence research contracts. Chapter 13 1 would like to dedicate tbis ehapter to the swamp farmers of Maputaland, especially David Mathenjwa, Joshua Mathenjwa aud Charles Mlambo from whom I Iearnt aboul farmíng in Maputaland. 1. 'Umnotho wethu amadobo' translated from Zulu means 'the swamp/swampy areasJ swamp forests are our fertílityJwealth'. The oId man who made this statement was referring to the swamp forest, which he had ulilized in the past and whieh had been used by his father before him. 2. KwaZulu was the Bantustan administered in a form of quasí-independence by the KwaZulu Government and lbe elhníe base for lhe Inhtha Freedom Parly, both headed by Chief Buthelezi. It is a fragmented lerritory sprcad throughout Natal and includes Maputaland, Zululand is the area north uf the Tugela River and was lraditionally the area of the Zulu kingdom. Presently Zululand is composed of parts of KwaZulu and parts of NataL The new region incorporating Natal and KwaZulu i5 now known as KwaZulu-NataL 3. 1 refer to people as Thonga and Zulu to emphasize cultural and linguistic differences and not to perpetuate ethnicity_ 4. According lo Bryanl (1949), uNyawoti (pearl millet) was formerly widely grown by the coastal Zulus having been replaced during the last century by the more proJific amaBele (sorghum) whieh in lum has been largely replaced by umBíla (maize). Pearl milJet is almost unknown in Zululand now. But he 15 in all likelihood incorrect, for both mHlets and sorghums would have formed pan ofthe farmer's repertoire in both Zululand and Maputaland. References Introduction Adams, William M. and L Jan Slikkerveer, eds. (1991) ¡ndigenous Knowledge and Change in African Agricultur• . Studies in Technology and Social Chango, No. '26. Ames: CIKARD, lowa State lJniversity. Adegboye, Rufns O. and 1. A. Akinwumí (1990) 'Cassava Processíng lnnovations in Nigeria: In Matthew s. Gamser, Helen Applelon. and Nicola Carter, eds. Tinkef, Tiller, Technical ehange. London: Intermediate Technology PublicalÍons. Amanor, Kojo S. (1989) 340 abstraets on farmer participatory researeh. Agricultural Adminislration (Research and Extension) Network Paper 5. London: Overseas Devel­ opment Institute. -- (1991) 'Managing the Fallow: Weeding Technology and Environmental Knowledge in the Krobo Distriet of Ghana: Agriculture and Human Values 8 (1 +2): 5-13. Ashby. Jacqueline A., Teresa Gracia, Maria del Pilar Guerrero, Carlos Arturo Quiros, Jose Ignacio Roa and Jorge Alonso BeJtran ([996) arvaez Carvajal, G. and Paredes Hernandez, E. (1994) El Pica·Pica (Mucuna pruriensj: ,\.fas que un abono verde para maíz. en el norte del Istmo Oaxaqueno. Oaxaca de Juarez, Mexlco: Universidad Autónoma Chapingo, Dirección de Centros Regionales.. Norman, D., Baker, O., Heinrich, G., Jonas, e., M.skiara, S. and Worman, F. (1989) Farmer groups for technology development: Experience in Botswana, In R. Chambers;. A, Pacey~ and L.A. Thrupp eds. Farmer First: Farmer mnovation and agricultural research. London: Intermediate Technology Publications. 136--146. Pare, L. Blanco R .• Buckles. D., Chevalier, 1. M .. , Gutíerrez Martinez, R.) Hemandez D" Perales Rivera, H. R., Ramirez R. and Velazquez H., (1993) La Sierra IÚ! Santa Marta: Hacia un desarrollo sustentable. Unpublished MS. Perales Rivera, H. R. (1992) El autoconsumo en la agricultura de los Popolucas de Sotea­ pan, Veracruz, MSc. thesis, Montecillo. F40. de Mexieo: Colegio de Postgraduados. Perales Rivera, H. R. and BuckJes, D. (1991) ExpenmentaCÍón campesina con el bejuco de abono (M.cuna sp.) en la lona indigena de la Sierra de Santa Marta. Veracruz: Una propuesta de ínvestigación. Unpublished MS. Seolt, J. M. (1919) Velver Bean Varieties. Lniversity of Florida Agricultural Experiment Station, Bulletin 152. Stnart, J. W. (1978) Subsistenee ecology of the Isthmus Nahuat lndians of Southern Veracruz, Mexico. PhD. dissertation. Riverside. California: Universíty of California, Riverside. Tr.cy, S. M. and Coe, H. S. (1918) Velvetbeans. FarmelS' Bulletin 952. Washington, O.e.: United States Departmcnt of Agriculture. 208 Bi%gical and Cultural Díversity Triomphe. B. (1995) Fertilidad de los suelos en la rolación maízlmucuna, Costa Norte de Honduras: Resultados preliminares, Paper presenled al Ihe XLI PCCMCA Annual meeting, 26 March-I April 1995, Tegucigalpa, Honduras, Wilmot-Dear, e M, (1984) A revision of Mucuna (Leguminosae-Phaseolease) in China and Japan, Kew BuUefin 39: 23--{Í5, Wilmot-Dear. e M, (1987) A version ofMucuna (Leguminosae-Pha,eoleae) in the lndian suhcontinent and Burma, Kew Bulletin 42: 23~6, .lea. 1. L. (199)) Efecto residual de intercalar leguminosas sobre el rendimiento de maíz (Zea ma"vs L.) en nueve localidades de Centro America. In Sin/esis de resultados experimentales, 199J. GuatemaJa City, Guatemala: Programa Regional de Maíz para Centro America y el Caribe. 97-103. Chapter 3 Chaudhary R. e and Fujisaka S. (1992) Farmer-participatory rainfed lowland rice vadetal testing in Cambodia. lntern Rice Res Newsl 17: ~)7. Fujisaka S. (1990) Rainfed lowland rice: building research on farmer practice and tech­ nical knowledge. Agric Ecosyst Enviran 33: 57~ 74. Fujísaka S, (1991) A diagnostic survey of shifting cultivation in northern uos: targeting research to improve sustainability and productivity. Agr%r Sysl 13: 95-109. Fujísaka S. (1993) A case of farmer ad.ptation aud adoption of eontour hedgerows fm :mil conservatiOll. Experiment Agric 29: 97~ 105, Fujisaka S" lngram, K. T. and Moody. K, (I991a) Crop establishment (beusanr) in Cuttack District, India, lnternational Rice Research Institute (1 RRIJ Res Paper Series 148, IRRl: Los Baños Philippines, Fujisaka, S., Kirk, G" Litsinger, 1. A., Moody. K., Hosen, N" Yusef, A., Nurdin, F" Naim, T" Artati, F. Aziz, A., Khatib, w., aud Yustisía (1991 b) Wild pigs, poor soils, and upJand rice; a diagnostic survey of Sitiung, Sumatra, Indonesia. InternationaJ Rice Research Institut. (IRRI) Res Paper Series 155. IRRl: Los Baños Philippines. Fujísaka, S., Moody, K. and survoy team, (1992) Raínfed lowland. deepwater, .nd upland rice in Myanmar: a diagnostic survey, Myanmar J Agric Sci 4: 1~13. Haugerud, A. and Collinsoo, M, P. (1990) Plants, genes and people: improving the relevance of plant breedíng in Africa. Experiment Agrie 16: 341-62, IRRI (Internattonal Rice Research Institute). (1988) Agricultura! econamics database. IRRI Los Baños Philippines. IRRI (Intemational Ríce Research Institute). (1989) lRRI loward 1000 and be}'Ond. IRRI: Los Baños Philippines, Maurya, D. M" Bottrall, A. and Farrington, 1. (1988) Improved livelihoods, genetic diversity, and farmer participation: a strategy for rice breeding in rainfed areas of India. Experimenl Agric 24: 311 ~20, Richards P. (1986) Coping with hunger, AlIen and Unwin: London. Simmonds N. W. (1991) Seleelion for local ad.ptation in a plant breeding programme, Theor Appl Genet 82: 363-67. Simmonds, N. W. and Talbot, M, (1992) Analysis of on-farm rice yield data from India, Experiment Agrie 28: 325-329, Chapter 4 Ausenda. G, (1987) 'Lei,urely Nomads: The Hadendowa (Beja) of the Gash Delta and Their Transition to Sedentary Village Life', Ph,D. thesis. Columbia University, Baker, S, W. (1867) The Ni/e Tributaries of Abyssinia, and the Sward Hunrers allhe Hamran Arabs. London: Macmillan and eo. Brouwers. 1. (1994) 'Réaction des Paysans Adj. á la Baisse de Ferlilité du sor, in p, Ton and L De lIaan (ed,,) A la Recherche de I'Agriculture Durable au Olmin, PI', 55-60. Amsterdam: Instituut voor Sociale Geografie. Universiteit van Amsterdam, Craig, G. M. (ed,) (1991) The Ag'ieulrure 01 the Sudan. Oxford: Oxford University PTeS,. Referenees 209 Critchley, W, C. Reij and A. Seznec (1992) Water Harvesting for Plant Production; Par! l/: Case Studies and Conc/usions for Sub-Saharan Africa. (World Bank Technical Paper No. 157), World Bank. Dangbégnon, C. (1994) 'L'lmportance des Connaissances Endogenes pour une Agricul­ ture Durable au Sud-Bénin', in P. Ton and L. De Haan (eds.) A la Recherche de I'Agriculture Durable au Bénin, pp. 61-64. Amsterdam: Instituut voor Sociale Geo­ grafie, Universiteit van Amsterdam. Hassan Mohammed Salih (1980) 'Hadanduwa Traditional Territorial Rights and Inter­ population Relations Within the Context of the Native Administration System (1927- 1970)'. Sudan Notes and Records 61: 118-33. Hillel, D. (1992) Out ofthe Earth: Civilization and the Lije ofthe Soil. Berkeley: University of California Press. Hjort af Omas, A. and G. Dahl (1991) Responsible Man; the Atmaan Beja of North­ easlern Sudan. Uppsala: Stockholm Studies in Social Anthropology and Nordiska Afrikainstitutet. Jefferson,1. H. K. (1949) 'The Sudan's Grain Supply'. Sudan Notes and Records 30(1): 77-100. Johnson, A. W. (1972) 'Individuality and Experimentation in Traditional Agriculture'. Human Ecology 1(2): 149-59. Leakey, L. S. B. (1936) Kenya: Contrasts and Problems. London: Methuen. Lebon, 1. H. G. (1965) Land Use in Sudan. London: Geographical Publications Limited. McLoughlin, P. F. M. (1966) 'Labour Market Conditions and Wages in Ihe Gash and Tokar Deltas, 1900-1955'. Sudan Notes and Records 47: 111-126. Newbold, D. (1935) 'The Beja Tribes ofthe Red Sea Hinterland', in 1. A. d. C. Hamilton (ed.) The Anglo-Egyptian Sudan from within, pp. 140-64. London: Faber & Faber Limited. Niemeijer, D. (1993) Indigenous RunoJ! Farming in a Changing Environment: The Case oJ Kassala's Border Area, Sudan. Landscape and Environmental Research Group, Vni­ versity of Amsterdam and Department of Irrigation and Soil and Water Conservation, Wageningen Agricultural Vniversity. Niemeijer, D. (1996) 'The Dynamics of African Agricultural History: Is it Time for a New Development Paradigm?' Development and Change 27(1): 87-110. Niemeijer, D. (1998) 'Soil Nutrient Harvesting in Indigenous Teras Water Harvesting in Eastern Sudan'. Land Degradation & Development 9 (4): 323-30. Óstberg, W. (1995) Land is Coming Up: The Burunge of Central Tanzania and Their Environments. Stockholm: Almqvist & Wiksel1 IntemationaL Palmisano, A. L. (1991) Ethnicity: The Beja as Representation. Berlin: Das Arabische Buch. Paul, A. (1954) A History of the Beja Tribes of the Sudan. Cambridge: Cambridge University Press. Reij, C. (1991) Indigenous Soil and Water Conservation in AJrica. London: International Institute for Environment and Development. Reij, c., P. Mulder and L. Begemann (1988) Water Harvesting for Plant Production. (World Bank Technical Paper No. 91), World Bank. Rhoades, R. and A. Bebbington (1988) 'Farmers Who Experiment: An Untapped Resource for Agricultural Research and Development', Paper presented at the Inter­ national Congress on Plant Physiology, New Delhi, India. Richards, P. (1985) Indigenous Agricultural Revolution: Ecology and Food Production in West AJrica. London: Hutchinson. S0fb0i, G. M. (1991) 'Systems ofPastoral and Agricultural Production in Eastem Sudan', in G. M. Craig (ed.) The agriculture of the Sudan, pp. 214-29. Oxford: Oxford Uni­ versity Press. Tothill,1. D. (ed.) (1948) Agriculture in the Sudan. Oxford: Oxford University Press. Van Dam, A. 1. and 1. A. Houtkamp (1992) 'Livelihood and RunoffFarming: the Case of tbe Border Area Vil1age of Hafarat, Eastem Region, Sudan', Master's thesis, Vniver­ sity of Amsterdam. 210 Biological and Cultural Diversity Van Dijk, 1. A, (1991) Water Spreading by Broad-based Earl" EmbankmenlS, WARK Final Report: Technical Performances and Socio-Economic Aspecls in the Border Area, Easl­ cm Region, Sudan, National Council for Research & The Ford Foundatian, Van Dijk, 1. A, (1995) 'Taking the Waters: Soil and Waler Conservation among Seltling Beja Nomads in Eastem Sudan', Ph,D, Ihesis, Uníversily of Am,lerdam, Van Dijk, 1 A. and Mohamed Hassan Ahmed (1993) Opportunitiesjor Expanding Water lfarvesting in Sub-Saharan Africa: The Case o/ (he Teras 01 Km'sala, London: lnter­ nationa! rnstitute ror Environment and Deve1opment. Wes!off, L. (1985) 'Socío-economic Development in Easlern Sudan: a Case Study of Ihe Kassala Rural District', Master's thesis, Univers.ity of t:"treche Chapter 5 Adepetu; A. A' j (1985) Farmers and their farms on four fadamas on the Jos Plateau. Jos Platellu Eni'irQnmenlal Resources Development Programme /nterim Report, No. 2, Durham; University of Durham. Ake, e" (1987) Sustaining De't'elopment on the Indigenous. Loltg-Term Perspectives Stud): Special Economie. Offiee. Washington, D.e. The World Bank. Atteh, o. D., (1992) Indigenou. local knowledga as key to local level development: possibílities. eúnstraints and planning issues in the context of Africa. Studíes in Technology and Social Change, No. 20, Technology .nd Social Change Program, Iowa State University, Ames. Chambers, R" (1983) Rural Development.· PUlling the Last FirSI. Harlow: Longman. Chambers, R" (1989) Reversals, instÍtulÍons and change. In R. Chambers, A. Paceyand L. A. Thrupp, (Eds.), Farmer First: Farmer lnnovation a/ld Agricultural Research. London: Intermediate Technology Publications. Cottingham, R" (l988) Dry-season gardeníng projec!s, Niger. In e. Conroy and M. Litvinoff(Eds.), The GreeningofAíd. London: Earthscan, pp. 69-73. IDRC, (1993) Background to the internalÍonal symposium on indigenous knowledge and suslaioable developmeot. Indigenous Knowledge and Development Monitor. 1, (2), 2-5. Morgan, W. T. w., (Ed.) (1979) The Jos Plateau: a survey of environment and ¡and use, Occasional Publicaríans (New Seríes) , No. 14, Durham, Departrnent of Geography, University of Durham. Morgan, W. T. w., (1985) Forward. In farmers .ud Iheir farrns on four fadamas 00 tbe Jos Plateau, by A. A. Adepetu. Jos Plateau Eni'ironml!/ltal Resources Deve/opmen! Pro­ gramme Interim Reporl, No" 2~ Durham. University of Durham. Phillips. A. O., (1990) Economic /mpact 01 Nigeria's Structural Adjuslment Programme. NISER Monograph Series No. L Ibadan. NISER. Phillips-Howard, K. D., A. A. Adepetu and A D. Kidd, (1990). Aspects of change in fadama fanning along lhe Delim; River, Jos L G. e (1982-1990). Jos Plaleau Envir­ onmental Resources Del'elopmetlt Programme Inrerim Report, No. 18. Durham, Uni­ versity of Durham. Phillips-Howard, K. D. and A D. Kidd, (1990) Significance of indígenous knowledga systems to the improvement of dry-season farming on the Jos Plateau, Jos Plateau EnvironmentaJ Resources Deve/apmen! Prograrnme Interim Report, No. 17. Durham, University of Durbam, Phillíps-Howard, K, D. and A. D. Kidd, (! 99 1) Knowledge and man.gamen! of soíl fertility among dry~season farmers on the Jos Plateau, Nigeria. Jos Plateau Environ­ mental Resources Deve[opment Programme lnterim Report. No. 25, Durham. University ofDurham. Porter, G. and K. D. Phillips-Howard, (1993) Bitter brew? Barley growing in Nigeria, Geography Review, 6, (5), 34--37, Slikkerveer, L. J., (1993) 'Transformation-of-technology and indigenous approaches lo innovation: towards a new paradigm of sustainable development.· Paper presented at lhe Pithecanthropus CentenniaI1893-1993, 'Human Evolution in its Ecological Con­ texl', held in Leiden, The Netherlands, 26 June-l July 1993. References 211 Titilol •. S. 0., (1992) 'The role of indigenous knowledge in rural development activities for sustainable development in Nigeria.' Paper presented at the InternatíonaJ Sympo~ sium on Indigenous Knowledge and Sustainable Development. held at the lnterna~ lional Institute for Rural Reconstruction, Silang. Cavite, Philippines. 21-25 September. 1992. Chapter 6 Altieri, M. A. .nd Merríck, L C. (1988) Agroecology .nd In Si/u Conservation ofNative Crop Diversity in the Third World. In Wilson, E. o.. (ed) Biodiversi/y. Washington: N.tional Academy Press. Boneodin, R. and G. Proin (1997) The Dynamics of Biodiversity Conservation by Home­ gardeners in Bukidnon, Southern PhiHppines. In Local R&D: lnstitutionalizing and sustaining agricultural innovation in Asia. Les Baños: UPWARD. Brush, S. (1991) A farmer-based approach to conserving crop germplasrn. Economic Botany 45(2) 153--{J5. Brush, S. (1992) Ethnoecology, biodiversity, .nd modernízation in Andean POtoto Agrí­ culture. J Elhnobiol. 12(2): 161-85. Brush, S. (1993) 'In situ conservation of landraces in Centers of Crop Diversíty.' Paper presented for the Syrnf.0sium on Global Implications of Germplasm conservation and utiJizatíon at the 85 Annual Meetings of the American Society ol' Agronomy, Cincinnati, OH. Bru,h. S. B., H. 1. Comey and Z. Huaman (1981) 'Dynamics of Andean potato agrienl­ ture', Economic Botan)' 35(1): 7~8. Chambers, R. (1992) Rural Appraisal. Relaxed and Partícipalory. IDS Discussion paper No. 331 October, Brighton, Sussex In,titute of Development Studies. Castillo, G T. (1995) Secondary crops in primary functions: The search for systems, 'ynergy, and su,tainability. UPWAR D Working Paper Series #2. Los Baños, Laguna: UPWARD. Edgerton, Ronald K. (1982) Frontier Society on the Bukidnon Plateau. In McCoy, A. W. and Ed. C. de Jesus (ed ..) Philippine Social Hislory: Global Trade and Local Trallsfor­ mations. Manila: Ateneo de Manila University Press., Geertz. C. (1963) Agricultural involution Tire processes o/ ("(ologica! change in Indonesia. London: University of California Press, Ingold, T. (1993) Globes and spheres: (he topology of environmentalism in Miltan, K. (00). Environmentalism: the vicw ¡rom anthropology, ASA Monographs 32 London: Routledge. IPGRI (1993) Diversity for Development. The Strateg)' 01 the International Plan! Genetic Resource.f Institutc. Rome: Intemattonal Plant Genetic Resources Institute. Joshi, A and 1. R. Whitcombe (1996) Farmer Particípatory erop Improvemen!. n. Partícipatory Varietal Selection. a case study in India. Experimental Agriculture Vol. 32, pp. 461-77. Nazarea-Sondov.l, V. (l991) Memory Banking of Indigenou. Technologies Associated with Traditional Crop Varieties: a focus on sweet¡ otatoes, In Sweetpotato Cultures 01 Asia alld Soulh Pacifico Proceedings of the 2" . Annual UPWARD Intemational Conference. Los Baños, Philippines. Nazarea~Sadova1, V. (1994a) Memory banking: The conservation of cultural and genetic diversity in sweet patato production. In: Prain, G. D. and C. Bagatanon (eds), Local knowledge, global science and plant genetic re$ources: fOwards a parrnership. Proceedings orthe lnternational Workshop on Genetic Resources. Los Baños, Laguna: UPWARD. Nazarea-Sandoval, V (1994b) Memory banking protúcol: A guide for documenting indigenous knowledge associated with traditional crop varieties.. In Prain. G. D. and C. Bagalanon (oos). Local knowledge, global science and plan! gene/ir: resources: towards a partnership. Proceedings of the Intemational Workshop on Genetic Resources. Los Baños, Laguna: UPWARD. Prain, G. (1995) Sweet potato in Asian Production Systems: an overview of UPWARD'g 212 Biological and Cultural Diversily firsl phase research. In Taking Root: Proceedings of the Third UPWARD Review aud Planniog Workshop. Los Baños: UPWARD. Prain, G., 11 Gin Mok, T. S.wor, P. Chadikun, E. Almodio .nd E. Relw.ly Silmorang (1995) Interdisdplin.ry collecting of lpomoea batatas germplasm .nd .ssodated indi­ genou. knowledge in Irian Jaya. In Guarino, L., V. Ram.n.lh. Rao and R. Reíd (eds). Colle(:ting Plant Genetic Diversity_ Oxford: CAD InternationaL Prain, G. and M. Piotero (1998) Cúrnrnunal Conservation of Rootcrop Genetic Resourccs in Southern PhiHppines. In Prain, G. and C. Bagalanon (eds) Conservarian and Change.' farmer management 01 agro-biological diversity in ¡he Philippines, Los Baños: UPWARD. Sth.pit, B. R., K. D. Joshi and 1. R. Whítcombe (1996) Farmer Particípatory Crop Jmprovement. lII. Participatory Plant Breeding, a case study tor rice in 'KepaL Experi­ mental Agriculture Vol. 32, pp. 479-96. Whitcombe,1. R., A. Joshi, K. D. Joshi and B. R. Sth"pit (1996) Farmer Partkipa!ory Crop Improvement. 1. Varietal Selection and Breeding Me!hods and their Impact on Biodiversily. Experimental Agricu/¡ure Vol. 32, pp. 445-60. Yen, D. (1974) The Sweet patato and Oecania, An essay in EtitrwbMany. HonoJulu, Hawaii: Bishop Museum Press. Vega, Amihan Belita and José L. Bawsmo (1998) Community-based Sweet Potato Gene­ banking and Distribution System. (n Prain, G. and e, BagaJanan (eds) Conservation and Change: farmer managemenf of agro~biol()gical diversity in fhe Philipplies. Los Baños: UPWARD. Chapter 7 Altieri. M. (1987) "'The significance of diversity in the maíntenance of the sustainabiJity of tradítional agroecosystem: [LElA Newsletrer (2): 3~7. Bodri, B. and A. Badri. (1994) 'Women and Biodiversity.' Development 1: 67-71. Bentley, J. W (1990) 'ConocÍmÍento y experimentos espontáneos de campesinos hondureños sobre el maíz muerto' [Honduran peasants' knowledge and spontancous experiments about dead-maizeJ, Manejo Integrado de Plagas (Costa Rica) 17: J6--26. Berkes, F. and C. Folke. (1994) 'Linking social .nd ecologieal systems for resilienee .nd sustainabHity_' Paper presented at the Workshop on ;Property Rights and the Perfor­ mance of Natural Resource Systems', Stockholm (Sweden): The Beijer lnternational Instítute of Ecological Economics, The Royal Swedish Acaderny of Seienees. Chambe", R. (1994) 'Afterword', pp. 264-65 in l. Scoones .ud 1. Thompson (eds). Beyond Former First. Rural people 's knowledge, agricultural research and extension practice. London: Intermediate TechnoJogy Publicatiúns. Cunningham, A. B. (1991) .périence pour l'ide!1lification des connaissances locales sur le plateau Adja, Report Agricultural Faculty, Beninese National Univershy, Cotonou, Darré, 1. P. {l985) La Paro/e el la Technique. L'univers de pensie des éleveurs du Ternois. Paris: I'Harmattan. Giddens. A. (I987) Socia/ Theory avd Modorn Sociology. Cambridge: Polily Pres>o Giddens, A. (1990) 'Jurgen Habermas'. Ch. 7 in: Skinner 1. Q. (Ed.) The r.'um ofgralld theory in the human sdences. 6th oo .. Cambridge: Cambridge lJnivershy Pres:;. Haverkort, A .. 1. Van der Kamp, and A. Waters-Bayer (Eds.). (1991) Joining formers' experiments. Experiences in Participaior}' Technology Development, London: Inter­ mediate Technology Publicarion&. Jiggins. J. L. S. and H. De Zeeuw. (l992) Participatol)! technology development in practice: preces!> and mcthods. In: e, Reijntjes., B. Haverkort and A. Waters~Bayer, Farming jor Ihe PUlure,' an introduclion lo low external input agriculture, london: MacMillan and Leusden: ILEIA. 13S~62. Kline, S. • and N, Rosenberg. (1986) An Overview of ¡nnovation. In: Landa u, R. and N. Rosenberg (Eds.). The Posaive Sum Strateg}~ Harnessing Technology for Economic Growlh. Washington (DC): Natlonal A,ademie Press, pp 275-306. Koudokpon. V., J. H. A. M. Brouwcrs. M. N. Vcrstecg, A. Budelman. (in press) 'Priority setting in research for sustainable land use: The case of the Adja Plateau. Bento.' accepted by Agrofores!r)' Systems. Lionberger, H. and C. Chango (1970) Farm Informationfor Afodernising Agrü.:ulture/ The Taiwan System. New York: Praeger. Matleson, P. • K. D. Gallagher, and P. E. Kenmore. (l992) Extcnsion and inrcgrared pest management for planthoppers in Asian irrigated rice. In Denno~ R. E and T 1 Perfect (Eds.). Ec%gyand Management of Plani Hoppers. Landon: Chapman and Hall. Rogers, E. M. (1995) Diffusion of Innovations. (4th edition) New York: Free Pless. Róling, N. G, (1988) Extension SCU!ftcc,' lnformation syslems in agricultural development. Cambridge: Cambridge University Press. Roling, N. G. (1993) Agricultural knowledge and environmental regulation: the Crop Protection Plan and the Koekoekspolder, Socí%gia Ruralis, 33(2), June: 21 2-3l. Róling. N. G. (1994) Platforms for declsion mabng about eco~systems. Chapter 31 of L O. Fresco el al. (Ed,). FUlure of Ihe úmd: Mobilising and ¡nlegrating Knowledge for Land Use Options. Chichester: John Wiley and Sons. Ltd, pp 386-93. Róling, N. G. and P. Engel. (1989) IKS and knowledge managoment: utílising knowlcdge in instttutional knowledge systems. In: Warren, D. M., L 1. Slikkeveer, and S, Ogun~ 216 Bi%gica/ and Cultural Diversity tunji Tititola (Eds.). Indigenous Knowledge Syslems." Implications for Agricullure ami [meroalional Developrnenl. Ames (Iowa): Iowa State University: Technology and Social Change Programma, Studies in Technology and Social Change, No 11, pp 101-116. Roling, N. G. and P. EngeL (1991) The development of the concept of Agricultural Knowledge and Information Systems (AKIS): Implications for Extension. In: Rivera. W and D. Gustafson (Eds.). Agricultural Extens;Qn,' Worldwide ¡nstitutional Evolulion and Forees for Change. Amsterdam: Elsevier Seicnee Publisher" pp 125-139. Roling, N. G. and E. Van de FlíerL (1994). Transforming extemion for sustainable agriculture: the case of Integrated Pest Management in rice in Indonesia, Agriculture and Human Values Vol 1I (2+3), Spring and Summer, pp 96-108. An adapled version of thi, article appeared in: Roling, N. G. and E. Van de Fliert (1998). Transforming Extension For Sustainable Agrículture: The Case of Integrated Pest Management in Rice in Indonesia. In: N. Roling and A. Wagemakers (Eds.). Facilitatmg Sustainable Agriculfure. Participatory Learning and Adaptive Managemem in Timesof Environmenlal Uncertainty. Cambridge: Cambridge University Press. SomerS, B. M .. and N. Roling. (1993) Ontwikkeling van kennís voor duurzame landbo"w: een verkennende studie aan de hand van enKe/e experimente/e projekten. Den Haag: NRLO. Van de FHert, E (l993) lntegrated pest managemenl: farmer field schools generale suslain­ ab/e praclices. A case sludy in Central Java evaluating IP M training. Wageningen: Wageningen Agricultural University Paper. 93-3. Van de Fliert, E .. K. van Elsen .nd F. Nang,ir Soenanto. (1993) Integrated Rat Manage­ ment: A Communíty Actívíty. Results of a Pilot Programme in Indonesia, FAO Plant Prol. Bult., 41(3). Van der Ley, H. A. and M. D. C. ProosL (1992) Gewasbescherming met een toekomsl: de visie van agrarische ondernemers: een doelgroepverkennend onderzoek ten behoeve van voorlichting. Wageningen: Agricu1tllral University, Department of Communication and Innovation Studies. Van der Ploeg, 1. D. (1991) Landbouw als mensenwerk. Muiderberg: Countinho. Van Kessel, 1. (1988) Aymara technologie: een culturele benadermg. Workbook for the course SociologylAnthropology of Development Poliey, Amsterdam: Free University, Dept. CASNW Van Weeperen, w., J. Proost and N. Roling (1998). Introducing integr.ted arable farming in the Nelherlands. Chapter 6 in: N. Róling and A. Wagemakers (Eds.). Facílitating SustainabJe Agriculture. Participatary Learning and Adaptive lkfanagement in Times 01 Enl'ironmental Uncertain(v, Cambridge: Cambridge University Press, Chapter 11 Boster, J. (1985), Selectíon for perceptllal distinctiveness: evidence from Aguaruna culti­ vars of },.{anihot ésculema. Economlc Botany 39: 3l0-25, Bulmer, R, (1965}. Belíefs concerning the propagation of new varieties of sweet potato in two New Guinea HighJand societies, JQurnal 01 the Polynesian SocielJ' 74: 237-39. Golson, 1. and D. Gardner (1990). Agriculture and sociopolitical organization Ín New Guinea Highlands prehistory. Annual Revíew of Anthropology 19: 395~17. Heider, K. (1969). Sweet potato notes and lexical queries, or, the problem of all Ihose names for sweet pútato in the New Guinea Highlands. Kroeber Anthropological Socíel)' Papers 41: 78-86. Matanubun, H. et al. (1991). Eksplorasi [uhi-uhian di Dabaupalen Pania; don Kahupaten Biak-Numfar, Iran Jaya. RTCRC Research Publications, No. 3. Manokwari: PSlJ UNCEN. Sehneider, 1. et al. (1993). Sweet polato in the Baliem Valley area, Inan Jaya: A repor! on ealleelian aud s{Udy ofsweet potato germplasm, April-May 1993. Bogor: CIP/CRIFC/ RTCRC. References 217 Yen, D. (1974). The sweel potato and Oceania: An essay in erhnohOlan}t Honolulu. Hawaií: Bishop Museum. Chapter 12 Gubbels, Peter. (1988) Pe.sant farmer agricultural self-development. ILEIA Newsletter 4 (3): 11-14. Gubbels, Peter. (1992) Farmer-firsl research: Populis! pipedream or practical paradigm? Prospects for indigenous agricultural development in fti?st Africa. MSc Dissertation, School of Development Studies, University of East Anglia. ETe. (1991) Learning for people-cenlered technology developmem. A Iralnlng gulde. Draft version 1.1 Part 4: Farmers' experimentation. Leusden. The Nether1ands, Hoveyn, e. A. (1991) Proeven me[ planten. Van opzet tot ana/yse. Wageningen, CABO­ DLO. Richards, Paul. (1987) Agriculture as a performance. In: Chambor:;, R., A. Pacey, and L. A Thrupp (eds.), 1989: Farmer first. Farmer innovQtiOll and agricultural rescarch, London, Intermediate Te chnology PublícationK Schon, Donald A. (983) The reflective practitü.mer. How professionals thínk in action. New York. Basic Books.. Stolzenbach, Arthur E V. (1992) Learning by improvisation: The lagic olfarmers' experi­ mentation in Ma/í. Paper prepared for tbe IIED/IDS Beyond Farmer First: Rural People's Knowledge Agricultural Research and Extensíon Practice Workshop. Insti~ tute of Development Studies, University of Sussex, 27-29 October 1992. StoI7.enbach, Arthur E V. (1993) Farmers' experimentation: what are we talking about? ¡LElA Newslmer 9 (1): 28-29. Vao der Ploeg, Jan Douwe. (1987) De verwetenschappelijking vall de landhouwbeoefening. Wageningen, Medelingen van de Vakgroep Sociologie, Landbouwuniversiteit. 21, Chapter 13 Association for Rural Advancement. (1990) Maputaland: Conservation and Removals Special Report No. 6. AFRA, Pietermaritzburg. Begg, G. W. (l980) 'The kosi system: aspects of its. biology, managemenl and research.· In: Bruton, M. N. and Cooper, K. H. (Ed.) Sludies on the ecalagy oI Maputaland. Rhodos University, Grahamstown and Natal Branch of the Wildlife Society of Southern Africa, Durhan. Bryant, A. T. (1949) The Zulu people: As they lVere before rhe while man carne. Shuter and Shooter, Pietermaritzburg. Centre for Cornmunity Organisation Research and Development (CORD) (199la) 'Regaining Control'. In: Ramphele. M. ReslOring che land, environment and thange in post-apartheid South Africa. Panos Publications, London. CORD. (I991b) Overooming apartheid's land legacy in Maputaland (Northern Natal). Working Paper 3. University of Natal, Durban. Cleveland, D. A. and Soleri, D. (1991) Foodfrom drpJand gardens. An ecological, nutri­ tional and social approach lo small-scale household foad production. Centre for People, Food and Environment, Tucson. Felgate, W. D. (1982) The Tembe Thonga of Natal and Mazambique: An ecologlcal appraach. (Edited .nd arranged by E. J. Krige) Dopartment of Afriean Studies. Uni­ versity of Natal. Durban. Granger,1. E., Hall, M., Mckenzie, B. and Feely, 1. M. (1985) Arehaeologlcal researcb on pJant and animal husbandry in Trans.kei. Sourh African Journal of Science. Vol. 81. 12-15. Junod, H. A. (1927) The life of a Sourh African Tribe. Vols, 1 & 2. MacMiIlan, London, Kwazulu Government (1992) Draft white paper on development poliey in lbe Ubombol Ingwavuma Regíon. Macvicar, C. N., De Villecs, J. M., Loxton, R. E, Verster, E. et al. (1977) Soil classification. 218 Biological and Cultural Diversity A binomial system for South Africa. Department of Agricultural TechnicaJ Services, Pretoria. N.ttrass,1. (1977) Migrant labour and underdevelopment. The case of KwaZulu. BI.ckl Vlhite incorne gap project. lnterim Research Repon No. ], Department of EconomÍcs. University of Natal, Durban. Richards, P. {1985} lnJigenous agricultura! revolution.' Ecologyandf {J()d production in ~Vest Africa. Hutchinson, London. Rhoades, R. (1990) Models, means and methods: Rethínking ruml development research. In: Rhoades, R. E., Sandoval, V. N. and Bagalanon, C. P. (Ed.) Use,', perspeclive wilh agricultural research and development: Asian trainlng af trainers on farrn diagnostit skills. Laguna. 10 .. ·14. Saterson, K. A. (1990) [ntegration of Bíologícal Conservatíon with Development Policy: The Role of Ecological Analysis. In: Goodland, R. (Ed.) Race /0 save the tropies. ecology and (!'conomics for a sustainable ¡Ulure. Istand Press~ \Vashington. 141 ~ 159. Taylor. D. (1988) Agricultural Practices in Eastern Maputaland. Development Southern Africa. Vol. 5, No. 4. 465~48L Taylor, D. (1991) Towards Sustainahle Agriculture: A People Orientated Approach. In: African Agr%restry: Emphasis on Southern Africa, Environmental Forum Report, Foundation for Research Department, Pretoria, Taylor, D. (J 993) The Consequence. of lnterventions on Local Agricultural Devclopment in South Africa. In: De Boef, w., AmaDor, K aud Wellard, K. with Bebbington, A. (Eds.) Cu/tivating KnOJ-llledge: Gene/ie Diversit}~ Fatmer Experimentation and erop Research. lntermediate Technology Publications, London. Tinley, K. L and Van Riet, W. F. (1981) TongaJand-Zonal Ecology and Land Use Proposals. Unpublished Report on Proposals Towards an Environmental Plan for KwaZulu. Nature Conservation Division, KwaZuJu Department of Agriculture and Forestry. Ulundi. Vandeverre Aspey Robinson and A.sociates (VARA). (1989) Ubombo-Ingwavuma Struc­ ture PI.n. Unpublíshed Report lor Department of Development Aid .nd KwaZulu Government. Van Riet, W. F. aud Begg, G. (1988) M.putaland: Ecological and Conservation PotentiaL In: VARA (Vandeverre Aspey Robínsnn .nd Assrn:í.tes) L'bombn-Ingwavuma Stru.­ lure Plan. Unpublished Roport for Dep.rtment of Development Aid and KwaZulu GovemmenL Webster, D. (1987) Family .nd Household in KwaDapha, KwaZulu. Mimeograph. Department of Social Anthropology, University of Witwatersrand, Johannesburg. Webster, D. (1991) Abqfazi Ba/honga Bafihlokala: Ethnicity .ud Gonder in a Kw.Zulu Border Community. In: Spiegel A. D. and McAlIister P. A. (Eds). Traditioo aad Transition in Southern Africa. African Studies Fiftieth Anni¡,'crsary Vo{umc, Vol. 50, Nos. I & 2. Witwatersrand University Press, Johannesburg. Zululand Land, Delimitation Commission. (1905) Report by the Joínt Imperial .nd Colonial Commissioners Award 1902~1904. Ann ..u re G. No. 38: 288.