SoH FertHity Research for Maize and Bean Production Systems of the Eastern Africa Highlands CA 1291 ;j 22 SET. 1993 Proceedings of a Working Group Meeting Thika, Kenya 1-4 September, 1992 CIAT African Workshop Series No. 21 Edited by: Charles S. Wortmann Joel K Ransom Organization: Centro Internacional de Agricultura Tropical (CIAT) Centro Internacional de Mejoramiento de Maíz y Trigo (CIMMYT) Sponsors: Canadian International Development Agency (CIDA) United States Agency for International Development (USAIDl ~ I Correct cítation: (1993). In: C. S. Wonmann and J.K. Renaom (Eda.) Soil Fertility Researeh for Maíze and Bean Production Systems of the Eastem ACrica Highlands: Proceedings of R Working Group Meeting, Thika, Kenya, 1-4 September, 1992. Network on Bean Research in ArrieR, Workahop Serie. No. 21, CIAT, Dar es Salaam, Tanzania. PREFACE This volume reports the proceedings of a working group meeting on soil fertility research for maize and bean production systems ofthe high altitude areas ofEastem Africa. The meeting was held in Thika, Kenya 1-4 September, 1992 with the objective ofimproving the effectiveness ofresearch through prioritization of research topics, improved collaboration between concemed research institutions, better focussed training and specialization, and increased availability of resources for research. The working group meeting was organized by the CIAT Regional Prograrnme on Beans in Eastern Africa and the CIMMYT East Mrican Cereals Programme. Funding for the meeting and this publication was provided by the United States Agency for International Development (USAID) and the Canadian Intemational Development Agency (CIDA). Further information on regional research activities on bean in Africa that are part of these projects is available from: Pan-Africa Coordinator, CIAT, P.O. Box 23294, Dar es Salamm, Tanzania. Coordinateur Regional, CIAT, Programme Regional pour I'Amelioration du Haricot dans la Region des Grands Lacs, B.P. 259, Butare, Rwanda. PUBLlCATIONS OF TBE NETWORK ON BEAN RESEARCH IN AFRICA Workshop Series No. l. Beanfly Workshop, Arusha, Tanzania, 16-20 November, 1986. No. 2. Bean Research in Eastern Africa, Mukono, Uganda, 22·25 June. 1986. No. 8. Soil Fertility Research fOT Bean Cropping Systems in Mrica. Addis Ababa. Ethiopia, 5-9 September, 1988. No. 4. Bean Varietal Improvement in Mrica, Masero, Lesotho, 80 January· 2 February, 1989. No. 5. Troisieme Seminaire Regional sur l'Amelioration du Haricot dans la Regian des Granda Lacs, Kigali, Rwanda, 18-21 Novembre, 1987. No. 6. First SADCClCIAT Regional Bean Research Workshop, Mbabane, Swaziland,4-7 October, 1989. No. 7. Second Regional Workshop on Bean Research in Eastern Mrica, Nairobi, Kenya, 5-S March, 1990. No. 8. Atelier sur la fuation Biologique d'Azote du Hericot en Afrique, Rubona, Rwanda, 27·29 Octobre, 1988. No. 9. Quetrieme Semina1re Regional sur l'Amelioration du Haricot dens la Region des Grands Laca, Bukavu, Zaire, 21·25 Novembre, 1988. No. 10. National Reaearch Planning for Bean Production in Uganda, Makerere University, Kampala, Uganda, 28 January. 1 February, 1991. No. 11. Proceedings of the First Meeting of the Pan-Africa Working Group on Bean Entomology, Nairobi, Kenya, 6·9 August, 1989. No. 12. Ninth SUAlCRSP Bean Research Workshop and Second SADCClCIAT Regional Bean Research Workahop. Progress in Improvement of Common Beans in Eastern and Southern Mrica, Sokoine University of Agriculture, MorogoTO, Tanzania, 17-22 September, 1990. No. 13. Virus Diseases orBeans and Cowpea in Mrica, Kampala, Uganda. 17·21 January, 1990. No. 14. Proceedings afthe First Meeting of the SADCClCIAT Working Group on Dronght in Beans, Harere, Zimbabwe, 9-11 May,1988. No. 15. First Pan-Africa Working Group Meeting on Anthracnose ofBeans. Ambo. Ethiopia, 17-23 February. 1991. No. 16. Cinquieme Seminaire Regional sur l'Amelioration du Hericot dans la Region des Grands Lacs, Bujumbura, Burundi, 13-17 Novembni. 1989. No. 17. Sixieme Seminaire Regional sur l'Amelioration du Haricot dans la Region des Grands Lacs. Kigali, Rwanda. 21-25 Janvier, 1991. No. 18. Conference sur Lancement des Varietes, la Productíon et la Distributíon de Semaineg de Haricot dans la Region des Grands Laca, Goma, Zaire, 2-4 Novembre, 1989. No. 19. Recommendations ofWorking Groups on Cropping Systems and SOU Fertility Research for Bean Production Systems, Nairobi, Kenya, 12-14 February, 1990. No. 20. Proceedings ofthe First African Bean PathoJogy Work.shop, Kigali, Rwanda. 14-16 November, 1987. No. 21. Soil Fertility Research for Maize and Bean Productíon Systems of the Eastern Mrice High1ands - Proceedings of a Working Group Meeting, hika, Kenya, 1-4 September, 1992. IV Occasional Publieations Series No. l. Agromyzid Pesta ofTropical Food Legumes: a Bibliography. No. 2. CIAT Training in Africa. No. 3A. First Amcan Bean Yield and Adaptation Nursery (AFBYAN 1): pan I. Performance in Individual Environments. No. SB. First African Bean Yield and Adaptation Nursery (AFBYAN 1): Parl II. Performance Across Environmenta. No. 4. Assessment ofYield Losses Caused by Biotic Stress on Beana in Africa. No. 5. Interpretation oC Foliar Nutrient Analysis in Beans-the Diagnosis and Recommendation Integrated System No. 6. The Banana-Bean Intercropping System in Kagera Regíon ofTanzania _. Resulta of a Diagnostic Survey Reprint Series No. 1. Bean Production Problcms in the Tropics: Common Beans in Africa and their Constrainta. No. 2. Rean Productíon Problems in the Tropics: Insects and other Pesta in Amcs. v TABLE OF CONTENTS Introduction 1 Farmer partlcipation in soll management research in Matugga village (Mpigi District) oCUganda - an alternatíve approach M. A Ugen, P. K. Jjemba and M. Fischler PrincipIes and methods oC BOíl Certility research in maize based cropping systems of tbe Eastern African Highlands P. Woomer, J. Lekasi, R. Okelabo and C. Palm 21 Low input alternatives for ímproved soH management M. K. O'Neill, F. K. Kanampiu and F. M. Murithi 35 Agricultura technology evaluation .- sorne cousiderations C. S. Wortmann Fertilizer trial resulte of maize and maize-bean intercrop trials in Kenya B. O. Mochoge 59 . Research for sustainable agricultural produetion systems in high altitude zones of Eastem Afriea C. S. Wortmann 79 What beyond sustainability? P. Woomer 91 Discussion on papere presented 101 Priority setting and identifieation oC research topies 107 Reeommendations oC small working groups 112 List of Partlcipants 114 VII. INTRODUCTlON Beans (Phaseolus JJulgaris L.) and maize (Zea mays, L.) are important food crops in the medium and high altitude zones of eastem Mrica and are often grown in association. Productivity of maize and beans in these systema is often constrained by soil fertility problems. In sorne cases, the soil fertility problems are easily managed, such as in sorne high potential areas where nitrogen and phosphorus deficiencies can be alleviated with fertilizer use. In other cases, the soil fertility problema are difficult to manage due to the problem's complexity, inadequate technologies or inadequate infrastructure. Research on nutrient use efficiency, including fertilizer use efficiency, is important for areas with easily manageable problema in order to improve sustainability and increase profitability. The research needs for the low potential areas are often great and complex: the problem is often poorly understood; soils may be fragiJe and the problem complex; input use is not profitable; and appropriate solutions are variable. While research needs are great, resources available for this research are scaree. To improve the effectiveness ofresearch in increasing production in a sustainable matter, available resources must be used more efficiently or the availability of resources must be increased. The Centro Internacional de Agricultura Tropical (CIAT) and the Centro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT) organized this working group meeting to address issues of soil fertility research in the Eastem Africa Highlands. The working group consisted of bean agronomists, maize agronomists, soil scientists and socio-economists from the national research organizations ofEthiopia, Kenya, Tanzania and Uganda. Regional staffofCIAT, CIMMYT and the Intemational Center for Research on Agroforestry (lCRAF) and staff ofTropical Soils Biology and Fertility (TSBF) also participated. ' The working group sought means to improve the efficiency of resources available for research as welJ as to increase these resourees so that research might have a greater impact on production and soil management. Problema and their probable solutions were reviewed and prioritized. Alternative research approaches, including approaches with greater farmer participation, were considered. ColJaboration between the various institutions and programmes involved in soil fertility research for maize-bean systems in the highlands of eastem Africa was addressed. The need for cornmon strategies and methods, and for specialization and training, was discussed. In order to solve problems or fi11 information gaps of most concern to the participants, research topies were identified and preparation of proposals initiated. This document is a compilation ofthe papers presented during the working group meeting and the results ofthe working sessions. FARMER PARTICIPATION IN SOIL MANAGEMENT RESEARCH IN MATUGGA VILLAGE (MPIGI DISTRICT) OF UGANDA •• AN ALTERNATIVE APPROACH 1292 O M. A Ugen', P. K Jjemba' and M. Fischler. 22 SE1,1993 , Agronomist and Soll Míerobiologiot, Ka_da Researth Station, P. O. Box 7065, Rampal .. , Uganda. Agronomist, ClAT Regional Sean Pro¡¡ramme olE. Africa, Kawanda Researth Station,. P. O. Box 6247, Kampala, Uganda. , INTRODUCTION Human agricultural activities shouId basicaIIy aim at exploiting the environment efficiently. If not properly planned, these activities can lead to soil degradation. Low soil fertility imposes serious constraints on erop productivity, even on fertile Boils ofUganda. Soil fertility continues to decline due to severe erosion, overgrazing, cultivation on marginal lands and exportation of nutrients without adequate replacement. Such degradstion is most prevalent in the less developed countries where agriculture is mainly by amall scale farmera producing nudnly for suhsistence amidst mounting population pressure. Conventional agricultural research has, in most cases, tended to greatly benefit the economicaIIy well-offfarmers in these countries, who malte up a small minority. For example, with all the resources and time put into research on inorganic nitrogen use and crop response in Africa. the average application rate for trua nutrient is a meager 5kg/ha (Hauck. 1988). But through generations of ohservations as well as trial and error. Carmera have learnt a great dea1 ahout theír soiIs which relate to colour, soíl depth, crop performance, existing natural vegetation etc. Likewise, management of these locally different types oC soila may ditfer in as far as trying to sustain erop production is concemed. Therefore, if research is to adequately address farmen' needs. it has to consider their traditional knowledge as ver¡ valuable. Farmers have a role in problem identification. determination of causes, evaluation of potential solutiona. and in the development and implementatíon of a research plan. Without their involvement, there is a greater probability that researchers will en in prioritization and selection of problems for research. in identificstion of causes and most Iíkely solutions, or in deciding how and where OOst to conduct the research. Lightfoot et al., (1987) indicated the importance of involving farmera in identifying, analyzing and solving systems problema in the Philippines. Ravnhorg (1990) showed the need to involve farmera in soíl management dísoussions in arder to set prioritiea for reaearch work in Tanzania. The involvement of farmers in research work have a1so been emphasized by Ashby(990). Fujisaka (1989,1990), and Tripp and Woolley (1989). AlI have shown that a OOtter understanding oC farmers' prohlems in managing their soils and production constramts has to start with serlous farmer involvement. Therefore, a starting point oC research on soil management is to underatand the farmers' present management practices; identü;y problema together with the farmera; understand causes and solutions to these problema; and set priorities ror future research work to solve the problems. Tripp and Woolley (1989) suggested a six step format in identifying ractors for experimentation and Lightfoot et al. (987) propased a three step approaeh in identifying problems affecting farmera. This papar reporta prelim!nary findings about a study done in Matugga village (Mpigi district. Uganda) following the format by Tripp and Woolley (1989). The objectives of the study were: (i) to understand .the predominant farming systems in the area; (ii) to identü;y farmers' problema (particularly soil-related) as to laya foundation for research towards improving the present soil management practicas; 3 mi) tCl identifY relatiolU!hips between the fanners perceptions, knowledge. and practices relating ro soil fertility; and (iv) tCl develop a research plan for future research in the area. MATElUALS AND METHODS The study was conducted in Matugga village, Uganda (latitude 0.44 N, longitude 35.5 E. and altitude 1200 masl)) with a total of twenty !lb!: farm units selected randomly. The selected farmers were subjected to open-ended, briefinterviews fol' passport and BOíl related infonnation (Appendb!: 1). Farmers were encoureged tCl relate the locally identified soil types to suitability of particular erops. Furthermore questions abeut soiJ management practicas were asked. The holdings were revisited ro collect composite soíl samples and to determine soil depth using a soil auger frero each of the identified soU types in March, 1992. The pH of the top soil (water ssturation method), wet soU coloul' (using s Munsell colour chart), and texture (by feel) were determined for al! soil samples. In addition s complete anaIysis includíng organic mstter (organic carbon" 1.7; Wallcley and Blsck oxidation), available P, K and Ca (Ammonium lactate extraction, pH 3.8; Foster, 1971) was carried out for the "problem" BOils, namely "Lunyu' and "Zibugo", as well as for some "Lidugavu' soils (control). Participant farmers were subsequently invited for a series of meetings ro further identify agricultural problema, map their area and draw the predominant soil catena. The identified prohlems were ranked according ro arder of importance by open voting. Farmers participated in diagramming the causes oftbe problems and identification ofpotential BOlutions. A research plant was prepared for experimentation tCl begin in September 1992. RESULTS AND DISCUSSlON Farmera' perception oftheir soils and observations Oftbe farmera interviewed, 69% were women and in 4% ofthe cases the husband and wife were joíntly interviewed. The majority offue farmera interviewed (69%) indicated to be fue head of the household. The average farm aÍze was 6.1 acres ranging from 1 te 18 acres. Regsrding Isnd ownership, 58% ofthe farmera questioned indicated te have no land title ("Kibanja" type of ownership, land inherited), whereas 39% held a land title ("Maílo' type of ownership). Lease ofland was reported in one case only. Only 38% oC the farmer. had livesteck ranging from 1 to 7 head oC cows and 2 to 9 head of sheep or gaste. On the average, there was 2.6 cows and 3.1 goate or sheep per fann having Iivesteck. Fifteen soil types were identified by the farmers interviewed. Table 1 liste the soil types in the Luganda vernacular wiili approximate translatiolU!. The criterla used by farmers for soil classification were soil color (5 types). texture (5 types). fertílity status (3 types). vegetation (1 type), and consistency (1 type). On a single farro, up to five different soil types were identified. On the average there were 2.3 soil types per farro. The criterla were very similar to those found by Fujisaka (l989) who listed slope, color, fertiJity, texture. acidity, and friability as criteria used by farmera for soíl classification in the Philippines. The predominant soil types were "Lidugavu", "Luyinjayinja", "Limyufu", and "Lunyu' which together accounted for 67% oí the fields surveyed. "Lunyu" and "Zibugo' accounted for 16% oC the fields and were classified as "problem" saíla associated with low soil fertility. 4 Table l. Farmers' Boil classifieation and obaervat.ion oC Boila at Matugga. W&t ..,il color Soil depth(em) Topeoil Subeoil Range Mean Ran¡¡e TrlUlll- OCC\llT"""" ¡atíon ToteJ of fa....,.. Lidngavu black/ dark 13 50 darle reddi.h brown or very dark grey 20-55 28 31->90 5.1-6.7 S.O LuyilliayiDja gravell)' 12 490 4.2-6.4 5.8 Limyufa(yofu) red(dish) 8 31 dark reddiab brown, reddi.h brown, dU8kY red or yellowish red 15-35 24 >90 5.0-6.4 5.4 Lunyu salty/ infertíle black 7 'J:/ dark reddi.h brown very darle gray or 17-35 26 40->90 4.8-5.9 5.2 Lukusikusi browni.h 5 19 dark reddi.h brown or reddiah brown 25-53 32 >90 5.3-6.0 5.5 Luoenyuoenyu Z,bugo aandy dead, kills erope 3 3 12 12 dark reddi.h brown dark reddish brown 28-40 29 30 45-85 40->90 4.7-5.5 5.0-5.6 5.0 5.3 f.rtil. l! 2 8 8 dark reddiab brown very dark grey or 25-28 26 32 >90 >90 5.4-5.5 4.0-4.8 5.4 4.4 1 4 (not determinedl 30 5.4 1 " dsrk reddi.h brown 20 >90 >90 Soil type (v.macular) Gimu Bumba !'Tooi) Lwazi L)'skibira clay/muddy blaek ro90 5.6 Ligonvu soft 1 4 dsrk raddi.h brown 30 >90 5.6 Kiwngankofu aandyloam, ailty 1 " very dark grey 45 >90 6.0 The typical soil catena for Matugga as described by the farmera ia ahown in Appendix 2. Soils on the hilltep were generally described as steny and shallow with a high infiltration rate but a low water-klding capacity. "Lunyu' soíls are oommon on eroded sites. Other soíl types frequently occurring are "Luyinjayinja- and "KiwugankoCu". Soils on the hilllílope were described as being deeper than on the hilltop with better soil moisture. Generally, aIl soil types except the cJayey "Bumba" type were found on the hillslope with the more fertHe soila such as "Lidugavu' and uGimu· being on the lower slopas {foothill}. The valley soils generaJly have a dark tep soil, a cJay subsoil and are underlain with sand. The vslley IIOila were descrlbed as difficult te till when wet. When farmers were asked te enumerate good or bad soil characteristics they usually indicated high or low crop yield as the good or bad soil festure, respeetively. Further probing ("What soil characteristiea are responsible for the good or bad crop yield?") was necessary to get the farmers' pereeption ofthe characteriBties relatad to the Boíl iteelf and not the crop. The most frequently mentioned lIoil cháracteristic was nutrient supply followed by watar holding capaeity (Table 2). Other important soil characteristics citad were soH depth, intiltration rata, erodibility, compaction and graveVstenes_ Tablea 3 and 4 list cited Boíl eharacteristics for the different soil types. (OnJy the soíl types mentioned at least three times were taken into consideration). Apart from "Gimu· (fertile), 85% of the "Lidugavu· and 80% of the "Lukusikusi" soils, but none ofthe "Lunyu· and ·Zibugo· soils, were classified as soHa having a good nutrient supply. The latter two were most frequently considered te be soíl types with a low nutrient supply 5 Table 2. Positiva and negativa soíl characterilltíca cited on one or more soils by 26 farmera interviewed in Matugga Soil cbaracteristics %ofCarme ... Positive Nument supply Negative 69 Water holding capacity 62 46 'l:I 1.2 Erodibility Soil deptb 23 1.2 8 19 23 Infiltratian rate Compaction 11) 4 Gravellstones 11 Table 3. Positive soH eharacteristies for major Boíl types mentioned by the farmera interviewed in Matugga. Soil type Lidugavu Luyinjayiqja Number of times mentioned Positive Nutrient supply 18 1.2 8 7 Limyufu Lunyu Lukusikuai Lusenyuaenyu 3 3 WHC' Soil deptb IR' 23 o o 8 38 26 85 23 26 38 17 80 33 20 33 o o o o 33 o o 13 o 1) Zibngo son charaeteristics mentioned (Frequency (%) of mention) o 14 Erodibility o o o o o o o o o Table 4. Negative aoil characterilltica for major soil types mentioned by tbe farmera interviewed in Matugga. Soil type Lidugayu Luyinjayi'1ie Lunyufu Lunyu Lukosikusí Lusenyusenyu Zibugo • • Number oftime. mentioned 18 1.2 8 7 5 8 3 Negativo.OO cbaracteristi.. mentioned (Frequency (%) oC mentían) Nutrient supply WHC' Soil deptb IR' Erodibility O 25 11) 8 O 33 60 57 O 33 67 58 17 O 8 O O O 29 O O O O 38 43 60 67 O O O 13 14 O O O Gravel Stones O 42 O O 20 O O WHC • Water·boldin¡¡ eapaclty IR .. lnfíltration rate A majority offarmers indicated a low water holding capaeity for "Lunyu·, "LuyiI\Íayh\la", Lukusikusi" and "Lusenyusenyu". None ofthe soíls were frequentIy mentioned to have a high water holding eapacity. Soil depth was not a eriteria used by farmera for soil clsssification. It was mentioned as a positive eharactenstic for 23% of the Lidugavu soils and as a negative charaeteristic for 17% of the stony "LuyiI\ÍayiI\Ía' soila. For 42% of the "Luyinjayinja" Boila the oecurrence of gravel and stones wss mentioned as negative feature. The farmers easily Usted the preferred crops for the different soH types. Generally al) the crops were preferred on the "Lidugavu' soils whieh were classified 8S Boils having a good nutrient supply (Table 5). "Lunyu' and "Zibugo' soils had the highest percentage oC crops mentiimed to be unadapted, whereas none of the etOPS were cited to be unadapted on the sandy"Lusenyusenyu' soils (Table 6l. 6 Cooking banana was preferred on 39% of tite "Lidugavu' and on 40% of tite "Lukusikusi" soilll but it was never preferred on "Luuyu", "Zibugo" and "Lusenyusenyu', It was cited as unadapted on 71% oC tite "Lunyu· and on 67% of the "Zibugo" soils, TIlese findings underlíoe the faet that eooking banana, tite main staple foad, has bigh priority 00 too more fertile soils sueh as "Lidugavu· and "Lukusikusi", Cassava was s bigh1y preferred erop on a1l oftlte son types except "Lukusikusi", altltough the reason for tltis exception was oot determined, lt was cited as an unadapted erop on some of tite "Luyinjayinja", "Lunyu· and "Zibugo' soils, Cassava grows relatively welJ on acid and highly infertile Boils (Howeler, 1981), TIlerefore it does not have a bigh priority to be grown on tite most Certile Boilll. Maize and sweet potato were preferred crops on all the soil types except "Lunyu· and "Zibugo". AlI three Carmers having "Zibngo· considered 8weet potato as an unsdapted crop for this soíl type. Beans was a preferred erop on 64% of tite "Lidugavu· soíl type but was never cited as a preferred crop on "Lukusikusi" and "Lusenyusenyu·, Groundnut was cited as a preferred crop only on 17 % oC tite "Luyinjayinja" Boils but as an unadapted crop on all oE too "Zibugo" soils. Soil color was determined using tite HUE 5YR Munsell color chart for 95% of tite soils. TIle predominant soil color was dark reddish brown accounting for 70% of the soilll surveyed (Table 1), OtIlar soil colora found were very dark grey (13%), reddish brown (10%), black (3%), yellowish red and dusky red (2% eaeh). TIle soil color generally corres- ponded witlt the Boíl color for wbieh Carmers named tite soHs, TIle top.. soíl textura of the surveyed fields was sandy elay loam, sandy loam and elay loam in 64%, 33% and 13% oftlte cases, respectively. AlI "Lukusi· kusi", 54% oftlte "Lidugavu·, snd 50% ofthe "Limyufu· soils were sandy clay ¡oams whereas 59% of the "Luyinjayinja" were sandy loams. AH sons were we1l·drained except tite clayey "Bomba" soillocated in the valley wbieh had a moderate to poor drainage. Table5. P.referred CfOps on mBjor son typea as mentioned by farmera interviewed in Matugg&. Soiltype Lidugavu Luyil\iayirij. Limyufu Lunyu Lukuoikus. Lu...nyuaenyu Zibugo No,oC times mentioned Cooking banana 13 39 39 64 O 17 64 50 46 12 42 25 O 40 O O 63 13 57 O 33 33 O O 20 20 33 O 33 38 29 O O 33 17 B 7 5 3 3 25 25 P",C.tTed crop (% oC time. mentioned) Sweet Canava Maiu Bea... potato 67 O Groundnut O O O O O Tabl.6. Uoadapted erops 00 major soíl types as mentiooed by farmera iotervíewoo in Matugga. Soiltype Lidugavu LuyirijayiJUa Lirnyufu Lunyu Lukuaikuai Lu••nywoenyu Zibugo N ..mber oftimes menboned Cooking banana 13 15 17 12 8 7 5 3 3 Unadaptcd eropo ('lb oC tim.. mentionedJ Sweet Maiu Beans Cassava Groundnut potato O 17 25 71 O O O 14 67 67 O O O 8 23 O 28 20 O 33 O 43 17 O O 100 O O O 14 20 O 33 8 O 13 O 20 O 100 7 The top80iI depth ranged from 8 cm for "Luyinjayinja" to 55 cm for "Lidugavu". On the average, topsoil depth was 24 to 32 cm. The range of the subsoil depth was from 25 cm for "Luyinjayinja" to over 90 cm for the majority ofthe soils. However, as there was a wide range in both top and sub soil depth for each soil type identified, soil depth did not explain farmers' classification of soils. In respect ofthe topographic position 83%, ofthe fields were located on a hillside, 10% on the hilltop and 7% in the valley. The slope ofthe fields on the hillsides ranged from 7% to 19%. The pH ofthe top soil was generally below 6.0. Only "Lidugavu", "Luyinjayinja" and "Limyufu" occasionally reached near neutral pH values. The most acid soil was the "Bumba" valley soil having pH as low as 4.0 to 4.8. The soil analysis ofthe "Lunyu" and "Zibugo" soils showed low levels for P, K and Ca (Table 7). Compared with the "Lidugavu" soil, the differences were significant. These results strongly confirm the farmers' perception of"Lunyu" and "Zibugo" being infertile soils with low crop yields. Table 7. Mean values for pH, Boil organic matter, available P, K and Ca for "Lunyu", ·Zibugo" and "Lidugavu" soils at Matugga. SoiI type pH OM P (%) (ppm) K (mel100g) Ca (mel100g) Lunyu Zibugo Lidugavu 5.0 5.2 6.2 2.5 3.7 3.8 7 14 50 0.20 0.41 1.61 2.50 3.84 6.99 Mean 5.5 0.89 3.3 24 n.B. 17.9 0.74 0.56 4.44 2.45 Uln (P. ~ potato, ooions, maize and tomatoea. Bananas are poorly adapted. Soll typee: KiwugankoCu, Luyinjayinja Lunyu. The latter ia common on eroded lites. Eetimated mean land cost la USha. 50,000 per acre. Deeper soil tban on hilltop which la more difticult to till but with a batter soil moiature. More C.rtile aoila on fue lower elopes. Preferred trops are banana, C8ssava, baans, maize, eweet potalo and tomate. Manure i8 sometimea applied. Homesteada are concentrated on th. lower elopes. Soil typea: Limyofu, Luyinjayinja, Lidugavu, Lukusikusi. Estimated mean land ecat is USha. 100,000 per acre. The valley 8011. generally have a dark top soil, a c\ay sub soil snd are underlain witb sand, Difllcult te till when wet. Good fertilitydue lo organic matter. The heet grazlng lend. Preferred crops are cabanga, 8weet potalo, eggplant, dodo, yams and supreme. Too wet and too sandy for banana, More important than previoUllly to get dry sesaon production. Soíl types: Bomba (Tosi), Lusenyusenyu. Estimated mean lsud coat;e USba. 200,000 per acre. .... '" APPENDIX 3: Causes of low soil fertility (LSF) as perceived by farmers Lack oC knowledge about aoil conservation metbods Low soiI. organi<: matter erosion Little use oC farmyard manure Lack of crop rotatíon LOW SOIL FERTILITY (LSF) Lack oC planting leguminoua planta Lack of fertilízer use Low watar holding capacity CLeaching) Gaseousloues ck statistics. Rome, ltaly. H~rrison, P. 1987. Ths greening of &frica: Breaking through in the battle for land and food. Penguin Boob. Ireri, L.W. 1990. Adoption atady on the use of organic farming practices in Embu and Nyeri Distriets, Kanya. Emba Regional Researcb Centre Report. Embu, Kanya. Jaetzold, R. and H. Schmidt. 1983, Farm managament handhook ofKenya. Natural conditiOllB and farm managament information. voJ.n, Part C. Eaat Kanya. Ministry of Agricultura, Kanya. . Kamau, CK and M.K. O'NaiU. 1990. Sorghum lmprovement in Kenya. Seventh EARSAM Regional Worksbop. 24 June - 2 July 1990. Nairobi, Kanya. Kanampiu. F.K., J.K. Ransom and M.K. 0'Neill. 1991. Plant nutrient flux between entarprise¡l in .. maize based farming &y8tam ín central Kanya. Second Arrican Boíl Science Society Conference. Cairo. Egypt. 4-12 Nov. 1991. Kanampiu. F.K. and J.W. Irungu. 1992. An informalsurvey on organie farming praetices in Nyeri and Embu DilItricta ofKenya: A me81l8 ro improve IIOiI fertility. Embu Regional Reeeareh Centre Report. Embu. Kanya. Karanja, G.M. andJ.O. Odour.198S. A report On Carming systema ofGaturi Bouth Loclltion. Runyenjea Division, Emba District. Embu Re¡!ional Rllllearcb Centre Report. Emba, Kanya. Lal, R., D.J. Eekert, T.J. Legan, ..lid D. Myrea. 1988. Environmentally 8ustainable farming syatems. In: P. W. Unger, W.R. JonIan, and T.V. Sneed (ed) ChaUengllll ín dryland farming - A global perspeetíve. Intematlonal Conference un DryJand Farming. 15-19 Aug. 1988. Busbland. Texa.a. p.661-667. Long, s.P. and N. Pera.ud. 1988. Influenee ofNeem Cktuüt<:ht4 indiaJ) windbreaka on millet yield, microc:limate, and watar use in Niger. Weat Arrica. In: P.W. Unger, W.R. Jordan, and T.V. Sneed (ed) Challengea in dryland farming - A glob,,¡ perspeetíve. Intemational ConCerenee on Dryland Farming. 15·19 An¡. 1988. Bush1and, Texas. p. 313-314. M'Ragwa, L.R. and R.M. Kanyel\ii. 1987. Stratagies for the improvement oC 80rghum and millet in semi·arid Kanya. In: Food Grain p,..lduction in Semi-arid Arríe... (eda) In: T. Beluneh and A. Youdeorveí. OAU! STRC-SAFGRAD Intem ..tional Drcught Sympoaium. 19 • 23 May 1988. Nairohi. Kenya. 45 MaeKay, K.T. 1989. Sustalnable farming systems issues fOl farming systems res.areh. In: S. Sukman., P. Amir, and D.M. Mulyadi (ed) Developments in proeedures for farming systems researeh: Proceedings of an Interoational Workshop. 13-17 Msreh 1989. Puncak, Bangor, ludonesia. p. lQ5.11S. MeWilliam, J.R. 1986. Btrivmg Cor sustsinability in O> ,,' ....- . ~.'\ ••* \ ...... N···. AA""'··,. I II Z0IA .. ". 1ft': J: ..~ " \...,.t..... -I ....:~··... .. ). ;o•• 111'. :- ../ >'. • ~~,f -·.(USHU ""\ ". " ~ N ": (' h..,N .N·.N. P. 'IATA /I O.... PIl'"" ( . ' . . . / '-.. 1-1 1 "-• . ..T\JI \ O I ~:> 1 t"" "-~. -.....".. "".. .I\r.fA .\ o ' .", \. MACHA_OS í . .. ",,". \ ./'~::, "\ NP •• ".: \ tI.l \ I \.. • <. • ". NARO_ A Figure 1. \ "o " .. / ~ .... > ..t.. :~ KIAMaiJ'··;::. t....,I .•• \ .t ;"" .. 1 '. ", \.-1--_ ........ ........ ."• 'l-- . /' \ H" .•K"" I •••• -'p .~""\' :,. :', .. ~ ..•• '.•••.,: . , ~o -, H"tEIU • , .\ .. ". ';, , N .. .J." " XlL1FJ .. nAJH\ \: ~~ P " N • .. NAXU'./U, \....... ,.11 .......... .,........ ~• • ... * • • '. ... ... o'" .KIJlUCHq •• ' : •••• '71':" IMlnl 'N ~",. ........ ~..;.I' N • : P ~ ~ J.:MVAANO'A ),'''',0'0 • tUSII ,NP ! • t' \ ' .... NP." \'.. ¡ .' •• ."..... " p.. ., PN. • •• -.....,) ••••••••••••• H. .... j' 1 \., . .l;.f * qL.~""'\\.'_'-::~"\. _._" ! ¡..r......,<. ':..... ¿ (' ! N . ) : ·1 0,,' ...... : .... ........ " NI . . Iy:l',. ....,....." •. u...... ." ..... 'UASllf"" 1'" 1 : .' '\'> ", .." I--'''NGO "->f._-.. ...... ,; . ·'lJNNO~,iD N ,r'~¡ .~ 'i:. t. IAMIU1tV" ," /'11'., 0 0. 1.·-... \.,........, I tOa .' .: . . . . . . '!n1MOMIASA NP ItW.l.U: ,....•.,.... /' N •••••• t ••• \ , _._, .M..=~;:·e~YW=·~~~______________________________~Be~M~~~·~ ro- 10 50 "." 8 40 .................... . ........ 7 3OL----------L----------~--------~L----------J6 ~ O 40 50 00 . +-Malze. sale crop +Maiza.lntarcrop ..... BeM. intercrop Response curves to applied P at Otambo. Figure 2. 70 Maize yield Baan yiald 60 ................ . 50 ............ . 40 ....................... . 3OL-----~- O ~ 9 B ·············7 _ _ _- - l_ _ _ _ __L__ _ _ _.....J6 40 60 00 -Maize. sola crop Figure 3. . . . .. 10 + Maize. intercrop "*-Bean. intercrop Response curves to applied N at Otambo. 67 Maize yield Sean yield 00r-~------------------------------------~-,5 4 40 --.----- .. - ---.- 2 00~--------~----------~----------~--------~1 00040 ... Maize, sole crop Figure 4. 00 + Malze, Intercrop 00 00 .... Seen, intercrop RespollBe curves te applied P at Mau Summit. Maize yieId Seen yield 5 50 _ ... - .. - - - . - - - - - - - - ------ 'T 4O+----------------+--~---.7.~.~.~.~.~.-.~.~-~~~~~~~-~-+:. - 3 30 - •• - •. - . . . . • . . - - - - - - - - - - - . - 2 00 '--_ _ _ _ _-1,_ _ _ _ _ _ _-'-,_ _ _ _ _ _"--,_ _ _ _ _...) 1 00040 00 00 - Maize, &ole crop Figure 5. 68 + Malze, intercrop .... Seen, Intercrop Response curves te applied N at Mau Summit. Mab:e yleld Sean yleld ~r-~------------------------------------~-,5 4()_ . . • . • . . . • . • . . . . . . . . . . • • • . . . . . . • . . . . . . . . . . _.¡ .. - 2 10 L-_ _ _ _ _..L, _ _ _ _ _ _ _,L-_ _ _ _ _..L, _ _ _ _ _-...JI 1 O W ... MaIz&, sole crop Figure 6. ~ ro 4() + Maize. intarcrop 00 "* Sean, intercrop Response curves to P at Kaguru (Qlhal. Maize yíeId Bean yield 5 40 .. 4 30 . . . . . . . . .~ . . . ~~.~ .... ~ 3 20 ................. 2 ....................... 10~----~------L-----..L-------~1 O ~ 40 00 ro -Malze, sole crop Figure 7. + MeI%s, Intercrop "* Bean, intercrop Response to applíed N at Kaguru (QlhaJ. 69 Bean yiekl Maize yieId 7Qr---~------------------------------------~-.7 60 _. .. 50 . ....... . ~ . .... - 6 ............... . ... ·········~·-5 .... ~"'*----_ ~ ¡- . . . • . • . . . . . . . . .. - ..¡ 30 '--________--1__________- ' -__________-,' - - - - - - - - - - - ' 3 O ~ ~ 60 00 - Malze, sole crop Figure 8. 7Q + Maize, Intercrop .... Bsan, Intercrop Response curves to applied P at Vihiga-Maragoli. Bean yield Maize yield 7 6 50 ...... . 5 ~ ................. . <4 30'--------'------.1..-------'--------'3 O ~ - Maize. 80le crop Figure 9. 70 ~ 60 + Malze, Intercrop .... Bean. Intercrop Response curves to applied N at Vihiga·Maragoli. 00 9OrM_aa_e~y~~ ____________________________________~ __~~_',7 80 .......... ...... ... . 70 ... 60 .............. . 6 ............ ... . . . . . . . . . . . . . . . . . .. . . . . .. 4 60'---------"--------'-------"-------13 O ~ ~ 60 00 ... Maae, sola crop +Maae, Intercrop ..... ~, Inlercrop Figure 10. Response curves to applied P at OI Joro Orok. MaIze y~ Sean yleld 90,----''----------------------------------''--17 00 ....... . 70 . . . . . . . . . . .. 00 6 . . . . . . . .. • 5 • ............................. ........... 4 50 ,'--_ _ _ _ _--1._ _ _ _ _ _-"-_________..1..-_ _ _ _--1 3 O ~ ~ ... Malze, sola crop + Malze, Intercrop 60 00 "* Sean, Intercrop Figure 11. Response curves to applied N at 01 Joro Orok. 71 22 Gross retums (Thousands) 18 -- - --- --- - - -- - 16 - - - - - - _ .. - - . . . . . . - . - - - .. 14 _. - . - .. - - - ------- ... - - . ~--:- 12L-------------~------------~------------~ O 25 50 75 N rata ... Maiz&. sola + Inlercrop Figure 12_ Gross returns (KSh) ro maize and maíze--bean intercropping in response ro applied N in Otamba. 20 -- 16 - - - - - - - - - - 14 - - - - - - - - - - - - .. - - - - - - - - . 12~------------~------------~------------~ O 25 50 P rate ... Maize, sola + Intercrop Figure 18. Grou returns (KSh) ro mair.e snd maize-bean int.erc:rop in response ro applied P in Otamba_ 72 75 , • f ______________________________--, 2O;G~roa~~==ms~~Ob~m=='u=n=~~) 15 - - - - - 10 . - 5 . -- - -- - . - . - - - - - - . - - - - - - -- - - - - - - - - - - - - - - - - . ......... - ...... - .-- . . . . . - .. - . . . ---------- OL-------------~------------~------------~ O 75 25 50 P rate F:igure 14. a.... returrul (KSh) te maize and maize-bean intercrop in reaponse te applied P at Mau Summit. 18 - - - - - - - - - - - - - ... - - - .... - - . 16 . - . - - - - -- - . - - - -- . - - - . - - - . -- 12 - - - - - . - . - 10L---------------L---------------~--------------~ O 25 50 75 N rate ... Maize. $Ole F:igure 15_ + Inlllrcrop a.... retunul (KSh) te maíze and maiz....bean intercropping in Teeponse f.o applied N at Mau Summit. 78 Gross retums (Thousands) 15r-------~----~--------------------------_. 13 -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 11 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 9 -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - - - -- -- 7~-----5~------------~1------------'~----------~ ~ O n 00 P rate - Maize. sale + Intercrop Figure 16_ Ol'O88 returna (KSh) f.o maize and maize.bean intercrop in responae f.o P at Kaguru. _____ ____ 15~Gross ~ ~~ ~) 13 __________________________~ - --- --- -- -- 5~-- O __________L -____________L -_ _ _ _ _ _ _ _ _ _ ~ 00 N rate .. Maize. sole + Interorop ~ Figure 17_ Oroaa retums fKSh) f.o maize and maize.bean interc:rop in responae f.o N at Kaguru. 74 n 16 _ .. - . .. - ..... . 14 _ . . . . . . . . 12 _ . . . 10L-------------~'---------------L-'------------~ O 25 50 75 P rate - Maize. soIe Figure 18. + Intercrop Grosa returns (KSh) te maize and maize-bean intercrop in response te> P at Vihiga-MaragoU. 16 14 12 - - . - - - . - . . . - .. - . 10L-------------~L-------------~--------------~ O 25 50 75 N rate -Maize. soIe +Inlercrop F"¡gure 19. Gross returns (KSh) to maiu and mane-haa" intercropping in response te N at Vibiga-MaragoU 75 Gross retums (1ñOl.lsands) ~,-----~----~- -------------------------~ 23 ------ 21 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 19 r-- - - - - - - - 17 '- - - - - - - - . - - - 15~------------~--------------~------------~ 50 O 75 P rata ... Maiza, sois + Intercrop Figure 20_ Grou returns (KSh) ro mahe and mahe-bean inl.erl:rop in response ro P at OI Joro Orok. ~r~-oss--rewnm--~(1ñousands----~)--------------------------__, 23 - - - "' 21 -------- 19 .. --- - - - - - - - - - .. - .. -- -. - . - - - - - - - - - - - - 17 - - - - - - . - - - - - - - - - 15~--------------~1--------------~1--------------~ O 50 ~ N rate ... Maiza. sola + Intercrop Figure 21. Gross returns !KSh) ro maize and maíze-bean intercrop in response 10 N at OJ Joro Orok. 76 ~ REFERENCES Agboola, A. A. and A. A. Fayemi, 1971. Preliminary triala en tbe intercropping oC maize witb difl'erent tropical legumes in Westem Nigeria. Jouma! oC Agricultura! Scienc:e 77: 219-225. Chui, J.N. and H. M. Nadar, 1984. Effeet of apatíal arrangemente un tbe yield and other agronomic cha:racters oC maize and legume intercropa. In: Dryland Farming Researcb in ReDya. Eaat AJlica Agriculture and Furestry Juumal Special leaue 44: 137·147. FAO., 1988. FAO-UNESCO Soíl Map oftbe World. Reviaed Legend. World Soíl Heeources, Report SO. FAO, Rome, lta1y. Francis, C.A., 1982. MuItiple cropping potential ofheane and maize. Horticulture Science 13: 12-17. FURP, 1987. Fertilizer Use Recornmendation Project. Final Report Phaee 1. Annex 1. Fertilize .... trial. documentation (FERDOC). Ministry of Agriculture, Nairobi. FURP, 1992. Fertilizer Use Recommendatiun Project. Quarterly progreea report 1 Jan. tu 31 Mar., 1992. National Agricultural Research Labs, Nairobi. FURP, 1992. Fertilizer Use Recornmendation Project. Final Report Phase n. Renya Agricultura! Research Institute, Nairobi. Gardiner, T.R. and lo. E. Craker, 1979. Development and light interception in a besn-maíz. intercropping system. Agronomy Abstracta 102. Jet;oId, R. and H. Kutech, 1982. Agro.ecological zonea oC tbe tropics witb a sample from Kenya. , Tropenlandwirtachaft 83: 15-34. Janssen, B. H., F. C. T. Guiking, D. Van de Eijk, E. M. A. Smaling, and H. Van Reuler, 1986. A ne.... approach to evaluate tbe <:heroica! fertility of tropical soils. TransaetiOIl3. XIIi Congress ISSS, Hamburg, 3: 791-792, Mocboge, B.O. 1990. Tbe behaviour oC nitrogen minereIization and ite avalIability for plant use in an aeid BOÍl. Paper presentad at the AGM oCtbe SoiI Science Society oC Eaat Africa, 3-7 December, 1990, Amaha, Tanzania. Mocboge, B.O. and S. M. Mwonga, 1991. Differences between soi1 properties ofland unde. continuous cultivation and fallo.... in three soil types in Renya. Diacovery and lnnovation Journal 3(3): 77-83, Nadar, R.M., 1984. Effeet of relay planting on maíze as inlluenced by cropping .yatems, row spacinga and I'''pulatlons. Dryland Farming Research in Kenya. East AJlica Agricultur. and Forestry Joumal SpeciaJ lasue 44:122-126. Nadar, R.M., J. N. Chui, E. S. Waweru, N. Bendera, and W. A. Faugbt, 1983. Agronomy research for marginal ralnfall """as. In Annual Report KARl, 1977-1980, Nairobi, Renya. Renya Agricultural Research Institute. Nadar, R.M. ano W. A. Faught, 1984. Malze yield response tu different levela oC nitrogen and phosphorus ferti\izer application: A seven-fiason study. Dryland Farming Research in Renya. Eaat Afriea AgricuIture and Forestry Joumal Special Issoe 44: 147·156, Okigbo, B.N., 1978. Crupping eystems snd relatad research in Africa. Special ¡ss"e on the occasion of 10th anniversary oi tbe MASA. Oceasiona! Publication Serie. 07·1. Qurashi, J.N. 1987. The cumuIative effecte ofN-P fertilizera, manure and crup residueIl on maíz. grain yields,lesf nutrlent contente and sorne chemical propertiea at Kabste. Paper presented at the Natienal Maize Agronomy Workshop organized by CIMMYT. Silver Springa Hotel, Nairubl, 17-18 February, 1987. Ris, J., K. W. Smilde, and G. Wijnen, 1981. Nitrogen fertilízer racommendatlollll for arable cropa as hased un aoiI analysia. Fertilizar Research 2: 21-32. 77 Shulcla, a.c., 1972. Efiect of difierent levels of nitrogen and pboapborus on yield. soil propertiea tmd JWl::rienta on coro. Agronomy Journal64: 136-141. &malin,. E.M.A., 1989. Towards area-specilie fertilizer recommendations Cor rainfed annual <:rOps in 'Keap. , . Researeb propoaal. Stiboka/Winand Staring Centre, Wageningen, Tbe Netberlands. Smaling, E.M.A. and R. F. Van dar Weg, 1. Using soiI and climate mape and associated date aeta,1o selec:t'iíítea far fertilizar trials in fulnya, AgriculturaJ Economícs and Environment 31: 263-274. Stocking, M.. 1986. Tropícal red soiIa: fertiUty manapment and degredation. In: Tbe Red Soils of Eaat Mrica and Soutbem Añica. 24-55. Proceedinga of an International Symposium, Harsre, Zimbabwe 24-2'1, Febtuary, 1986. Stocking, M., and L. Peake, 1986, Crop yield losa88 from tbe erosion oC A1fisols. Tropical Agriculture 63: 41-45. WiIley, R. W., 1979. Intercropping: ita importsnce and researcb needs. Competition and yield advanteges. Field CropAbstrects 32: 1-10. 78 RESEARCH FOR SUSTAlNABLE AGRICULTURAL PRODUCTION SYSTEMS IN mGH ALTITUDE ZONES OF EASTERN AFRICA Charles S. Wortmann Regional Programme on Beans in Eastern Africa 1292:' 22. SET, 1993 P.O. Box6247, Kampala, Uganda ABSTRACT Demanda for food and tiber are increasing at a faster rata than supply in the medium and high altitude zones oC Eaatern Arrica, while the productivity ofthe land ia apparently decJining. In many cases, productivity ia constrained primarily by insufficient nutrient supply. To meet future demanda. for food, the !and's productivity wiIl need to be maintained, or rejuvenated, and managed ror greater productivity. Sustainable agricultural aystems which allow fOl increased production need to be developed and implemented. This paper explores a number orissues relevant to research Cor austainable agriculture in Eastern Arrica and propos.es an alternative research approach. The need and place Cor both high-input and low-input production systems ia explorad. Roles oC alternative research methodologies, including commodity and disciplinary, farmer participatory, agro-ecological and eco-regional methodologies are diacussed. The challenge of improving agricultural production in a sustainable manner fOl a wide range of micro-environments ia address.ed. Promotion of technologies, eapecially wben packages of practiees or novel systems are needed, ls discussed. , A research approach ls proposed which ia basad on intensive Carmar psrticipatery research in carefully selected farming communities. The communities serve as benchmark res.earch "sites' of largar agro-ecological zones, with consideration of socio-economic factora. Various research metbodologies are applied at the various atages of the research process. INTRODUCTION Demand for fuod iB increasing at a fastar rate than supply in the medium and high altitude zones of eastern Arrica, while the productivity of the !and la apparently declining. Anthropogenic pressures are increasing due to a population growth rate ofS.1% for Sub-Saharan Arries (World Bank, 1989) and due to increas!ng demands for an improved diatribution of income. A¡riculturaJ production has decreased in sorne parta ofthe Region (Jain, 1988). Tbroughout most ofthe eastern Arries, the farming systemlinput levela might be classed as low or moderate traditional such that arable land requirements for subsiatence exceed 0.5 hectare per espita (Buringh, 1989). In many cases, productivity ia constrained primarily by insufficient nutrlent supply. At the same time, estimated ratas ofnet nutrient depletion are high, exceeding 40 kgIyr ofN and K, and 15lq¡1yr ofP, par hectare of arable land in Kenya and Ethiopia (Stoorvogel and Smaling, 1990). Lower, but stiU higb ratas, are est!mated for Tanzania and Uganda. Rate of fertilizar use is low throughout this ares and probably near the average of five kilograms of fertilizer per hectare of arable land used in Sub-Saharan Arries in 1988 (Plucknett, 1991). In places fertilizer use is deelining because of insufficient profitabiJity, inconsistent crop response, amllor failure te sustain high crop yielda with continued fertilizer use (Brossier, 1991). Much ofthe negative nutrient balance is due te loss of nutrients because of Boil erosion and leaching with detrimental effects on the regional environment. To meet future demands fOl food, the land's productivity will need te be maintained, or rejuvenated, and managed for greater productivity. Sustainable agricultural systems which allow for increased production need te be developed and implemented 79 SUSTAlNABLE AGRICULTURE Definitions of sustainability are numerous but the following elements are commonly suggested (Okigbo, 1989): - adequate economic retums to Carmers; maintenance of natural resourees and productivii3' indefinitely; minimal adversa environmental impaets and preferably enhancement of the environment; optimal production with efficient use, and maybe minima! use of non-renewable interna! and externa! resources; satísfaction of human needs for food, fiber and ineame; plovision for the socia! and psychoJogical needa offann familias ami communities; and eeonomical viabilii3' of the farro enterprise, in that it eams a fair return on farm investments. Sustainability is needed at the levels offield, farm, region (ecology), and macro-economy (Lowrance ct al., 1986). i} lt is needed at the field level for continued productivity ofthe land. Factors tbreatening sustainability at this level ineJude soil erosion, degradation of soil structure, soil organic matter loss, nutrient loases, depletion or contamination of water supplies, sa!inization, pest build-ups, and limitad bio-diversity. ii) Farro-Ievel sustainability is needed for economic viability ofthe production system. lt may be threatened by unsusteinable field-level systems, incompatibility of enterpríses, low profitability, high risk, insufficient supply ofrequired resources, infrastructural inadequacíes, and unhealthy living and working conditions. üi) Sustainahility at the eealogical or regionallevel is needed for the health of the 80ciety and oommunity, the environment and the resource base. PoJlution, resource depletion, inadequate supply offood snd fiber may threaten sustainability at this leve!. Sustainable egriculture sbouJd contribute tu the economic well-being of the region by creating or reducíng employment opportunities as nceded and by stimulating economic activii3'. iv) Macro-economic sustainabilíty is eantrolled by ractora such as fiscal policíes and interest rates which affect the viability oC national agricultural systems. Some production systems are more Hkely than others to be sensitive tu macro·economic fluctuations. SUSTAINABLE AGRICULTURE AND RESEARCH IN EASTERN AFRICA J..il order tu meet future food and fibar needs in castern Arrice, production will nead to inerease. Currently, both high potentia! (or favorable) lands' marginallands (due to unfavorable climate, poor soils, or maybe the socio-economie status of the farmers) and those areas of intermediate potential are fanned. The question arises as tu where tu invest available research resourcas. Some say tbat there are adequate technologies on·the-shelf for the bigh potential areas and relatively little investment in research is needed. Thersfore, they argue emphasis sbould be placad on the more problematic areas. Another side argues that the greatest retums to research wilI be achieved in the high potential areas, and that efficiency ofinput use and avoidance ofpollution nead the attention ofresearchers. This side argues tbat the marginallands are oRen too fragile to sufficiently intensify tbe agriculture in a sustainable manner, and tbe retums tu research will be relatively small (Plucknett, 1991). More production;8 needed from the high potential areas tu alleviate the pressure on the more fragile areas which may be irreparably damaged with improper intensification. The Green Revolution approacb has not been successful in eastem Afriea, but ample evidence indicates that a moderately high input approach is technicaJly feasible in favourable parta of the eastem Arrica highlanda. Global 2000 has auccessfully assisted Carmers in tbe Arusha-Kilimanjaro and the Southem Highland areas ofTanzania. to achieve very good responses ofmaize tu applied. inputs. Researchers and farmers in Kenya (e.g. the Kitale area) and in Uganda (e.g. the Mbale area) 80 have been auccessful with a moderately high-input approach. Inadequate infraatructure for transport, provision of inputs and marketa is a major hinderance te further intensification in many areas (Bose et al, 1991). Facilities for absorbing the increased risk associated with such systems are olten laeking, ineluding eredit and crop insurance facilities. Distances of production areas from the source of the inputs and te the markets are oRen great resulting in lésa profit Cor the producer. The demand for the produce, especially for production areas which are far from population canters, is oRen variable adding te the risk of production. Those areIl.S where intensification is not much hindered and the infraattucture is adequate might be high'priority areas for research (Jain, 1988). Research te increase yield may be needed, but olten much ofthe needed inforroation is already available. More research is needed on input-use efficiency, stability ofproduction and prevention ofpol1ution ofground and surfaee water. Work te maintain bio..ruversity in the system ia Iikely te be important in order te avoid creating an essentially mono-culture production system which will be especially susceptible te breakdown of genetic resistances to pesta, but also provide an environment where minar pests can develop te be mejor peste. Maintenance of soil organic matter for ita role in stabilizing production is a mejor concern. Most oftbe arable lande ofEastern Africa probably are not currently suitad for high-input agriculture, eitber because the land is marginal for production due to soil or elimate relatad factors, tbe soíl i8 fragile (due te easy erodibilíty, low fertility, sensitive te soil structure changea, etc.), or because tbe socio-economic conditions are inappropriate. In such areas, moderate and sustainable increases in output may be tbe objective, probably throngh combining use of low-input alternatives with regenerative agriculture. Okigbo (1991) suggests tbat the development of such systems may involve a combination oC elements of traditional systems and their component technologies that maximize on use oflocally availab1e biological inputs, with affordable external inputs. Increased efficiency in the use of locally available renewable resoutees may be an important goal. Improved soíl and water conservation, management oC organic materlals, enhanced nutrient cycling with reduced nutrient losses, biodiversity Cor improved resource-use efficiency and stebility, improved pest management, and improved compatibility oC crop and livestock production systems are likely features of such systems. Agricultural research has acbieved much in developing varieties and yield increasing and protection practices which are potentially useful te improved. 8ustainable systems. but it should also be pointed out that some achievements in increased productivityeame with increased threat to sustainability. Adoption oC new technologiea has in many cases led to fragile monoculture production systems witb little biodiversity, high pest levels, serious nutrient losses and pollution problems. Research must continue to playa major role in sustainable agriculture. However, the effective"ess of researeh must be improved. Conventional approaches to agricultural research Agricultural research in eastern Africa. and in most placea oC tbe globe, ia generally commodity or discipline oriented and aimed at variety development, testing for response te tbe use of inputs, and studies of narrowly defined components of systems. The effects of two or three components and their interactions are typically studied in factorial triala. Most research is on sole crop systems and seldom are the interactions ofmore than two _pacies evaluatad in intercropping trial_. The degree offarmer involvement Or consideration offarmer's circumstances varies considerablY but has probably increased recently through farming systems research and on-farro research. Still farmer participation in research is very minoro Conventional agricultural research approaehes have been successful in improving our knowledge about crop production systems, solving production problems, and developing technologies for increaaing productivity through the use of inputs. The approach has contributed primarily to the development ofhigh input systems. lt is well suitad to study oC the input-output aspects of sueh systems and will play a major role in development of high input sustainable Sy8tems (Lockertz. 1988). 81 lt íB useful for example in the study oC nutrient-use efficiency in narrowly defined beundaríes, but another approach may be needed te reach a high level of nutrient use management in which the interactions between the management ofthe crops and soils with soil macro- and micro-organisms are managed te improve the synchrony oC nutrient supply and demand. In low input systeme, we may envisíon the management oC considerable biodiversity, including numerous crops, weeds, soil fauna, etc.. Again, the interactions between the numerous companents may be important te maintaining sustainability while achieving increased production, but the study becomes complexo Agricultural research will probably need te be more inter-disciplínary, more eco-systemoriented, and te have more farmer participation te achíeve the development of improved, sustainable systems which are acceptable te farmera. Many opportunities of the future for sustainable agriculture are likeIy te be Cound at the interfaces oC disciplines (Francis et al., 1988, Jain, 1988) through inter-disciplinary research, for example, with weed scientísts and entemologists eollaborating in hígh quality research to control and insect pests or agronomists and soil mícro-bíologists collaborating te achieve improved symbiosis between cropa and soH micro-organísms. A merger oC agronomy and eeology, í.e. an agroeeological approach, may be necessary to adequately understand the dynamics oC systems and to define their crltical componenta (Altleri and Anderson, 1986). Once those critica) components are defined, McCalla (1991) suggests that the investigation of these crítical eomponents and their interactions will most likely be addressed through the application oftraditíonal approaches te breeding, agronomy, economics, etc.. Othera suggest that agro-ecology will have a major role in the study oC components and their interactions (Gliessman, 1987, Hendrix, 1987,). Greater farmer involvement in identification of researchabJe questions, selectíon of potentlal solutions, evaluation on Carro, and validation of results will give a more efficient research process. but will also lead to creative combinatlons of farmer wisdom and technical expertíse (Francis ct al., 1988). Farmer participatory research Farmers can contribute to most stages oC the researeh process. Their indigenous knowledge and understandingofthe technical and socio-eeonomic condítions oftheir situation can be usefulat all stages oC the research process. Over centuries, Carmera developed production systems which are often difficult te improve upon given their circumstances and resource availability, and an objectlve oC achievement oC ¡¡hort terro benefits. Many farmers are ready investigators oC alternative technologies and conduct simple trials on theír farms. Farmer participation can be valuable in better underatanding the Carmera' situation through: learning and analyzing the technical and socio-eeonomie aspects oC their living and agricultural production environment; diagnosis of production problems and opportunities; understsnding management practices 1Uld how these vary with soH types; and learning the history of changes which have occurred. They can eontribute to the prioritlzation of problems and identification oflikely solutions. Farmer evaluatíons of technologies are essential to successful adoption and the research process can be improved by early involvement of farmers in the evaluation process. Farmer participation in research on low-input alternatives is likely te be more important than with hígh-input alternatives. The low-input systems are generally more complex due to greater bíodiversity oflivestock. crop, weed and insect species, soíl mícrobiallife, etc.. Low-input alternatives are Jikely ro be more influenced by socio-economic factors. Efficiency of use of locaIly available and regenerative reaources ia ímportant, and farmer knowledge of the avaílability, present uses and restrictlons on use of such resources can be valuable in investigating alternative practices. Often a technology te be tested may be one which was "diseovered" on a Carmers' field but in need of further evaluation or modífication to apply it in another environment. As low-input alternatives are often líkely ro be location-specific, research requirements to adequately serve all of the varied agro-ecosystems will be very high, and it is probable that farmera wiIl continue to be major playera in the development of their production systems. 82 The Carmen' role in farming sy¡¡tems researeh and on-farm research has generally been of minor importaru:e, and often limitad to providing information during the diagnostíc phase and providing fields on which to conduct on-farm triala. ']'he work has ofien been done extensively to test. technologies on many farms, with some input mm farmers in trial management and assessment. A successful approach to development or improved low-input sustainable systems will require a more intensiw approach with research, extenaion an.d representative farmera working together in-long term oo11aboration at a small nwnber of sitas. ']'he expectation is that a1\ parties will g&in a deep understanding ofthe dynamics ofthe systams, while with experience the oollaboration will become more effective. Opportunities to involve other disciplines will develop. The participating farmera will be key to the promotion ofthe new systems by demonstrating these on their own farma to visitora from elsewhere. Not alI farmera ara potentially valuable collaberators in researeh. Researchera experienced in working with farmers bove undoubtadly eneountered disappointments due to farmar disinterest or their failure to mske a useful oontribution. Selection of partieipating farmera is importent as some are not suitable for co1laboration. either hecause oflaclt of interest, mental or physieal deficiencies, or limitad abilities in observation, anaIysis or articulation. Observation of a farmete household and fields may give clues ofhislher potential in participatory research: a good oollaborator may be one who triea different enterprises, but who manegea them well. A good collaborator should be of a typieal socio-economie stetus, observant and capable of artieu1ating observations and opinions. An important role far egricultura1 anthropologists is to further develop proeedures for farmer participation in research: to better utilize indigenous knowledge and farmera' skills; to better select collaborating Carmera; and to better understand farmers' research procesa and how to improve their capacities in improving their production systems. In any case, the skilla as collaborators in research may need time to develop tbrougb experience. Research for Diche fllrJllhig With smaIl-scale farming there generally is much variation within and amongst farme. Farming in response to the various soils, 01' other conditions, on a farm has been calIed niche farming, precision farming, aquare-Coot farming and prescriptive farming. Tbe farmer attempts to maximíze resource use and produetivity of the farm by managing each of the soila specifically. Niche farming ofien has not been feasible with large-acale farming because of insufficient flexibility in equipment use. In many developing cauntries, Carmera are advised to follow blanket recommendations for their management practicas. However, sma1l-sca1e,low-input farmers do already consider the variation in their land and farm it accordingly. Large-sca1e niche farming i8 expected to be commercially feasible within a few years. The equipment is becoming available which enable tite use of computer-guided equipment to operate at variable ratas, including plantera, fertilizar and pesticide applicators, and tillege eqmpment which is variable far residue incorporation. Such equipment is guided by either ground (beacons) or satellite (global positioning Bystems) bosed systems and computer software with detailed mapa ofthe field showing variation in soila, weed infestation. etc.. Research on niche farming is nesded. On both Jarge- and amaJl-seale high input farms, farmers need to optimiza use ofresources on their various soils. Site specific information is needed. Simplified land classificetion systems, 8uch as tbe Fertility Capability Classification system (Bou! et al, 1975), need to be confirmed for Eastern Africa soils and applied. Research for niche farming may imply the need to develop numerous alternatives from which farmers can select (Sperling and Steiner, 1991). Deve10pment ofnovel systems A common approach to agricultural research is to make step-by-step improvements on farmers' production systems. Novel systems are Jess easilyadopted tbon are single practice modifications. 83 Howaver, researdl on novel systems may be justified for both high- and low-input sUBtainable agriculture. In the USA, strip intercropping with crop rotation ia showing promisa as an alternative to simple rotationa of 801e-crops (Francis et al., 1986). Typically, the advantage of mixad 01' row intercropping lesgena as lavala ofproduetivity increase. Strip intercropping may be an alternative to small-scale farmers who are using moderateJy high lavels of inputs a8 it mayease the input application, increase the efticiency ofinput use and profit, and possibly increase productivity. SimiJarly, novel rotations, relay intercropping, agroforesUy 01' inc1usion oí new crops may present opportunities for novel systems. In low input systems, typicalJy numerous crops are produced with a valÍety of management practíces. In sorne cases a novel system roay be a modificstion of a production system alreedy in use by farmera eJsewhere, but undar slightly different conditions. A novel approach may involve the management of certain weed species to accomplish more than weed suppression, but ro provide a suitable habitat for beneficial insects or to enhance nutrient cycling. In Tanzania and Zaire, farmera have been observad te m 8D age certain weeds as green manure cropa to compliment the current crop. In Uganda on the westem slopas ofMt. EIgon, grassy weeds appear to play an important roJe in stabilizing tbe soil, protectíng it from erosion and improving permeability. Low input novel systems may be needed te increaBe the favorable activity of soil microbial activity and improve the soil microbe-crop symbiosis. Utility of on-the-shelftechnologies Much progress as been achievad by research in developing teehnologies, but in many cases the ' rate of adoption is low, i.e. the technoJogies are still on the shelí. Most. ¡¡f these teehnologies were developed for moderstely high-input systems, and many are probably appropriate where the use of higher levela ofinpute ie economically feasible. With small-scale, low-input systems, however. such on·the-shelf technologies frequently are not /iUperior to technologies whieh have evolvad in trsditional systems. or are inappropriate for other reasons and generally have had little impact (Altieri and Anderson, 1986). Complexity of systems and information needs It is frequently said that improvad sustainable. low-input agricultural systems need ro be more intensive than conventional high-input systems (Lynam and Herdt. 1988 and Loclteretz. 1988). The sustainable low-input systems are viewed as complex ecosystems with many interspecific interactíons and their interactions with the environment. Information needs for research are undoubtedly great, however, the information needs to mansge these systems may not be so great once they are deveJoped and establishad. They are expected ro be more atable snd better buffered than simpler, high-input systems snd therefore should require lees attention and fewer therapeutic actíons ofthe €armer. informl>~ion Research for high-input aystems, on tbe other hand, may require less information but fue manager must be relatively better informad overall, though the information requirements will differ from tbat oC tbe low-input manager. The high-input manager needa to be well informed of the valÍOU! products available and to use them in a sustaineble matter. In high·input systems of limited diversity. the manager is probsbly working with a relatively fragile system and must be preparad for frequent instances of opposing nature with therapeutic treatment. Computer-run modela are becoming increasingly important in the spplication of information on crops and soils to understanding the dynamics oC systems. for pre-testing technologies. and for extrapolation ofresults. The maize model ofthe DSSAT (IBSNAT, 1989) crop growth simulation 84 models has been used succell8fully in Kenya (Wafula and Cornor, 1992), and BEANGRO ís being evaluated for eastern Mriea. WEPP (NSERL, 1989) and RUSLE (Renard et al, 1991) appear te be useful in eva!uatíng soil erosion problems and potential solutions. EPlC is useful in tIle evaluation of agricultura! Bustainability and has been usad successfully to measure crop fertilizer requirements, nutrient transport in runoff, soiI and fertilizer phosphorus dynamics and tIle effect oflow-input legume-based crop rotations (Jones et al, 1991). ADSS (TropSoils, 1991) and otller expert systems are UBeful decision making tools for farmers and technical advisors. While tIle utility oC many of these modela for eastern Mrica is currently limited due to inadequacy oC infonnation for many of tbe production environments, researchers as well as managers oflarger farms in eastorn Mriea and extension staff should be considering tIleir use. Technology promotion Ease oC demonstration and adoption are two important criteria for t.echnology evaluation. Promotion of tIle use of chemical inputs, new varieties, or alternative planting methods is relatively easy, especially when there is a significant yield inerease. The use oC method and result demonstrations conducted through tIle extension serviee can be affactive. The promotion of SOIne alternatives aimed at improving Ilustainability, especially with low input systema, is likely to be more diffieult. The impaet may not be obvious in tIle short term, and severa! aspeets oC the system may be affeeted witbout a majar impaet on any one aspect. Result demonstrations may not be eft"ective as the resulta are not very obvious and they may have to run for severa! to many ·seasons to demon&trate their ful! benefits. Alternative approaches to result demonstrations may be needed. Farmers are Iikely to need considerable faith in the researchers and extensionists, and in the basic principIes of the t.echnology, in order to try 8uch a technology and te continue using it. More effactive than result demonstrations may be tbe use of show-case villages where farmers who participated in the research apply the resulta on tbeir own farma and eagerly discuss it witb visiting farmera. Eco-regional research Research for agricultura! sustainabllity requires an improved understanding and information on farming systems, agroclimatie constraints, and the dynamics of each (Lynam and Blackie, 1991). However tIle information base on Mrican agricu1ture ia very weak witb a lack of detailed information on soils, climate, and socio-economic charaeters of farming systems. A strategy for tIle development of improved and sustainable agricultural production systems requires planning based on stratification of environments at macro-levels (primarily rainfall and temperature) and micro-Ievels (soils differences, farmar and consumar preferences, socio-economic variables), snd then developing technologíes and designing alternativo IDsDagement systems within each strstum (Okigbo, 1991). Such a strategy will rely on well-struetured, geo.referenced, relational databases with tbe eapacity to anaIyze the date using geographic information systams (GIS) software, statistical packages, and increasingly sophisticated erop, disease, and biological and edephie proceSlles modela (Lynam and Blackie, 1991, Plucknett, 1991). The eco-regional approach should aid in focusing on critical problema sud allocating resources more efliciently but it will not reduce the need for high quality efforts of more traditional agricultura! scientists BpollBe farming for maize production in ehe dry tropics. p 251n Abatraets ofehe First International Crop $cience Con~ees, 14·22 July, 1992, Amee, Iowa, USA. WorId Banl<. 1989. Sub-Sabaran Afri.ea: From crisis to suatainable gramb. World Bank, Waabington, D.C., USA. 89 WHAT BEYOND SUSTAINABILITY? PaulWoomer Tropical Soil Biology and Fertility Programme UNESCO-ROSTA, NAlROBI, KENYA ABSTRACT Improved sustainability oftropical agroecosystems must not be viewed by scientists and decision-makers as an end in itself, but as a means of improving the quality of human lives while protecting the environment. For many years, agricultural development was guided by the principals of economic viability and technical feasibility. More recently, environmental soundness and social acceptability have been identified as equally important criteria. Suddenly, the impact of all criteria have been grouped into the catch all of"sustainability". Meanwhile, individual research objectives continue to be addressed within single disciplines, or occasionally by teams representing a few disciplines. Perhaps, the definition of sustainability needs to be better elaborated through an interdisciplinary approach directed toward the maintenance of the agroecological resource base. The resource base consists of renewable and non-renewable requirements for plant productivity, labour and capital. These resources interact with farming systems through a series of cropping cycles. Because the farmer seeks to recover part of the resource base as yield, changes in the sizes of individual components ofresource pools are an inevitable consequence ofland management. However, the sizes ofthe individual components of the resource are partially interchange-able. For example, capital and labour may be combined as a variable input and substituted for plant nutrients removed as yield or lost to leaching and erosiono pfthe many mineral nutrients required for plant growth, only carbon, nitrogen and sulphur are biogeologically recycled, with readily available and manageable atmospheric reserves. All other plant nutrients are subject to long-term sedimentary processes, and are concentrated and recycled over geological time. This disparity in nutrient recycling processes must be considered within the context of developing sustainable agroecosystems reeeiving little or no external inputs. A key component to the development of more sustainable agroecosystems is the reduction ofloss from the non-renewable resource base, primarily soils. While it can be argued that soils are a renewable resource, and that soils lost from an location ofien accumulate at another, it must be remembered that soil formation also occurs over geological time, and that soils lost from a single storm event can result in losses equivalent to the amount formed over centuries. Inereased removal of the resource base as yield per unit land area is an important means of addressing the food requirements of inereasing populations. Meanwhile, the improved sustainability of agroecosystem supplied with little or no external inputs approach a yield potential proportionately lower than these population increases. The solutioli to this dilemma lie in the resource base itself. Cash generating agriculture must be promoted as a means to increase capital, and in turn used to resupply renewable resources and better reward labour. Human resources must also be more effectively applied to limit the losses of non-renewable resources; particularly soil degradation due to erosional processes. The development ofhigh value/low volume agricultural products is an important means to reduce the international export of plant nutrients from lesser to more developed economies. Soil conservation and the development and promotion ofhigh value/low volume export eommodities must remain important national priorities as a means of reducing plant nutrlent 1088es that otherwise must be replaced with costIy, importad fertilisers. Improved sustainability ofthe resouree base at the farm-Ievel is not the solution but on1y the first step to improving human lives in developing nations of East Afriea. INTRODUCTION Given the present state of affairs, the attainment of sustainable development in Arrica is an attractive ideal. Population increases, coupled with the diminished availability of unexploited lands with agricultural potential have resulted in a decline in per capita food production. The threat of continued desertification and the extent of soil erosion and fertility decline has led to the rethinking of 91 agricultural development strategies. At tIlla assembly, a series of presentations will taIte place where sustainability iB used as a rationale for continued researcb objectives. Once again, we wíll be told that research and development activities willlikely result in improved sustainability of agroecosystems, but additional, longer-term researcb efforts will be necessary in order to guarantee tbat, indeed, sustainability has been acbieved. Too often, sustainability is referred to in the titIe and througbout a paper, yet remains undefined within tbe texto Less often. a serious attempt will be made to define sustainability in a n.ew and revealing context witbout regard for tbe confusion resultant from constant redefinition. Sustainability reina in the universities and institutes from tbe most to the lcast developed nations and rings &om lips of deana and donors, scientiste and public servante, schoolchildren snd graduate studente and someday will be taught to young, innocent minds on tbe knees of grandfathers. It 8eema everyone ia talking sbout sustainability issues except for the silent majority most affected. 1 have never heard a farmer say tbis word; perhaps they are too busy. SUSTAINABILITY DEFINED In aIl fnimess, tbe recognition of the need for agricultural sustainability has developed from the good intentions of a great many seetora. The World Commission ofEnvirorunent and Development (WCED, 1987) defined sustainable deveIopment as meeting "the needs and aspirations of the present without compromising the ability offuture generations to meet their own needs", Agricultura! sustainability may be viowed as having evolved from the pubJic awareness that agricultura! systems must meat the food needs of present as well ss future generations and to represent a refinement of agricultura! development policies that consider tbe need. and potentials of small-scale farmera to contrlbute to nationa! and internationeI food production (Brady. 1990). This awareness was in marked contrast to previous paradigma manifest in the "Green Revolution" of the 1960's and 1970's, marked by lsrge-seale internationaI efforts to develop increased food production in the tropics through modification of tecbnologies in the developed countries during previous decades (sea Harwood, 1990). Harwood (1990) defined tbe framework for sustainability as "an agricultura tbat can evolve indefinitely toward greater human utility, greater efficiency of resource use and a balance with the environment that is favourable both to humana and to most other species". Okigbo (1991) atatad that "A sustainable agricultural produetion system ls defined as one which maintains an acwptable and increasing level of productivity, that satisfies prevailing nesdo snd i8 continuously adapted to meet the future needa for increasing the csrrying capacity of the reseurce base .." The aboye definitions are holístic, and reflect the awareness that agricul-tural development must address more than immediate food needs snd current world msrkets. A great many fsrmera are resource poor, and are likely to remain so; agricultura! development "solutions' that fail to address the present and future nesds ofsmallholdera in the tropics are in faet missing a large segment oftarget populations. Othera have taken a more agroeoologiceI spproseh to sustainability lssues directed toward the identification of environmental parameters useful in sustainability assessment. Conway (1985) defined sustainability as the ability of a system to maintain productivity in spite of larger disturbances such as repaated stress or a major perturbation._. n Similarly, Young (1989) recognized sustninability as the maintenance of production over time, without degradation of the resource base on wbieh that production is depandent. Recently, 8wift 1992 (personal communication) has examined tbe feasibility of identifYing spacific soíl parameters that may serve as indicators of the soíl resouree base. These indicators inelude organic matter fractions, rations of those fractions, indicator groups of soH fauna, and N-mineralisation eapacity. These definitions of sustainability obviously span a great deal oC spacial and temporal boundaries ss well as reflect the wide range of disciplines that have implicit interest in tbe subject. When regarded in tbis light. the diversa and occasionally contradictory definítions of sustainability become less confusing, allowingme to (apologetically) contribute one oCmy own. "Agroecosystems can 92 never be entirely sustainable due to the inevitability of resource removal as yield, however, eomponents of the homeostasis associated with natural eeosystems can be promoted within managed eeosystems resulting in reduced depletion of non-renewable and greater resílience of renewable resources.- More simply stated, 80mething need not be completely sustainable to provide long-term benefit to hl'manki nd in a more sustainable fasbion. It la the purpose of thís paper to justifY this definition, and to demonstrate that improved sustainability at tIle cropping aystem level is but a first step to meeting the needs of present and future generations. BEFORE SUSTAlNABILITY The emergence of sustainability as an agricultural philosophy la the culmination of a series of previous deve10pments in agricultura and international aid. successful traditional agricultural aystems in Aftica serviced local communities, with few components of yield removed frem the immediate vicinity. In part, these syatems were displaced by colonial plantations designed to develop export products intended ror European markets. Immediately following World War n, western agriculture underwent drastic transformation. Improved crop varieties in conjunction with broad based use of petroleum fuelled farm vehlcles and fertílisers, followed by the widespread availabílity of pesticides (Edwanls, 1988) led to what may be termed ·conventional high energy input agricultura". Agriculturalists were focused upon the optimalisation of short-tenn gama. Intemational agricultura! development during the 1960's and 1970's 80ught to horizontally transfer these developments frem developed to developing countries (Okigbo. 1990). In what la termed ~e Oreen Revolution", research was conducted to develop improved larger-scale rarm technologies. develop infrastruetural capabilities to import or produce agricultural supplies, and create progres8Íve rural structures 4. 7 8 5 3 9 18 10 10 18 " N-DEFlCIENCY EVIDENCE 1. 2. 3. 4. Symptoms obaerved Experimental resulte Soil and tissue testing Soil suzvey resulta. ADDlTIONALEVIDENCE REQUIRED l. Improved soilsurvey reporte í.e. better information on distribution of problem. POTEl~ SOLVTIONS 2. 8. 4. 5. 6. 7. 8. Efficient ferti.lizer use Efficient N-use cultivare Control of soD erosian Proper crop residue management Proper llIllIl8gI!DlImt oC FYM Breed or identif.y bean Oegume} varieties for improved N.fixatíon. Identif.y superior atrains ofRhizobia Proper weed management 9. Improve plant nutrition for enhanced BNF 10. 11. 12. 13. 14. 15. 16. Improve extrapolation ofresearcll resulte Improve syncbrony of nutrient suppJy with demand Improve credit facilítíes Improve extrapolation for research resulte to varied environments Improve reeommendations for the maize-bean intercropping system Hedgerows or alJey cropping Leguminous ¡reen manure crop 1. 107 P·DEFICIENCY EVIDENCE AVAILABLE 1. 2. 3. 4. Symptoms observed Experimental results SoH and tissue testing SoH survey reports ADDITIONAL EVIDENCE REQUIRED 1. Improved soH survey reporte, te. te know distribution of problem, POTENTIAL SOLUTIONS AND TBEffi lMPORTANCE AS RESEARCH TOPICS