lessons for farmer-oriented research: experiences from a ... · lessons for farmer-oriented...

Lessons for farmer-oriented research: Experiences from a West African soil

fertility management project

E. Suzanne Nederlof1 and Constant Dangbegnon21Royal Tropical Institute (KIT), Amsterdam, The Netherlands; 2International Center for Soil Fertility and Agricultural Develop-

ment-Africa (IFDC-A), Lome, Togo

Accepted in revised form May 4, 2006

Abstract. Donors, scientists and farmers all benefit when research and development projects have high impact.However, potential benefits are sometimes not realized. Our objective in this study is to determine why resource-poorfarmers in Togo (declined to) adopt recommended practices that were promoted through a multi-organizational projecton soil fertility management. We examine the processes and outcomes related to the adoption process. The project wasundertaken in three villages in the Central Region of Togo in West Africa. The development and research processesthat took place during the implementation of the project were critically analyzed using a conceptual framework thatmay be useful for improving the impact of future participatory projects. At the macro level, opportunities forinnovation were not deliberately explored with participating farmers and other village members; consequently‘‘pre-analytical choices’’ made during the planning phase resulted in practices that resource-poor farmers were, for avariety of reasons, unable or unwilling to adopt. From the outset, donors and scientists focused on soil fertilitymanagement, but failed to take into account the wider economic context within which soil fertility management tookplace. This was a major obstacle to the subsequent adoption of recommended management strategies. Scientists anddonor partners measured the success of the Project in terms of crop productivity, but farmers’ choices were influencedby a complex mix of socio-economic, political and technical factors. We also illustrate the importance of selectingappropriate categories of farmers for a particular experiment. We conclude that for participatory research anddevelopment projects to be successful, it is not enough to develop technologies that ‘‘work’’ in a technical sense. Inorder to be scaled up and widely implemented, such technologies must also meet a variety of needs of resource-poorfarmers and be acceptable from a socio-cultural point of view.

Key words: Agricultural research, Farmer livelihoods, Participatory development, Pre-analytical choice, Resource-poor farmers, Soil fertility, Technology development, Togo, West Africa

Abbreviations: CoS – Convergence of Sciences; ISFM – Integrated Soil Fertility Management; NGO – NonGovernmental Organization; PCC – Project Co-ordination Committee; PTD – Participatory Technology Development

Suzanne Nederlof holds a PhD in Communication and Innovation Studies from Wageningen University in theNetherlands. After finishing her MSc. degree in Development Sociology at the same university she worked at aresearch project in Burkina Faso and at the International Center for Soil Fertility and Agricultural Development(IFDC) in Togo. Subsequently, she was based at the FAO regional office for Africa in Ghana, assigned to theConvergence of Sciences Program. She conducted the fieldwork for this article during that period. She completed herPhD research on agricultural research with a scholarship from the Netherlands Organization for Scientific Research.She is currently employed as an advisor in rural innovation at the Royal Tropical Institute (KIT) in Amsterdam, TheNetherlands.

Constant Dangbegnon worked for several years (1990–1998) on Inter-University Cooperation (Benin, the Net-herlands and Israel) in the area of indigenous knowledge and the soft side of natural resource management. Thiswork constituted the basis of his PhD from Wageningen University in The Netherlands. He joined IFDC – AfricaDivision in Lome (Togo) in June 1999 as a socio-economic extension specialist in the Input Accessibility Program(IAP).

Agriculture and Human Values (2007) 24:369–387 � Springer 2007DOI 10.1007/s10460-007-9066-0

Introduction

Several authors have expressed skepticism as to whe-ther attempts made by agricultural research to improvethe sustainability of land use for West African farmershave been successful (Chambers and Jiggins, 1987b;Pretty, 1995; Mutimba, 1997; Bie, 2001; Stoop, 2002).The ‘‘Green Revolution’’ in irrigated rice in Asia wassuccessful mainly because it was based on the intro-duction of high-yielding varieties, which needed inputssuch as irrigation, pesticides, and fertilizers in order tobe productive. In this particular context it was possibleto make the environment fit the genotype (Castillo,1998). However, in many African contexts, agricultureis too heterogeneous for such an approach. Africanagriculture has been described as risk-prone, highlydiverse, and rain-dependent (Reijntjes et al., 1992). Inaddition to biophysical constraints, African farmersface low market prices. The availability of cheapagricultural products from regions using Green Revo-lution technologies and from industrial countries hasdriven prices down, worsening the situation for Afri-can smallholders.

Under such conditions, what benefits may researchrealistically contribute to agricultural development? Theimpressive literature on the non-adoption of researchoutputs (e.g., Rogers 1995; Scheuermeier et al., 2004)suggests that non-adoption may be explained by farmers’lack of knowledge and scarce resources, non compati-bility of the technologies promoted with farm conditionsand farmers’ goals, and the limited political influence ofresource-poor farmers on the research process. Theseproblems have been addressed by the Training and Visitapproach (Chambers and Jiggins, 1987a), FarmingSystems Research (Collinson, 2000; Dixon et al., 2001),on-farm research (Chambers, 1990, 1994, 1997; Werner,1996; Mutsaers et al., 1997), and participatorytechnology development methods (Rhoades and Booth,1982; Pretty et al., 1995; Vereijken, 1999). A more recentexplanation for the limited impact of research on thelivelihoods of resource-poor farmers is that researchersalone cannot grasp the complexity and dynamics of localsituations (Pound et al., 2003; Scheuermeier et al., 2004).This recognition resulted in the development of differentapproaches, for example, Participatory InnovationDevelopment (Scheuermeier et al., 2004), EnablingInnovation (Douthwaite, 2002), partnership building foradvancing Participatory Technology Development (PTD)(Van Veldhuizen et al., 2003), Local AgriculturalResearch Committees (Ashby et al., 2000), Farmer FieldSchools (van de Fliert, 1993; Pontius et al., 2002; CIP-UPWARD, 2003) and the Convergence of Sciences(CoS) approach (Hounkonnou et al., in press).1 Van deFliert and Braun (2002) also address this issue and state

that it is increasingly accepted that farmers should playan important role in research, development, andextension.

The present study aimed to fine-tune a client-orientedframework for agricultural research, developed within theCoS program, in order to support the development of cri-teria and procedures for better stakeholder collaboration.Deliberate, careful negotiation and interaction with allstakeholders was considered necessary for agriculturalresearch outcomes to be efficient and fit the needs andopportunities of farmers. We assumed that the context forresource-poor farmers cannot be changed in order tomeaningfully apply an innovation; therefore, it was ines-capable that innovations produced by science needed to fitthe existing ecological, social and economical context.

The research and development project on IntegratedSoil Fertility Management (ISFM) carried out in Togo wasused as a case study to analyze the impact of choices onresearch design and processes.2 Giampietro (2003, 30)calls these choices ‘‘pre-analytical’’ and defines them asthe ‘‘choice of relevant goals, variables, and explanatorydynamics for the selection of an explanatory model.’’ Toexplain the concept, he uses Mandelbrot’s (1967, inGiampietro, 2003) example: if you want to know thelength of Britain’s coastline, you had better agree on thescale of the map that is to be used, because the scale willstrongly affect the result. Critical issues emerging from thecase were analyzed in order to gain insight into the rela-tionships and dynamics underpinning the effectiveness ofagricultural research that was intended to enhance thelivelihoods of resource-poor farmers. These critical issuesprovide micro case studies within the larger case study,revealing the often transformative effect of the relation-ship between the context, the activities, and outcomes.

We deliberately emphasize farmers’ perspectives be-cause farmers were the intended beneficiaries. The casereported in this article aims to draw lessons regarding thefactors needing to be considered in the design of researchintended to benefit resource-poor farmers; hence, thisarticle reports ‘‘research on research.’’

Conceptual framework

The focus of this study is the research process rather thanthe research outcomes. An initial conceptual framework(Nederlof, 2003) was developed on the basis of a liter-ature review. To validate the initial framework, tools suchas brainstorming, validation workshops and discussionsessions, both with individuals and with groups, wereorganized with CoS and other scientists. In order todevelop a research process that benefited resource-poorfarmers, we proposed five criteria and procedures (seealso Roling et al., 2004).3

370 Suzanne Nederlof and Constant Dangbegnon

1. Research takes into account existing opportunities orpotential for innovation at a macro level. This impliesa thorough understanding of the context and relevantstakeholders before the start of the project.

2. Research is grounded in the opportunities, needs andexpectations of intended beneficiaries. This meansanchoring research activities in local conditions,including stakeholder demands and needs, specificallythose of resource-poor farmers.

3. Research results in design systems that work under theconditions facing farmers. Hence, the (agricultural)innovation or technology developed should be feasi-ble and efficacious given the farming system and fieldconditions.

4. Research results in design systems that are acceptableand appropriate for resource-poor farmers. This im-plies that the innovation or technology not only yieldsthe desired results but also fits the culture, prefer-ences, traditions, personal circumstances and priori-ties of resource-poor farmers.

5. Research develops innovations that can be scaled up.Thus, the technologies or innovations developedshould potentially have an impact beyond that of thefarmers initially involved and past the duration of theoriginal project.

In discussing these criteria, certain proxies or indicatorswere used because the criteria in and of themselves werenot ‘‘measurable.’’ These proxies are crosscutting, mean-ing that they are relevant to more than one of the criteria:

1. Participation of stakeholders in ‘‘platforms’’ toengage in collaborative learning. Pretty (1994, 1995;Pretty et al., 1995) developed a ladder distinguishingdifferent types of participation. He argued that forsustainable development, nothing less than interactiveparticipation was required.4 Participation is a tool tofacilitate learning that increases the countervailingpower of farmers over the research process (thedemocratization of science, c.f. Funtowicz andRavetz, 1993). Johnson et al. (2003) assessed theimpact of participatory methods on research useful-ness and found that participation resulted in greatereconomic impact and more relevant innovations,mainly when participation was implemented at anearly stage in the research process. In the lateseventies, Morss (1976) made similar observations.

2. Socio-cultural factors that are pertinent to commu-nities, the production system, and technical aspects.

3. There is an interface between technical / biologicaland social issues.

4. Issues related to the wider context affecting farmers’livelihoods (e.g. marketing possibilities).

5. The assumption that farmers have veto power (Rolinget al., 2004); therefore, research must be negotiated

with farmers. The design of research processes is atthe interface between science and the veto power offarmers. Hence, the pre-analytical choices made (e.g.,regarding hypotheses, topic, type of benefit to beachieved) can hamper the impact of the research if notnegotiated with farmers.

In our case study, stakeholders, including farmers,were asked to evaluate the project and related issues.This allowed the framework to be fine-tuned as the studyprogressed. The proxies thus emerged during the casestudy and should therefore be considered an outcome ofthe present study.

Methods

We opted for a qualitative research approach due to theexploratory, conceptual and constructivist character ofthe study (see Guba and Lincoln, 1989; Denzin andLincoln, 1994). Two principles guided the collection ofdata for the case study: (a) include rich detail in recordingevents, interviews and observations; and (b) use trian-gulation of methods in data collection to allow cross-validation of information.

First, we conducted a desk study of the soil fertilitymanagement literature. We also analyzed the project’sarchival records, such as informal reports, notes, corre-spondence, and so forth. Most documents concerning theproject were confidential; therefore, no references arecited in this article. Data on the context – both the areaand the villages in which the project carried out itsactivities, and participants and stakeholders - weregathered through the desk study.

Second, between October 2003 and April 2004 severalvisits were made to the three villages in which the projectwas conducted. The first author was not involved inproject implementation; however, the second authoractively facilitated the research and development pro-cesses associated with implementation of the project. Thecombination allowed both an outsider- and an insider-perspective of the project to emerge. The second authorhas both an agronomic and an extension studies back-ground, while the first author is a social scientist(including anthropology), allowing for a diversity ofperspectives. During the field visits, the authors madeparticipant observations during (a) field days organizedby national research and extension services, (b) partici-patory evaluations of the strategies proposed by farmersand organized by the international institute, (c) meetingsof the regional platform, and (d) data collection byresearch partners in farmers’ fields.

Third, the authors conducted semi-structured inter-views with the stakeholders. All interviews were con-ducted by the authors. Approximately thirty farmers were

Lessons for farmer-oriented research 371

interviewed individually in their fields. Three farmergroups also were visited for discussions. Three repre-sentatives of the international institute involved in theproject were interviewed and joint field visits were alsoorganized. In addition, we interviewed directors of theresearch, extension and agricultural policy analysisinstitutes that were involved in the project, as well as theprofessionals delegated to the project by each institute.During the interviews, the strengths, weaknesses,opportunities and constraints of project activities wereanalyzed. By the end of this study, all individualsinvolved in the project had been interviewed.

Description of the project

The Integrated Soil Fertility Management (ISFM) project,carried out between 1999 and 2003 in Central Togo, aimedto ameliorate the livelihoods of resource-poor farmersthrough soil fertility improvement. This focus was basedon the assumption that growing crops ‘‘mines’’ the soil,leading to nutrient depletion, decreased soil fertility, anddeclining production; hence poverty (Stoorvogeland Smaling, 1990). The project focused on the adoptionand maintenance of ISFM practices. The main resultsexpected from the project were the identification andadaptation of strategies for different farming circumstancesby using systematic learning with the stakeholders; makingfertility management strategies, methods and data availableto various stakeholders; strengthening the capacities ofresearchers and extension agents to successfully facilitatefarmer innovation; and strengthening partnershipsthrough project co-ordination and management.

The project facilitated negotiation among scientists,farmers, and national partners. It not only systematicallytested technologies but also engaged in developmentactivities, mainly through providing advisory services tofarmers. Extension mainly offered demonstration plots,and the research was based on Participatory Technology

Development (PTD). PTD is a process of purposeful andcreative interaction between farmers and outside facili-tators to develop technology options. It involvesprocesses such as gaining understanding of eco-specificand cultural contexts, defining priority problems andlocal experimentation, involving farmers in generatinglocally adapted technologies, and evaluating whetherfarmers internalized these technologies (Jiggins and deZeeuw, 1992; van Veldhuizen et al., 2003).

Stakeholders

The project involved various stakeholder organizations,ranging from an international institute, decentralizeddepartments of national extension, and research andpolicy analysis agencies; as well as individual resource-poor farmers. The extension service’s main activity in theproject was to demonstrate innovations to stimulate dif-fusion of technologies to farmers through such tech-niques as demonstration plots, farmer field days andIndividual Farm Management Advice.5 The major inno-vations introduced were the use of the leguminous covercrop Mucuna pririens var utilis and manure pits. Bothinnovations were demonstrated in farmers’ fields. Inaddition, cereal banks were set up.6

The main research activity of the project was to set upexperiments with farmers based on endogenous prac-tices. As shown in Table 1, experiments undertaken in-cluded determination of optimal chemical fertilizersdoses for a leguminous cover crop-cereal system, deter-mination of optimal chemical fertilizer doses in a rota-tional leguminous cover crop-cereal system, andimprovement of fertilizer efficiency in a combinedleguminous cover crop-cereal system.7

Agricultural policy analysts were responsible forgeneral coordination of the project in the CentralRegion and monitoring and evaluating the researchprocesses; this included field visits for discussion withfarmers and the facilitation of meetings between the

Table 1. Research activities in the soil fertility project.

Village Activity year 1 and 2

Ababa Determination of optimum rates of mineral fertilizer applications on maize in the cowpea-maize relaycropping systemDetermination of optimum rates of mineral fertilizer applications on maize in the soybean-maize rotationsystem

Figigi Improving the endogenous system of Egusi melon-maize relay cropping system through mineral fertilizerapplications on maizeImproving the endogenous system of Egusi melon-Cowpea and maize rotation system through mineral fer-

tilizer applications on maizeCedede Determination of optimum rates of mineral fertilizer applications on maize crops in the soybean-maize

rotation systemDetermination of optimum rates of mineral fertilizer applications on sorghum crops in the cowpea-sorghummix-cropping system


different organizations (see Figure 1). The internationalinstitute supported these activities through training andmonitoring.

Project location

The Project took place in the Central Region, one of fiveregions in Togo. This region is characterized by annualrainfall of between 1000 and 1200 mm. There is onerainy season from April to October and harvesting takesplace in November and December. Soil types includetropical ferruginous and ferralitic soils. Causes of soildegradation are sheet erosion, reduction of organic matterin the topsoil, and loss of plant nutrients (Brabant et al.,1996).

The Project had three research sites: Affem-Kabye,Sessaro and Goubi. The three villages were selectedduring a workshop with the main national partners.Criteria provided by the project for site selection wereaccessibility (practical element); level of soil degrada-tion, crops cultivated, and level of intensification,including fertilizers use (situational elements); andreceptivity to innovation, socio-economic and culturalcontext, and access to innovations (social elements).8

The main cropping systems in each of the threevillages were maize-based, cotton-based and yam-based,respectively. Legumes such as cowpea, soybean,groundnut and Cajanus cajan were used in rotation, andrelay cropping was practiced. Although cattle were usedfor animal traction, the main livestock component in theagricultural system was small ruminants (sheep and goat)and poultry. The dominant ethnic group in both Affem-Kabye and Sessaro was made up of Kabye migrants fromthe northern part of the country. Affelees are the domi-nant ethnic group in the third village, Goubi. Mineralfertilizers were used, but less so in Sessaro because thevillage tended to have even fewer resources than theothers. In Goubi, farmers tended to be less concerned

about soil fertility due to the abundance of fallow land. Inboth Affem-Kabye and Sessaro, compound farming wasvery important; but in Goubi the farms and householdareas were spatially separate.

Findings

We sought to assess each of the criteria discussed in theconceptual framework using some of the proxies dis-cussed above. Discussion of each criterion starts with aquestion related to the key concepts used for analysis andends with concluding remarks.

Macro-level opportunities for innovation

Did the project take existing macro-level opportunities orpotential for innovation into account? Research planningideally is negotiated among the different stakeholdersconcerned, including the ultimate beneficiaries, in thiscase, the farmers (e.g., Johnson et al., 2003). Did thishappen in practice or were project contents established ata central level?

Initial choices could have been influenced by factorssuch as donor requirements, the personal preferences ofscientists, and the personal convictions of peopleinvolved. The quality and acceptability of these pre-analytical choices (Giampietro, 2003) depended on theprocess through which the choices are made.

Pre-analytical choicesDuring the first negotiation phase between the donor andthe international institute that initiated the project, severaldecisions were made that did not involve farmers. A firstchoice was that soil fertility management was to be thefocus, rather than rural credit, which had initially beenproposed by the international institute. The internationalliterature (Steiner et al., 1988; Pieri, 1989; Stroosnijder,

Agricultural policy analysts

coordinate

Site selection

Training of partners

Participatory Diagnosis of

village situation

Extension activities

Problems with available

solutions

Participatory Technology

Development

Problems withoutavailable solutions

• Mucuna demonstration fields• Manure pit• Farm Management Advice• Cereal bank

• Optimal doses of fertilisers• Cereal- cover crop rotation• Cereal- cover crop

intercropping

Training of partners

= Results in

Figure 1. Sequence of activities in the soil fertility project.


1992; Swift, 1996; Mazzucato and Niemeijer, 2000;Stroosnijder and Van Rheenen, 2001) suggests that soilfertility management is indeed a major issue in this re-gion. A second decision, based on the ideas of special-ized scientists within the international institute, was thatagriculture needed to be intensified in order to improveproduction; it was assumed that an increase in productionwould benefit farmers and reduce poverty. A third deci-sion was the focus on technology development ratherthan on rural development, which would have includeddevelopment of infrastructure and markets. Such deci-sions can have major effects in on operationalizingproject objectives.

Problems, solutions and criteria for activities that werelisted during the initial donor meeting had to be groundedin the demands of country partners in order to fit thespecific context. The international institute thereforeorganized a meeting in-country with potential stake-holders, including the governmental research and exten-sion organizations, the agricultural policy analysisorganizations, and a non governmental organization(NGO) working on credit, to present the outcomes of thedonor workshop and discuss the way forward. The deci-sion to base the Project in the Central Region of Togo wasbased on the need to improve the rural credit scheme,which was the initial topic proposed. This decision wasnot reconsidered by the partners even after the focus ofthe project changed. Thus, the choice of region was madebefore the final topic was known. It soon became clearthat the NGO initially involved was not interested in thenew direction that the project had taken, which in itsopinion could not be linked to credit, and so decided towithdraw. Other partners, however, decided to joinbecause of the likely importance of the subject.

Concluding remarksThe research topic, soil fertility, and general contents(problems, solutions and criteria for activities) werenegotiated at a higher level of decision-making, and itrespected the donor’s requirements. The outcomes of themeetings with donors were discussed with local partnersin the pre-selected province; however, the partners werepresented with the choice to either join or withdraw, butthey were not offered an opportunity negotiate theactivities, objectives and conditions, since these hadalready been established during the donor meeting.Indeed, one prospective partner, who was not dependenton the anticipated project resources, withdrew. The otherpartners chose to join the project because they consideredsoil fertility to be a problem. No deliberate attempt wasmade by the international institute to understand thereasons behind partners’ participation. It was assumedthat the decisions made at the higher-level meetingswould benefit the farmers, although they themselveswere not directly consulted.

A dilemma for numerous (participatory) projects ishow to involve farmers in the project planning andnegotiation phase. What alternative methods for involv-ing farmers in project development might be possible,especially in West Africa where farmers have limitedpolitical clout? In our study, it was difficult to evaluatewhether soil fertility was indeed the most relevant issueto be tackled from the perspective of farmers. Despite themany justifications found in the international literaturethat supported the decision to focus on soil fertilityimprovement, it was important to verify such assump-tions with farmers and to even change the topic, if nec-essary. Roling et al. (2004) argue that it is at least asimportant to provide farmers with the opportunity toinfluence the topic as it is to ground research in inter-national literature. However, the fact that farmers werenot involved did not mean that the topic was thereforenot of interest to them. The point is that soil fertilitymanagement was chosen as the focus of the project basedon donor requirements and the ideas of expert scientists,and did not take into account farmers’ perspectives.

Needs and expectations of resource-poor farmers

Were the project activities grounded in the needs andexpectations of resource-poor farmers? Before answeringthis question we need to understand who the farmerswere. Only then the process used to negotiate experi-ments and trials can be reviewed.

Types of farmersDuring a community meeting following a participatorydiagnosis, farmers volunteered to participate in the pro-ject’s activities. The project worked intensively with 60households equally distributed across the three villages.Twenty-six of these households had relatively large farms(more than 5 hectares). From field observations, discus-sions with the farmers, and from data available within theproject about farmer characteristics, it became clear thatparticipation was predominantly from the relativelyresource-rich, better-informed and well-educated malefarmers in the community. A critical-minded staff mem-ber of the international institute expressed his concern. ‘‘Itis worrying that only a small group of farmers is involved,and that those farmers are the ones who have relativelybig fields, and much labor available.’’

Johnson et al. (2003) argue that farmers choosing toparticipate in projects are unlikely to be the poorest in thecommunity or from a marginalized group. In addition,chances are that these farmers have had previous expe-rience with experimentation. The agricultural extensionagent also suggested that some ‘‘lead’’ farmers would beincluded. Roling (1988) stated that extension workerstend to interact with the top 10–20% of the farmers, andthat in the absence of explicit efforts to involve the


poorer farmers, it is safe to assume that the relativelybetter-off are involved.

Gender issuesWomen were not involved in the project, despite the factthat farming at the intervention sites was largelywomen’s work. Women sow and harvest, while menplow the land, and together they weed and maintain theplots. The neglect of gender aspects was confirmed inproject documentation, which reported as early as thediagnosis phase that ‘‘a major shortcoming was the weakparticipation of women.’’ The partners claimed this wasdue to the cultural context in which they were operating.The organization of the village was based on a hierar-chical patriarchal system. Women did not generallyparticipate in important meetings, including communityand project meetings, when men were involved. As aresult, only men participated in the meetings andsubsequent project activities. Even when women werepresent at a meeting they were expected to remain quietwhen in the company of men. Some women cooked foodfor the men attending the meetings, but this could hardlybe considered participation.

The following critical incident concerning a soybeandensity trial demonstrates the importance of includingwomen in experiments. A woman farmer explained thatshe obtained a higher yield using her traditionalproduction method than was produced with the intro-duced practice. According to the researcher who wasworking in the area, this could be explained by the wayin which the plot was sown, as follows:

When the extension agent went to the local pub to havea tchouk, he met one of the farmers participating in thetrial on soybean density. Under the joy of the tchouk,the extension agent explained to the farmer exactly howto sow the plot. The farmer was to divide the plot intofour quarters, sow using two different densities; twoplots were to be lines and the other two at random, andapply varying fertilizer applications. Since his wife wasresponsible for sowing the fields, when the farmer ar-rived home, he immediately explained to his wife whatthe extension agent had just told him. A few days laterthe woman went to the field and sowed the plot fol-lowing her husband’s explanations. The woman did notfully understand the second-hand explanation and as aresult there were few differences in sowing practicesbetween the four plots. The plot did simply not respectthe extension agent’s protocol for research.

Because women are responsible for sowing, theproject could not succeed without their involvement.

Negotiating experimentsExperiments can be based on ‘‘contracts’’ made betweenfarmers and scientists. In Affem-Kabye and Sessaro a

participatory diagnosis resulted in a contract to developcollective solutions for soil fertility improvement basedon local practices. In Goubi, it led to an agreement toraise awareness about the importance of maintaining soilfertility.

To ground the project in the needs, demands, andconditions of farmers, the national partners undertook adiagnostic study of soil fertility management issues withfarmers using participatory methods.9 A team of agron-omists and socio-economists carried out a diagnosis byvisiting the villages. The study generated an under-standing of community processes, as well as traditionalpractices and previous experiences in soil fertility man-agement. It also allowed the partners to be properlyintroduced to the local authorities, and to obtain generalinformation about the village.

The diagnosis identified concrete problems related tosoil fertility decline experienced by farmers. In bothAffem-Kabye and Sessaro, the farmers believed lowproduction to be largely the result of soil depletion.However, from the process used it cannot be concludedthat soil fertility depletion was indeed a major preoc-cupation for farmers. Farmers may well have deducedthat assistance might be forthcoming if they focused onthe topic evident on the logo of the Land Rovervehicles. However, a clear indication that the farmersconsidered soil fertility to be a problem was theirdevelopment and practice of soil fertility improvementmeasurements.

In Goubi, soils are still relatively fertile, and almostno chemical fertilizer was used except in cotton culti-vation. The availability of unexploited land in forestareas surrounding the village allowed farmers to prac-tice shifting cultivation. The diagnosis in this villagewas to take a first step, towards that of raising aware-ness about the threat of soil depletion in the future. Thefarmers indicated interest in further exploration of thesituation.

Depending on the type of problems that farmersidentified in relation to soil fertility, different strategieswere followed. For problems that had already-existingsolutions, and that were considered suitable for extensionat the national level, the extension service was the mainnational stakeholder. A total of 114 farmers were trainedin Mucuna fallowing, 30 in composting techniques, and91 demonstration plots were prepared across the threevillage sites. Following information sessions on issuesrelated to develop of cooperatives, eight farmer groups inAffem-Kabye and two groups in Goubi started cerealbanks. In addition, several field days were organized toinform non-participating farmers of these activities. Theinternational institute also reported that 14 farmers in thethree village sites received individual farm managementadvice. Success was measured in terms of the number offarmers reached.


Ownership of the researchTo produce innovative solutions, endogenous practicesidentified during the diagnosis formed a basis forimproving practices through PTD. Researcherssuggested a number of possible experiments based onthe results of the participatory diagnosis to improveupon existing practices. The farmers in turn prioritizedthe options and researchers focused on about threeexperiments, taking into account both feasibility andfarmers’ preferences.

According to the international institute, the researchconducted is farmer-led, on-farm research. However,ownership was not always well negotiated, and in somecircumstances the experiments turned out to beresearcher-led.10 Another example emphasizes theimportance of ownership of the research relates to thesoybean density trial in Affem-Kabye. When discussingthe trial, researchers repeatedly stressed that theexperiments belonged to the farmers, who worked forthemselves and not for the researchers. Some time latera farmer involved in the soybean trial explained that hedecided to dig up the soybean plants that had notreceived fertilizers because they were not doing well.As a result the difference in yields was not large. Oneof the researchers got mad at the farmer and asked himwhy he had not done what he was told to do. Theresearcher then explained to the farmer that the resultswere no longer comparable due to his intervention andhad become useless. A bit later the researcher calmeddown and requested that the farmer would not act insuch a way again. The farmer replied, ‘‘Yes sir, I willnot deceive you again, and I will not spoil yourexperiment again.’’

During the first two cropping seasons no controlplots were set up because farmers did not see thebenefits. As such it was impossible for biological sci-entists to make sound statements. During the thirdcropping season, the trial design was therefore re-negotiated and control plots were included in order tomake the study scientifically sound according to thecriteria of biological science. The researchers managedto convince the farmers that they needed a control plotin order to be able to draw conclusions. In one of thevillages it was very difficult to convince the farmers ofthis need. Farmers in Affem-Kabye argued that theyalways used fertilizer in maize production; therefore,they did not think it sensible to compare a fertilized plotto a plot where no fertilizer had been applied (i.e., thecontrol). In another case, a farmer in Goubi appliedfertilizer to a control plot because he was not willing torisk production losses.

At times, it appeared illogical to farmers to compare anintroduced practice with what they already did. Theexperience of Bjornsen Gurung (2003) confirmed that thescientific approach of comparing two systems that

differed by only one factor was indeed illogical forfarmers. Because farmers seek benefit rather than proof,it is incomprehensible for them to leave parts of theirplots untreated when there is another production methodthat is more likely to succeed (ibid).

During the first two cropping seasons, different trialswere conducted in the three villages depending on farmerpreferences. However, biological scientists required thattreatments be replicated so that they could say somethingabout the effect of an introduced practice (because ofpossible side effects). It was therefore necessary to con-duct the same trials in all three villages for severalcropping seasons. This design allowed biological scien-tists to gather sufficient data. As such, farmers’ needswere given less priority during the third year and theconditions for the experiments proved to have beeninsufficiently negotiated. The farmer in Goubi whohosted the soybean varietals trial explained that otherfarmers rarely came to see the experiment. He explainedthis by repeating what he had already indicated duringthe diagnosis phase, that local farmers do not have a soilfertility problem. Hence farmers participated by provid-ing their fields and labor to carry out experiments forscientists on a topic in which they had already indicateddisinterest during the diagnostic phase.

Concluding remarksIdeally, the diagnostic phase provides ample opportunityfor negotiating between farmers and researchers so thatthe interests of both are met. However, the internationalinstitutes made several pre-analytical decisions aboutproject design that farmers and partners could not influ-ence but instead were compelled to accept. First, it wasassumed that (a) solutions for some problems werealready available and could be brought to farmers viaextension services and (b) some solutions were not yetavailable and demanded exploration with farmers. Sec-ondly, ‘‘productivity’’ was chosen as the criterion forsuccess. This favored a focus on large farmers for whomuptake and impact generally happen more quickly. Whatthe participatory diagnosis did not address was the con-text within which the project strategies were embedded,such as market outlets, availability of fertilizers and seed,different soil fertility management approaches by land-owners and migrants (Adjei-Nsiah et al., 2004; Saıdouet al., 2004).

After the donor withdrew and the ISFM project wasintegrated into another project of the same internationalinstitute, the role of biological scientists becameincreasingly more important. As a result, the requirementto produce scientifically sound data prevailed. Scientificrigor made it necessary to compare two systems thatdiffered by one factor and were replicable. Experimentalscience requirements – such as the ability to replicate andthe need for controls in order to draw scientifically sound


conclusions – were not sufficiently negotiated withfarmers.

So, why did the farmers agree to the proposedexperiments? One group of farmers answered this ques-tion with respect to their readiness to participate in theresearch trials thus, ‘‘If you give us the inputs we want toparticipate, but if we have to buy them ourselves thatchanges everything!’’

Benefits for resource poor farmers

It is important to ask whether the project designed sys-tems that worked under farmers’ conditions and wereacceptable and appropriate for resource-poor farmers.The project delivered extension messages about prob-lems for which solutions were considered available andengaged in further trials with farmers on other issues.Have these two types of efforts benefited resource-poorfarmers? Did the context allow the innovations intro-duced to benefit resource-poor farmers? Below, theperspective of each of the stakeholders is discussed.

Evaluation of project results by the international instituteBased on an evaluation of the project’s research activi-ties,11 the international institute concluded that the sys-tems developed were effective in the farming systemunder study. The most successful participatory trial wasthe relay cowpea-maize cropping system. The yields inthis system were comparable to the optimal yieldsobtained in a simulation model. Average maize grainyields doubled in two years. The farmer who had thehighest level of production – seven tons per ha – wasbased in Affem-Kabye. The trial in which he wasinvolved allowed farmers to harvest two different cropson the same plot in a single cropping season. The averagemaize grain yields in the Egusi melon-maize relaycropping system also increased, but not impressively.The use of fertilizers in the endogenous Egusi melon-Cowpea and maize rotation system improved maize grainyields. The effect of both fertilizers and a precedingcowpea crop improved yields. Also, the yields werehigher where the residues of soybean were incorporatedinto the soil in a soybean-maize rotation system, althoughthe difference was not large. Another result was that animproved sorghum variety was found not suitable for thecowpea-sorghum mixed cropping system. The maincriterion for the international institute was yield; hence,the results of the PTD activities seemed promising.However, how many farmers can achieve these results?And if they can, is there a viable strategy to market thesurplus production?

Evaluation of project results by national partnersBoth researchers and extension workers expressed theirsatisfaction with the project’s results. They most appre-

ciated the project’s participatory approach. Researchersand extension workers felt that farmer knowledge wastaken seriously and that farmers were involved in theexperiments. However, researchers and extension work-ers seemed reluctant to openly criticize the project. Aninitial explanation for their reticence might be that theproject provided researchers and extension workers withsupplementary revenue (in the form of a daily subsis-tence allowance for field trips and remuneration in thecase of extension workers) and as a result they had aninterest in sustaining it. A second explanationmight be thatresearchers did not think criticism would make a differ-ence. Researchers themselves have become skepticalabout the impact a project can have on sustainable devel-opment. Indeed, some researchers have lost confidence inwhat can be achieved through a research project.12

From the interviews it appeared that agricultural policyanalysts appreciated the project results very much be-cause they felt the results contributed to the developmentof options for increased production. Increased yields alsoimproved the analysts’ credibility vis-a-vis the farmers.Due to increased production, (urban) consumers couldpurchase produce at a lower price. In an informaldiscussion, the director of the agricultural policy institutecourageously admitted that farmers then would receivelower prices. So, farmers do not share the same interestsas (urban) consumers and alternative marketing optionsfor farmers must be found. A major drawback of theproject, according to the director, was the limited fundsavailable; consequently, there were limited possibilitiesfor scaling up to other districts. Understandably,researchers and extension workers supported this view.

Evaluation of project results by farmersThe soil fertility improvement technologies tested wereoften positively evaluated by farmers and project partnersalike. These evaluations focused on the technical perfor-mance of the innovation under specific conditions; how-ever, the technical performance of an innovation is justone reason for adoption. The critical incidents reported inTable 2 demonstrate the mix of criteria used by farmers.

Table 2. Farmers’ indicators for assessing an innovation.

Indictors in thetechnical domain

Indicators in thesocio-economical domain

Yields Land tenure arrangementsOrganic biomass content Integration in farm

(Multi-purpose crop)

Climatic conditions Farming calendarMarketing possibilitiesCost efficiency

(investment needed)Short-term benefitLabor needs


Results from extension messagesAn important technology introduced by extension was theuse of Mucuna puriens as a green manure cover crop toincrease soil organic matter content. It was observed,however, during the farm visits that few farmers usedMucuna. In general the farmers at the research siteacknowledged the effect ofMucuna to be remarkable dueto its capacity to increase biomass, hence organic mattercontent. Yields of maize increased considerably ifMucuna had preceded it. It was reported by scientistsinvolved in the project that one year of Mucuna produc-tion could be equated with five years of fallow. However,the use of Mucuna for improved soil fertility had majordrawbacks, according to the farmers interviewed such as(a) slow decomposition of stalks in the field; (b) inflam-mability of the dry Mucuna crop; (c) the likelihood oftenants losing their plots because owners withdraw themfrom the tenant’s use after a Mucuna season; high laborrequirements; and (d) the fact thatMucuna cannot be usedfor other purposes such as seed consumption, animalfodder or building material. Also, this part of the countryhas only one rainy season, so that farmers reported thatthey would ‘‘lose the season’’ if they cultivated Mucuna.One farmer explained, ‘‘If the rains come you have tochoose your crops, and since we cannot eat Mucuna, thechoice is quickly made.’’ Other technical problems in-cluded that Mucuna is difficult to plough under withanimal traction; it provided favorable environments forscorpions and snakes; it germinated abundantly the yearafter sowing due to bursting of pods, and it needed to be

protected from bush fires. The farmers explained their lowuptake of Mucuna through reasons other than technicalperformance and yield increases. Deffo et al. (2002,2004), Tarawali et al. (1999), and Manyong et al. (1996)reported similar problems with Mucuna. Several farmersexplained that they had already tried Mucuna before theproject came to their village and that they planned toabandon it once the project ended. The Mucuna storyshows that available technology that is ‘‘on the shelf’’according to scientists is not necessarily acceptable tofarmers. Figure 2 provides an overview of farmer reasonsinfluencing Mucuna uptake.

Extension activities also triggered individual discov-ery learning, demonstrated in this account of a criticalincident.

Some of the farmers who did not want to lose an entirecropping season to Mucuna tried intercropping Mucunawith maize. Others preferred intercropping sorghumwith Mucuna because of the cultural and social valueattributed to sorghum. Despite the project’s warningsthat Mucuna would overrun the sorghum, one farmer inSessaro decided to try this practice in his field. In fact,he had already concluded that his sorghum productionyield was lost and therefore could afford using that plotfor experimentation. During a field day a group offarmers visited his plot and concluded that Mucuna hadindeed largely overrun sorghum. Although scientistsconcluded the practice was not successful, one farmercommented, ‘‘Even though I agree that Mucuna

= Missing link

Introduction of Mucuna for ISFM

Improved Soil Fertility

Higher yields

Slow decomposition of stalks

Incidence of snakes and scorpions

Technical space

Land tenure arrangements

Labor availability

Economic profitability

Multi Purpose Crop

Socio-economical space

= Link

Figure 2. Linking technical and socio-economical space: the case of Mucuna.


overruns sorghum, the farmer improved his soil fertilityand in addition he has at least some sorghum which isalready better than Mucuna alone!’’The example demonstrates that stakeholders use dif-

ferent indicators to measure an innovation’s success.The project aimed to improve soil fertility through

ISFM strategies such as organic manure, chemicalfertilizers and cover crops. As a result of these practices,the production of maize, for example, increase asreported earlier. During our field visits, the major pre-occupation of farmers was the slump in maize prices.Even though farmers had a surplus of maize, it was notpossible for them to actually sell it. One farmer said, ‘‘Wefollowed your advice to improve our soil fertility, andnow here we are with our surplus. What are we supposedto do with our maize? You should not advise people toproduce more if you have no market for the produce.’’

Consequently, the Project introduced cereal banks.Cereal banks allow storage of produce immediately afterharvest when the price is low, to be held until the pricegoes up again. However, this final year, farmers com-plained that the price remained low year-round.13 Afteranalyzing the statistical data from FAOSTAT,14 it was

indeed confirmed that the price decreased by 25%compared to the previous year and 20% compared to thefive-year average. The fact that production had onlyincreased slightly (i.e., 1% compared to last year and12% compared to the five year average) was an indi-cation that more maize was available on the market. Thereason was likely due to the fact that maize production inneighboring countries increased by 21% compared withthe previous year. Also, cheap maize and other foodproduct imports year-round have undermined the sea-sonal rhythm on which the cereal bank was based.

Fertilizers are generally expensive,15 if available at all,and as a result production costs are high. In such a sit-uation, fertilizer application may increase yields, buthigher yields may not improve resource-poor farmers’livelihoods, if not accompanied by institutional or socio-economic development. One farmer stated, ‘‘We do notneed your help anymore to increase our production lev-els, because we attained that objective and know how todo that, but we want you to change your objective andhelp us with better markets for our yields.’’

The introduction of cereal banks did not prove to besufficient to overcome the marketing problem. Figure 3

Political space


Higher Maize yields

Technical space Socio-economical space

Marketing possibilities

Availability of fertilisers

Information aboutmarket prices


Higher maize yields

Credit availability

Cereal bank

Soil fertility Improvement through organic manure, fertilisers and cover crops

= Missing link= Link

Figure 3. Linking socio-economical, technical and political space.


gives a schematic overview of how different factors areinterlinked. One of the staff members of the internationalinstitute argued ‘‘The Project should not take increasedproductivity out of its context, but look at the entirecontext prerequisite for rural development that is net-working, knowledge and information, market develop-ment and credit.’’

Giller et al. (2003) discuss the different impacts thatpolicy can have (see Figure 4). In a discussion with agroup of farmers and the agricultural extension agent inSessaro about individual farm management advice, thefollowing concern was raised:The farmer you see over there does not want otherfarmers to visit his field during the farmer field day thisyear. During last year’s farmer field day he got a lot ofattention when visiting farmers were impressed by thehigh yields he was able to produce due to the soilfertility improvement strategies he had practiced.Shortly after the farmer field day he fell ill. The farmerclaims that one of the visiting farmers has bewitchedhim out of jealousy about the high yield he obtained.He dropped all the practices the individual farm man-agement advisor had recommended even though he wasconvinced of their benefit. The farmer spent all hismoney on medicines and has ,no force’ left to work inthe field. The farmer argues that it does not make senseto become rich if it means losing your health or maybeeven your life.

This example demonstrates that the often-madeassumption that individuals can progress if they make aneffort can be wrong in another social context. Levelingmechanisms that aim to ‘‘even out’’ the distribution ofwealth (Nanda, 1990) are very strong and imply thatpeople do not want to publicly display their wealth(Breusers et al., 1998). Farmers believe witchcraft to bereal in its consequences and capable of having an impacton their well-being. Leveling mechanisms can take manyforms (Nanda, 1990) and in this case resulted in a lack offollow-up on project recommendations.

Results of research experimentsExperiments on soybean to increase organic matter con-tent were conducted. Soybean is a non-traditional cropand is not often consumed in the rural communities in-volved in the study.16 An NGO contracted farmers toproduce soybean and guaranteed a ready market for theproduce. The presence of a market outlet motivatedfarmers to engage in soybean production. Moreover, theypreferred soybean to Mucuna. The advantage of soybeanis that it is a multi-purpose crop (food, fodder, organicmatter). In addition, soybean had market potential.Farmers were however primarily interested in soybeangrains and considered the soil fertility improvement effectas an advantageous side-effect, whereas soil scientistsmainly evaluated soybean on the basis of the biomassavailable for soil fertility improvement. A farmer

Figure 4. Political space further explored (adapted from Giller et al., 2003).


commenting on the soybean varietals trial observed, ‘‘Thesoybean variety soil scientists prefer is indeed high andgreen but that is not what we look for. The variety theychoose has only two grains per pod and the pods are onlynear the top, not covering the whole stalk.’’

One farmer who participated in a soybean trial had leftthe residues in the field for distribution over the land thenext day. In the evening his brother passed by andnoticed the heap of soybean residues. He decided toclean the plot for his brother and burned it. This illus-trates that it is not only the owner who makes decisionsabout the plot, but also other family members.

Farmers prefer cowpea, on which experiments werealso conducted. Cowpea can be planted very early androtated with maize during the same cropping season.Cowpea does not only have a positive effect on maizeyields but, more importantly, can be used to bridge thehunger gap (it can be harvested very early) and can alsobe sold on the local market. In addition, cowpea is amulti-purpose crop that fits more easily into the farmingcalendar due to its early harvest. The problem withcowpea is its susceptibility to pests. However, for theyear we reported on, very few farmers planted a cowpea-maize rotation trial because the rains came too late,leading the farmers to decide that it was impossible tosow cowpea before maize. In this example, climaticconditions led farmers to decide that the trial would notbe interesting. This illustrates, once again, that technicalsolutions are not fixed blueprints but depend on a com-plex mix of factors.

A farmer participating in the relay cropping Egusi-Maize on-farm trial in Goubi did not respect the fertilizerrate application stipulated in the protocol. He boughtadditional fertilizer, which he applied both to the treat-ment plot and to the control plots. Maize yields on thedifferent treatments plots motivated the researchers toinvestigate the situation. The farmer explained that hewanted high yields on his plots and applied what he hadlearned in the previous year from the project’s activities.

Concluding remarksScientists in the project focused on designing systemsthat were effective and focused less on ‘‘appropriatetechnologies,’’ The experiences with Mucuna probablyprovided the clearest example. The high yields obtainedwith Mucuna convinced scientists that farmers wouldaccept the crop, overlooking the fact that farmers usedadditional criteria to determine acceptability. Therefore itis necessary to consider farmers’ conditions for accept-ability. Another example was the early indication fromfarmers from Goubi during the diagnostic phase that soilfertility degradation was not a problem. Even when soilfertility is an issue, farmers may not use fertilizers if theyare either unavailable or considered to be too expensive.The case study demonstrates that farmers had veto power

(Roling et al., 2004) and will not use technologies that donot benefit them. Scientists determined what was (or wasnot) a solution for the farmers, but did not take intoaccount that farmers may have had other priorities andthat increased production was not necessarily amongstthem, depending on other conditions. Whereas the soilscientists involved in the project used yields and bio-mass content as the main indicators of success, farmersused a mix of criteria both in technical and socio-eco-nomic domains in which yield was only one factor, albeitan important one.

The farmers selected for participation in the projectwere predominantly the relatively resource-rich farmersin the community due to increased chances that suchfarmers would adopt the strategies introduced. Thesefarmers complained that their main problem was thecommercialization of their produce. It could well be thatthe project targeted the wrong farmers.17

The farmers we spoke to complained about the lack ofmarket and low prices. This also demonstrates that newproblems arise as a result of solving old ones. Therefore acontinual assessment of the context is required if the aim isto design research that is useful for resource-poor farmers.

Innovations that can be scaled up

‘‘Scaling up’’ usually means moving beyond the local orpilot level to include more beneficiaries and to involveother agencies in terms of management and funding, witha view to making a larger impact that is beyond theoriginal project’s duration (while continuing to fosterparticipation of intended beneficiaries and other stake-holders involved) (Uvin and Millar, 1994; Douthwaiteet al., 2003). What lessons can be learned about what theProject has done to scale up its pilot activities?

Intra-institutional collaborationA platform was constituted to involve different institu-tional actors and thereby influence the political scene forscaling-up purposes. A Project Co-ordination Committee(PCC) was established to facilitate the implementation ofthe project, guide the research process, and make itpossible to continue activities in the project area. It washoped that this PCC would eventually carry out theactivities by itself. A platform perspective was used(Dangbegnon et al., 2001) to set up such a collaborativedevelopment and research framework. The PCC con-sisted of the national partners in research, extension, andagricultural policy analysis; and the internationalorganization. As such, the PPC helped to inform andinvolve policy makers at the level of the Central Regionof Togo. Farmers and their organizations were not rep-resented. The PCC was designed as an inter-institutionalplatform and functioned as a forum for negotiationbetween the national stakeholders (see Figure 5).18


A difficulty encountered by the project was the highturnover of the PCC members; in the end, none of theoriginal members remained involved. The new membershad to be briefed about the history, norms and rules of theplatform. Another problem was the lack of communica-tion between the people on the platform and the membersof the organization they represented. Collective com-mitment to continuation of the platform and the projectactivities was not evident.

Spontaneous scaling up‘‘Informal’’ or ‘‘spontaneous’’ horizontal scaling upthrough diffusion had taken place. Such a process isalso referred to in the literature as ‘‘scaling out’’(Douthwaite et al., 2003). Several farmers fromsurrounding villages or non-participating farmers fromthe same village reported, for example, having receivedseed from fellow farmers or having heard about theproject from a neighbor. These farmers also experi-mented before adopting the practices demonstrated. Inaddition, during the farmer field day, it was observedthat farmers from other villages who had not receivedinvitations also participated. There had been noincentive such as transport reimbursement. A criticalstaff member of the international institute wonderedwhether they should continue to invest in trials,given that farmers are capable of experimentingindependently.

Other dimensions of scaling upScaling up has, however, another dimension. The casestudy clearly showed that technical aspects (e.g., soilfertility, productivity per ha) were but one bottleneck indevelopment. Once they were overcome, other issues

such as marketing become the limiting factor. Sponta-neous diffusion of technical innovations, the greatmultiplier of research impact (e.g., Rogers, 1995),requires a favorable context. This includes availabilityof and access to credit facilities, security of land use,timely access to high quality fertilizers, appropriate andfunctioning extension services, and most importantly theavailability of and access to marketing channels andacceptable prices.

Concluding remarksFor scaling up, the notion of replicability is essential.One cannot expect the project process to be replicated ifthe conditions created by the project were artificial.Several factors impact on scaling up. In the first place,partnerships that allow for interaction and ownershipamong different stakeholders need to be built so as tofoster internalization of the project objectives. Secondly,the political context must be favorable. Finally, scalingout requires that diversity in the farmer population betaken into account. Resource-poor farmers do not nec-essarily have the same options as large farmers. Whenwomen do the farm work, it is important that they beincluded.

Emerging lessons

The case study analysis, based on five criteria for farmer-oriented research, allowed us to draw a number oflessons regarding agricultural research that is effective inameliorating the livelihoods of resource-poor farmers.During fieldwork, farmers mentioned on severaloccasions that the research conducted does not always

PCC

Extension service - focus on extension activities for innovations already

available

Research service - focus on explorationof new innovations

Agricultural policyanalysts- Monitoringand coordination of

activities

Internationalinstitute facilitating

the processes

Figure 5. An inter-institutional platform in the Soil Fertility Project.


address their needs and that research results do notroutinely benefit them. What explains the (lack of)impact of the ISFM Project on farmers’ livelihoods? Byway of conclusion we assert the following.

Pre-analytical choices determined the scope and possibleimpact

Pre-analytical choices were made by donors and theinternational research institute but were insufficientlynegotiated with resource-poor farmers. The project wasnot used to develop effective farmer clout over theproject’s activities. Consequently, ownership of theactivities remained with the scientists and planners, withthe result that the technologies introduced and tested didnot adhere to the criteria outlined at the beginning of thearticle and will not autonomously diffuse amongstfarmers.

Solutions that are technically sound are not necessarilyacceptable to resource-poor farmers

The assumption that science has technologies on theshelf that can solve farmer problems is particularlydangerous. In this case study, scientists decided thatexisting technologies (for example the use of Mucunato increase organic matter content) provided the besttechnical means to solve soil fertility problems. Thisdecision was based on a test of ‘‘what worked’’ andnot ‘‘what was acceptable.’’ However, scientists deci-ded what works on the basis of narrow criteriae.g., Mucuna’s contribution to organic matter contentand yield. Farmers used a more complex set of criteriasuch as labor costs, production of edible grains, and soforth.

Solving a problem generated other problems

When some farmers improved soil fertility and increasedproduction, they could not market their produce. Thismeans that agricultural research must continually engagein diagnosis of the relevant problems. Also, the contextchanges (e.g., prices for agricultural products rise andfall), affecting the relevance of project activities.

Appropriate categories of farmers need to be addressed

Farmers are not homogenous in terms of needs andperceptions e.g., soil fertility is an urgent problem forresource-poor farmers who do not have the labor or thefinancial means available to improve soil fertility, hencehave limited ‘‘space for innovation.’’ However, the pro-ject seemed to focus on farmers with the means and laborto deal with soil fertility problems, but who instead facedmarket challenges (for their surplus production). In

addition, other procedures to involve women need to beexplored.

The farmers revealed that a complex mix of socio-eco-nomic, political and technical factors influenced uptakeor rejectionFarmer uptake can only be understood if the socio-cul-tural context is taken into account. Cultural values are toooften overlooked, yet have a major impact on howresource-poor farmers perceive agriculture and theusefulness of the technologies proposed.

Summary

Many questions remain unanswered when it comes to thecritical issue of making agricultural research relevant toachieving the Millennium Development Goals. But ourcase study demonstrates that it is not enough to developsystems that work. Farm innovation needs to beembedded in macro-level opportunities, grounded inresource-poor farmers’ needs, be acceptable to them, andallow for scaling up.

Acknowledgements

We would like to acknowledge the different stakeholdersand farmers who dedicated their precious time in order todo interviews with us. We would also like to acknowl-edge the International Institute for allowing us to lookcritically at its work and access confidential documents.It is because of this open attitude that we managed tolearn so much. We are pleased to know that an earlierdraft of our work also led to constructive discussionwithin the Institute. The first author was employed by theFood and Agriculture Organization (FAO) through theDutch Ministry for Development Cooperation (DGIS) atthe time the study was carried out and she gratefullyacknowledges this support. We acknowledge the con-structive comments of Niels Roling, Arnold van Huis,Aubert Ouango, Thom Kuyper and Ken Giller on earlierdrafts. The views expressed are nevertheless our own.

Notes

1. Convergence of Sciences (CoS) is a research program(2002–2006) that has been executed by a consortiumof the Universite de d’Abomey-Calavi, Benin; theUniversity of Legon, Ghana; and WageningenUniversity in the Netherlands. Within the programeight African students worked with a group of farmerswho developed technologies in such areas as land useand soil fertility, weed management and plant genetic


diversity for food crops, and integrated pestmanagement. They also experimented with ways ofimproving the framework conditions for technologicalinnovation (Roling et al., 2004; Hounkonnou et al.,2006; A. van Huis et al., submitted). The first authorwas conducting PhD research within the frame of theCoS program to draw comparative lessons from theeight farmer-scientist interactions, in order to throwlight on research procedures, methodologies, andprocesses that were assessed for their effectiveness inimproving the livelihoods of resource-poor farmers.The approach CoS used is based on a convergencebetween scientists and farmer and between biologicaland social scientists and comprises different phases(see also E. S. Nederlof et al., submitted).

2. For purposes of confidentiality the name of the projecthas been changed.

3. Roling used insights from Van Schoubroeck (1999)and Tekelenburg (2001).

4. People participate in a joint analysis that might lead toaction planning and the formation or strengthening oflocal institutions. It tends to involve interdisciplinarymethodologies that seek multiple perspectives andmake use of systematic and structured learning pro-cesses.

5. Individual Farm Management Advice is better knownunder its French name: Conseil de Gestion. It is aholistic extension tool for an individual householdand aims to improve production levels and soil fer-tility strategies. It consists of a series of visits by theextension agent to the farmer aimed at giving her orhim farm-specific advice. Results are encouraging butthe scope for scaling up is limited due to the highcosts involved and the low farmer-extension workerratio.

6. A cereal bank (Kpaikpai, 2003) consists of a group ofpeople who agree to jointly store their products whenprices are low and the product is abundant. Whenthe product becomes scarce and prices increase, theproduct is sold. This not only allows selling theproduct at a higher price, but since prices are ingeneral higher just before the next season starts, italso makes cash available at a time it is needed for thestart of the next season (e.g., for purchasing agricul-tural inputs, hiring labor). Well-known drawbacks ofcereal banks are that the moisture content of thegrains, and as a result the weight, decreases; indi-viduals might want to sell the bags at other times;costs for a care taker; and other additional costs thatare not foreseen by farmers and may result in accu-sation of stealing (Roling, pers. comm.).

7. These experiments were based on contracts that weremade between farmers and scientists during thediagnosis.

8. Given the institutional complexity of the Project, the

selection could have been influenced by any of anumber of factors, including personal reasons ofofficials.

9. Tools used were community mapping, transect walk,semi-structured interviewing, flow diagram, wealthranking and Venn diagramming.

10. When the agricultural extension agent was harvest-ing the soybean density plot together with the farmeron whose land the trial had been installed, the farmerbecame tired and left. He told the agriculturalextension agent he would take a rest and return buthe never came back and left the agricultural exten-sion agent alone with the work. This was the farm-er’s way of saying he did not agree with the mannerin which they had to harvest the field for the scien-tists. The agricultural extension agent followed theevaluation form handed out by the research service,which included an indicator on the number of grains.To estimate the number of grains, the soybean had tobe threshed, what usually is not done. The farmer didnot understand the logic of doing this additionalwork and decided to withdraw. It later appeared thatthe agricultural extension agent indeed misunder-stood the requirement. The case shows that farmersnot only negotiate through discussion but alsothrough action. In this example indicators forevaluating the trial were not negotiated, and as aresult the experiment turned out the agriculturalextension agent’s responsibility rather than thefarmer’s.

11. Plots of similar size and different applications wereharvested and yields put in a bag. The bags werecompared to determine best practices.

12. During a field trip to discuss some preliminary find-ings of the project activities of the preceding seasonwith the farmers, a young researcher was discouragedby the low uptake of some of the strategies proposed.The discouragement was very evident from theexpression on his face. When the other researchersreturned to the car, we overheard them commentingon the young researcher. One of them said: ‘‘poorguy, he still believes farmers will one day adopt whatresearchers introduce’’, and another added: ‘‘once hehas more experience he will learn that making achange for farmers is no more than a dream’’. Theresearchers continued commenting on the young re-searcher stressing that no research project they haveever seen has made an impact.

13. The problem of maize price slump seemed to be aregional one. Farmers had the same complaints inthe Central region of Benin (observation inS. Aliou’s field on 17.10.03) and in Burkina Faso.

14. FAOSTAT data available online at http://fao-stat.fao.org/site/340/default.aspx

15. The cost is the equivalent of 15.5 US dollars,


e.g., 7500 FCFA for a bag of 50 kilos. See http://www.oanda.com/convert/fxhistory.

16. At them time of writing, however an NGO demon-strated how soybean cheese and mustard can beproduced and soybean is gaining popularity forhome consumption.

17. Thom Kuyper (personal communication).18. The concept of ‘‘platform’’ originally emerged in the

context of natural resource management (Roling andJiggins, 1998). Currently it has, however, a muchwider applicability.

References

Adjei-Nsiah, S., C. Leeuwis, K. E. Giller, O. Sakyi-Dawson, J.Cobbina, T. W. Kuyper, M. Abekoe, and W. van der Werf(2004). ‘‘Land tenure and differential soil fertility managementpractices among native andmigrant farmers inWenchi, Ghana:Implications for interdisciplinary action research.’’ NJAS -Wageningen Journal of Life Sciences 52: 331–348.

Ashby, J. A., A. R. Braun, T. Gracia, M. P. Guerrero,L. A. Hernandez, C. A. Quiros, and J. I. Roa (2000).Investing in Farmers as Researchers: Experience withLocal Agricultural Research Committees in Latin America.Colombia: Centro Internacional de Agricultura Tropical(CIAT). Retrieved from http://www.ciat.cgiar.org/down-loads/pdf/Investing_farmers.pdf on 2 August 2006.

Bie, S. W. (2001). ‘‘The 1990s: An important decade foragricultural research in developing countries.’’ In Interna-tional Service for National Agricultural Research. AnnualReport 2000: Reflecting on an Important Decade for Agri-cultural Research in Developing Countries. The Hague:International Service for National Agricultural Research..

Bjornsen Gurung, A. (2003). ‘‘Experiments in store: Participa-tory Action Research with women farmers of Nepal.’’ Bera-terinnen News. Lindau, Switzerland: LandwitrschaftlicheBeratungszentrale 1: 31–38.

Brabant, P., D. Darracq, K. Egue, and V. Simonneaux (1996).Togo: Etat de Degradation des Terres Tesultant des ActivitesHumaines. Paris: Editions de l’ORSTOMCollection NoticeExplicative No. 112.

Breusers, M., E. S. Nederlof, and T. van Rheenen (1998).‘‘Conflict or symbiosis? Disentangling farmer-herdsmenrelations: The Mossi and Fulbe of the Central Plateau,Burkina Faso.’’ Journal of Modern African Studies 36(3):357–380.

Castillo, G. T. (1998). ‘‘A social harvest reaped from a promiseof springtime: user-responsive, participatory agriculturalresearch in Asia.’’ In N. Roling and M. Wagemakers (eds.),Facilitating Sustainable Agriculture: Participatory Learningand Adaptive Management in Times of EnvironmentalUncertainty, (pp. 191–214). Cambridge, United Kingdom:Cambridge University Press.

Chambers, R. (1990). Developpement Rural: La PauvreteCachee. Paris: Editions Karthala and Wageningen, TheNetherlands: CTA.

Chambers, R. (1994). ‘‘Participatory Rural Appraisal (PRA):Analysis of experience.’’ World Development 22(9): 1253–1268.

Chambers, R. (1997). Whose reality counts? Putting the lastfirst. London: Intermediate Technology Publications.

Chambers, R. and J. Jiggins (1987a). ‘‘Agricultural research forresource-poor farmers. Part I: Transfer-of-Technology andFarming Systems Research.’’ Agricultural Administrationand Extension 27: 35–52.

Chambers, R. and J. Jiggins (1987b). ‘‘Agricultural research forresource-poor farmers: A parsimonious paradigm.’’ Agricul-tural Administration and Extension 27: 109–128.

CIP (International Potato Center) (2003). Farmer FieldSchools: From IPM to Platforms for Learning and Empow-erment: Users’ Perspectives with Agricultural Research andDevelopment. Los Banos, Laguna, Philippines: InternationalPotato Center.

Collinson, M. P. (ed.) (2000). A history of Farming SystemsResearch. New York, New York: CABI Publishing andRome, Italy: Food and Agriculture Organization,.

Dangbegnon, C., A. Blum, E. S. Nederlof, N. Roling, andR. C. Tossou (2001). Platforms for Sustainable NaturalResource Management: The Case of West Africa. Amster-dam: NIRP, KITResearch for Policy Series No. 10.

Deffo, V., S. Hounzangbe-Adote, R. Maliki, M.E.H. OuldFerroukh, and E. Torquebiau (2004). ‘‘Contraintes contin-gentes a l’adoption d’une technologie. Evaluation contin-gente des contraintes a la fertilisation des sols par le systememaıs-mucuna-engrais mineraux au Sud Togo.’’ Tropicultura22(1): 19–25.

Deffo, V., S. Hounzangbe-Adote, R. Maliki, M. E. H. OuldFerroukh, E. Torquebiau, and H Van Reuler (2002). ‘‘Optionsd’intensification durable des cultures vivrieres au Sud duTogo.’’ African Crop Science Journal 10(3): 239–249.

Denzin, N. K. and Y. S. Lincoln (eds.). (1994). Handbook ofqualitative research. Thousand Oaks, California: Sage.

Dixon, J., A. Gulliver, and D. Gibbon (2001). Farming systemsand poverty: Improving farmer’s livelihoods in a changingworld. Rome: FAO and Washington, DC: World Bank.

Douthwaite, B. (2002). Enabling Innovation: A Practical Guideto Understanding and Fostering Technological Change.London: Zed Books.

Douthwaite, B., T. Kuby, E. van de Fliert, and S. Schulz (2003).‘‘Impact pathway evaluation: An approach for achieving andattributing impact in complex systems.’’ Agricultural Systems78: 243–265.

Funtowicz, S. O. and J. R. Ravetz (1993). ‘‘Science for thepost-normal age.’’ Futures 25(7): 739–775.

Giampietro, M. (2003). Beta-Gamma Science for SustainableAgriculture: Taking the Implications of Complexity Seriously.PhD dissertation. Wageningen University, The Netherlands.

Giller, K., N. Koning, and P. Richards (2003). ‘‘Bumping alongthe bottom: The lack of resilience in smallholder Africanfarming.’’ Paper presented at the International Conference onLocal Land Use Strategies in a Globalizing World: ShapingSustainable Social and Natural Environments. August 2003,Copenhagen.

Guba, E. G. and Y. S. Lincoln (1989). Fourth GenerationEvaluation. London: Sage.


Hounkonnou, D., D. Kossou, T. W. Kuyper, C. Leeuwis,P. Richards, N. Roling, O. Sakyi-Dawson, and A. vanHuis (2006). ‘‘Convergence of Sciences: The manage-ment of agricultural research for small-scale farmers inBenin and Ghana.’’ NJAS - Wageningen Journal of LifeSciences 53(3–4): 343–367.

Jiggins, J. and H. de Zeeuw (1992). ‘‘Participatory technologydevelopment in practice: Process and methods.’’ In C.Reijntjes, B. Haverkort, and A. Waters-Bayer (eds.),Farming for the Future: An Introduction for Low-ExternalInput and Sustainable Agriculture, (pp. 135–162). London:Macmillan and Leusden, Belgium: ILEIA.

Johnson, N. L., N. Lilja, and A. Ashby (2003). ‘‘Measuringthe impact of user participation in agricultural andnatural resource management research.’’ Agricultural Systems78: 287–306.

Kpaikpai, K. K. (2003). Guide pour la Constitution et laGestion d’une Banque de Cereals. Togo, West Africa: Institutde Conseil et d’Appui Technique.

Manyong, V. M., A. V. Zoundekon, A. Gogan, M. N. Versteeg,and F. van der Pol (1996). ‘‘Determinants of adoption for aresource management technology.’’ In Z. Senwen and W.Yunlong (eds.), Proceedings of Advances in Agricultural andBiological Environment Engineering. August 15–19, (pp.186–193). Beijing, China: China Agricultural UniversityPress.

Mazzucato, V. and D. Niemeijer (2000). Rethinking Soil andWater Conservation in a Changing Society: A Case Study inEastern Burkina Faso. PhD. Dissertation, WageningenUniversity, The Netherlands.

Morss, E. (1976). Strategies for Small Farmer Development:An Empirical Study of Rural Development Projects. Boulder,Colorado: Westview Press.

Mutimba, J. K. (1997). Farmer Participatory Research: AnAnalysis of Resource-poor Farmer Involvement in, andContribution to, the Agricultural Research Process inZimbabwe. PhD dissertation, Faculty of Agriculture,University of Zimbabwe, Harare, Zimbabwe.

Mutsaers, H. J. W., G. K. Weber, P. Walker, and N. M. Fischer(1997). A Field Guide for On-Farm Experimentation. Ibadan,Nigeria: International Institute of Tropical Agriculture.

Nanda, S. (1990). Cultural Anthropology. 4th edition, Belmont,California: Wadsworth Publishing Company.

Nederlof, E. S. (2003). Proposal: Research on AgriculturalResearch: Enhancing Innovative Performance in SustainableLand Use Management in West Africa. Case Studies fromGhana and Benin. [Granted by the Dutch Organisation forScientific Research NWO]. Unpublished report.

Pieri, C. (1989). Fertilite des Terres de Savanes: Bilan deTrente Ans de Developpement Agricoles au Sud du Sahara.Paris, France: Ministere de la Cooperation et CIRAD-IRAT.

Pontius, J., R. Dilts, and A. Bartlett (2002). From Farmer FieldSchool to Community IPM: Ten Years of IPM Training inAsia. Bangkok, Thailand: Food and Agriculture Organizationof the United Nations, Community IPM Program.

Pound, B., S. Snapp, C. McDougall, and A. Braun (2003).Managing Natural Resources for Sustainable Livelihoods:Uniting Science and Participation. London: EarthscanPublications.

Pretty, J.N. (1994). ‘‘Alternative systems of inquiry for a sus-tainable agriculture.’’ In S. Davies (ed.), Knowledge isPower? The Use and Abuse of Information in Development,(pp. 37–48). Brighton, UK: University of Sussex, Institute ofDevelopment Studies.

Pretty, J. N. (1995). Regenerating Agriculture: Policies andPractice for Sustainability and Self-reliance. London:Earthscan.

Pretty, J. N., I. Guijt, I. Scoones, and J. Thompson (1995). ATrainer’s Guide for Participatory Learning and Action.London, United Kingdom. International Institute forEnvironment and Development Brighton, United Kingdom:University of Sussex, Institute of Development Studies, IIEDParticipatory Methodology Series.

Reijntjes, C., B. Haverkort, and A. Waters-Bayer (1992).Farming for the Future: An Introduction to Low-ExternalInput and Sustainable Agriculture. The Netherlands: ETCand ILEIA.

Rhoades, R. E., and R. H. Booth (1982). ‘‘Farmer-back-to-farmer: A model for generating agricultural technology.’’Agricultural Administration, London: Overseas DevelopmentInstitute 1: 127–137.

Rogers, E. M. (1995). Diffusion of Innovations. 4th edition,New York: Free Press.

Roling, N. (1988). Extension Science. Cambridge, UnitedKingdom: Cambridge University Press.

Roling, N. G. and J. Jiggins (1998). ‘‘The ecological knowledgesystem.’’ In N. G. Roling and M. A. E Wagemakers (eds.),Facilitating Sustainable Agriculture: Participatory Learningand Adaptive Management in Times of EnvironmentalUncertainty, (pp. 283–311). Cambridge, UK: CambridgeUniversity Press.

Roling, N. G., D. Hounkonnou, S. K. Offei, R. C. Tossou, andA. van Huis (2004). ‘‘Linking science and farmers’ innova-tive capacity: Diagnostic studies from Ghana and Benin.’’NJAS - Wageningen Journal of Life Sciences 52: 211–236.

Saıdou, A., T. W. Kuyper, D. Kossou, R. Tossou, andP. Richards (2004). ‘‘Sustainable soil fertility manage-ment in Benin: Learning from farmers.’’ NJAS - Wa-geningen Journal of Life Sciences 52: 349–370.

Scheuermeier, U., E. Katz, and S. Heiland (2004). Finding NewThings and Ways that Work: A Manual for Introducing Partic-ipatory Innovation Development (PID). Lindau, Switzerland:Swiss Center for Agricultural Extension (LBL)..

Steiner, J. L., J. C. Day, R. I. Papendick, R. E. Meyer, and A. R.Bertrand (1988). ‘‘Improving and sustaining productivity indryland regions of developing countries.’’ Advances in SoilScience 8: 79–122.

Stoop, W. A. (2002). A Study and Comprehensive Analysis ofthe Causes for Low Adoption Rates of Agricultural ResearchResults in West and Central Africa: Possible SolutionsLeading to Greater Future Impacts: The Mali and GuineaCase Studies. Rome: FAO Interim Science Council andCGIAR.

Stoorvogel, J. J. and E. M. A. Smaling (1990). Assessment ofSoil Nutrient Depletion in Sub-Saharan Africa: 1983–2000.Volume 1: Main Report. 2nd edition, Wageningen, TheNetherlands: The Winand Staring Centre.

Stroosnijder, L. (1992). ‘‘Desertification in Sahelian Africa.’’The Courier 133: 36–39.


Stroosnijder, L. and T. van Rheenen (eds.) (2001). ‘‘Agro-silvopastoral land-use in Sahelian villages.’’ Advances in Geo-ecology 33. Catena Verlag, Reichskirchen, Germany.

Swift, J. (1996). ‘‘Desertification: Narratives, Winners andLosers.’’ In M. Leach and R. Mearns (eds.), The Lie of theLand: Challenging Received Wisdom on the African Envi-ronment: African Issues, Portsmouth, New Hampshire:Heinemann.

Tarawali, G., V. M. Manyong, R. J. Carsky, P. Vissoh, P. Osei-Bonsu, and M. Galiba (1999). ‘‘Adoption of improved fal-lows in West Africa: Lessons from Mucuna and stylo casestudies.’’ Agro-forestry Systems 47: 93–122.

Tekelenburg, A. (2001). Cactus Pear and Cochineal inCochabamba: The Development of a Cross-epistemologicalManagement Toolkit forIinteractive Design of Farm Inno-vation. PhD dissertation. Wageningen University, The Neth-erlands..

Uvin, P. and D. Miller (1994). Scaling Up: Thinking Throughthe Issues. New York, New York: Global Policy Forum.Retrieved from http://www.globalpolicy.org/ngos/role/intro/imp/2000/1204.htm#1 on December 5, 2003.

van de Fliert, E. (1993). Integrated Pest Management: FarmerField Schools Generate Sustainable Practices. A Case Studyin Central Java Evaluating IPM Training. PhD dissertation,Wageningen University, The Netherlands.

van de Fliert, E. E. and A. R. Braun (2002). ‘‘Conceptualizingintegrative, farmer participatory research for sustainableagriculture: From opportunities to impact.’’ Agriculture andHuman Values 19(1): 25–38.

van Schoubroeck F. (1999). Learning to Fight a Fly: Devel-oping Citrus IPM in Bhutan. PhD dissertation, WageningenUniversity, The Netherlands..

van Veldhuizen, L., A. Waters-Bayer, and C. Wettasinha(2003). ‘‘Advancing PTD: Making our way towards institu-tional integration.’’ In C. Wettasinha, L. van Veldhuizen, andA. Waters-Bayer (eds.), Advancing Participatory TechnologyDevelopment-Case studies on Integration into AgriculturalResearch, Extension and Education, Wageningen, TheNetherlands: Technical Centre for Agricultural and RuralCooperation (CTA).

Vereijken, P. A. (1999). ‘‘Methodological way of prototypingintegrated and ecological arable farming systems (I/EAFS) ininteraction with pilot farms.’’ In C. Leeuwis (ed.), IntegralDesign: Innovation in Agriculture and Resource Manage-ment, Mansholt studies 15, Wageningen/Leiden, TheNetherland: Mansholt Institute/Backhuys Publishers.

Werner, J. (1996). Developpement Participatif d’InnovationsAgricoles: Procedures et Methods de la Recherche en MilieuPaysan. Germany: Deutsche Gesellschaft fur TechnischeZusammenarbeit (GTZ).

Address for correspondence: Suzanne Nederlof, Julianastraat78, 6707DH, Wageningen, The NetherlandsPhone: +31-0317420237; Fax: +31-0205688444E-mail: [email protected]


lessons for farmer-oriented research: experiences from a ... · lessons for farmer-oriented...

Documents