11

Processes Influencing the Successful Adoption of New Technologies by Smallholders

R.A. Cramb1

Abstract

Many studies have examined the particular farm-level factors affecting the adoption of newtechnologies by smallholders. Using an ‘actor-oriented perspective’, this paper focuses on the largerprocesses at work in the development, dissemination and adoption of improved technologies. Threecase studies of upland soil conservation projects in the Philippines are used to illustrate theargument that successful adoption depends on more than careful planning in research and the useof appropriate methodologies in extension. It depends on the timely formation of coalitions of keyactors whose interests converge sufficiently that they can focus their resources and efforts onachieving change in agricultural systems. It also depends on critical external factors that are largelyunpredictable. Newer approaches such as ‘participatory technology development’ are based on anappreciation of the evolving, adaptive and inherently participative nature of agricultural develop-ment processes. However, a broader, more flexible approach is needed which gives explicit recog-nition to the personal, cultural and political dimensions of coalition-building for technologydevelopment.

AN UNDERSTANDING of the processes leading to theadoption of new technologies by smallholders haslong been seen as important to the planning andimplementation of successful research and extensionprograms. This paper draws on observations andexperiences gained during the SEARCA-UQ UplandsResearch Project (ACIAR Project 9211), which usedsurveys, case studies, participatory appraisal tech-niques, and bio-economic modelling to investigatethe factors affecting the adoption of recommendedsoil conservation technologies by upland farmers invarious locations in the Philippines (Cramb in press).The principal technologies encountered in the fieldwere variants of Sloping Agricultural Land Tech-nology (SALT) (Partap and Watson 1994), involvingthe cultivation of annual crops such as maize in alleysbetween contour hedgerows, usually of multipurposeshrub legumes such as Leucaena and Gliricidia. Theprincipal means of promoting these technologies wasthe Integrated Social Forestry Program (ISFP) of theDepartment of Environment and Natural Resources

(DENR), the agency which has jurisdiction over theextensive Public Forest Lands where most uplandfarming occurs.

At one level, that of the individual farm household,the results of the SEARCA-UQ project were unsur-prising. A number of farm-household factors wereassociated with adoption, such as the age, education,and personal characteristics of the household head;the size, location and tenure status of the farm; theavailability of cash or credit for farm investment;access to urban markets; and so on (Cramb andNelson 1998; Cramb et al. 1999; see also Garcia1997; Pandey and Lapar 1998). However, at thevillage level and beyond, more interesting and signifi-cant issues arose: Why was there widespreadadoption in one village but not others in the samegeneral location? Why did one project lead toapparently successful adoption, but another,following the same procedures and promoting thesame technologies, result in failure? Answers to thesequestions are likely to be more useful in achievingwidespread agricultural development, particularly inwhat have been termed the ‘complex, diverse andrisk-prone’ farming systems of the uplands.

1University of Queensland, St Lucia, Qld 4072, Australia.Email: r.cramb@mailbox.uq.edu.au

12

The approach taken in this paper is to focus onthese higher-order factors affecting successfuladoption of technologies. Drawing on the ‘actor-oriented perspective’ in rural sociology (Long andLong 1992), it is argued that successful examples ofadoption at this higher level are not merely a functionof the technology, nor of the research and extensionmethodology, but result from a complex conjunctionof people and events, with outcomes which may havebeen quite unanticipated at the outset. From thisperspective, research and extension projects and pro-grams are viewed as arenas in which social actors –village leaders, farmers, researchers (local and inter-national), aid officials, municipal agents, extensionworkers, traders, etc – each pursuing their own short-and long-term objectives and strategies, manoeuvre,negotiate, organise, cooperate, participate, coerce,obstruct, form coalitions, adopt, adapt, reject, allwithin a specific geographical and historical context.1

Out of this process, improved technology may bedeveloped, disseminated, and incorporated infarming systems, and many of the actors may bemade better off as a result (farmers may earn moreincome or gain more prestige, landowners may con-serve their soil, extension agents may be rewarded,researchers may be applauded, published, even pro-moted, traders may do more business, mayors may bere-elected, aid officials may achieve their targets).However, there is nothing predetermined about thisoutcome. Hence a detailed, ‘all things considered’,case history approach is needed to understand andexplain the patterns of success in achieving beneficialtechnical change.

The paper aims to illustrate the fruitfulness of thisapproach with three case studies undertaken duringthe SEARCA-UQ Project. The case studies all relateto village-level upland development projects whichhad been previously identified as ‘successful’ interms of farmer adoption of a variant of SALT orother conservation measures. The projects werelocated at Domang, Nueva Vizcaya, in NorthernLuzon; Guba, Cebu, in the Visayas; and Managok,Bukidnon, in Mindanao.

Before proceeding to the case studies, however, itis useful to review briefly the way we think about thedevelopment, dissemination, and adoption of agricul-tural technologies, from an actor-oriented perspec-tive. This review highlights some important themeswhich help in the interpretation of the case studies,and hopefully of other experiences reported in thisvolume.

Rethinking the Development, Dissemination and Adoption of Agricultural Technologies

(a) Technology developmentIn the conventional or ‘central source’ view of agri-cultural research and development, technologyemanates from ‘upstream’ activities in the formalresearch system and is adapted by ‘downstream’research until it is ready for dissemination to farmers(Biggs and Clay 1981; Biggs 1990). Anderson andHardaker (1979) use an analogy from homeeconomics rather than hydrology, speaking ofquarter-baked (notional), half-baked (preliminary)and fully-baked (developed) technology. In a similarfashion, Scherr and Muller (1991) refer to thedevelopment of agroforestry technologies in terms ofexperimental, prototype and off-the-shelf tech-nologies. All of these analogies imply a linearprocess of technology development and dissemi-nation, culminating in the adoption of new tech-nologies by farmers.

However, in practice, as Biggs (1990) demon-strates, agricultural innovations derive from multiplesources, not only from the laboratories and researchstations of the national and international centres.These sources include research-minded farmers,innovative research practitioners at the local level,research-minded administrators, NGOs, private cor-porations, and extension agencies. In the ‘multiplesource’ model ‘technology is made up of many oldand new components; it has evolved and beenmodified over time, and will continue to do so’(Biggs 1990:1487). Consequently, and in contrast tothe notion of ‘transfer of technology’, there is no‘unambiguous, one-way progression in the research,extension and adoption process’ (Biggs 1990:1481)

One implication of this perspective is that theprocess of technology adaptation cannot be separatedfrom the process of technology adoption by farmers(discussed below). As Anderson (1993) notes, adop-tion and adaptation are intertwined in that adaptationof the technology frequently occurs in the process ofimplementing it on-farm – a phenomenon whichRogers (1995) terms ‘reinvention’. Sumberg andOkali (1997) go further, arguing that such adaptationis the norm, resulting from an on-going process of‘farmer experimentation’. This experimentation isnot confined to a few research-oriented farmers, butis the process by which almost all farmers incor-porate technology into their farming systems. Tech-nology supplied by the formal research andextension system thus becomes ‘raw material’ forfarmer experimentation (Sumberg and Okali 1997).In terms of the above analogies, rather than acquiringa ‘fully-baked’ technology ‘off-the-shelf’, farmerscan be viewed as shopping around for ‘ingredients’

1In this sense, research and extension interventions havealways been ‘participatory’.

13

or technological components which they incorporateinto their own recipes. In other words, technology isonly fully developed or adapted as part of a specific,operational farming system.

Thus, from an actor-oriented perspective, tech-nology development is a complex, multi-stranded,and multi-directional process, involving many actorsother than scientists in the formal research system.Moreover, as Biggs and Smith (1998) argue, theemergence of a particular technology depends notonly on its scientific merits but on the actions of whatthey term ‘development coalitions’ – loose groupingsof actors who combine their resources to push for aparticular path of technical change. Hence, while it isappropriate to evaluate a given technology in itself,‘the result is a necessarily incomplete account of therequirements for ‘successful’ technology develop-ment and dissemination. In addition to good luck, thelatter typically involves networking, advocacy,lobbying and other activities – here called coalitionbuilding – which are mainly excluded from conven-tional technology development narratives’ (Biggsand Smith 1998:6). These authors cite the example ofthe wheat ‘green revolution’ in India, based on thehighly risky importation of quantities of largelyuntested exotic wheat seed from Mexico. There wasdebate in the 1960s concerning the desirability of thisapproach, but a ‘coalition of scientific, farmer, donor,administrator, and political actors who were com-mitted to advocating and promoting a particular typeof science and technology strategy, focused on dwarfwheats,’ prevailed (Biggs and Smith 1998:5).

The complex pathway of technology developmentis evident in Harold Watson’s first-hand account ofthe development of SALT, a long and involvedprocess which began with the attempts of a church-based NGO to find practical solutions to farmers’problems in the remote uplands of Mindanao:

Obviously, a farming technology that could con-serve the topsoil and, if possible, improve its fertilityand productivity was needed for these uplands [ofMindanao]. Recognising this problem, from 1971 the[Mindanao Baptist Rural Life Centre] started to con-ceptualise a system now known as Sloping Agricul-tural Land Technology, or SALT. After testingdifferent intercropping schemes and observingLeucaena-based farming systems, both in Hawaii andat the Centre, efforts to develop the first-ever SALTprototype model commenced in 1978 … By 1980,MBRLC had acquired the confidence that SALTcould fulfil the objectives adequately. But, it tookabout four more years of testing and refining beforeSALT could be heralded as ‘applicable’(Partap andWatson 1994:29,30).

Partap and Watson go on to acknowledge that thetechnology has been further modified and adapted tosuit individual farmers’ conditions; in fact, they state

that ‘on-farm experimentation with SALT is anessential element’ (Partap and Watson 1994:91). Thesubsequent promotion of SALT throughout thePhilippine uplands by DENR and other agencies canbe seen as resulting from a loose but effectivecoalition of actors, including university scientists,DENR officials, international and local NGOs, andpublic and private donor agencies. The factorsleading to the success of this coalition warrantfurther study.

(b) Technology disseminationConventional extension theory, based on the centralsource model of technology development anddiffusion, examines the role of various organisationalarrangements and communication techniques in per-suading farmers to adopt a recommended technology(Van den Ban and Hawkins 1996). The Training andVisit System, promoted extensively (and expen-sively!) by the World Bank in the 1970s and 1980s,exemplifies this approach (Antholt 1998). The‘transfer of technology’ view of extension has beensuperseded (in the literature, if not widely in practice)by more participatory, community-based method-ologies, reflected in the currently fashionableapproaches of Participatory Rural Appraisal (PRA),Farmer Participatory Research (FPR) or, more gener-ally, Participatory Learning and Action (PLA)(Chambers et al. 1989; Okali et al. 1994; Scoones andThompson 1994; Chambers 1997).

Such participatory methodologies have now beenincorporated in development agency manuals andtraining courses world-wide. Biggs and Smith (1998)quote a recent set of guidelines for watershed develop-ment produced by the Ministry of Rural Developmentin India: ‘[Project staff] need to be trained in the toolsand techniques of project management, ParticipatoryRural Appraisal (PRA) methods, community organ-isation and other administrative and accounting pro-cedures’. Such statements hint at the rigid, top-downenforcement of ‘participatory’ procedures. TheDENR has also incorporated such community partici-pation techniques into its official procedures for theISFP (Gerrits 1996). While institutional endorsementof innovative participatory approaches is to be wel-comed, there is a concern that a preoccupation withmethods (described as a ‘manual mentality’) and, inparticular, their institutionalisation within bothgovernment and non-government agencies, will leadto unrealistic expectations of their general efficacyand distract attention from the complex requirementsfor successful research and extension projects (Biggsand Smith 1998).

Tendler (1993) examined a dozen success storiesin agricultural research and extension in poverty-stricken Northeast Brazil. She found that success was

14

not well correlated with adherence to ‘best practice’with regard to, for example, an emphasis on client-oriented, participatory research, or close coordinationbetween research and extension institutions. Agencieswhich had not been performing well, and which sub-sequently lapsed into poor performance, weretypically galvanised into effective action by a par-ticular set of circumstances, including a strongdemand from farmers or local development agenciesfor a solution to a particular problem (such as a pestoutbreak) within a limited time frame. Other elementsin the success stories were that ‘localised credit sub-sidies played a surprisingly important role in bringingabout rapid and widespread adoption by smallfarmers in a short period of time’ and that ‘municipal-level actors and institutions played important roles …unexpected because they were not included in theproject design’ (Tendler 1993:1577). However, not-withstanding this apparent support for decentralisa-tion, Tendler notes that ‘standing behind each stronglocal actor was a more centralised government agency– offering financial incentives, talking up the desirednew approaches, providing technical assistance,rewarding the good performers and keeping fundsaway from the bad ones’ (Tendler 1993:1577).

These observations fit well within an actor-oriented framework. As Long and Van der Ploeg(1989) argue, rural development interventions, suchas agricultural extension projects, involve a variety ofsocial actors, with diverse histories and agendas,from both within and beyond rural communities.Hence a project intervention needs to be recognisedas part of ‘an ongoing, socially-constructed andnegotiated process, not simply the execution of analready-specified plan of action with expected out-comes’ (Long and Van der Ploeg 1989:228). More-over, as Biggs emphasises, ‘all technology generationand promotional activities take place in a historicallydefined political, economic, agroclimatic, and institu-tional context’ (Biggs 1990:1487). The influence ofthese contextual factors may be crucial in deter-mining the outcome of a particular extension project.

Thus, for example, in evaluating the role of par-ticipatory technology development in the Forages forSmallholders Project in Malitbog, Bukidnon, inMindanao, it is important to consider the role oflocal and extra-local actors such as a progressive(cattle-owning) mayor, an entrepreneurial farmer andvillage leader, an experienced and well-regardedextension worker, and visiting national and inter-national scientists, as well as the apparently decisiveinfluence of other development initiatives, notablyvarious livestock dispersal programs which are con-tingent on the establishment of forage plots.

(c) Technology adoptionConventional research into farmer adoption of newtechnology explains the adoption-decision and itstiming (early or late) primarily in terms of thedecision maker’s perceptions and inherent character-istics, with ‘innovators’ at one extreme and ‘laggards’at the other (Rogers 1995). However, farmerdecision-making is generally more complex than thisimplies. As Scherr (1995) emphasises, farmers havemultiple objectives (including food security, adequatecash income, a secure asset or resource base, socialsecurity) and select ‘livelihood strategies’ to pursuethese objectives with the resources available to them(Ellis 1997). Both the objectives and the availableresources vary between farmers and change over thelife-cycle of the farm household (e.g. some farmers atsome times may rely on off-farm work as a majorsource of livelihood, restricting their capacity toinvest in labour-intensive conservation measures).Thus farmers in the same environment may have dif-ferent objectives and livelihood strategies, and sorespond differently to a given technology. Hence,Biot et al. (1995:24) suggest that ‘different behaviour[with respect to soil conservation] may be as much afunction of different opportunities and constraints asof different perceptions’.

The conventional adoption framework furthersimplifies the analysis of the adoption-decision by itsimplicit assumption of an individual ‘decision-maker’. Within the farm household, the ability tomake decisions regarding resource use and tech-nology varies according to age, gender, and othercategories, and actual decisions can depend on acomplex bargaining process among householdmembers (Ellis 1993, ch. 9; Jackson 1995; Biot et al.1995). Beyond the household, group processes andthe ability to harness them can play a crucial role inadoption decisions, particularly with regard to con-servation practices (Chamala and Mortiss 1990;Frank and Chamala 1992; Pretty and Shah 1994;Chamala and Keith 1995). Moreover, decisionsabout new technology are frequently prompted by anintervention of some sort, typically in the form of aproject. As discussed above, such interventions drawfarmers into a wider arena in which various socialactors are pursuing their personal and institutionalstrategies. Hence the outcomes in terms of adoptiondecisions will be highly contingent on the interplaybetween these actors, including such factors as thecreation of a sense of obligation to a respected exten-sion worker, or the development of conflict betweencontending factions within a community.

Thus an actor-oriented perspective leads us toexpect a range of responses to the promotion of anagricultural technology such as SALT, not merely aclear-cut decision to adopt or not. Differences

15

between the environment in which the technologywas developed and the environment of the ‘target’community will prompt farmers to adapt the tech-nology in the process of adopting it. Differenceswithin a given community in farmers’ goals and cir-cumstances, hence livelihood strategies, and thecomplexity of intra-household, group, and projectinteractions and decision-making, will result in avariety of adoption-adaptation behaviours, whichshould be investigated on their own terms and notpre-judged by labelling them as ‘poor adoption’ or‘non-adoption’.

Three Case Studies

(a) The Domang story2

Domang is a sub-village or hamlet (sitio) of 87households located in Kasibu Municipality in thenortheast of the province of Nueva Vizcaya innorthern Luzon. Rainfall in the uplands of NuevaVizcaya averages about 2400 mm and occursthroughout the year, with a somewhat drier periodbetween December and March. Regular typhoonsresult in intense rainfall events. Sitio Domangoccupies about 200 ha on a ridge descending fromMt Gusing (1455 m), at elevations ranging from 50to 1000 masl. The topography is heavily dissectedand gently to steeply sloping, with dominant slopesbetween 30% and 50%. Soils are predominantlyacidic clay-loams and erosion is moderate to severe.

The population density is relatively low at around50 persons per sq. km; hence, the mean farm size isclose to 4 ha and tenancy as such is non-existent.Domang is only moderately accessible, requiring ajourney on foot or carabao of up to an hour to reacha roadhead where jeepneys can provide transport tomarket towns. Market access has improved sinceDomang was settled in the 1970s and the farmingsystem now includes both subsistence and com-mercial crops – rainfed bunded rice on terraced land,upland rice, maize, a variety of vegetable and fieldcrops (beans, tomato, ginger, taro, etc.), and bananaand other fruit crops.

The Domang area is Public Forest Land and waslogged commercially in the 1950s and 1960s. Ifugaoand other migrants from the Central Cordillera beganarriving from the early 1970s, practicing shiftingcultivation of upland rice in the logged-over lands. Inthe mid-1970s, the community came into conflictwith officers of the then Bureau of Forest Develop-ment, who charged the settlers with illegal occupation

and sought forcefully to evict them. Six memberswere arrested and others took refuge in surroundingareas, until the local mayor intervened on their behalf.In the early 1980s, the community had to contestseveral claims to the land and repeatedly sought tohave the land reclassified as Alienable and Dis-posable, with full titles issued. However, a presiden-tial decree in 1984 prevented any further release ofPublic Forest Lands in the province.

A local forester advised the community to applyfor inclusion in the government’s Integrated SocialForestry Program, enabling them to be issued withCertificate of Stewardship Contracts (CSCs), a con-ditional 25-year lease of Public Forest Landrequiring farmers to establish agroforestry measuresfor soil conservation. A minority faction opposedthis move as it undermined their campaign for fulltitle to the land. Nevertheless, Domang became anISF project site, and by 1986 CSCs for 179 ha wereissued to 64 residents.

Extension activity under the ISF began in 1986 –a nursery was established and training was con-ducted in SALT and other conservation farmingpractices. However, there was little or no adoptionuntil 1990 when the site was selected as a ModelSite. This involved higher levels of funding andextension support – an energetic and well-regardedextension worker visited frequently, staying at thesite for up to three days per week, and farmers werepaid P6 per metre of hedgerows established. Oneparticipant’s farm was used as a demonstration farmand training site. By 1991 the majority of residentshad adopted contour hedgerows. After this, a changein policy meant that ISF projects no longer paidfarmers to plant hedgerows.

The project recommended using Leucaena leuco-cephala and Gliricidia sepium as hedgerow species.Inadequate local supplies of planting materialsforced farmers to approach lowland farmers forcuttings, but they met with resistance because low-land farmers were using their limited stocks as asource of fuelwood and wood for fencing. Also, theydisliked the fact that the ISF participants were usingthe cuttings for hedgerow development and receivinga monetary incentive to establish them. The limitedavailability of planting material for the recom-mended species induced farmers to look for alterna-tives. They adopted Hibiscus sp. (gumamela) as themajor hedgerow species and, to a lesser extent,banana. Hibiscus was locally available as it wascommonly used as an ornamental plant aroundhomes as well as around the school. The use ofhibiscus as a hedgerow species resulted from theexperimentation of one of the early adopters and theencouragement of the ISF extension worker.Cuttings struck easily and quickly when planted in

2This section draws on results of a survey conducted inMay 1996 which are reported in more detail in Garcia et al.1996.

16

moist soil. Nonetheless, the supply of plantingmaterials was still limited and farmers were expectedto use in-fill planting to increase plant density withinthe hedgerow once the first cuttings had becomeestablished.

The Domang ISF site was devolved to the localgovernment in 1993, after which extension activitypractically ceased. However, at the time of theSEARCA-UQ survey in 1996, there were 78 adopter-households or 90% of the Domang population. Non-adopters included those (dubbed pilosopo or ‘recalci-trants’) who had refused to join the ISF project onprinciple. Hedgerows were being maintained butthere was no expansion onto additional land. Thealleys were being used for maize, upland rice, and arange of commercial vegetable and field crops. Dif-fusion beyond the village was almost non-existentand where adoption did occur it was not well imple-mented due to poor understanding of the principlesand techniques involved. It should be noted thatbunded rice terraces (an indigenous technology forthe Ifugao members of the village population) werebeing constructed before the project began and con-tinued to be developed at the time of the survey.

Successful adoption of contour hedgerows inDomang occurred due to a set of circumstances at aparticular juncture – the dependence on CSCs fortenure security (after a decade or more of harassmentand threat of eviction), the allocation of an energeticextension worker on almost a full-time basis for aconcentrated period, and the payment of a subsidyfor hedgerow establishment. This combination of cir-cumstances induced rapid and widespread adoptionwithin the community. Farmer experimentationhelped resolve the problem of shortage of preferredplanting materials, resulting in successful adaptationof the recommended technology. The impetus givenby these circumstances appeared to be sufficient toget farmers to the point where they were prepared tomaintain the hedgerows, indicating ‘genuine’ adop-tion. Thus the ISFP, generally regarded as an unsuc-cessful program, was galvanised into making a briefbut significant impact in this location (consistentwith Tendler’s [1993] findings in Northeast Brazil).However, once the conditions which gave rise to theinitial wave of adoption ceased to exist, the wider‘diffusion’ of the technology did not occur.

(b) The Managok story3

Barangay Managok is a village located in MalaybalayMunicipality in the centre of Bukidnon Province in

Northern Mindanao. The annual rainfall of 2500 mmis fairly evenly distributed throughout the year, apartfrom a relatively dry period from December to March.The village occupies 1872 ha at elevations rangingfrom 400 to 1000 masl. The area encompasses anarrow plain lying between two ranges of hilly tomountainous terrain with slopes of 10% to 70%. Thesoils are predominantly acidic clays showingmoderate to severe erosion.

The population density of the upland sitio ofManagok is around 100 persons per sq. km (twicethat of Domang). The population comprises dumagator immigrant groups (75%) and lumad or indigenousgroups, namely the Tala’andig and Manobo (25%).The opening of the upland areas to in-migrationthrough logging resulted in rapid population growthin the late 1970s and early 1980s. Most of the land inManagok is Alienable and Disposable, though in theuplands farmers typically have no formal title and30% of farmers are share-tenants or mortgagees.Farm size averages 2.9 ha. Farmers have moderateaccess to markets; the Managok township is 14 kmby gravel road from the main north-south highwaythrough Bukidnon, and 24 km from Malaybalay, theprovincial capital. Smaller feeder roads provideaccess to upland sitio, and farmers use carabao andhorses to transport produce to the road.

The farming system in the uplands is based ontwo crops of maize per year, which is both the staplecrop and the main source of cash income. Inaddition, farmers cultivate vegetables, field cropsand tubers, perennial cash crops, and fruit and foresttrees. Livestock are raised for draught, transport andsale. Erosion control measures used by some farmers(apart from the introduced measures discussedbelow) include cultivating and planting across theslope, piling of stover in furrows across the slope,strip planting, tree planting, placing debris in gullies,and construction of rock walls.

At Managok, an Integrated Social Forestry Project(ISFP) of the DENR began in 1983 and theMUSUAN project was implemented by a team fromCentral Mindanao University (CMU) and XavierUniversity between 1988 and 1992. Initially, DENRactivities were limited to surveying farms, issuingCSCs, and conducting lectures on the conditionsattached to the CSCs and on recommended tech-nologies for farming the uplands, principally SALT.By the time the MUSUAN project began, the DENRhad issued CSCs to 106 farmers, 10% of whom hadplanted Gmelina arborea on their land, but none hadadopted SALT.

The MUSUAN project was implemented in foursites in Bukidnon Province with funding from theFord Foundation (US$200 000). A consultant fromthe College of Forestry, University of the Philippines,

3This section draws on results of a survey conducted inAugust-September 1994 which are reported in more detailin Garcia et al., 1995.

17

Los Baños, who helped prepare the MUSUANproposal to the Ford Foundation, recommended theuse of ISFP sites for the project. With the help oflocal DENR staff a short-list of possible sites wasidentified, including Managok. The project initiallyintended to focus on one sitio in Managok whereslopes were steep and farming was more intensive.However, a desire to include more lumad farmerssaw expansion of the project to include two moreremote sitio where shifting cultivation was stillpractised.

Recommended technologies were derived fromthe Mindanao Baptist Rural Life Centre (MBRLC) inDavao del Sur and the Mag-uugmad Foundation Inc.(MFI) in Cebu (see below), and were collectivelylabeled Slopeland Technologies for AgroforestryResources (STAR). These included contour hedge-rows (SALT), contour canals, contour rock walls,and bench terraces. Various combinations of thesetechnologies were presented as packages for selec-tion and adoption by farmers. Most opted for contourhedgerows and cultivation of fruit and timber trees.

The recommended hedgerow species were foragetree legumes such as Leucaena leucocephala, Leu-caena diversifolia, Flemingia macrophylla, Desmo-dium cinerea (= D. rensonii), Gliricidia sepium,Bauhinia monandra, and the forage grasses, napiergrass (Pennisetum purpureum) and guinea grass(Panicum maximum). However, in practice, wildsunflower (Tithonia diversifolia), a common weedon fallowed lands and roadsides, was the majorhedgerow species used. Its use developed fortui-tously. Farmers first used sunflower cuttings asstakes to mark the contour along which recom-mended hedgerow species (especially Flemingia andDesmodium) were planted. Hedgerow establishmentfailed but the sunflower cuttings struck and formedhedgerows themselves. This then became the recom-mended practice.

Preliminary project activities included construc-tion of the nursery site, introduction of the project,and selection of lead farmers. Initially there were 22lead farmers of whom 10 were dumagat and 12lumad. Lead farmers attended various seminars andworkshops and were taken on three cross-farm visits,including one to the MBRLC in Davao del Sur andone to the MFI in Cebu. Once the lead farmer hadselected a technology package, the farmer and theMUSUAN team prepared individual farm plans anda schedule for implementation. In addition, the teamassisted lead farmers (and subsequent adopters) withfield implementation of contour hedgerows, pro-vided planting material for hedgerows and fruit andforest trees, and distributed limited quantities of pro-duction inputs such as maize and vegetable seed.

Once lead farmers had successfully establishedcontour hedgerows on their own landholdings, theywere encouraged to act as extension agents and wereprovided with a monetary incentive for successfulestablishment of contour hedgerows on otherfarmers’ land. Farmers who had not adopted thetechnology would be approached by MUSUAN staffand/or lead farmers. If the farmer agreed, MUSUANstaff, the lead farmer, and the farmer concernedwould establish contour hedgerows on the farm.Some farmers commented during the survey thatthey had limited knowledge of the hedgerow tech-nology because the lead farmers took responsibilityfor implementing it. Farmers were initially organisedinto small labour-exchange groups of three farmerswith adjacent lots to facilitate layout and establish-ment of contour hedgerows, but these dispersed afteran initial learning period. Farmers cited differencesin their work habits as reasons for disbanding thegroups.

The DENR continued to provide funding duringthe MUSUAN project, for lead farmers to participatein the cross-farm visit to Cebu, for construction of anaccess trail and a reservoir, and for extension of thenursery. DENR staff also assisted in seedling pro-duction and extension of the technology. Between1989 and 1991, it was DENR policy to provide ISFproject participants with a subsidy for the establish-ment of conservation measures (P6 per metre ofdouble-row hedgerows and P2 per metre of single-row hedgerows). However, the MUSUAN projectopposed these payments, citing the detrimentaleffects of providing monetary incentives to farmers.Farmers claimed they signed papers acknowledgingreceipt of payment but never actually received anyfunds. Moreover, confusion regarding the ownershipof most of the land over which CSCs had beenissued resulted in their cancellation in November1993, though few farmers surrendered their docu-ments.4 The DENR simultaneously transferred thestatus of Model Site to another barangay.

By the end of the MUSUAN project in 1992, 60farmers were reported to have implemented contourhedgerows throughout their maize farms. However,at the time of the survey in 1994 there were only 47adopters (several farmers indicated that hedgerowshad been ploughed in or abandoned after the projectceased to operate). This represented 20%–40% ofupland households within the project’s target area;there was little adoption in the more remote and

4It was discovered that an area of 200 ha for which CSCshad been issued had been reserved for an agricultural schooland in fact belonged to CMU. The affected farmers have aninformal agreement with CMU allowing them to remain onthe site, but legally they are deemed to be squatters.

18

extensively farmed sitio where lumad farmerspredominated, and there was no evidence of widerdiffusion. Farmers’ stated reasons for adopting thetechnologies included increasing yields, supply offorage, and control of erosion. Non-adoption wasexplained in terms of farmers’ lack of interest, lack ofplanting material (in the case of leguminous hedge-rows) and lack of land ownership (i.e. tenancy).Farmers felt that access to extension, secure landownership, and a sole dependence on upland hold-ings contributed to higher rates of adoption.

The MUSUAN project in Managok (and elsewherein Bukidnon) was reported to be a successful exampleof a participatory approach giving rise to the develop-ment and adoption of a new package of agroforestrytechnologies (STAR). An actor-oriented perspectivehelps account for this success, beyond merely attrib-uting it to ‘best-practice’ methodology and tech-nology. At the same time, this perspective highlightsthe limited prospects for ongoing adoption. Theimpetus for the project came from a coalition of uni-versity scientists (including the UPLB consultant),local DENR staff, and the Ford Foundation. Theinterests and backgrounds of the actors in thiscoalition accounted for the emphasis of the project onparticipation (including the attempts to involve lumadfarmers) and agroforestry (which was seen to bescientifically acceptable, in line with DENR’s man-date for Public Forest Lands, and supported by anestablished NGO network). In reality, the STARpackage was a repackaging of existing technologies,principally SALT, and thus, in Biggs’s (1990) terms,was an example of the generation of ‘new’ tech-nologies through ‘labeling’. Serious problems withhedgerow establishment, which may have been fatalfor the project, were forestalled by the fortuitous dis-covery of Tithonia hedgerows – the one distinctiveelement in STAR. Notwithstanding the participatoryrhetoric, the workshops to identify lead farmers andget them to choose their preferred technologies, andthe formation of farmer groups to implement the tech-nologies, the project involved strong elements of the‘transfer of technology’ approach, with project staffand lead farmers actively persuading other farmers to‘adopt’ and taking a major role in the establishmentof contour hedgerows on their farms. This activistapproach proved quite successful, even in the absenceof DENR subsidies, but in the end only a minority offarmers responded – excluding most lumad and tenantfarmers – and the prospects for diffusion (or evenmaintenance) were not good. Confusion over thetenure status of much of the land led to conflictbetween the DENR and many project farmers, thecancellation of CSCs, and the loss of Model Sitestatus. This, combined with the inability of CMU to

gain further funding for the project, meant that exten-sion activities at Managok lapsed.

(c) The Guba story5

The project site considered in this case includesBarangay Guba and nine other upland barangay(villages) in the hinterland of Cebu City. Togetherthese barangay occupy an area of 78 km2 on theisland of Cebu in the Central Visayas Region.Annual rainfall is 1600–1800 mm, with no distinctdry season. The site is 200–600 m above sea leveland the terrain ranges from rolling to very steep.Soils are heterogeneous but primarily acidic, heavyclay-loams with slight to severe erodibility and sus-ceptibility to water-logging. The area is about 25 kmnorthwest of Cebu City centre and can be accessedby a network of gravel roads.

The population density in 1990 was 239 personsper sq. km, considerably higher than at Domang orManagok. Average farm size is 1 ha with a range of0.25 to 2 ha. Lands in the project area are a mixtureof Alienable and Disposable Lands and Public ForestLands. Most land is privately owned (with or withouta formal title). However, about 30% of farmers rentpart or all of their land under a share-croppingarrangement. Most of these have stable, long-termtenancy agreements with local or absentee landlords.

In the early 1980s, the farming system was domi-nated by the cultivation of maize for subsistence andthe production of perennials (mango and banana) andlivestock (cattle, goats, pigs and chickens) for sale.There were two croppings of maize per year. Theincreasing accessibility of the site has seen farmersrestrict maize cultivation to the first cropping andutilise the second cropping period for the cultivationof vegetables and flowers. Some are plantingvegetables in both seasons and purchasing most oftheir maize requirements.

In 1981, an expatriate staff member of WorldNeighbours, an international NGO focusing on ruralcommunity development, approached officials fromthe Department of Agrarian Reform (DAR) and theDepartment of Environment and Natural Resources(DENR) for assistance in selecting upland siteswhich would benefit from a soil and water conser-vation program. Barangay Guba was one of thoserecommended. Preliminary meetings introducedWorld Neighbours and discussed farmers’ problemsand their underlying causes, i.e. loss of soil fertilitydue to soil erosion and lack of nutrient cycling. As aresult of these discussions, the Cebu Soil and Water

5This section draws on results of a survey conducted inDecember 1996 which are reported in more detail in Gerritset al. 1997. See also Cramb et al., in press.

19

Conservation Program (CSWCP) was initiated.Funding for the CSWCP was initially provided byWorld Neighbours. Most of the on-going fundinghas come from World Neighbours and the FordFoundation, as well as from training programs andconsultancies. In 1988 participants in the CSWCPand a related coastal development program com-bined to form the Mag-uugmad Foundation Inc.(MFI), mag-uugmad being a Cebuano term for tiller,farmer, or an advocate for change or development. In1989, the MFI established a training centre at Gubaand, largely as a result, it continues to have an activepresence at the site.

CSWCP activities in Guba focused on the intro-duction of soil conservation technologies which hadbeen implemented in World Neighbours projects else-where. Contour canals, contour bunds and contourhedgerows were the main soil conservation technolo-gies promoted. Using the A-frame, farmers identifiedand marked contour lines at regular intervals downthe slope. These were ploughed and the loosened soilremoved and placed above the contour line therebyforming the contour canal and the contour bund. Next,grasses and leguminous shrubs were planted on thebund to form contour hedgerows. Initially mostadopters planted napier grass (Pennisetum pur-pureum) hedgerows. The vigorous growth of napiergrass led to its partial replacement with leguminousshrubs. Trimming of hedgerows was to occur everyone to two months. Trimmings were to be fed to live-stock or applied as a mulch to the alleys.

The method of extension of conservation tech-nologies in Guba was based on the organisation offarmers into work groups (alayon) and the utilisationof successful adopters as part-time farmer instructors.Farmers interested in implementing soil and waterconservation technologies on their farm organised orjoined a work group which then worked on each oftheir farms in rotation. On each farm the owner andfarmer instructor designed a suitable farm plan, andthe latter then demonstrated how the technologyshould be implemented. Each group and its indi-vidual members initially received some material orfinancial support, which formed the basis of arevolving fund.

The process of adoption began in 1981 when aleading farmer at Guba formed a five-memberalayon with his siblings to implement some of theconservation techniques promoted by World Neigh-bours. In 1982 the group expanded and establishedthe new technologies on 23 farms. In 1983increasing farmer interest led to the formation ofthree more groups. As farm development occurred,farmers in neighbouring barangay also enquiredabout the technologies. Maximum expansion of theproject occurred during 1983 and 1984. The project

supported 20 farmer-instructors in eight of the tenupland barangay. Each farmer-instructor managedthree to four alayon groups. At the peak of theproject there were over 70 alayon groups operatingin the project area and over 1000 farmers hadadopted the hedgerow technology. Thus the projectbecame widely known as the most successful conser-vation farming project in the Philippines.

Adoption of contour bunds, canals, and hedge-rows, and ploughing of the resultant alleys, resultedin rapid development of terraces. Once terraces hadformed, contour canals were no longer necessary andconsequently were not maintained. Terraces wereused for maize and vegetable cultivation and, to alesser extent, cut-flower production. The effects ofhedgerow adoption and terrace development asreported by farmers were: reduced soil erosion,better maintenance of soil fertility, a consequentreduction in inorganic fertiliser requirements, cropdiversification, increased crop production, a supplyof fodder for livestock, and an overall increase inhousehold welfare.

Following the end of the project in Guba and sur-rounding barangay, farmer-instructors no longerreceived wages and their activities correspondinglydecreased. Similarly the activities of the alayongroups diminished, mainly because most, if not all ofthe alayon members had established the soil conser-vation measures on their farms. Farmers reportedthat the alayon groups had no further tasks to accom-plish, although some continued to assist members ineveryday farm operations.

Several problems relating to the on-going utilisa-tion of the technologies (particularly contour hedge-rows) had emerged at the time of the SEARCA-UQsurvey in 1996. The development of contour canalsand terraces resulted in greater retention of water onthe field. In other contexts the resultant increase insoil moisture would be considered desirable becauseit extends the cropping season and reduces pro-duction risk. However, the project area had heavyclay soils. During the second cropping when rainfallwas higher, greater retention of water resulted inwater-logging problems throughout the area, particu-larly in Barangay Guba which had the heaviest soils.Some farmers dealt with the problem by constructingtemporary drainage canals which were subsequentlyploughed out. Other farmers suggested the need forpermanent drainage canals. MFI staff were recom-mending the use of raised beds.

A more general issue was that, throughout theproject area, hedgerow quality was observed to be indecline. Where hedgerows still existed, weed specieswere becoming dominant. In many farms hedgerowsno longer existed and, while some terraces werestable, others were reverting to the natural slope.

20

Several fields appeared to be in long-term fallow orabandoned. The explanations offered by farmerswere as follows: (a) Hedgerows were difficult to maintain. This was

because some hedgerow species were short-lived; hedgerows were adversely affected by twodroughts (1987 and 1989); large ruminants werefrequently tethered on the terraces during the dryseason, often without the landowner’s permis-sion, resulting in overgrazing of the hedgerowsand damage to the terraces; and hedgerows weredifficult to re-establish once cogon grass(Imperata cylindrica) had taken hold.

(b) Some farmers felt that hedgerows no longerneeded to be maintained once soil erosion hadbeen controlled and the terraces had formed.

(c) At the peak of adoption, farmers establishedcontour hedgerows over their entire farm. Later itwas found that land was required for alternativepurposes, e.g. tethering of livestock, and farmersconsequently abandoned some of their hedge-rows.

(d) In some cases land was left idle because ofhousehold labour shortages, old age, illness,death, or out-migration.

The widespread adoption of conservation tech-nologies at Guba clearly demonstrates that projectinterventions need to promote appropriate tech-nologies (those that address the farmers’ needsdirectly) at the appropriate time (as determined bythe stage of evolution of the farming system) andwith the genuine participation of farmers (utilisingfarmer-instructors and small, close-knit workinggroups), though not necessarily of the whole com-munity. That is, the ‘right’ technology and extensionmethodology are important. Success was also due toa number of site-specific factors, including goodcommunications, close community interaction, stableland tenure, increasing accessibility and market link-ages, and the evolution of the farming systemtowards new enterprises (Cramb et al. in press).

Within this context, however, successful adoptionat Guba was ultimately due to the effectiveness of acoalition of actors, including staff from WorldNeighbours, DENR and the Ford Foundation, andleading farmers at Guba, coming together to addressa pressing problem and focusing resources, ideas,and energy from within and beyond the community.The formation of the Mag-uugmad Foundationhelped to keep this coalition in place, and extend it,as new technical and institutional problems andneeds emerged. This does not mean that all emergingproblems were being dealt with as rapidly or effec-tively as was the soil erosion and fertility problem inthe 1980s, but the Guba story clearly indicates thelong-term, evolving nature of farm technology

problems, hence the need for on-going linkages ifinitial success with technology adoption and adap-tation is to be sustained.

Conclusion

The successful development, dissemination andadoption of improved technologies for smallholdersdepends on more than careful planning of researchand the use of appropriate methodologies in exten-sion. It depends on the timely formation of coalitionsof key actors – including key farmers as well as arange of key outsiders, researchers and others –whose interests converge sufficiently that they canfocus their resources and efforts on achieving changein agricultural systems, if only locally and only for aperiod. As Biggs and Smith (1998:10) emphasise:

The development coalition is a curious, opportun-istic grouping, loosely constructed through friendshipand other ties, reflecting both idealistic and self-interested impulses. It is pervasive enough to passunnoticed but remains remarkably significant inaffecting outcomes of development processes, as wellas in influencing the way those processes and theirhistories are seen.

Successful adoption of technology also dependson critical external factors – climatic events, marketfluctuations, the availability of subsidies for plantinghedgerows, livestock dispersal programs, edicts onland tenure – which enhance (or undermine) theeffectiveness of a development coalition in pursuit ofits strategy. It also depends on a good deal of luck.

Recognising the role of diverse social actors, thecoalitions they form, and the critical elements (manyof them unique to a given context) conditioning theirsuccess, helps our understanding of the social dimen-sion of technical change in agriculture, particularlyfrom a historical perspective. Harnessing this under-standing to enhance the process of future technicalchange is more problematic. Many of these factorsfall outside the scope of conventional planning andmanagement concepts and frameworks. Newerapproaches, such as participatory technologydevelopment and particpatory monitoring and evalu-ation, while giving greater recognition to theevolving, adaptive, and inherently participativenature of agricultural development processes, canthemselves become ‘manualised’, causing us to over-look the importance of personal, cultural, and polit-ical factors which may be crucial to success (Biggs1997). A broader and more flexible approach isneeded which gives explicit recognition to the role ofdevelopment coalitions and to the personal, cultural,and political dimensions of coalition-building fortechnology development.

21

References

Anderson, J.R. 1993. The economics of new technologyadaptation and adoption. Review of Marketing and Agri-cultural Economics, 61, 301–310.

Anderson, J.R. and Hardaker, J.B. 1979. Economic analysisin design of new technologies for small farmers. In:Valdes, A., Scobie, G.M. and Dillon, J.L. ed. Economicsand the Design of Small-Farmer Technology. Ames,Iowa State University Press, 11–26.

Antholt, C.H. 1998. Agricultural extension in the twenty-first century. In: Eicher, C.K. and Staatz, J.M. ed. Inter-national Agricultural Development, 3rd Edition,Baltimore and London, Johns Hopkins University Press,354–369.

Biggs, S.D. 1990. A multiple source of innovation model ofagricultural research and technology promotion. WorldDevelopment, 18, 1481–1499.

Biggs, S.D. 1997. Livelihood, coping and influencingstrategies of rural development personnel. ProjectAppraisal, 12, 101–106.

Biggs, S.D. and Clay, E.J. 1981. Sources of innovation inagricultural technology. World Development, 9, 321–336.

Biggs, S.D. and Smith, G. 1998. Beyond methodologies:coalition-building for participatory technology develop-ment. World Development, 26, 239–248.

Biot, Y., Blaikie, P., Jackson, C. and Palmer-Jones, R.1995. Rethinking Research on Land Degradation inDeveloping Countries. World Bank Discussion Papers289, Washington DC, The World Bank.

Chamala, S. and Keith, K. ed. 1995. ParticipativeApproaches for Landcare, Brisbane, Academic Press.

Chamala, S. and Mortiss, P. 1990. Working Together forLandcare: Group Management Skills and Strategies,Brisbane, Academic Press.

Chambers, R. 1997. Whose Reality Counts? Putting the FirstLast, London, Intermediate Technology Publications.

Chambers, R., Pacey, A. and Thrupp, L.A. ed. 1989. FarmerFirst: Farmer Innovation and Agricultural Research,London, Intermediate Technology Publications.

Cramb, R.A. in press. Conservation Farming Technologiesfor the Philippine Uplands: Socio-Economic Perspec-tives. ACIAR Monograph Series. Canberra, AustralianCentre for International Agricultural Research.

Cramb, R.A., Garcia, J.N.M., Gerrits, R.V. and Saguiguit,G.C. 1999. Smallholder adoption of soil conservationtechnologies: Evidence from upland projects in thePhilippines. Land Degradation and Development, 10,405–423.

Cramb, R.A., Garcia, J.N.M., Gerrits, R.V. and Saguiguit,G.C. in press. Conservation farming projects in thePhilippine uplands: Rhetoric and reality. World Develop-ment, 28.

Cramb, R.A. and Nelson, R.A. 1998. Investigating con-straints to the adoption of recommended soil conservationtechnology in the Philippines. In: Penning de Vries,F.W.T., Agus, F. and Kerr, J. ed. Soil Erosion at MultipleScales: Principles and Methods for Assessing Causes andImpacts, Wallingford, UK, CABI and IBSRAM, 99–120.

Ellis, F. 1993. Peasant Economics: Farm Households andAgrarian Development. 2nd Edition, Cambridge,Cambridge University Press.

Ellis, F. 1997. Household strategies and rural livelihooddiversification. Paper presented at Annual Conference ofDevelopment Studies Association, University of EastAnglia, Norwich, England, 11–13 September 1997.

Frank, B.R. and Chamala, S. 1992. Effectiveness ofextension strategies. In: Lawrence, G., Vanclay, F. andFurze, B. ed. Agriculture, Environment and Society:Contemporary Issues for Australia, Melbourne,Macmillan, 122–140.

Garcia, J.N.M., Gerrits, R.V., Cramb, R.A. and Saguiguit,G.C. 1995. Soil Conservation in an Upland FarmingSystem in Bukidnon: A Socio-Economic Survey,SEARCA-UQ Upland Research Project, Survey ReportNo. 3. Los Baños, Philippines.

Garcia, J.N.M., Gerrits, R.V., Cramb, R.A. and Saguiguit,G.C. 1996. Soil Conservation in an Upland FarmingSystem in Nueva Vizcaya: A Socio-Economic Survey,SEARCA-UQ Upland Research Project, Survey ReportNo. 6. Los Baños, Philippines.

Garcia, Y.T. 1997. Analysis of Decision Models for UplandSoil Conservation in Argao, Cebu. PhD dissertation,University of the Philippines, Los Baños.

Gerrits, R.V. 1996. The Philippine Government’s Approachto Upland Development: The Integrated Social ForestryProgram, SEARCA-UQ Uplands Research Project,Working Paper No. 16. Los Baños, Philippines.

Gerrits, R.V., Garcia, J.N.M. and Cramb, R.A. 1997.Hedgerow Adoption in the Hinterland of Cebu City: AReconnaissance Survey, SEARCA-UQ Uplands ResearchProject, Working Paper No. 20. Los Baños, Philippines.

Jackson, C. 1995. Environmental reproduction and genderin the Third World. In: Morse, S. and Stocking, M. ed.People and Environment, London, UCL Press, 109–130.

Long, N. and Long, A. ed. 1992. Battlefields ofKnowledge: The Interlocking of Theory and Practice inSocial Research and Development, London, Routledge.

Long, N. and Van der Ploeg, J.D. 1989. Demythologisingplanned intervention: An actor perspective. SociologiaRuralis, 29, 226–249.

Okali, C., Sumberg, J. and Farrington, J. 1994. FarmerParticipatory Research: Rhetoric and Reality, London,Intermediate Technology Publications.

Pandey, S. and Lapar, M.A.L. 1998. A microeconomicanalysis of adoption of contour hedgerows in the Philip-pine uplands. In: Penning de Vries, F.W.T., Agus, F. andKerr, J. ed. Soil Erosion at Multiple Scales: Principlesand Methods for Assessing Causes and Impacts, Walling-ford, UK, CABI and IBSRAM, 83–98.

Partap, T. and Watson, H.R. 1994. Sloping AgriculturalLand Technology (SALT): A Regenerative Option forSustainable Mountain Farming, ICIMOD OccasionalPaper No. 23. Kathmandu, International Centre for Inte-grated Mountain Development.

Pretty, J.N. and Shah, P. 1994. Soil and Water Conservationin the Twentieth Century: A History of Coercion andControl. Research Series No. 1, Rural History Centre,University of Reading.

22

Rogers, E.M. 1995. The Diffusion of Innovations, 4thEdition, New York, Free Press.

Scherr, S.J. 1995. Economic factors in farmer adoption ofagroforestry: patterns observed in Western Kenya. WorldDevelopment, 23, 787–804.

Scherr, S.J. and Muller, E.U. 1991. Technology impactevaluation in agroforestry projects. AgroforestrySystems, 13, 235–257.

Scoones, I. and Thompson, J., ed. 1994. Beyond FarmerFirst: Rural People’s Knowledge, Agricultural Research

and Extension Practice, London, Intermediate Tech-nology Publications

Sumberg, J. and Okali, C. 1997. Farmers’ Experiments:Creating Local Knowledge, Boulder and London, LynneRienner.

Tendler, J. 1993. Tales of dissemination in small-farm agri-culture: Lessons for institution builders. World Develop-ment, 21, 1567–1582.

Van den Ban, A.W. and Hawkins, H.S. 1996. AgriculturalExtension. 2nd Edition, Oxford, Blackwell Science.

23

Participatory Approaches to Forage Technology Development with Smallholders in Southeast Asia

P.M. Horne1, E. Magboo2, P.C. Kerridge3, M. Tuhulele4, V. Phimphachanhvongsod5, F. Gabunada Jr.6, Le Hoa Binh7 and

W.W. Stür6

Abstract

Conventional approaches to forage technology development have not resulted in substantialadoption of forage technologies by smallholder farmers in Southeast Asia. This paper describesalternative approaches which were developed and implemented by the Forages for SmallholdersProject in four countries since 1995. The approaches resulted in a sounder understanding offarmers’ perceptions of their problems and opportunities, greater adaptation and adoption of foragetechnologies than had been previously achieved and gave the project a much stronger bridge intopoor farming communities. The challenge now is how to expand the benefits of locally-successfulforage technologies to more people more quickly, whilst maintaining active farmer involvement inthe expansion process.

THE PHYSICIST Niels Bohr once remarked ‘There areonly two real sciences – physics and stampcollecting’. His quip (and the elegant response of theNobel Prize Committee in awarding him the NobelPrize in Chemistry) has lessons today even for thoseof us working in agricultural research and develop-ment. As with Bohr’s work on subatomic structure,where making a distinction between physics andchemistry was neither easy nor useful, unnecessaryboundaries have existed between agriculturalresearch and agricultural development which havehindered adoption of agricultural technologies bypoor farmers in developing countries.

An example of this is the track record of forageresearch and development in Southeast Asia. Despitesubstantial research over the past 40 years, the impactof improved forage technologies on smallholder live-stock systems has been generally negligible. Many ofus who have been involved in this process have beenasking ‘Why?’ and ‘What can we do to improveadoption of promising forage technologies by small-holder farmers in future?’ A common conclusion hasbeen that the lack of linkages between technologydevelopment on research stations and technologyadoption on farms has often resulted in research pro-grams developing technologies that offered few ifany solutions to the major problems that farmersfaced. This paper describes some approaches thathave been taken over the past five years to overcomethis problem and the major lessons learned, not onlyfor forage technology development in future, but foragricultural development in general.

Conventional Approaches to Forage Technology Development

‘Forage technologies’ are combinations of adaptedforage varieties and ways of integrating them intofarming systems that provide tangible benefits to

1Forages for Smallholders Project, Vientiane, Lao PDR.Email: P.Horne@cgiar.org2Livestock Research Division, PCARRD, Los Baños, Phil-ippines. Email: ECMAGBOO@PCARRD.DOST.GOV.PH3CIAT, Cali, Colombia. Email: P.Kerridge@cgiar.org4Directorate General of Livestock Services, Jakarta, Indo-nesia. Email: fsp-indonesia@cgnet.com5National Agriculture and Forestry Research Institute,Vientiane, Lao PDR. Email: FSP-Lao@cgiar.org6Forages for Smallholders Project, Los Baños, Philippines.Email: F.Gabunada@cgiar.org7National Institute of Animal Husbandry, Hanoi, Vietnam.Email: fspvietnam@hn.vnn.vn

24

farmers (see Gabunada Jr. et al. 2000, these Proceed-ings). The conventional approaches that were used todevelop forage technologies were similar to thoseused in field crops. From interviews with keyfarmers, development workers (often researchers andextension workers) identified problems to research,developed technical solutions to these perceivedproblems on research stations and demonstratedthese ‘solutions’ through well-supported modelfarmers, expecting that adoption would occur sponta-neously from these carefully managed demonstra-tions (Figure 1). In most cases this approach, whichhad worked well with crops such as rice, resulted inpoor adoption of forages. There appeared to be manyreasons for this poor adoption, including:

• growing forages was a new concept for mostfarmers (unlike growing rice);

• the need for forages may not have been great inthe areas selected for extension;

• forages provide longer-term benefits to farmersthat are not as immediately obvious as, forexample, the direct cash benefits of improved ricevarieties;

• often, planting material of the promising foragevarieties was not locally available;

• researchers were often developing feeding tech-nologies based on concepts such as ‘increasedlivestock productivity’ and ‘improved efficiencyof utilisation of feed resources’, where farmersdecision-making was based on a much more com-plex set of constraints, opportunities and goals.

In 1995, AusAID provided funding for a newapproach to address these limitations. The Foragesfor Smallholders Project (managed by Centro Inter-nacional de Agricultura Tropical (CIAT) and theCommonwealth Scientific and Industrial ResearchOrganisation (CSIRO) Division of Tropical Agricul-ture) worked:

• in seven countries of Southeast Asia to evaluateforage varieties for their adaptation to climate,soils and potential uses on farms in Southeast Asia(Stür et al. 2000; Gabunada Jr. et al. 2000);

• with more than 1700 farmers at 19 sites in four ofthese countries to develop, implement and modifyparticipatory approaches to forage technologydevelopment on smallholder farms (Tuhulele et al.2000). The participatory approaches used by theFSP were adapted from methods developed byCentro Internacional de Agricultura Tropical(CIAT) in Central and South America (Ashby1990). These approaches are described in the restof this paper.

Participatory Approaches to Forage Technology Development

Upland farming systems in Southeast Asia arecharacterised by great diversity in opportunities andlimitations for forages, not only between villages butbetween individual farmers within villages. What issurprising about this statement is that we are oftensurprised by it! For example, the FSP worked in atransmigration village (Marenu) in North Sumatra,Indonesia, where two years earlier all farmers hadbeen given equal areas of similar land, an equalnumber of sheep and a house. Two years later, whatwas striking was the great divergence of activitiesthat farmers were involved in. Some had plantedforages and expanded their sheep numbers, somewere concentrating on cash crops and others weresupplementing farm incomes with jobs in town. Con-ventional approaches to forage technology develop-ment could not deal with this kind of diversity. Asstudents of statistics, researchers are trained to mini-mise variation in experiments and compare averagesof treatments. ‘Average’ forage solutions developedon research stations for ‘average’ farmers, however,fail in diverse environments.

The approach of the FSP to working within thesediverse situations was to identify robust, broadlyadapted forage varieties, which we regarded as thebasic building blocks of promising technologies(Stür et al. 2000). Through participatory approacheswe encouraged farmers to evaluate these buildingblocks and experiment with them in developing inte-grated forage systems for the particular conditions ontheir farms. The underlying principle was to givefarmers ‘ingredients and information, not recipes’that would allow them to make decisions and takeaction in response to changes in their own farmingsituations.

The right hand column of Figure 1 summarises theprocess used by the FSP. Some of the steps involvedfarmers making the decisions (facilitated by thedevelopment worker) and one step was the responsi-bility of the development worker. However, the mainstep (evaluating and adapting technology options)required a robust partnership between farmers andthe development worker to be successful.

Before applying these approaches, it was fre-quently necessary to carefully select communitieswhere there seemed to be greatest potential for adop-tion and impact.

Where should We Work to Maximise the Potential for Impact?

Many development workers and projects are respon-sible for all agricultural development in their area

25

Figure 1. Conventional and participatory approaches to forage technology development.

Select a community

Collect and interpretdata and keyinformation

Inteviews with key farmers

The Development Worker identifiesproblems

The Development Worker looks forbest ‘solution’ (or researchers

develop ‘solutions’)

‘Model Farmers’ are selected todemonstrate the ‘solutions’

Other farmers encouraged to adoptthese ‘solutions’

The community identifies andprioritises the main problems facing

farmers in their area

The Farmers discuss the strategiesthey already have tried to address

these problems

The Farmers decide whichproblems are serious enough forthem to commit their own time to

finding better solutions

Development Worker searches foradditional technology options and

offers these to farmers as a startingpoint to finding better solutions

The Farmers decide which optionsto test and adapt to their situations

The Farmers and the DevelopmentWorker together evaluate and

adapt the technology options todevelop solutions which providesignificant impact and have the

potential to be adopted more widely

Conventional approachesto developing agriculturaltechnologies

Participatory approachesto developing agricultural

technologies

= Development worker makes the decisions

= Farmers make the decisions

= Farmers and the development worker make the decisions together

Par

ticip

ator

yD

iagn

osis

26

including agricultural commodities from rice to live-stock to fruit trees. Their main priority is to identifythe broad range of problems and opportunities forimprovement in their area and respond withpotentially useful technologies. Other developmentworkers are specialists in a particular field. Thismeans that they have to look for places where theirparticular technologies have the greatest potential forimpact. If the particular technology is a new irrigatedrice variety, for example, then this site selection islikely to be relatively easy, targeting farmers whoalready grow irrigated rice and could benefit fromthe particular characteristics of the new variety. Withforage and livestock technologies, however, thebasic concept of the technologies is very often newto farmers. As a result, careful area and communityselection was a critical first step in our participatoryapproaches. It did not matter how good our partici-patory approaches were if there was no potential forthe technologies to have an impact in a particulararea!

How can you identify sites and communitieswhere your technology options are likely to have thebest potential for impact? Learning from fieldexperience, the FSP developed a set of criteria tohelp select areas and communities where foragetechnologies had a greater chance of adoption andimpact. These are framed as questions in Table 1,which guided our early work in site selection.

At first glance some of these questions appeartrivial and self evident, and yet it is surprising howoften on-farm technology development fails becausethese issues have not been addressed. This is bestillustrated using some examples from forage tech-nology development in the FSP:

1. Is there a genuine (livestock feeding) problem?

The FSP was asked to work with a group of farmersin one area to evaluate forages for feeding theircattle. When we investigated more-closely, we foundthey didn’t have any cattle! The motivation of thefarmers in wanting to plant forages was that it wasthe requirement of the local cattle credit scheme thatthey must plant a certain area of forage before theywould receive animals from the cattle bank. Withouthaving managed cattle before, these farmers hadlittle idea if they were going to have any problemsfeeding their cattle. They planted forages, receivedcattle and abandoned their forage plots. They did nothave a genuine problem.

In other situations, we have been directed byplanners to areas where there are large numbers oflivestock or natural grassland areas where they see apotential for increasing livestock production.Diagnosis with the farmers in such situations hasoften shown that there is not a gross shortage of feedand farmers see no need to expend effort in devel-oping more intensive forage systems.

2. Do farmers think that this problem is important enough to commit their time in working towards a solution?

One group of cattle farmers asked the FSP to assistthem with developing better feeding systems, but byproviding seed, a tractor, better cattle breeds, fer-tiliser and fencing material. There is no doubt that ifthese had been provided they would have developedan impressive livestock system in the short term, butthey would have failed in the long-term when thesubstantial external inputs dried up. It would also nothave been possible for farmers from other areas toadopt the systems that were developed without thesesubstantial inputs. The farmers did not regard theproblem as being sufficiently important to committheir time to working towards a more-sustainablesolution.

In another area, the farmers clearly had a majorproblem feeding their cattle in the dry season andwere interested in looking at forage options. How-ever, there were so many other more-serious prob-lems with their cropping systems and human health(which were receiving outside support) that the live-stock feeding problems were not a high enoughpriority for them to commit any time in workingtowards a solution.

Table 1: Key questions to ask when selecting communi-ties and farmers for on-farm agricultural technology devel-opment.

Questions For successful sites we need

1) Is there a genuine problem?

‘YES’is essential

2) Do farmers think that this problem is important enough to commit their time in working towards a solution?

3) Are there many farmers and other com-munities who have the same problem

4) Do we have potential solutions to offer which can provide substantial benefits?

5) Are there active local individuals or groups who are committed to working with farmers to solve this problem?

6) Are there farmers who are already trying to solve this problem? ‘YES’

is desirablebut not

essential7) Is there a local enthusiast who would

‘champion’ the resulting technologies in future?

27

3. Are there many farmers and other communities who have the same problem?

At several sites, the answer to questions 1 and 2 was‘Yes’ but we found that the farmers we were talkingto were a small group and there were no otherfarmers in the area with similar problems and moti-vations. The potential for impact and expansion wasvery limited.

4. Do we have potential solutions to offer which can provide substantial benefits?

Even if there are many farmers with livestockfeeding problems that they could alleviate withforages, we have to be sure that we have promisingoptions to offer. In one instance, many farmers in alowland rice cropping area had started usingirrigation to move from growing one rice crop a yearto two. These farmers were heavily dependent ontheir buffaloes for draught power, manure and liveli-hood security. Previously, the buffalo had grazed onthe rice paddies for most of the year and were fed cutgrasses from the river banks during the rice croppingseason. With two rice crops, there was not enoughgrass to sustain their buffaloes throughout the year.However, they had no vacant land which could beallocated to growing enough forage to have a signif-icant effect on feed supply during the wet season.We had no potential solutions to offer.

5. Are there active local individuals or groups who are committed to working with farmers to solve this problem?

This was perhaps the most critical factor governingthe success of on-farm forage technology develop-ment. Active and enthusiastic local developmentworkers, with an empathy for the participatoryapproaches, can inject a great deal of momentuminto the forage technology development even in anarea where the potential may initially appear to belimited. This is particularly true for forages, whichare a new concept for most farmers and have slowearly growth compared with many crops. At onevery successful site the development worker said ‘itwas hard to start with because the farmers neededregular encouragement to look after the small seed-lings, but once the plants were tall and they couldstart to see the benefits, my job became easier’.

6. Are there farmers who are already trying to solve this problem?

All institutions and individuals working in ruraldevelopment want to work with innovative farmers.We found that identifying and working with farmerswho had already been working actively, often on

their own to solve their problems, increased thechances of successful technology development.These farmers usually identify themselves duringdiagnostic sessions with the community.

7. Is there a local enthusiast who would ‘champion’ the resulting technologies in future?

At several FSP sites, the momentum of local expan-sion of promising forage technologies came from theinvolvement of a local enthusiast (sometimes afarmer, sometimes a development worker, sometimesa trader or local official) who actively ‘championed’the technologies.

To help answer these seven questions and select‘possible candidate sites’ for developing agriculturaltechnologies with farmers, there is a need to collectsecondary information. However, it is important tocollect only the minimum data set necessary toanswer the questions. There is a tendency at thisstage of site selection to collect too much irrelevantdata that becomes so daunting to interpret that all ofthe information is ignored. In the FSP, we used:

1. Environmental and census data (from local government statistics)

These helped us understand which forage varietiesmight be adapted to the area. These data can some-times be inaccurate and difficult to collect. The mostimportant data for forages were:• Climate data (average monthly rainfall and max./

min. temperatures for the past ten years);• Livestock data (numbers, location, type);• Soils data (broad fertility status and pH ranges for

the soils in which forages are likely to be planted).

2. Key information and observations (from discussions with key individuals and from field visits)

This information helped us to understand what live-stock feeding problems might exist and what possibletechnology options we might be able to offer. It wasgenerally more useful and easier to collect than theenvironmental and census data. For forage tech-nology development, there were some key questionswe would ask local government officials, agriculturalextension workers and individual farmers:• What are the key problems affecting agricultural

development in the area? • What is the major land use and it constraints?• Who owns the land?• How big are the farms?• Why do farmers keep livestock?• Who keeps livestock (only certain ethnic groups?

only the wealthy?)

28

• What type of animals and how many do farmersraise?

• What are the main problems with keeping live-stock?

• Who is responsible for looking after the livestock(the men? the women? shared?)

• How are the animals fed and managed?• How do farmers overcome these problems now?• How are the farming and livestock systems

changing?Having used this information to identify com-

munities where the forage technologies might havepotential, we started the process of ParticipatoryTechnology Development.

Participatory Technology Development

‘Participatory technology development’ (or PTD) isa broad concept referring to development approacheswhich have ‘active farmer participation in all stagesof the development process’ as a central principle.

‘If you give a man a fish you feed him for a day, if you teach the man to fish you feed him for a lifetime . . . but if you ask the man about fishing he may prefer to diversify into tree crops and

livestock production.’

There is a bewildering diversity of tools andapproaches that fall under this concept of PTD, butan intriguing convergence of experiences and under-pinning principles. This paper does not attempt tocover the many tools and approaches which are wellcovered in other publications (some recent examplesbeing van Veldhuizen et al. 1997 a, b; and Hagmann1998) but to summarise the approach used by theFSP and the lessons learned from this.

Figure 2 summarises the PTD concept used by theFSP. The key to this process was active involvementof the community at all stages of technology devel-opment (prioritising problems, identifying possiblesolutions to test, experimentation and evaluation).

Having identified a community where foragesappeared to have potential, the first step in PTD wasto confirm this potential through a village meetingwhere the community diagnosed and prioritised theproblems they experienced in their farming and live-stock systems. This was a process similar to Partici-patory Rural Appraisal (PRA) and used many of thetools of PRA, but was generally much quicker thanPRA as it is usually implemented (taking no morethan one day) and was directed towards ‘Action’ not‘Appraisal’. It moved beyond simply gathering listsof problems into problem diagnosis and actionplanning.

Figure 2. The Participatory Technology Development approach used by the FSP.

Identify and selectforage varieties/technologies

to test

Test foragevarieties/technologies

Participatory technology development

Feedback

On-station research

Evaluate foragevarieties/

technologies

Expansion(Adaptation and

Adoption)

Diagnose andprioritise problems

Active FarmerParticipation

29

There were several common steps in the diagnosis:a. Familiarisation with the area (for example, a

village walk to allow the development workers tobecome familiar with the landscape and farmingsystems and get some ideas about opportunitiesand limitations);

b. Resource inventories (This was often the firstprocess and involved farmers describing theirvillage resources through mapping, seasonalcalendars and historical calendars. It not onlyserved as a basis for discussion of problems in thefarming systems but broke down communicationbarriers in a participatory process that wasinitially strange to farmers);

c. Problem identification (With active facilitation,often using cards on which to write problems,farmers identified the major problems they face intheir agricultural and livestock systems);

d. Problem analysis (The community would iden-tify causal linkages between these problems);

e. Current management of these problems (Thecommunity would describe how they have copedwith these problems until now. This not onlyshowed which problems they regarded as beingimportant but identified innovative, motivatedfarmers who have been actively trying to solvethese problems in the past);

f. Prioritisation of problems the farmers want towork towards solving;

g. Agreement on a plan of action between thedevelopment worker and the community toevaluate a range of technology options that havepotential to alleviate the priority problem(s).Following diagnosis, individual farmers and, in

some cases, farmer groups, would begin planting andevaluating forage varieties, many of which had beenselected from on-station research (Figure 2). Thesefarmers were encouraged to plant forages whereverand in whatever ways they wished to test them,although the development workers collaborating withthe FSP would provide some suggestions and infor-mation on establishment and management. Thesedevelopment workers would then follow-up with par-ticipatory evaluations in which the farmers describedwhich forage varieties/technologies they preferredand why. In many areas the development workerswere able to work with existing farmer groups orform special interest groups of farmers who wereinterested in evaluating forages. Ultimately, however,each farmer wanted to grow forages on his or herown land and individual follow-up was veryimportant to support on-farm development at all sites.Farmer groups do not always exist or in such casesmay not be appropriate for particular situations, andthe FSP worked with individual farmers.

It was common that some farmers would quicklysee benefits of particular varieties and expand theirplots and neighbouring farmers would spontaneouslystart planting forages (‘Local expansion’ – Figure 3).Others would maintain their plots without expandingor would abandon them if the benefits were notimmediately obvious. Regardless of the farmingsystem, however, we found that farmers alwayswanted first to test varieties in small plots, usuallynear their houses, and only with time (usually afterone or two seasons when some farmers becameconvinced of the potential benefits of particularvarieties) would they begin to experiment with novelways of integrating forages onto their farms (‘foragesystems’ – see Figure 3). These are the farmers whobecame the ‘core’ that drove the technology develop-ment process of the FSP.

In some cases, as their experience with foragesincreased, farmers found new problems to solve withforages or developed new opportunities for foragesby changing their farming systems. An example ofthis kind of innovation occurred in northernVietnam, where we started working with farmers toevaluate forages for feeding their buffalo in the dryseason. Most farmers were initially impressed withthe forage varieties that they tested but did not wantto allocate a lot of valuable crop land to plantingforages. The FSP almost withdrew from the sitebecause of the lack of potential when several farmersstarted feeding their fish (carp) with forage grasses.This caused great interest among other farmers in thearea since most farmers had fish ponds and there wasa severe shortage of feed available for fish. Foragesin backyard plots for cut and carry feeding of fish arenow expanding rapidly in this area. Some farmersare replacing dryland crops with forages to expandtheir fish production. The farmers taught us avaluable lesson: ‘If you don’t feed your buffalo inthe dry season, they get thin, but they get fat again inthe wet season. If you don’t feed your fish, they die’.Some of the earlier farmers are now planting foragesto fatten cattle, another source of cash income.

Expanding to New Areas – More Benefits for More People More Quickly

One measure of success is when farmers expandtheir areas of forage and neighbouring farmers spon-taneously adopt the new forage technologies. Wehave seen this occur in many sites and examples aregiven in other papers in these Proceedings (e.g.Gabunada et al. 2000; Nacalaban et al. 2000; Ibrahimet al. 2000). For example, at Malitbog, Philippines,the number of farmers evaluating forages expandedfrom 15 farmers in the first year to more than170 farmers in the third year, and almost 50 farmers

30

had expanded their forage areas considerably. Ourpresent concern is how we can most rapidly expandto new areas.

The participatory approaches promoted by theFSP are not part of a process that has a clear endwhere development workers eventually withdraw,but are part of a process that demands changing rolesof farmers, development workers and institutions inan evolving partnership. As local expansion starts tooccur, the roles of development workers evolve moretowards that of being facilitators of farmer-to-farmerexchange of ideas and planting material. The chal-lenge facing us now is how to expand more benefitsof locally successful forage technologies to morepeople more quickly. How can we expand impactsbut maintain the participatory approaches and prin-ciples of active farmer involvement that are ‘inherentin smallness and cannot simply be replicated’ (IIRR1999). It is evident that this has to involve not only ascaling-out of geographic impact but a scaling-up ofinstitutional capacity and the formation of strategicalliances or development coalitions.

Some Lessons Learned

The participatory approaches being developed by theFSP are an on-going learning experience for allinvolved, based on a cycle of field activitiesfollowed by reflection on outcomes and modificationof the approaches. There are no shortcuts to thislearning process. The approaches do not constitute aseries of steps that can be rigidly followed but are

more ‘a complex conjunction of people, events andluck, often with unanticipated outcomes’ (Cramb,2000). As such, it was often difficult for develop-ment workers to adopt the new approaches. Withtime, however, they found that, although ‘the partici-patory approaches are time consuming and require alot of legwork, … [it is] … through participatorymethodologies we have finally found a way to bringforage technologies to the farmer’ (Staples andRoder 1998).

Despite these benefits, the participatory approachesare currently fragile. This is partly because there arefew experienced development workers who canconfidently operate within such an unstructuredframework. Training courses do not necessarilychange this. What is needed is a program of fieldexperiences and mentoring that will give thosedevelopment workers who already have an empathyfor participatory approaches, the decision tools,methods and skills they need to work more con-fidently with farmers in the field. ‘We shouldn’t tryto standardise innovation but systematise it’ (Uphoff,quoted in IIRR 1999).

The participatory approaches are also fragilebecause of pressures within national organisationsand projects for ‘quick results’. This results inpressures to reproduce or ‘photocopy’ locally suc-cessful technologies from one site to another usingthe conventional extension approaches. This is rarelysuccessful (as has been widely experienced bydevelopment projects that have tried to promotecontour hedgerows in the uplands of Southeast Asia).

Figure 3. The process of development of integrated forage systems on farms.

Represents the PTDprocess from Figure 2

Some farmers drop out or donot expand their areas

Local expansion

Testing forage systems

Testing varieties

Time

Some experienced farmersstart testing

new species and systems

Some farmers find new problemsto solve with existing forages or

find new opportunities for forages

Indi

vidu

al fa

rmer

’s e

xper

ienc

ew

ith fo

rage

s

31

Farmers in a new area will want to experiment with‘strange’ new varieties just as much as did the orig-inal farmers. However, new farmers may start imme-diately evaluating forage systems (not simplyvarieties) in an area where there are many experi-enced farmers already using forage systems. Findingways of scaling-up locally successful technologieswithin a participatory framework is the next majorchallenge in methodology development (see IIRR1999; Cramb, 2000; Connell, 2000).

There is no doubt that the participatoryapproaches have given FSP a much sounder under-standing of farmers’ perceptions of their problemsand opportunities, and given us a much strongerbridge into poor farming communities … a bridgebased on trust and mutual respect. Many otherresearchers, extension workers and project personnelin Southeast Asia are familiar with the benefits ofthese approaches and the concepts of implemen-tation, but are looking for practical guidelines onhow to implement them in the field.

ReferencesAshby, J.A. 1990. Evaluating technology with farmers. A

Handbook, Colombia, CIAT, IPRA Project, 95 p.Connell, J.G. 2000. Scaling-up: The roles of participatory

technology development and participatory extensionapproaches. In: Stür, W.W., Horne, P.M., Hacker, J.B.and Kerridge, P.C. ed. Working with Farmers: the Key toAdoption of Forage Technologies. ACIAR: These Pro-ceedings, see Contents.

Cramb, R.A. 2000. Processes influencing the successfuladoption of new technologies by smallholders. In: Stür,W.W., Horne, P.M., Hacker, J.B. and Kerridge, P.C. ed.Working with Farmers: the Key to Adoption of ForageTechnologies. ACIAR: These Proceedings, see Contents.

Gabunada F. Jr., Heriyanto, Phengsavanh P., Phimphach-anhvongsod V, Truong Thanh Khanh, Nacalaban W.,Asis P., Vu Thi Hai Yen, Tugiman, Ibrahim and Stür W.2000. Integration of adapted forages on farms in South-east Asia – experiences from the Forages for Small-holder Project. In: Stür, W.W., Horne, P.M., Hacker, J.B.and Kerridge, P.C. ed. Working with Farmers: the Key toAdoption of Forage Technologies. ACIAR: These Pro-ceedings, see Contents.

Hagmann, J. 1998. Learning together through ParticipatoryExtension. A Guide to an Approach developed inZimbabwe, Harare, Zimbabwe, AGRITEX/GTZ,. 59 p.

Ibrahim, Truong Tan Khanh and Heriyanto. 2000. Foragetechnologies for smallholders in grassland areas. In: Stür,W.W., Horne, P.M., Hacker, J.B. and Kerridge, P.C. ed.Working with Farmers: the Key to Adoption of ForageTechnologies. ACIAR: These Proceedings, see Contents.

IIRR 1999. Scale-Up! Highlights and Synthesis of Proceed-ings of the CGIAR NGO Committee Workshop onScaling Up Sustainable Agriculture Initiatives, held inWashington, October 22-23 1999, Cavite, Philippines,International Institute for Rural Reconstruction, 56 p.

Nacalaban, W., Le Van An, Asis, P., Moneva, L. andBalbarino, E. 2000. How do forages fit into smallholderfarms in mixed upland cropping systems? In: Stür,W.W., Horne, P.M., Hacker, J.B. and Kerridge, P.C. ed.Working with Farmers: the Key to Adoption of ForageTechnologies. ACIAR: These Proceedings, see Contents.

Staples, R. and Roder, W. 1998. Report of the externalmid-term review of the ‘Forages for SmallholdersProject’, Vientiane, April 1998, Philippines, CIAT.

Stür, W.W., Ibrahim, T., Tuhulele, M., Le Hoa Binh,Gabunada, F. Jr., Ibrahim, Nakamanee, G., Phimphach-anhvongsod, V., Guodao, L. and Horne, P.M. 2000.Adaptation of forages to climate, soils and use in small-holder farming systems in Southeast Asia. In: Stür,W.W., Horne, P.M., Hacker, J.B. and Kerridge, P.C. ed.Working with Farmers: the Key to Adoption of ForageTechnologies. ACIAR: These Proceedings, see Contents.

Tuhulele, M., Le Van An, Phengsavanh, P., Ibrahim,Nacalaban, W., Vu Thi Hai Yen, Truong Tan Khanh,Tugiman, Heriyanto, Asis, P., Hutasoit, R., Phimmasan,H., Sukan, Ibrahim, T., Bui Xuan An, Magboo, E. andHorne, P.M. 2000. Working with farmers to developforage technologies – field experiences from the FSP. In:Stür, W.W., Horne, P.M., Hacker, J.B. and Kerridge,P.C. ed. Working with Farmers: the Key to Adoption ofForage Technologies. ACIAR: These Proceedings, seeContents.

van Veldhuizen, L. Waters-Bayer, A and de Zeeuw, H.1997a. Developing technology with farmers: a trainer’sguide for participatory learning, Netherlands, ETC, 224 p.

van Veldhuizen, L. Waters-Bayer, A., Ramirez, R.,Johnson, D. and Thompson, J. 1997b. Farmers’ Researchin Practice: Lessons from the Field, London, Inter-mediate Technology Publications,320 ph.

32

Farmer Participatory Research in Latin America:Four Cases

A.R. Braun1 and H. Hocdé2

Abstract

Farmer Participatory Research (FPR) emerged in response to limitations of top-down R&Dapproaches. In Latin America, the principles and concepts of FPR are rooted in earlier participatoryresearch experiences in fields such as education, sociology and health, usually played out within acommunity-development context. Contributions of Paulo Freire and Orlando Fals Borda are dis-cussed briefly. To analyse these experiences, a typology based on decision-making locus inresearch, farmers’ and scientists’ roles, and the style of research conducted was used. Threeapproaches were distinguished: scientist-led, farmer-led and interactive research. Four cases areanalysed: (1) Farmer-to-Farmer program, Nicaragua, founded in 1987 by the National Farmers andRanchers Union (UNAG) based on volunteer farmer-promoters. The focus is on low external-inputagriculture. (2) Diagnosis, Investigation and Participation (DIP), formed in 1994 by a multi-disciplinary team with linkages to the Veterinary Medicine and Animal Science Faculty at theAutonomous University in Yucatan, Mexico. Their objective is to improve the quality of life ofindigenous communities at the forest-agriculture interface through participatory innovation basedon local resources. (3) Farmer Experimentation, initiated by PRIAG (Regional Program forReinforcement of Agronomic Research on Basic Grains) in Central America, in 1991. Theobjective is to increase the self-reliance of small- and medium-scale producers in generating anddisseminating technology. (4) Local agricultural research committees (CIALs), first launched byCIAT in Colombia in 1990, to strengthen rural communities’ capacity as decision-makers andinnovators of agricultural solutions and to exert demand on the formal R&D system. The discussionfocuses on similarities and differences in the processes, principles, roles and relationships under-lying these experiences and key lessons learned.

IN THE development context, participatory researchmay be defined as a process whereby a group or acommunity identifies a problem or question ofinterest, reviews what is known about it, conductsresearch on it, analyses the information generated,draws conclusions and implements solutions (modi-fied from Selener 1997). In this definition, the locusof decision-making rests implicitly within the groupor community involved. Farmer ParticipatoryResearch (FPR) is understood by many to be oneelement in a larger participatory development agenda

that aims to change the orientation of existingresearch and development (R&D) structures, developsustainable community-based research capability andcreate new social and political institutions (Okali etal. 1994).

Intellectual Roots of FPR in Latin America

Participatory research approaches have been devel-oped and applied in four broad areas: (1) communitydevelopment, (2) action research in business andindustry organisations, (3) action research in schoolsand (4) farmer participatory research (Selener 1997).FPR emerged as a response to the limitations ofearlier top-down agricultural research approachesthat often failed to deliver significant improvementsin levels of well-being for the poor in complex, risk-prone environments (Chambers et al. 1989; Conway

1Project Manager, Approaches to Participatory Research,CIAT, Apartado Aéreo 6713, Cali, Colombia. Email:a.braun@cgiar.org2CIRAD-Tera, Centre de Coopération Internationale deRecherche en Agronomie pour le Développement, ApartadoAéreo 855, 2150 Moravia, Costa Rica. Email: henri.hocde@cirad.fr

33

1997). FPR draws upon concepts from earlierparticipatory research experiences, involving a broaddiversity of disciplines including non-formal adulteducation, sociology and health (e.g. Gianotten andWit 1985; Lammerink and Wolfers 1994). FPR inLatin America is particularly strongly rooted inconcepts that emerged from participatory researchexperiences in community development. PauloFreire and Orlando Fals Borda rank importantlyamong the protagonists of participatory approachesand made major contributions to the communitydevelopment field in the late 1960s and early 1970s.

A major break from traditional concepts of adulteducation occurred in Brazil, catalysed by Freire(1970, 1973), who was appalled by the situation ofthe illiterate poor. Freire conceived education to be atool for changing society structurally. Knowledge,which he equated with power, had traditionally beenmonopolised by an elite that sought to protect itsinterests. The learner was considered a passiverecipient of knowledge, deposited by an educator,just as money is deposited in a bank.

Freire organised a multidisciplinary team todevelop educational materials that would stimulatethe poor to reflect on their lives and on the under-lying causes of their conditions. People were organ-ised in ‘cultural circles’ in order to recover theiridentity and indigenous knowledge. Dialogue oncontroversial issues such as land tenure rights wasthe central process, followed by reflection and thenaction. Thought-provoking photographs were used toinitiate this process, which Freire called concienti-zação, a Portuguese word that implies a liberatingprocess, whereby oppressed people evolve towards astate of critical consciousness. Threatened by hisefforts to change society structurally, the Braziliangovernment jailed Freire, who eventually soughtexile in Chile. He later worked in Portuguese-speaking African countries.

Another militant who sought to construct a newsocial order was the Colombian sociologist, OrlandoFals Borda (Fals Borda 1985). Coming from anacademic setting, Fals Borda was isolated from therealities of rural life. Motivated by the desire to finda balance between reflection and action, he and agroup of university intellectuals went out into thefield to bring science to rural people. At first, com-munications barriers and differences between theirconcepts of reality caused the farmers to reject FalsBorda’s group. Moreover, the technological solutionsdeveloped by the university researchers were notapplicable to rural conditions. After a process of deepreflection, Fals Borda radically reoriented his work,no longer treating rural people as passive ‘objects’,but encouraging them to become active ‘subjects’ oragents of their own liberation. For a period, he was

involved in militant action research, collaboratingwith an aggressive association of small farmers thathad invaded lands belonging to large cattle ranches innorthern Colombia. Fals Borda continued to try toimprove local access to technical and scientificknowledge by providing systematic feedback toresearchers about farmers’ needs. The establishmentof feedback mechanisms between farmers andscientists, and the awakening of intellectuals to newperspectives on the realities of rural life were amongthe most valuable outcomes of his work (Fals Borda1987).

FPR draws upon a number of concepts that origi-nated in the work of Freire and Fals Borda and otherearly pioneers – the most notable being:• the iteration or repeated cycling of reflection and

action;• the breakdown of subject-object polarity;• the rejection of passive knowledge banking in

favour of active knowledge acquisition andgeneration through participation in research andanalysis, and application of the results;

• facilitation of the development of critical con-sciousness by external actors.

The Emergence of FPRThe term Farmer Participatory Research was coinedby Farrington and Martin (1987). FPR emerged as aresponse to the limitations of earlier agriculturalresearch and extension approaches such as on-farmand farming systems research and the ‘Training andVisit’ extension model. In these earlier approaches,farmers were often considered as research subjects,components of the system under investigation, orpassive recipients of extension messages. FPR hasreceived increased attention and recognition sincethe ‘Farmer First’ (Chambers and Ghildyal 1985;Chambers et al. 1989) and Participatory TechnologyDevelopment (Haverkort et al. 1988) concepts werefirst introduced in the 1980s. In contrast to earlieragricultural research and transfer-of-technologyapproaches, FPR advocates farmers’ involvement asdecision-makers at all stages of the process,including the early stages of problem definition,prioritisation and the setting of research objectives.

It should be stressed that there is no single model,prescription or recipe for the process that we arecalling FPR. Not surprisingly, an often-confusingarray of approaches, methods, platforms, tools,relationships and terms is associated with FPR. In hisanalysis of the historical development of partici-patory research, Selener (1997) emphasised:

‘The realisation that there is an array of participatoryresearch approaches may be disconcerting at first. Butit is clear that no single approach can address the mul-tiple realities existing within society.’

34

A Typology of Agricultural and NRM Research Strategies

A simple typology of agricultural and NRM researchapproaches1 was developed to facilitate the analysisof FPR experiences. The typology draws upon con-cepts from earlier classifications (Biggs 1989; Pretty1994; Mikkelson 1995; Ashby 1996; H. Hocdé andB. Triomphe, unpublished; S. Sherwood, unpub-lished) and on recent analysis by the CGIAR2

Systemwide Program on Participatory Research andGender Analysis (see Lambrou 2000; Probst 2000).It is based primarily on the locus of decision-makingwith respect to the research process and secondarilyon the respective roles of farmers and scientists andon the style of research conducted. Three broadresearch approaches are distinguished (Table 1):scientist-led, farmer-led and interactive. Rather thana rigid classification system, the typology is intendedas a conceptual tool for analysing the relationshipsamong different approaches, recognising that theyoften have overlapping boundaries.

Scientist-led research

In this approach, scientists determine the researchagenda and conduct formal research, employinghypothesis testing, controls and replication, anddrawing upon the full range of tools associated withthe scientific method. Scientists may consult withfarmers and may involve them as collaborators inexperiments. When consultation with farmers ispractised, it may be with groups or individuals, and itmay employ Participatory Rural Appraisal/Assess-ment (PRA) tools3 (e.g. focus-group discussions;semi-structured interviews; seasonal and historicaldiagramming; mapping and modelling; preferenceand wealth ranking; transect walks; institutionalmapping) to gather information and to improve com-munication with rural people (World Bank 1994).

Farmer-led research

Farmers – Individually or collectively – determine theresearch agenda and conduct the research. Farmers,scientists or other professionals may be involved asfacilitators in one or more steps of the researchprocess and often introduce and apply PRA tools tostructure communication with and among farmers. Ingeneral, farmer-led research is non-formal and doesnot apply concepts such as replication and control.

Interactive research

In this approach, both farmers and scientists areinvolved in determining research agendas, which aremutually integrated, complementary and synergistic.The interactivity arises because the scientists’research is demand driven and the farmers areevaluating the options under development by formalresearch centres alongside other alternatives.Scientists are involved as facilitators of the localresearch process with farmer groups or individuals,and this facilitation may include teaching farmers keyelements of the scientific method, such as replicationand control, offering ideas, information and access totechnology, and helping to establish contacts.

Scientist-led research is the least participatory ofthe three approaches. Although farmers may be con-sulted or involved as collaborators, they are notdirectly involved in decision-making about researchpriorities or about the research process itself. Thisapproach is complementary to farmer-led and inter-active research, and continues to be the principalstrategy for elucidation of fundamental biologicaland ecological principles and processes. Agriculturaland NRM research methods, tools and approachesare other important products of scientist-ledresearch.

It should be noted that the typology does notinclude participatory learning approaches such as theFarmer Field School (FFS) or CATIE’s4 INTA-IPM5

project in Nicaragua6 (CATIE, unpublished). Simi-larities, differences and complementarities betweenparticipatory learning and participatory research

1In this paper, an approach refers to a body of experiencesor platforms that share key principles and processes, butthat may differ in organisational form and deploy differenttools. The concept of platform was originally introduced tothe literature of sustainable agriculture R&D by Röling andWagemakers (1998). A platform is an implemented,coherent set of principles and processes, organisationalforms and tools.2Consultative Group for International AgriculturalResearch.3We define a tool as a conceptual or physical aid or devicethat facilitates the accomplishment of a specific partici-patory research objective.

4Centro Agronómico Tropical de Investigación y Enseñanza.5Instituto Nicaragüense de Tecnología Agropecuaria-Integrated Pest Management.6The CATIE INTA-IPM project has concluded that bothparticipatory and conventional (scientist-led) research areneeded to develop the agricultural technology and strategiesfor strengthened farm family management of ecologicalvariability. The project has identified the need to increasefarm families’ capacity to select and modify efficient andappropriate pest management practices. Farm families andIPM professionals associated with the project have under-taken a broad array of participatory evaluations of IPMmethods for coffee, tomatoes, cabbage and plantains.

35

approaches are discussed in an analysis of FFS andCIALs (Braun et al. 2000).

The remainder of this paper is devoted to compar-ative analysis of several FPR experiences from LatinAmerica. Our objective is to identify similarities anddifferences in the processes, principles, roles andrelationships underlying these experiences and tosummarise key lessons learned.

A review of formal and grey literature was under-taken; however, the scope was not exhaustive. Thereview was conducted in two stages. After compilingthe available information about each experience andconducting an email survey to capture additionalinformation wherever possible (including drafts andsubmitted versions of papers), each case was classi-fied according to the typology, and its essentialaspects were summarised (Table 2). Several of thecases had characteristics of both the farmer-led andinteractive approaches. These are indicated in Table2 as mixed-mode experiences.

Four cases were chosen for in-depth analysisbecause they offered:• systematised analysis of principles, processes,

roles, relationships and lessons learned;• evidence of a proven track record;• diversity and contrast.

The selected cases (Farmer-to-Farmer, DIP,PRIAG and CIALs) span the continuum of farmer-led and interactive approaches. The Farmer-to-Farmer approach clearly falls at the farmer-led endof the continuum. DIP and PRIAG are rooted in theFarmer-to-Farmer approach, but also share elementsof the interactive approach and are considered asmixed-mode cases. The final case, Farmer ResearchCommittees (CIALs)7, represents the interactive endof the continuum.

The FPR Landscape in Latin AmericaA farmer-led experience

Farmer-to-Farmer8

The Farmer-to-Farmer program was founded inNicaragua in 1987 by the National Farmers andCattle Ranchers Union (UNAG)9. It started withexchange visits between farmers from Nicaragua andMexico with the aim of promoting and diffusingappropriate technologies among poor farmers. Theprogram was a reaction to the top-down transfer-of-technology model that prevailed in Nicaragua duringthe 1980s, which promoted expensive technologypackages involving improved varieties, irrigation,imported chemical fertilisers, pesticides and agricul-tural machinery. The program sought to improve soilfertility, productivity and living standards, whilereducing production costs and external dependency.

Soon after their inception, Farmer-to-Farmerexchanges rapidly began to resemble classicaltechnology transfer, but with the role of the exten-sion agent filled by farmers. It became clear that pro-moting externally developed technologies, evenwhen they were of farmer origin, had its limits:hence there was a need for farmers to test candidatetechnologies in their own fields.

In this approach, volunteer farmers or promoters,who are willing to share their knowledge and experi-ence with others, conduct experiments in their own

7CIAL is the Spanish acronym for Comité de InvestigaciónAgrícola Local. Other names in English are FarmerResearch Committees, Local Agricultural Research Com-mittees and Committees for Local Agricultural Research.8Farmer-to-Farmer is known as Campesino-a-Campesino inSpanish.9Unión Nacional de Agricultores y Ganaderos.

Table 1. A typology of agricultural and natural resource management research approaches.

Research type Decision-makers Roles of farmers and scientists Research style

Scientist-led Scientists determine research agenda • Scientists conduct research • Farmers may or may not be

consulted; they may or may not collaborate

Formal

Farmer-led Farmers determine research agenda • Scientists or farmers may facilitate farmer research

Non-formal

Interactive Farmers and scientists develop interactive and integrated research agendas. Decision-making about local research is done locally. Decision-making by formal research services is influenced by their clients’ priorities

• Farmers and scientists conduct research interactively

• Scientists facilitate farmer research • Two-way communication is

organised to facilitate mutual feedback

Farmers may conduct formal or non-formal researchScientists conduct formal, demand-driven research

36

*PTD connotes the specific farmer-led approach developed by the consulting group, ETC International. It is also anumbrella term to describe an approach or activity that combines technology development with participatory methods(Sutherland et al. 1998)

Table 2. Selected Farmer Participatory Research experiences in Latin America.

FPR ExperienceInstitution/

Area of Influence Highlights

Farmer-led

Farmer-to-Farmer 1. UNAG/Nicaragua2. COSECHA/Honduras

1.See Programa Campesino a Campesino 1999, Hocdé in press; Website: http://www.copacgva.org/ifapidc97.htm#UNAG

2.Founded by R. Bunch, Regional Representative of World Neighbours and author of Two Ears of Corn (Bunch 1982). The primary objective of COSECHA is to disseminate the effective use of ecologically sound Agri-Cultural Development. It has training and rural development programs that serve as models in areas such as agroforestry and soil restoration (Bunch 1990; Bunch and Lopez 1994). Website: http://www.desarrollo-rural.hn/docs/curriculum/cosecha.html

Farmer inventors Panamerican School of Agriculture (EAP), Zamorano, Honduras

See Bentley 2000.

Participatory Technology Development (PTD)

1. Ideas/Bolivia & Peru 1.The PTD* concept was originally developed by ETC (Waters-Bayer 1989; Haverkort et al. 1991); Website: {HYPERLINK http://www.etcint.org/)}. Ideas works with Andean peoples, who conduct experiments by themselves. Indigenous knowledge is considered to be of great value and is the starting point for PTD. Emphasis is on low external inputs. Seeks to institutionalise PTD in communities. Role of institutions is to provide support in accessing information and in networking.

Mixed mode

Farmer Experimenters 1.PRIAG (CORECA/EU, IICA, KIT, CIRAD)/ Central America

2.World Neighbours/ Bolivia

3.UNICAM/Nicaragua4.As-PTA/Brazil

1.See text, Hocdé in press, and Jaén & Silva (1995)2.See Ruddell (1994) (Website: http://www.wn.org)3.This NGO has supported the farmer-led experimentation process

for 7 years. Emphasis on women, family and Farmer Experimenter groups. Collaborates with municipalities in some cases. Current emphasis is on introducing concept of Farmer Experimenters in producer organisations.

4.This NGO supports rural farmer associations in learning how to manage their own experiments to strengthen family farming within difficult agroecological systems. Emphasis on generation of local knowledge. Website: http://www.dataterra.org.br/aspta/faspta.htm

DIP FMVZ-UADY/Yucatan, Mexico

See text, Gündel (1998) and Anderson (1998).

Interactive

PPB INIAP, Ecuador; EMBRAPA-CNPMF, Brazil; EAP-Zamorano, Honduras; CIAT; CGIAR SP-PRGA

See CGIAR (1999), Sperling et al. (in press); Website: http://www.prgaprogram.org/

CIALs CIAT – 8 countries of Central & South America

See text and Braun et al. (2000), Ashby et al. in press; Humphries et al. (in press). Website: http://www.ciat.cgiar.org/cials

37

fields. Each takes responsibility for guiding a groupof experimenting farmers from his/her community,visiting them regularly to help with planning, imple-mentation and interpretation of their experiments.Promoters also organise and give training on topicsdetermined by their own accumulated experience andconcrete results, ranging from soil conservation,cover crops, husbandry, forestry and organic agricul-ture to cropping systems and diversification. Theyfocus their experiments on low external-input agri-culture10. Today in Nicaragua, there are 700 farmerpromoters working throughout the country in a widerange of agroecological and socio-economic con-texts. The approach has taken root throughoutCentral America and is applied by many NGOs andin some R&D projects.

The farmers themselves define the researchagenda, manage the experiments and interpret theresults. They may do this individually or in groups.Generally, they do not apply formal scientificmethods such as the use of control plots or repetitions.Two key elements in the Farmer-to-Farmer processare the farmer promoter and the mechanisms of com-munication employed (Hocdé in press).

The promoter’s basic function is to find technicalsolutions to problems in smallholder agriculture. Amain task is therefore to communicate solutions toneighbouring farmers who are also searching foranswers to their problems. In order to have credi-bility as a communicator, a promoter needs to havetested recommendations on his/her own land. Thetwo functions and processes – experimentation andcommunication – are therefore interdependent.Promoters do not recommend technical recipes orpackages, but rather suggestions or ideas to stimulateexperimentation by others. A promoter’s main toolfor convincing others is through mentoring andsetting an example, although workshops and trainingevents are also employed. The goal of the Farmer-to-Farmer movement is to promote a culture of inquiryand experimentation among smallholder farmers.

Sharing and disseminating knowledge horizon-tally is a central responsibility of each promoter.Each communicates intensively with other farmers,as well as with other promoters, using traditionalcommunication tools such as sociodrama, theatre,poetry and music. A diversity of mechanisms such asforums and exchange visits are used, employing abroad range of PRA tools.

Promoters organise exchange visits involvingfarmers and communities. They may involve smallor large groups and be of short or long duration, 1–4days. Experiments are exposed to the critical eye of adiversity of people, each with his or her own per-spective. These are intensive training and learningopportunities and their pedagogical content can beconsiderable. During exchanges, participants explainand discuss results, methods and procedures, oftenamid criticism, argument and debate. Each par-ticipant analyses the strengths and weaknesses of hisideas and results before the group. The atmosphereof mutual reinforcement and encouragement per-meating these events helps motivate farmers to con-tinue experimenting. Learning from mistakes isencouraged, as is the idea that each person followshis own problem-solving path.

Promoters facilitating these situations must beable to create a constructive and productive atmos-phere, and to bring out and synthesise the ideas ofothers to orient and guide the design of new exper-iments. This requires that promoters be highlyskilled in facilitation techniques.

Lessons learned

The Farmer-to-Farmer process can result in radicalchanges in farmers’ mental maps of their role in theprocess of technology generation and diffusion.Through involvement in the program, farmers realisethat they are capable of experimenting, offeringsolutions, communicating and transmitting techno-logical options to others (Merlet 1995).

The Farmer-to-Farmer process builds enthusiasm,self-confidence, pride and hope for the future(Programa de Campesino a Campesino 1999).Motivation grows as creative capacities are tapped,and the attitude of dependency on external actorsdiminishes as farmers begin to identify themselves asexperimenters. The most radical of the farmersinvolved in the program view it as a way to break themonopolisation of the technology-developmentprocess by agricultural professionals.

A number of initiatives within and outside ofNicaragua are applying this approach. Althoughmany of these conserve the horizontal communi-cation between farmers, which is the key methodo-logical element of the approach, they may omit theother key dimension, which is the creation of a socialmovement of innovators, making their own decisionsand supported by their own organisational structures.This has led to some confusion as to what thisapproach really consists of. When looking at theseexperiences, care should be taken to see whichdimensions are actually being applied.

10The rationale behind the focus on low external input agri-culture is that the farmers targeted by the program haveextremely limited access to purchased inputs, live in remotelocations and/or have little purchasing power.

38

Technological lessons

• Farmers’ research themes tend to concentrate onagronomic, animal husbandry and technical issues.

• In some cases, the advent of a solution generatedby promoters leads to excessive promotion of thetechnology over an ongoing search for solutionsto other limiting factors.

• The strong emphasis on low external-input tech-nologies can be a barrier dissuading some farmersfrom participating in the Farmer-to-Farmer move-ment, thus impeding its growth.

Methodological lesson

• Farmers’ concepts of the experimental process aredifferent from those of formal researchers. Forexample, farmers may not limit what they con-ceive of as experimentation to plots specificallydesignated for that purpose.The relationship between Farmer-to-Farmer initia-

tives and the formal research sector have tradition-ally been limited, with a few notable exceptions11.Advocates of the Farmer-to-Farmer approach con-tend that formal researchers tend to consider theexperiments conducted by farmer promoters as anextension mechanism rather than as bonafideresearch. They also complain that promoters havetended to find few useful elements in the technolog-ical solutions offered by formal researchers. Holt-Jimenez (in Scarborough 1995), who participated inthe creation of this program, argues that it is indeedpossible for professionals to participate in theprocess, but that most of them need to be trained bycampesino promoters in order to do so successfully.Overcoming the mutual reservations between pro-moters and professionals would undoubtedly consti-tute a leap forward, thereby improving and enrichingthe work conducted by both. Potential gains from thejoint development of realistic solutions to concreteproblems in farming lie not only in the better designand management of experiments, but also in theincreased diversity of options that would becomeavailable.

Despite its limitations, the Farmer-to-Farmerexperience constitutes an important reference pointfor both the farmers themselves and formal agricul-tural services, in terms of demonstrating the potentialof smallholder farmers as researchers and communi-cators. This approach is of historical significance,because it made a significant break with the conven-tional models of knowledge and technology transfer.

Mixed-mode experiences

Diagnosis, Investigation and Participation (DIP).

In 1991, participatory research and rural assessmentmethodology was introduced into the graduatecurriculum of the School of Veterinary Medicine andAnimal Science of the Autonomous University of theYucatan (FMVZ-UADY), Mexico, as part of anODA-sponsored project (DIP unpublished). The DIPgroup formed in 1994 as a multidisciplinary team of10 members and decided to conduct participatoryresearch on backyard animal production. Their objec-tive is to improve the quality of life of indigenouscommunities at the forest-agriculture interface in theregion through participatory innovation based onlocal resources.

DIP, which emerged in response to the need to fillthe gap between conventional research and farmers’needs that were not being addressed by governmentresearch and extension systems, proposes to facilitateendogenous change among small farmer families toimprove food security and family well-being throughthe sustainable use of natural resources. Equity issought both in terms of benefits from agriculturalinnovation and in the development of gender-differ-entiated knowledge systems. The resulting processcan be divided into four overlapping, interlockingphases (Figure 1), which can be repeated as often asnecessary and be constantly readjusted by the par-ticipants on the basis of new information (Gündel1998).

Appraisal

In this phase, researchers and farmers establish thecommon vocabulary essential for effective com-munication. This is accomplished by using PRAtools such as maps, calendars, ranking exercises andsemi-structured interviews in order to identify theproblems and potentials of the local agriculturalsystem and to understand farmer strategies for itsmanagement. Changes that occurred within thesystem in the past and existing sources of informa-tion on innovations are identified. The outcome isshared information on existing local knowledge andconcepts.

Convergence

In the convergence phase, the actors involved nego-tiate and agree upon shared objectives, activities tobe undertaken together, and the division of tasks andresponsibilities. Negotiation does not always resultin an agreement. Where there is no convergence ofinterests among the actors, the process is generallydiscontinued. In some cases, however, a partial con-vergence can be reached, making it possible to

11An example is the close articulation between the Farmer-to-Farmer program and the Soils Department of theAgronomy School of Nicaragua’s National AgrarianUniversity.

39

Reflection Appraisal

Convergence Experimentation

Processes

decisions made for next cycle

DIPfacilitates

SWOT analysispreference Ranking

future visioning

NGOsFMVZ-UAY

DIP articulateswith other actors

analysis &systematisation

Identification ofevaluation criteria

develop new technologies

modify/adaptexisting technologies

individual, informal trials

identify problems/potentials

establish a common vocabulary

understand farmerstrategies

mapscalendarssemi-structuredinterviewsranking exercisesforce field analysistransects

negotiateshared objectives

define responsibilityand assign tasks

access informationendogenousexogenous

farmer & researcherobjectives made explicit

Figure 1. Processes associated with the DIP experience, Yucatan, Mexico. The thick arrow indicates the initiation point.

40

initiate the process. Having started the process, theparticipants can then re-negotiate specific objectivesin subsequent cycles.

Innovation opportunities are explored and dis-cussed among the participants. Different sources ofinformation can be accessed, for example, from otherfarmer groups (as in the Farmer-to-Farmer approach),or information available from formal research. Anytechnology components introduced by researchersshould have the potential to be modified. Group dis-cussions and exchange visits are the principlemethods used to facilitate the negotiation anddecision-making. The outcome of this phase is theconvergence of different objectives into a sharedstrategy.

Experimentation

Experimentation in DIP is rooted in the Farmer-to-Farmer approach, but stresses greater involvement ofresearchers and other actors. During this phase, par-ticipants either develop new technologies or modifyand adapt existing ones (external or local) to localconditions and needs. The experimentation is carriedout by the participating farmers and is often organ-ised on an individual basis. Farmers determine theexperimental designs. This leads to non-formal trialswithout formal requisites such as replication andcontrol. The objective of the experimentation phaseis to become familiar with the innovations beingtested and adapt them to specific conditions. Theresearchers learn about qualitative issues such asfarmers’ strategies and objectives. The managementstrategies for individual experiments are sharedamong the participants during village exchangevisits, group discussions and frequent visits to eachother’s plots.

The researchers’ role in this phase is to facilitatethe exchange among the participants of the differentways the experiments are conducted and to provideinformation where required. The mechanisms usedinclude interviews, group discussions, field visitsand participant observation, combined with PRAtools. During the establishment of farmers’ exper-iments the researchers discuss their ideas and reasonsfor experiments with individual farmers and supporttheir decisions with additional information.

Reflection

The information generated during the experimentationphase is documented, systematised and analysed bythe participants. Evaluation criteria used by thefarmers are identified and established and are madeaccessible to both farmers and researchers. Decisionson the next experimentation cycle are made based onthe experiences of the previous cycle.

The role of the researchers in this phase is tofacilitate the documentation, systematisation andanalysis of the process. The researchers facilitategroup meetings, both within the communities andbetween communities. Two-day workshops are oftenused to share experiences and to evaluate findings.Tools used during this phase include analysis ofstrengths, weaknesses, opportunities and threats(SWOT), preference ranking and ‘future visioning’.

Linking the local and formal research

DIP facilitates the flow of information between therural communities and the FMVZ-UADY. Thisbegan once research was initiated at the FMVZ-UADY on questions identified from PRA and farmerexperimentation work (Anderson 1998). Thesequencing of the farmer- and scientist-led researchis very important: The starting point is the PRA workdone within the communities. This sets the researchagenda for the formal researchers and influences theway research questions are asked. The nature of thecontrol treatments, the level of significant difference,which response variables are chosen and what levelof response is sought, all depend greatly on infor-mation from the PRAs and farmer-led researchprocess. The DIP group has made some headway inreorienting conventional research work performed atthe FMVZ-UADY towards the researchable ques-tions encountered in the community-based appraisaland experimentation phases (Anderson 1998).

The DIP Group has also developed a strategy offorging partnerships with local NGOs to involveother informed actors in the identification ofresearchable questions. These alliances enable theNGOs to facilitate the uptake of the outputs of theparticipatory research by other communities wherethey are involved.

Based on their experiences, DIP has identified anumber of lessons with respect to methodology,technology options and scaling up:

Methodological lessons

• Food security is of top priority among the verypoor; thus, initiatives focusing only on NRMissues will fail to achieve convergence.

• Farmers’ innovative capacity, knowledge avail-able resources and their priorities will be hetero-geneous, resulting in different economic strategies

• When there are many individuals experimenting,the pooling of their information can fulfil thefunction of replication.

Technological lessons

• In smallholder systems, there is no one ‘best’practice, because farmers are always makingmodifications to suit their specific needs.

41

• Specific technologies have a life span. New tech-nologies substitute for former ones as conditionschange. This implies that farmers are involved inan ongoing search for suitable options.

• New options should be presented as modifiablecomponents rather than as fixed and final solutions.

Scaling up

• Sufficient resources for exploiting complemen-tarity between formal and farmer research need tobe committed from the onset in order to providecontinuity and follow-through on R&D priorities.

• Scaling out (i.e. starting new initiatives) may bemore advantageous than scaling up (i.e. replica-tion of a pilot-scale initiative). Small grants maybe a useful mechanism for stimulating scaling out.

PRIAG (Regional Program for Reinforcement of Agronomic Research on Basic Grains in Central America)

PRIAG is an example of various experiencesknown collectively as Experimentación Campesinain Spanish, hereafter referred to as FarmerExperimentation.

PRIAG emerged as a response to the erosion inpublic-sector capacity for research and extension. Aprimary objective is to increase the autonomy andself-reliance of small- and medium-scale producersin the generation and dissemination of technology,by strengthening farmer capacity to identify, obtain,modify, adapt, disseminate and use information,knowledge and technology (Jaén and Silva 1995).PRIAG’s mandate is to increase the responsivenessof national research and extension systems tofarmer’s needs. It approaches this mandate indirectlythrough two main strategies (Hocdé 1997). 1. enhancing farmer’s capacity to engage in dia-

logue with researchers and extension workers, sothat they have their own specific space and role inthe research-extension-farmer chain;

2. strengthening farmers’ capacity to investigate andinnovate.PRIAG strives to complement existing agricul-

tural technology generation and dissemination effortsthrough enabling more localised and independentinnovation, while responding to the need forachieving agricultural improvement among largerpopulations in a cost-effective manner (S. Sherwoodunpublished). Although it is firmly rooted in Farmer-to-Farmer methodologies, PRIAG has gone farther inarticulating farmer and formal research.

PRIAG focuses on human resource developmentprocesses to develop linked cadres of Farmer Exper-imenters (FE) and Farmer Communicators (FC)(Jaén and Silva 1995). A number of organisations in

various Central American countries take part inPRIAG (see Table 2), but discussion will focus onPRIAG-Panama and Guatemala12 13.

Agricultural professionals (facilitators) fromPanama’s national agricultural research institute(IDIAP) and the Ministry of Agricultural Develop-ment (MIDA) identify Farmer Experimenters (small-and medium-scale producers of basic grains) basedon the criteria in Figure 2. After training in basicpresentation skills, Farmer Experimenter participatein a workshop facilitated by professionals fromIDIAP or MIDA, where they present their experi-ences, identify the agricultural problems in each oftheir communities and brainstorm solutions. Afterthe workshop, training in basic project formulationskills is provided, and the Farmer Experimentersmeet with IDIAP and MIDA facilitators to prioritisethe problems identified in their communities and toformulate research projects.

The objective of FE experiments is to evaluatepotential technological alternatives. Each farmerexperimenter develops and presents his/her annualresearch work plan at a local planning and evaluationworkshop. At these workshops, which last 3–5 days,Farmer Experimenters, extension agents andresearchers present the results of their previousseason or cycle of research, so that it can be evalu-ated by all present. Then the three groups of actorspresent proposals for the upcoming cycle. TheFarmer Experimenters can reject the proposals offormal researchers and vice versa (Hocdé 1997). Theexecution of the work plan developed by the FarmerExperimenters is their responsibility; however,‘collegial’ support (Biggs 1989), oriented towardspromoting the non-formal, indigenous experimen-tation systems in existence, is available from IDIAPand MIDA staff.

A number of training ‘modules’ were developed byIDIAP and MIDA to meet demands expressed by theFarmer Experimenters and in response to knowledgegaps identified by the Farmer Experimenters and thefacilitators14.

Once the Farmer Experimenters have compiledresults from their experiments, training is providedin simple data analysis and interpretation methods.Each Farmer Experimenter presents results in a

12PRIAG teams are also found in Costa Rica, El Salvadorand Nicaragua. Each team innovates in response to thelocal context.13The agencies involved in Panama are the national agricul-tural research institute (IDIAP) and the Ministry of Agri-cultural Development (MIDA).14The training modules are Farmer Research; Methods ofknowledge transfer; Monitoring and Evaluation; Soil Con-servation, Crop Agronomy and IPM.

42

Figure 2. Processes associated with Farmer Experimenters in the PRIAG-Panama experience (Jaén and Silva 1995). The thick arrow indicates the initiation point.

basic trainingdata analysis

publication of resultspresentations in communities

presentations in local Planning& Evaluation workshops

new projects formulated

innovationscommunicators

implementorsevaluators

FE roles

colleagues of FEstechnical backstoppers of FEs

Scientist roles

varietiessoil fertility/conservation

peststhemes

innovationleadershipcommunication abilityinterest

criteria

Diagnose agriculturalproblems in FE’scommunities

Train in presentation skills

identify technologies withpotential for resolvingpriority problemstrain in project developmentpresent project/plans at localPlanning & Evaluation workshops

develop training modulesoriented towardsfarmers’ expressedpriorities

Processes

Execute

Analyse, discuss & disseminate results

Identify FEs

Exchange &systematise FEexperiences

Develop projects & work plans

projectsresearch

43

workshop held in his/her community and in a localworkshop. Both are attended by other farmers and byMIDA and IDIAP facilitators. New research projectsare formulated at these workshops, thereby con-tinuing the cycle and building upon previousexperience. This process allows experimentingfarmers and researchers to refine their research andprovides mutual feedback on relevant topics as inputto the agendas of Farmer Experimenters, scientistsand other agricultural professionals.

In the PRIAG approach, the FE processes arelinked to another set of processes that stimulate theflow of the research results though existing localinformation channels and networks. Farmer Exper-imenters with innate ability as communicators15 areidentified and trained in information diffusionmethods, with development and delivery of radioprograms by farmers, for farmers, as a centralelement in some cases. Other communicationmodalities include field days, exchange visits andcommunity meetings, planned, organised and facili-tated by Farmer Experimenters. Visiting otherfarmers’ experiments requires Farmer Experimenterswith different kinds of facilitation and communica-tion skills than those required by Farmer Exper-imenters working as radio correspondents.

In Guatemala, Farmer Communicators have beentrained as radio program developers and presentersin workshops, where they produced and broadcastradio shows. After the Farmer Communicatorsdeveloped the themes to be presented, these wererecorded and evaluated in plenary workshop sessionsand recordings were finalised based on the resultingfeedback. Farmer Communicators learned how tocombine the agricultural content with music, drama,advertising humour and other elements. Program-ming was developed in local languages in somecases. To scale up the process, successful FarmerCommunicators committed time to training newFarmer Communicators in regional workshops.

Evolution of PRIAG

The key steps in the PRIAG process are: • helping farmers to organise and communicate;• helping farmers to experiment;• negotiating concerted annual work plans among

interacting Farmer Experimenters, researchers andextension workers. As these processes have unfolded, capacity

building in areas critical to experimentation has takenon greater importance. New areas of capacity-building include the biology and dynamics of

managed and natural ecosystems, experimentaldesign and analysis of results. In addition, PRIAG isresponding to higher levels of information demandby Farmer Experimenters by encouraging the partici-pation of experts and the introduction of more formalscientific tools and methods of analysis (S. Sherwoodunpublished).

PRIAG has not invested in the formal training ofresearchers and extension agents in participatoryapproaches to working with farmers because itsprimary strategies for influencing the agendas ofnational agricultural services were to strengthen thefarmers’ capacity to engage in dialogue withresearchers and to improve their research capacity.Interaction with, and feedback from farmers, hasinfluenced the setting of research priorities by scien-tists. Changes in attitudes have occurred in farmersand scientists alike, with sharp rises in farmer self-esteem, as they become advisors to fellow farmersand to scientists. As scientists have experimentedwith the additional role of facilitation, many havediscovered the necessity to complement, update andimprove their professional capacity. Nevertheless,many have become discouraged by the magnitude ofthe changes required in their modes of working. Con-tinued strengthening of the whole PRIAG processrequires that the technology supply be increaseddramatically, and stronger participation of the formalresearch system is viewed as essential for the furtherdevelopment of the approach (H. Hocdé and B.Triomphe, unpublished). PRIAG now aims toenhance processes and mechanisms for moredemand-driven research and to introduce new actorsto agricultural knowledge and technology generation.

A key strategy used by PRIAG to foment FEprocess is the organisation of well-structuredexchange visits between Farmer Experimenters, sup-ported by teams of professionals from national agri-cultural research services, NGOs or other projects.Exchanges involving PRIAG teams composed offour Farmer Experimenters and one scientist orextension worker have proved an effective way tofoment innovation and share experiences. Theexchanges involve organisational as well as techno-logical issues (Hocdé in press).

PRIAG’s view to the future is that Farmer Exper-imenters need to be well organised in order to beable to pressure national agricultural services torespond to their needs effectively. Strong FarmerExperimenters are necessary, but not sufficient.PRIAG believes that producer organisations alsoneed to become better organised if they are torespond more effectively to the problems of theirmembers and to exert pressure on research services.

PRIAG has concluded that successful initiatives arethose that strengthen critical thinking, organisational

15The other criteria for selecting FCs were high interestlevel and leadership capacity.

44

and operational abilities, and other life skills. PRIAGhas identified the following lessons learned withrespect to global strategy and methodological issues:

Global strategy

• The capacity to experiment and innovate, and tomanage these processes (including the selection ofFarmer Experimenters) should be in the hands ofproducer organisations that can defend theinterests of smallholder families. This requiresstrengthening the negotiation skills of producerorganisations. If producer organisations do notexist, mechanisms for stimulating their formationshould be given priority.

• Relationships between producer organisations andagricultural research and extension servicesshould be formalised in order to foment a cultureof mutual accountability.

Methodological issues

• Farmer Experimenters should be considered notonly as users but also as generators of knowledgeand information.

• Farmer Experimenters should be selected by pro-ducer organisations based on their capacity forinnovation and communication.

• Effective communications mechanisms amongdifferent actors (e.g. farmer to farmer, betweenfarmer and scientist) are essential.

• A diversity of organisational and methodologicalapproaches should be encouraged.

• Linkages with a diversity of actors and sectorsshould be encouraged in order to avoid thecreation of self-limiting FE ‘ghettos’ and to diver-sify the scope of FE experience towards non-agricultural as well as agricultural problems.

• Entry points such as IPM, soil fertility and NRM,should be combined and integrated.

• Limited farmer research experience in animal hus-bandry has resulted in a paucity of methodologicalapproaches for this area. This is a gap thatrequires attention.

• It is critical that the information generated via theexperiments conducted by Farmer Experimentersshould be sufficiently rigorous and reliable for itto be transmitted confidently to other farmers. Complementary lessons from other farmer-led

approaches (Table 1) summarised by Larrea andSherwood (in press) include: • Technology as the exclusive or primary focus of

projects should be avoided. Technologies aretools, and tools alone are not enough to tackle thesocial issues underlying poverty and naturalresource degradation.

• Motivation to participate should be inspired byrecognisable successes, not by paternalism in theform of gifts or subsidies.

• Starting small is important. Projects should avoidunnecessary complexity and demands on time andresources. Starting with simple, manageableprojects permits people to build up their con-fidence and abilities without having to assumesubstantial risk.

• As participants gain experience, the scale andcomplexity of research projects can be increasedand new priorities can be taken on.

An interactive experience

Local agricultural research committees (CIALs)

A CIAL is a local research service belonging to andmanaged by a rural community. The research team ismade up of volunteer farmers, chosen by the com-munity for their interest in and aptitude for research.The main objectives (Figure 3) are to strengthen thecapacity of rural communities as decision-makersand innovators of agricultural solutions, and toincrease their capacity to influence and exert demandon the formal R&D system. CIALs link farmer-researchers with formal research services, therebyincreasing the capacity of local communities notonly to exert demand on the formal system but alsoto access new skills, information and researchproducts that could be useful at the local level.

The first CIALs were established in the Colom-bian province of Cauca in 1990 by the CIAT partici-patory research project. Since then the approach hasspread to other countries in South and CentralAmerica. Today 36 institutions, including govern-mental organisations, and universities in eightcountries, have established nearly 250 CIALs(Ashby et al. in press).

Successful CIALs require that facilitating organ-isations adhere to the following basic principles:• Knowledge is generated through learning by

doing, and building upon experience.• The foundations of the interactions among the

CIAL, community and external actors are mutualrespect, accountability and shared decision-making.

• Technologies are generated and/or modifiedthrough systematic comparisons of alternatives ina participatory process.

• The research products are public goods.• Partners share risks inherent in research.

The staircase (Figure 4) is a metaphor for theiterative process followed by the CIALs.

45

Facilitating the process

Facilitation, like monitoring and evaluation, isdepicted as a pillar that supports all the other CIALprocesses. The farmers providing the research servicehave a formal link with a research institutionmediated by a trained facilitator. He/she may be afarmer who has been a CIAL member or a pro-fessional from an NGO, research institution or exten-sion service. The facilitator initiates the CIAL processby convening a motivational meeting in the commu-nity and supports the CIAL members until they are

able to manage the entire process independently. Thefacilitator must respect local knowledge and acceptrisk as a normal characteristic of experimentation.

Training in CIAL processes is provided in thecommunity through regular visits by the facilitator. Itequips the local farmer research team to conductexperiments that compare alternatives with a controltreatment and that employ replication in time andspace. The training also familiarises the farmerresearchers with terminology that will give theirresults credibility with the formal research system,

Figure 3. The CIAL objectives.

Figure 4. The research staircase, a metaphor for the iterative CIAL process.

Feedback

Facilitation,monitoring

andevaluation

Analysis

Evaluation

Experimentation

Planning

Diagnosis

Election

Motivation

Objectives

develop pride in local values,capacity, and knowledge

link farmer experimentationto formal research & extension

improve farmer capacityto access servicesand negotiate withexternal agencies

increase diversityof new technologies

generate knowledge

stimulate and acceleratefarmer innovation

build management skills

increase, conserve anduse local biodiversity

improve capacity forcritical analysis

46

but that is communicable to local people. Thetraining also builds skills related to planning,management, meeting facilitation, monitoring andevaluation; record keeping and simple accounting.To reinforce the development of these skills, thefacilitator encourages regular communal or collec-tive reading aloud and group discussion of the CIALhandbooks by the farmer research team and as manyothers as are interested.

Facilitators are expected to respect the researchpriorities established by the community and thedecisions made by the farmer research team indefining experimental treatments and evaluationcriteria, generating recommendations and managingresearch funds.

Profound changes are needed in the relationshipsamong farmers, rural communities and agriculturalprofessionals if all the CIAL processes are to beexecuted successfully. A change of attitude is fre-quently required before a professional is capable offorming and facilitating a CIAL. The training offacilitators therefore begins with a sensitisationprocess and learning new communication skills. Thefirst lesson is to avoid the leading questions that sooften characterise a researcher’s interactions withfarmers. Instead, facilitators learn how to ask openquestions that permit true two-way communicationwith their clients. Another change that facilitatorsmust make—but one that is even more difficult toachieve – is to cease promoting their organisation’sagenda.

A facilitator begins training with a 2-week courseand continues throughout the formation of his or herfirst CIAL. During the first year of work with theCIAL, he/she has the support of a professionaltrainer with several years of experience as a facili-tator. The trainer visits the CIAL at key moments(diagnosis, planning and evaluation; see below),monitors processes and provides feedback to thecommittee and to the facilitator, and points outstrengths and weaknesses. After the first year, ayearly follow-up ensures that the facilitator and theCIAL have access to an expert with experience insubsequent phases of the process as the CIALsevolve.

Motivation

The facilitator invites everyone in the community toa meeting. Enough information about the nature andpurpose of a CIAL is provided for the participants toevaluate whether they want to establish one. Thefacilitator asks farmers to analyse what it means toexperiment with agricultural technologies or optionsfor managing local resources. Local experience inexperimentation and its results are discussed,

together with the possibility of accessing technologyfrom outside the community. If the communitydecides to form a CIAL, it elects a committee with aminimum of four members to conduct research on itsbehalf.

Managing risk

A CIAL fund is established to help absorb researchrisks. The fund is initiated from seed money, whichmay take the form of a one-off donation from thefacilitating organisation. Alternatively, it may beprovided from a rotating fund managed by an associ-ation of CIALs16 (Ashby et al. in press; Humphrieset al. in press), or it may be raised by the CIAL itself(N. Gamero, EAP, personal communication). Thefarmer research team uses the fund to procure inputsneeded for their experiments that cannot easily beprovided in kind locally, and to compensatemembers for losses incurred. The fund is owned byand is established in the name of the community.The CIAL and the community are jointly responsiblefor assuring that decapitalisation does not occur andare expected to contribute to building the fundthrough collective efforts.

Electing CIAL members

A key selection criterion for elected members is thatthey should be experimenting on their own, and areable and willing to provide a service to the rest of thecommunity. CIAL members agree to serve for aminimum of one year. Each elected member agreesto take part in a regular training and capacity-building process over at least one year. Each has aspecific role as leader, treasurer, secretary or com-municator. The CIAL is often assisted by severaladditional volunteers.

Diagnosis

The research topic is determined through a groupdiagnosis in an open meeting of the community. Thestarting point for the diagnosis is the question: ‘Whatdo we want to know about?’ or ‘What do we want toinvestigate?’ The objective is to identify research-able questions of priority to the community. Thetopics generated by the discussion are prioritised byasking questions about the likelihood of success,who and how many are likely to benefit, and the esti-mated cost of the research.

16Thus far, CIAL associations have been established inColombia (CORFOCIAL) and Honduras (ASOCIAL).

47

The research cycle

The iterative research process includes the followingsteps:• Planning

The experiments carried out by the farmer-researcher team generate information on techno-logical options of local or external origin that areof interest to the community. The experiments donot demonstrate technologies or teach principles.Alternatives from outside the community need notbe finished products. Offering access to tech-nology while it is under development, and makingadjustments based on the feedback obtained fromthe CIAL, is a powerful mechanism by whichresearch organisations can respond to farmerneeds and priorities.

The facilitator helps the farmer-researcher teamobtain the information required to plan the exper-iment. Other farmers and staff of formal researchand extension services are often consulted. If theinformation gathered indicates that the selectedtopic should be modified, this is discussed at acommunity meeting.

The facilitator helps the CIAL formulate a clearobjective for each experiment. The objectiveshould guide the CIAL in all the decisions itmakes from design to evaluation. Based on theobjective, the CIAL decides what, how and whento evaluate the trial. It also determines levels ofexperimental variables, criteria for evaluatingresults, comparisons to be made, data to be col-lected and measurement units to be used.

• Establishment and management of the experimentThe CIAL carries out the experiment as planned.The costs of the inputs are covered by the CIALfund.

• EvaluationThe farmer-researcher team meets with the facili-tator to evaluate the treatments, compare themwith the control and record the data.

• AnalysisThe CIAL draws conclusions from the exper-iment. Their analysis includes the question: ‘Whathave we learned?’ Analysis of the process isespecially important when an innovation is unsuc-cessful or when unexpected results are obtained.

Iteration of processes

The facilitator guides the CIAL through three suc-cessive experiments. In the first, known as theexploratory or preliminary experiment, the CIALtests innovations on small plots. There may beseveral treatments, such as different crop varieties,fertiliser amounts or types, sowing dates or densities.The exploratory trial is a mechanism for eliminating

options that are unlikely to succeed under local con-ditions. If the objective is to compare the perform-ance of different crop varieties, eight to ten materialsmay be planted including at least one local control,and the area planted may be in the order of three tofour replicates of eight to ten rows, each five metreslong. The treatments selected as the most promisingare then tested on larger plots in a second exper-iment. In a comparison of varieties, the secondexperiment might consist of five materials planted inten rows ten metres long. Finally, two or three top-performing choices are planted over a still largerarea in the third experiment, often called the pro-duction plot. A production plot for top choicevarieties might consist of three or more replicates of20 to 30 rows of 20 to 30 m. After this, the CIALmay continue with commercial production if itwishes, or switch to a new research topic.

To begin on a small scale is fundamental. Smallplots provide the CIAL with the experience ofapplying new concepts such as replication andcontrol and allow it to gain confidence beforemoving to larger and therefore riskier scales. Small-scale experiments allow the CIAL to screen outoptions that have little likelihood of success.

As the CIAL becomes proficient in managing theprocess, the facilitator reduces the frequency ofvisits. The number generally drops from two visitsper month for new CIALs, to one every three or fourmonths in mature CIALs (for a contrasting case seeHumphries et al. in press). The main purpose offacilitator visits to mature CIALs is to acquire feed-back on research priorities and results, and to pro-vide the CIAL with access to technology underdevelopment by formal research services.

Providing feedback

Open meetings are held with the community on aregular basis. The CIAL presents its activities,reports on progress and makes recommendationsbased on its experiments. It also reports regularly onthe state of its finances. This is essential in creating aclimate of accountability to the community andensuring that research products become publicgoods. In turn, the facilitator is responsible forensuring that CIAL research priorities and resultsreach the formal research system.

Monitoring and evaluation

Monitoring and evaluation is a mechanism forbuilding mutual accountability among partners in theCIAL process. The community evaluates the per-formance of the farmer-researcher team and maydecide to replace any member. The CIAL keepsrecords of its experiments, which belong to the

48

community and are available for consultation. Thefarmer-researcher team is also responsible for theappropriate use of the CIAL fund. It should informthe community of financial decisions, expendituresand cash inflow.

The CIAL formally evaluates the support receivedfrom the facilitator and shares these results with thecommunity and with the facilitating organisation.Experienced trainer-facilitators visit CIALs formedby new facilitators to monitor the evolution of theCIAL process and provide timely feedback to boththe facilitator and the CIAL members. They assessthe CIAL’s understanding of the research processand capacity for self-management.

Evolution of the CIAL process

Complexity of research issues. The majority of CIALs initiate their work as com-

munity research services with experiments involvinggermplasm. These CIALs are generally trying toincrease productivity of staples such as maize, beans,potatoes or cassava in order to improve foodsecurity. Women’s CIALs are often concerned withimproving family nutrition and may decide to inves-tigate production of protein sources such as soybeansor small livestock. As food security improves,CIALs generally begin to search for ways togenerate more income, often seeking to diversifyproduction to include non-staple species. Manyexperiment with fruit or vegetables at this stage.

Beginning with research on varieties, or new cropspecies, creates a firm basis for maturation and evo-lution. CIALs researching germplasm-related issuescan obtain useful results from small-scale experi-ments and thus build their confidence. Those thatbegin with a poorly defined or overly complexresearch objective often experience frustration; and,if the facilitator is not successful in helping themextract lessons from a ‘failed’ experiment, they maybecome demotivated and cease their activities. AsCIALs become more experienced, they are better pre-pared for and more apt to tackle more complexissues, such as pest or disease management or soilfertility problems. Nevertheless, pest and soilproblems generally require consideration of scaleissues, more sophisticated problem-solving capacitiesand integration of strategies. They also requireknowledge of biological and ecological processes.Finding ways to build this knowledge is a currentchallenge facing the CIAL approach. The solutionmay involve integrating participatory mechanismsfor building upon and enriching local knowledge.

One of the most difficult changes to achieve in afacilitator is that she/he is able to resist promoting,consciously or subconsciously, her/his organisational

agenda. Respect for farmer knowledge and prioritiesis therefore heavily stressed in facilitator training aspart of the effort to avoid biasing CIALs withresearchers’ priorities and concepts. The intention ofpreventing the intrusion of organisational prioritiesand agendas into CIAL decision-making may alsoexplain the lack of an explicit enrichment element inthe CIAL process. Nevertheless, CIALs can and domake requests for training to their facilitators, andtraining on Integrated Pest Management (IPM) is oneof the commonest. As enrichment processes areoptional, CIALs have not yet fully capitalised onavailable participatory learning approaches, such asthe Farmers Field School (FFS) (Braun et al. 2000) orthe CATIE-IPM experience. The result is that CIALsconfronting pest problems, for example, may not gobeyond comparing pesticides because they may lackthe biological and ecological knowledge required toformulate research objectives that reflect a search forecological alternatives and strategies for the manage-ment of ecological problems.

Because of the emphasis on risk management, theCIAL process is initiated with small-scale exper-iments. However, many biological or ecologicalprocesses occur over larger scales, and someproblems cannot be managed on a plot or field scale.Management of resources such as water, soil ornatural enemies, for example, often requires coordi-nation of action beyond the boundaries of a singlelandholding. Scale and related collective-actionissues need to be addressed explicitly when researchobjectives are formulated and in the planning, execu-tion and analysis of experiments. At present, theseissues are not explicitly addressed in the training ofCIAL facilitators; consequently, they are rarelyaddressed during the research process itself. Incor-porating participatory learning mechanisms, eitherdirectly within the CIAL or indirectly through estab-lishment of linkages to participatory learning initia-tives, is an evolutionary direction that could enhancethe capacity of the more mature CIALs to undertakemore complex research challenges.

Maintaining motivation in poor communitiesThe organisations working with CIALs have

generally focused their efforts in poor communities.Several interrelated management issues have emergedfrom this choice. Firstly, concrete improvements inwellbeing are often important for maintaining themotivation of the CIAL members and for retaining thesupport of the sponsoring community. CIAL partici-pation in community development activities has beenan effective way of increasing interest and participa-tion in CIALs in very poor hillside communities ofHonduras (Humphries et al. in press). Several typesof community-development projects have been

49

conducted by these CIALs in addition to theirresearch. Examples include: • training in food processing techniques;• germplasm multiplication initiatives;• formation and facilitation of new CIALs (often for

women);• vaccination campaigns;• campaigns to protect and conserve water resources;• education of schoolchildren in the research process;• establishment of rotating credit schemes.

Such projects create social cohesion and permitthe CIALs to undertake longer-term research whileensuring their sustainability as research services.

Secondly, where poverty is deep and socialcapital17 is low, CIAL members may be motivated toprivatise research products and other benefitsresulting from the CIAL process. There may also bea tendency for the CIAL to be composed of thebetter off members of the community. These andother factors can feed upon each other, causingfriction and resulting in violation of the CIAL prin-ciples by the CIAL itself. Monitoring by the facili-tator and the skilful and timely introduction ofcorrective actions is critical. In Honduras violationsof the CIAL principles related to public goods andaccountability have been successfully managed byinvesting effort in obtaining the participation of mar-ginalised individuals or groups and by opening upthe CIAL process to more members (Humphries etal. in press).

CIAL associationsAn analysis of failed CIALs (Ashby et al. in press)

showed that discontinuation of activities is oftenrelated to lack of continuity in program goals, staffingand funding among supporting organisations and topaternalistic policies, resulting in violations of the

principles of mutual accountability and risk-sharingby partners. In search of a more stable institutionalframework for the CIALs, CIAT facilitated the estab-lishment of CORFOCIAL, an association of the 46Cauca CIALs as a means of stimulating a higherdegree of self-management and autonomy. CORFO-CIAL has absorbed many CIALs that were inade-quately supported or abandoned by their originalcounterpart organisations. CORFOCIAL organisesannual CIAL meetings in Cauca and has sponsored alarge number of cross-training visits and other enrich-ment activities. It has also provided loans to launch anumber of agro-enterprise development projects andhelped obtain seed money for several more. TheAssociation has obtained its legal status, is learningto manage administrative and technical responsi-bilities, and is developing a solid bridge betweenmember CIALs and research organisations.

Taking the CORFOCIAL model as a starting point,Participatory Research in Central America (IPCA)18

facilitated the formation of ASOCIAL in Honduras in1999. One of ASOCIAL’s priorities is to providemechanisms for replenishing and incrementing theCIAL fund. A second related priority is the diversifi-cation of CIAL activities. Both strategies stimulatethe flow of benefits to the communities and help sus-tain motivation and participation in the CIAL process.

Many CIALs submitted projects to ASOCIALrequesting loans for commercially oriented produc-tion of maize, beans, pigs or chickens. Upon sale ofthe produce raised via the project, each returns itsloan plus a small amount of interest to ASOCIALand deposits the remainder in its own fund. An addi-tional benefit is the building of local capacity in theformulation and presentation of projects. Other bene-fits from ASOCIAL activities include:

• the establishment of community shops that reducethe cost of purchasing basic products, increaseopportunities for commercialisation of localproducts and reduce the time and money spent ontravel to commercial centres,

• the establishment of rotating savings and creditsystems to increase savings capacity and provide asource of credit that does not require extensivepaperwork or collateral.

These accomplishments testify to the capacity of asecond-order organisation to contribute to sustain-ability by promoting broader community developmentobjectives and overcoming the limitations of formalresearch organisations, the narrow mandates of whichconstrain their role in development.

17The concept of social capital was introduced by JamesColeman in 1988 and has been expanded upon by Bourdieu(1993), Putnam (1993) and Fukuyama (1995). In the 1950seconomists hypothesized that the key difference betweenrich and poor countries lay in the amount of physicalcapital. After disappointing foreign assistance experiencesin less-developed countries in the 1960s, the concept ofcapital was broadened to include human capital. Morerecently, the focus has broadened again to include institu-tional requirements for economic growth such as social net-works, legal frameworks and relations of trust, summed upunder the heading of social capital and reviewed by Harrissand de Renzio (1997) and by Wall et al. (1998). Putnam’sdefinition (1993) of social capital is: Features of socialorganisation, such as networks, norms and trust, thatfacilitate co-ordination and co-operation for mutual benefit.Wilson (1997) defines social capital as a propensity forindividuals to join together to address mutual needs and topursue common interests. 18Investigación Participativa en Centro America.

50

Lessons learned

Institutional issues• Institutional commitment to the CIAL approach

translates into investment in training facilitatorsand committing sufficient time and resources sothat facilitator responsibilities to the CIALs arefulfilled. Unless there is serious commitmenton the part of a supporting institution, the prin-ciples of the CIAL process will be violated.When this occurs, CIALs often discontinuetheir activities unless there is a CIAL associa-tion that can fill the gap.

The quality of facilitation is of paramountimportance. Facilitators must be able to guidethe CIAL process without dominating it and tocede responsibility to the CIAL as its capacityto manage the process independently develops.They must be flexible enough to deal withevolutionary changes as the CIAL matures.Additionally they must be able to monitor pro-cesses within the CIAL and relationshipsbetween the CIAL, the community and otherexternal actors and be prepared to negotiate orfacilitate solutions to the problems that willinevitably arise.

• The formation of regional and national associa-tions of CIALs greatly enhances their sustain-ability and potential as community services.The benefits of CIAL organisations are manyand include:1. Enhanced communication and exchange of

research products among CIALs;2. development of local management, organ-

isational and negotiation skills;3. building of financial capital that can be used

to undertake community developmentprojects, initiate new CIALs, build CIALfunds, decrease dependence on externalactors and thus ensure their sustainability;

4. increased credibility with governmental andother formal services and enhanced abilityto influence and exert demand on them.

Methodological issues• Establishing CIALs in very poor communities

requires particularly good facilitation, monitoringand evaluation skills. Literacy and other skills aregenerally at lower levels in poorer communities,hence more time may be required to master theCIAL processes. If the capacity for association(social capital) is low, greater emphasis on plan-ning, the establishment of norms and on negotia-tion of responsibilities will be required. LargerCIALs that increase representation of women andother marginalised groups have been found to

reduce the risk of privatisation of benefits undersuch circumstances.

• The length of time required to realise benefitsfrom the CIAL process is often important inpoorer communities. Research themes that bringabout concrete improvements in food security,such as evaluation of varieties of staple crops, canhelp build confidence and maintain the motivationlevel. Embedding CIAL research activities withina broader community development context isanother way of securing the flow of short-termbenefits. Credit/savings schemes, rural stores andproject development support, are examples oftypes of community development activities thatcan be catalysed by CIALs. The establishment ofCIAL associations increases local capacity forsuch undertakings.

• Second-order associations (ASOCIALs) can becreated once there is a nucleus of CIALs func-tioning in a given area. These associations canconsolidate efforts among member CIALs, pro-vide information and networking services, feed-back to institutions, training opportunities andfacilitate horizontal diffusion of technology.Ashby et al. (in press) provide information on the

institutional costs of establishing CIALs and analysesof the impact CIALs have had on local experimen-tation and innovation, diffusion of technologies andon levels of well being in poor communities.

Cross-cutting Lessons from Farmer-to-Farmer, DIP, PRIAG and CIALs

• All four experiences draw upon the concepts ofFreire and Fals Borda of rejection of passiveknowledge banking in favour of active knowledgeacquisition and generation. These experiencesview farmers not only as users but also as devel-opers and transmitters of knowledge and of tech-nology. Other concepts which have also beenmapped over to these FPR cases from earlier par-ticipatory research experiences include the inter-action of processes, the cycling of action19 andreflection20, and the importance of feedback askeys to ensure learning from the experimentationprocess.

• Each experience has generated diverse tools andmethods to support the practice of FPR. All fouruse PRA tools to facilitate communication withfarmers and rural communities; however, the

19Action refers to processes such as planning, experimen-tation and evaluation20Reflection refers to processes such as analysing resultsand also extracting what has has been learned from theprocess itself.

51

application of these tools is not synonymous withFPR, which has been defined as the involvementof farmers in decision-making at all stages of theresearch process.

• A common thread running through all four experi-ences is the importance of building confidenceand reducing risk when innovations are beingtried. Risk management usually involves startingon a small scale.

• The farmer-led and interactive experiences areconvergent, both recognising that the developmentof knowledge and technology is synergised whenit occurs in a process involving different types ofactors. This convergence is also manifested insimilarities in principles and processes. Thecoexistence of approaches at different pointsalong the continuum from farmer-led to inter-active invites a multi-tiered approach involvingthe integration of networks of rigorous farmerresearcher ‘experts’, less rigorous community-based research networks, and large-scale indi-vidual, non-formal experimentation.

• A conclusion common to all the cases presented isthat the essential factor in strengthening farmerinnovation capacity is not technology per se butrather the construction of social processes thatsupport experimentation and learning. This is anincentive towards moving beyond supporting indi-vidual innovators to support diverse forms ofexperimenter groups (e.g. isolated or inserted incommunities, producer organisations). The experi-ences of DIP, PRIAG and CIALs suggest thatthese processes should involve the articulation ofdifferent actors. In the cases of PRIAG andCIALs, the focus on multiple stakeholder linkageshas led to the formalisation of responsibilities (e.g.concerted annual work plans in PRIAG and defi-nition of roles and responsibilities for CIAL mem-bers and facilitators).

• A number of concepts emerge from this analysisthat were not common in the FPR literature some5–10 years ago. These include mutual responsi-bility, accountability and the convergence ofagendas or shared decision-making. This reflectsan evolution towards a more actor- and process-oriented perspective (see also Cramb 2000) asevidenced by:1. New criteria for identifying and prioritising

research themes such as likelihood of success,analysis of who benefits,

2. The offering of technological options that arein early stages of development by formalresearch services.

3. Evaluating the utility of experimentation notonly in terms of results (e.g. resistance to pestsor higher yields) but also in terms of what hasbeen learned through the process.

The importance of ‘outsiders’ with high-qualityfacilitation and interaction skills is a commonconclusion of the DIP, PRIAG and CIAL experi-ences. The outsiders may be individuals (as in theCIAL approach) or a team (DIP, PRIAG). Never-theless, in those situations where outside expertiseis unavailable and the small farmers cannot counton their valuable support, they themselves canfoment the farmer research process, as occurred inthe Farmer-to-Farmer approach in Nicaragua andin Central America.

• An interesting difference among the four experi-ences lies in the identity of the group thatcatalysed the process. In Farmer-to-Farmer, DIP,PRIAG and CIAL the original protagonists were,respectively, a farmer organisation, a university,an externally funded international cooperationproject and an international research centre.Future analyses of impact might do well to con-sider this difference and examine in detail howeach actor constructed strategies and mechanismsto promote sustainability.

• There is an interesting twist to the common con-clusion that capacity to innovate is the key ratherthan the development of specific technologies.The term Farmer Participatory Research, whichhas been so useful during the 90s, now seems tofall short of describing or representing the fullspectrum of the innovation processes that havebeen activated. Farmers involved in participatoryresearch approaches are not restricting their inter-actions and activities to agricultural research.NRM, education, health, local government andeven information and communication systemshave become arenas of interaction, involving newactors.

• The experiences analysed have generated anatmosphere of critical analysis, and each contri-butes important elements to the debate surroundingthe role of farmers and formal researchers in agri-cultural development and NRM.

• Each has incorporated internal mechanisms formonitoring and evaluation, and has made adjust-ments based upon these. This suggests that FPRapproaches themselves must be capable of evolu-tion if they are to respond to the rapidly changingcircumstances in agriculture and NRM.

• None of the experiences presented has been ableto accomplish fully the objectives for FPR sug-gested by Okali et al. (1994):1. changing the orientation of existing R&D

structures;2. developing sustainable community-based

research capability;3. creating new social and political institutions.

52

They have, however, demonstrated that it ispossible to increase community research capacity,and they have made some progress towardschanging the orientation of existing R&D struc-tures. Achieving sustainable community-basedresearch processes and organisations is still amajor challenge. Placing the community researchprocess in the hands of farmer organisations is animportant first step in this direction.

Acknowledgments

The authors wish to acknowledge with gratitudecontributions from the following individuals andorganisations:• Simon Anderson • COSECHA, Honduras• Jeff Bentley • DIP, Mexico• Trudy Brekelbaum • EAP, Honduras• Falguni Guharay • MIP-CATIE, Nicaragua• Sabine Gündel • CIAT-IPRA, Colombia• Kirsten Probst• Steve Sherwood• Charles Staver• Bernard Triomphe

We are grateful to the Forages for Smallholdersproject for giving us the opportunity to develop theideas presented in this paper. They are our own anddo not necessarily reflect the views of thosementioned here.

References

Anderson, S. 1998. Research centres and participatoryresearch; issues and implications. In: Blauert, J. andZadek, S. ed. Mediating Sustainability, Mexico,Kumarian Press, 304 p.

Ashby, J.A. 1996. What do we mean by participatoryresearch in agriculture? In: New Frontiers in Partici-patory Research and Gender Analysis. Proceedings Inter-national Seminar on participatory research and genderanalysis (PRGA), Cali, Colombia 9–14 September 1996.Cali. Centro Internacional de Agricultura Tropical(CIAT) Publication No. 294, 15–22.

Ashby, J.A., Braun, A., Gracia, T., Guerrero, M.P., Quirós,C.A. and Roa, J.I. in press. Investing in FarmerResearchers: Experience in Latin America. Centro Inter-nacional de Agricultura Tropical, Cali, Colombia.

Bentley, J.W. 2000. The mothers, fathers and midwives ofinvention. Zamorano’s Natural Pest Control course. In:Stoll, G. ed. Natural Crop Protection in the Tropics:Letting Information Come to the Life. Tropical Agro-ecology Series, Weikersheim, Margraf Verlag.

Biggs, S.D. 1989. Resource-poor farmer participation inresearch: a synthesis of experiences from nine NationalAgricultural Research Systems. The Hague, OFCORComparative Study Paper 3, ISNAR, 37 p.

Bourdieu, P. 1993. Sociology in Question, London, SAGEPublications, 192 p.

Braun, A.R., Thiele, G. and Fernández, M. 2000. FarmerField Schools and Local Agricultural Research Commit-tees: Complementary platforms for integrated decision-making in sustainable agriculture. AgREN NetworkPaper No. 105, ODI, UK. 20 p.

Bunch, R. 1982. Two Ears of Corn. A Guide to People-Centered Development. Oklahoma, World Neighbours,250 p.

Bunch, R. 1990. Low input soil restoration in Honduras.The Cantarranas Farmer-to-Farmer extension program,IIED, Gatekeeper Series No. 23.

Bunch, R. and López, G. 1994. Soil recuperation in CentralAmerica: Measuring impact 4 to 40 years after inter-vention. Honduras, COSECHA, 18 p.

CGIAR 1999. Program on Participatory Research andGender Analysis, Crossing Perspectives: Farmers andScientists in Participatory Plant Breeding. Cali,Colombia, CIAT, 46 p.

Chambers, R. and Ghildyal, B.P. 1985. Agriculturalresearch for resource-poor farmers: the farmer-first-and-last model. Discussion Paper #203. Institute of Develop-ment Studies. Brighton, England, University of Sussex,.Also in Agricultural Administration, 20: 1–30.

Chambers, R., Pacey, A. and Thrupp, L.A. ed. 1989. FarmerFirst: Farmer Innovation and Agricultural Research.London, Intermediate Technology Publications, 218 p.

Coleman, J.S. 1988. Social capital in the creation of humancapital. American Journal of Sociology, 94: 95–120.

Conway, G. 1997. The Doubly Green Revolution: Food ForAll in the 21st Century. Ithaca, NY, Cornell UniversityPress, 334 p.

Cramb, R.A. 2000. Processes influencing the successfuladoption of new technologies by smallholders. In: Stür,W.W., Horne, P.M., Hacker, J.B. and Kerridge, P.C. ed.Working with Farmers: the Key to Adoption of ForageTechnologies. These Proceedings (See Contents).

Fals Borda, O. 1985. Knowledge and People’s Power. NewDelhi, Indian Social Institute.

Fals Borda, O. 1987. The application of participatoryaction research in America. International Sociology,2(4): 329–347.

Farrington, J. and Martin, A. 1987. Farmer ParticipatoryResearch: A review of concepts and practices. Agri-cultural Administration Network, Discussion Paper 19,88 p.

Freire, P. 1970. Pedagogy of the Oppressed, New York,Continuum International Publishing Group. 164 p.

Freire, P. 1973. Education for Critical Consciousness, NewYork, Continuuy International Publishing Group, 180 p.

Fukuyama, F. 1995. Trust: The Social Virtues and theCreation of Prosperity, Toronto, Penguin. 457 p.

Gianotten, V. and Wit, T. de. 1985. Organizacióncampesina: el objetivo politico de la educación popular yla investigación participativa. Amsterdam, The Nether-lands. Latin America Studies Vol. 30. Centre for LatinAmerican Research and Documentation. Dordrecht, ICGPrinting, 531 p.

Gündel, S. 1998. Participatory Innovation Developmentand Diffusion: Adoption and Adaption [sic] of Intro-duced Legumes in the Traditional Slash-and-BurnPeasant Farming System in Yucatan, Mexico, Eschborn,GTZ, 79 p.

53

Haverkort, B., Hiemstra, W., Reijntjes, C. and Essers, S.1988. Strengthening Farmers’ Capacity for TechnologyDevelopment, ILEIA Newsletter. Issue on ParticipativeTechnology Development, 4, 3, 3–7.

Haverkort, B, Kamp, J. van der and Waters-Bayer, A. ed.1991. Joining Farmers’ Experiments: Experiences inParticipatory Technology Development, London, Inter-mediate Technology Publications, 269 p.

Harriss, J. and de Renzio, P. 1997. ‘Missing link’ or analyti-cally missing? The concept of social capital. An intro-ductory bibliography essay. Journal of InternationalDevelopment, 9: 919–937.

Hocdé, H. 1997. ‘No quiero plata; quiero conocimientos’;no equivocarse de planteamiento. San José, Costa-Rica,PRIAG technical document, 65 p.

Hocdé, H. in press. Los intercambios campesinos: más allade las fronteras …. Seamos futuristas! San Jose, CostaRica, IICA- CIRAD technical document, 200 p.

Humphries, S, Gonzalez, J., Jimenez, J. and Sierra, F. inpress. Searching for sustainable land use practices inHonduras: Lessons from a program of participatoryresearch with hillside farmers, ODI, U.K., AgREN.

Jaén, B. and Silva, A. 1995. De facilitadores a protagonistasdel proceso de generación y transferencia de tecnología.Segundo Simposio Latinoamericano sobre Investigacióny Extensión en Sistemas Agropecuarios, Santa Fe deBogotá, D.C., Colombia, 7–9 Nov. Reimpresión: Pro-grama Regional de Reforzamiento a la InvestigaciónAgronómico sobre los Granos en Centroamérica(PRIAG). San José, Documento técnico No. 18, 30 p.

Lambrou, Y. 2000. A Typology: Participatory Researchand Gender Analysis in Natural Resource ManagementResearch. In: Nina Lilja, N., Ashby, J.A. and Sperling,L. ed. Assessing the Impact of Participatory Researchand Gender Analysis. Cali, CGIAR Program on Partici-patory Research and Gender Analysis, CIAT, 3–18.

Lammerink, M.P. and Wolfers, I. ed. 1994. ed. Selecciónde algunos ejemplos de investigación participativa. Pro-gramas de investigación para el desarrollo financiadospor la DGIS, La Haya.

Larrea, S. and Sherwood, S. in press. Looking back to seeahead: Farmer lessons and recommendations after 15years of innovation & leadership in Güinope, Honduras.Agriculture and Human Values.

Merlet, M. 1995. Consolidación y ampliación del programade Campesino a Campesino Nicaragua. Paris, France,Comission des communautes europeennes, Institut deRecherches et d’amélioration des méthodes de dévelop-pement, 56 p.

Mikkelsen, B. 1995. Methods for development work andresearch. A Guide for Practitioners, New Delhi, SAGEPublications, 296 p.

Okali, C., Sumberg, J. and Farrington, J. 1994. FarmerParticipatory Research: Rhetoric and Reality, London,Intermediate Technology Publications, 159 p.

Pretty, J.N. 1994. Alternative systems of inquiry for sus-tainable agriculture. IDS Bulletin 25(2): 37–48.

Programa de Campesino a Campesino. 1999. La experi-mentacion campesina; algo más que una parcela deensayo. Managua, Nicaragua, UNAG, 56 p.

Probst, K. 2000. Success Factors in Natural ResourceManagement Research: Dissection of a Complex Dis-course. In: Nina Lilja, N., Ashby, J.A. and Sperling, L.eds. Assessing the Impact of Participatory Research andGender Analysis. Cali, CGIAR Program on ParticipatoryResearch and Gender Analysis, CIAT, 19–48.

Putnam, R.D. 1993. The Prosperous Community: SocialCapital and Public Life. The American Prospect, 13:35–42.

Röling, N.G. and Wagemakers, M.A.E. ed. 1998. Facili-tating Sustainable Agriculture, Cambridge, CambridgeUniversity Press, 230 p.

Ruddell, E.D. 1994. Documentation of farmer experiments:A key strategy for achieving food security on a sustain-able basis, Santiago de Chile, Vecinos Mundiales.

Scarborough, V. ed. 1995. Farmer-led approaches to exten-sion: Papers presented at a workshop in the Philipppines,July 1995. AgRen Network Papers Abstracts for papersNo. 59a, 59b and 59c. ODI. UK. (http://www.oneworld.org.od/agren/agabs59.htm)

Selener, D. 1997. Participatory action research and socialchange. The Cornell Participatory Action Research Net-work, Ithaca, New York, Cornell University, 358 p.

Sperling, L., Ashby, J.A., Smith, M.E., Weltzien, R.E. andMcGuire, S. in press. A framework for analysing partici-patory plant breeding approaches and results. Euphytica.

Sutherland, A., Martin, A. and Salmon, J. 1998. Recentexperiences with participatory technology developmentin Africa: Practitioner’s review. Natural Resource Per-spectives. No. 25. 15 pp. (http://www.oneworld.org/nrp/25.html)

Wall, E., Ferrazzi, G. and Schryer, F. 1998. Getting thegoods on social capital. Rural Sociology, 63(2): 300–322.

Waters-Bayer, A. 1989. Participatory Technology Develop-ment in Ecologically-Orientated Agriculture: SomeApproaches and Tools. AgREN Network Paper No. 7,63 p.

Wilson, P.A. 1997. Building social capital: a learningagenda for the twenty-first century. Urban Studies, 34:745–760.

World Bank 1994. Participatory Rural Assessment: Tech-niques to improve communication with the rural poor.Agriculture Technology Notes 6. (http://www-esd.world-bank.org/extension/key/agtech6.shtm).

54

Working with Farmers to Develop Forage Technologies – Field Experiences from the FSP

M. Tuhulele1, Le Van An2, P. Phengsavanh3, Ibrahim4, W. Nacalaban5,Vu Thi Hai Yen6, Truong Tan Khanh7, Tugiman4, Heriyanto4, P. Asis8,R. Hutasoit9, H. Phimmasan10, Sukan11, T. Ibrahim9, Bui Xuan An12,

E. Magboo13 and P.M. Horne14

Abstract

Participatory approaches to developing forage technologies present some major challenges tothe researchers, extension workers and project field officers whose job it is to implement them inthe field. New skills and methods are required as these development workers move from being‘implementers of government programs and promoters of technologies’ to partners with farmers inthe technology development process. There are many publications that describe the processes ofparticipation but few that provide guidelines of how to implement participatory approaches in thefield. This paper summarises the field experiences of development workers who implemented andadapted participatory approaches to forage technology development in Southeast Asia, and somelessons gained through the ongoing process of action-learning. The participatory approachesdescribed in the paper were the catalyst for strong new linkages between farmers and developmentworkers and for ongoing forage technology development on farmers’ fields.

A PREVIOUS paper (Horne et al. 2000a) described theparticipatory approaches that were developed by theForages for Smallholders Project (FSP) to try toimprove adaptation and adoption of forage tech-nologies on smallholder farms. One of the conclusionswas that ‘Many other researchers, extension workersand project personnel in Southeast Asia are familiarwith the benefits of these approaches and the conceptsof implementation, but are looking for practical guide-lines on how to implement them in the field’.

The authors of this paper were faced with the samechallenge when commencing participatory tech-nology development in 1995. This paper highlightstheir subsequent experiences from the field and somelessons for future on-farm development.

Field activities of the FSP

The FSP had two main goals:1. To identify robust, broadly-adapted forage varie-

ties that had the potential to provide substantialimpacts to smallholder farmers and

2. To develop and implement participatoryapproaches to demonstrate that the potential

1Directorate General of Livestock Services, Jakarta, Indo-nesia. Email: fsp-indonesia@cgnet.com2University of Agriculture and Forestry, Hue, Vietnam.Email: upland@dng.vnn.vn3National Agriculture and Forestry Research Institute,Vientiane, Lao PDR. Email: FSP-Lao@cgiar.org4Livestock Services, Samarinda, East Kalimantan, Indonesia5Agricultural Office, Malitbog, Bukidnon, Philippines6District Agriculture and Rural Development Office, TuyenQuang, Vietnam7Tay Nguyen University, Ban Me Thuot, Vietnam. Email:TanKhanh@dng.vnn.vn8City Veterinary Office, Cagayan de Oro City, Philippines9Assessment Institute for Agricultural Technologies,Medan, North Sumatra, Indonesia. Email: Tatang@indosat.net.id10District Agriculture and Forestry Office, Pek, XiengKhouang, Lao PDR11District Agriculture and Forestry Office, Xieng Ngeun,Luang Phabang, Lao PDR12University of Agriculture and Forestry, Ho Chi MinhCity, Vietnam. Email: an@hcm.fpt.vn13Livestock Research Division, PCARRD, Los Baños,Philippines. Email: ECMAGBOO@PCARRD.DOST.GOV.PH14Forages for Smallholders Project, Vientiane, Lao PDR.Email: P.Horne@cgiar.org

55

impacts described in Point 1 can be achieved onsmallholder farms in Southeast Asia.The processes that were followed to achieve these

goals are illustrated in Figure 1.Early activities focused on site selection (Horne et

al. 2000a) and species evaluation (Stür et al. 2000;Gabunada Jr. et al. 2000). As robust, broadly-adapted forage varieties emerged from these evalua-tions and promising sites were selected, the emphasisof field activities shifted more towards participatorytechnology development. This naturally led intolocal expansion of forage technologies, as farmersexpanded the areas of forage on their own land ornew farmers in the same area started plantingforages, often using vegetative planting material. Asthese developments started to occur, the projectimplemented a process of monitoring to understandthe process of adaptation and adoption of foragevarieties and technologies (Horne et al. 2000b).

Figure 2 shows where the on-farm technologydevelopment was implemented. At each of theselocations, there was at least one development workerwho implemented the process and monitored out-comes. Their on-going experiences were shared withpartners both within and between countries. Despitedifferent countries and different cultures, we foundthere were common livestock feeding problems withcommon opportunities for solving them.

These 19 sites represented six broad farmingsystems, most of them in the uplands (Table 1). Twosites were in the lowlands but these did not developbecause of a lack of potential for impact. At all sites,we were working with resource-poor smallholderfarmers. Although the farmers at some sites hadaccess to much better resources than at others, all ofthem were primarily subsistence farmers producing

their own food and some products for sale (e.g.cattle), but with little generation of cash.

Field Methods: Participatory DiagnosisWhen commencing participatory development offorage technologies, the ‘entry point’ into a rural com-munity was usually through the process of Participa-tory Diagnosis (Horne et al. 2000a). In this process,the people of the community were encouraged to:• describe their resources; • identify the main limitations in their farming

systems and their livestock systems;• prioritise the limitations in their livestock systems

that they would like to resolve;• describe how they have coped with these problems

in the past; and• discuss strategies for working on solving these

problems in future.

1Source: FSP Adoption Tree Database, 1999.2Stopped after the first year.

Table 1. Number of farmers evaluating forages in eachfarming system1.

Farming system Number of sites

Farmers evaluating forages on

farm

Short duration slash and burn 3 395Grassland 3 240Extensive Upland 3 268Moderately Intensive upland 5 450Intensive upland 3 385Rainfed lowland 2 192

TOTAL 19 1757

Figure 1. Field activities of the FSP.

Year

1995 1996 1997 1998 1999

Act

ivity

Monitoring

Local expansion

Participatory Technology Development

Site selection

Species evaluations

56

A large range of ‘tools’ is available for partici-patory diagnosis, most of them having been developedfor Participatory Rural Appraisal (PRA). We mostfrequently used village resource mapping, seasonalcalendars and long-term calendars to understandresource availability and card&chart methods (some-times linked with preference weighting) to analyseproblems. Through this process, we were able to con-firm whether there was real potential for forage tech-nologies (as had been indicated in the site selectionprocess).

From field experience with participatory diagnosis,we learned that:• Participatory Diagnosis was an important first step

in building trust with communities. It was the firsttime that we encouraged them to be partners in theprocess of solving their problems – a concept thatoften was met with initial surprise but was laterreadily adopted, especially after one or twofollow-up visits.

• While being aware of the potential of biasingfarmers’ responses in the diagnosis, it was essential

Figure 2. Location of field activities of the FSP.

On-farm sites

Forages for Smallholders Project

Southern ChinaLaos

VietnamThailand

Philippines

Malaysia

Indonesia

= On-farm evaluation sites

= Partner countries includedin networking component only

57

to clarify from the beginning exactly what kinds oftechnology options we had to offer. Farmers oftenhad unrealistic expectations of the project; expec-tations that were based on previous experienceswith projects and programs that provided creditand other inputs in a process of ‘technology injec-tion’ (which rarely worked) rather than ‘tech-nology development’.

• It was important for diagnosis to start with a‘broad scope’ before focusing on livestock systemproblems. If we only analysed the livestock systemproblems, we would not understand how importantthese were in the overall farming system.

• Farmers were sometimes confused and surprised atthe start, expecting that we had come to follow theusual procedure, where extension workers tellthem what they should be doing. The very activeprocesses of producing resource maps andcalendars invariably and rapidly broke down thesebarriers. This was especially true for the womenfarmers, who frequently were shy at the beginning,preferring to observe from behind the men farmers.With encouragement and the active methods, how-ever, they would quickly become lively partici-pants in the process. These active methods alsohelped overcome the common problem of one ortwo farmers dominating village meetings.

• The diagnosis process was necessarily fast, fre-quently taking no more than one day. In most ofthe cultures and communities where we worked,farmers could describe their complex farmingsystems and analyse their problems in detail in avery short time. It is our experience that PRA, asit is often implemented in the region, is too slowand too oriented towards ‘appraisal rather thanaction’. Farmers want quick action! They wanttechnology options to test. We do not have to waituntil we fully understand their problems and con-straints before we initiate this action in the field. Ifso, we may wait for several years, and even thenwe have little hope of understanding the com-plexity (and, sometimes, volatility) that influencesfarmers’ decisions in upland areas.

• For this reason, we found it useful to identify‘technology entry points’ in the various farmingsystems. These entry points were, in our case,forage technologies that we could confidently pre-dict would give some quick benefits for livestockfeeding and contribute to building trust with thefarmers. For example, an entry point in some ofthe upland villages of Laos and Vietnam wasStylosanthes guianensis ‘Stylo 184’ for feedingpigs. Often farmers in these areas kept pigs andwere spending large amounts of time collecting,chopping and cooking pig feed. In one report,‘Forage gathered from the forest provided a very

important source of protein and vitamins, makingfor a balanced diet of pigs. In most households inthe study area, women were the main foragegatherers … [spending] … from thirty minutes totwo hours to reach their gathering destinations’(Oparaocha 1997). Stylo 184 can provide largequantities of high-quality feed that can be feddirectly to pigs without any processing. Another‘entry point’ at several sites in Indonesia was fastgrowing grasses, providing cut feed for pennedsheep and goats in areas where farmers werealready spending a lot of time cutting nativegrasses each day.

• In some instances, it was necessary for thedevelopment worker to clarify the technologicallimits to development, so that farmers’ expecta-tions were not unrealistic. For example, in severalareas farmers were hopeful that forages could helpthem convert communal grazing land into pro-ductive pasture. We had to explain that this wasnot possible without first solving the problem ofuncontrolled grazing and land tenure.

• To be successful, Participatory Diagnosis requiresexcellent facilitation skills from the developmentworker, but not everyone has an aptitude forfacilitation. The skills cannot be learned in atraining course and cannot be learned overnight,but are developed and improved through an on-going process of action (field experience) andreflection (evaluating what methods worked well,why and how they can be improved).

• We had initially imagined that a distinct processof action planning with the community wouldfollow diagnosis. With farmers’ desire for quickaction and our identification of ‘technology entrypoints’ however, we found that action planningwas a quick process that was either part of thediagnosis session or conducted with individualfarmers or small farmers groups.

• The outcomes of Participatory Diagnosis are areflection only of the views of the active par-ticipants. For this reason, it is essential to firstgive careful consideration as to which group offarmers you want to reach (and therefore whoshould participate in the diagnosis). The questionto keep in mind is ‘Are we talking to the rightpeople?’ If your main target group is, for example,women who keep sheep or the poorest farmers inthe community, then they must be active par-ticipants in the diagnosis and subsequent tech-nology development. One example that illustratesthis well comes from a study of Hmong womenfarmers in northern Laos: ‘The women not onlyown the livestock resources but control decisionsmade on the final outcome. All the women I methad broad and comprehensive knowledge about

58

the common illnesses of pigs ... [but] ... when ananimal was treated with modern medicine, themedication was bought and administered by themen. With women nowhere in the extensionpicture, it is not surprising that the incidence ofdisease and epidemics in small livestock increasedin the last 10 years’ (Oparaocha 1997).

• The outcomes of Participatory Diagnosis need tobe confirmed with the individual farmers and/orparticular stakeholder interest groups, as theirneeds and opportunities vary greatly fromindividual to individual and from group to group.For example, if a diagnosis is conducted with adiverse group of farmers, including ‘wealthy’cattle farmers and ‘poor’ farmers who only keep afew poultry, the problems identified from thegroup diagnosis may not accurately reflect theneeds of either group.

Field Methods:Developing Forage Technologies

Two approaches to developing forages technologieson farms were used:

1. Researcher-designed and farmer/researcher-managed trials.These were mostly regional trials which weredesigned to confirm the broad adaptation of promisingforage varieties.

2. Farmer-designed and farmer managed trials.Most of the on-farm technology development con-sisted of development workers offering farmers arange of forage varieties that were adapted to thearea and had some potential to alleviate the problemsidentified during diagnosis. The main goal was toencourage farmers to innovate, adapt and integratethe forages on their farms, learning lessons andgenerating ideas throughout the process.

These are described in more detail in GabunadaJr. et al. (2000). From field experiences with devel-oping forage technologies, we learned that:• Across all sites, farmers usually wanted to evaluate

the new varieties in small monoculture plots, oftenclose to their houses, before evaluating ways ofintegrating the preferred varieties into theirfarming systems. In doing so, they very often pre-pared clean seedbeds and planted the forages inwell-managed rows. We had thought that wewould have to offer labour-saving ways of estab-lishing forages to interest farmers in evaluatingthem, but this proved not to be so.

• In most places, successful development occurredwhere individual farmers planted and developedforages on their own land rather than a group offarmers planting and managing a single plot oncommunal land.

• Regular visits by the development worker duringthe early stages of forage development, were veryimportant to encourage farmers to persist with theevaluation and provide them with basic technicalinformation about forages (Figure 3). Forageseeds are small and the plants grow more slowlythan many of the crops that farmers already grow.The young seedlings are easily damaged bygrazing or uncontrolled weed growth. Withoutearly encouragement from development workers,farmers would sometimes abandon the evaluationswithout having ever reached a stage where theforages were well enough established to providebenefits.

• Development workers should always offer farmersa broad range of forage varieties/technologies toevaluate at the beginning. At many sites in theFSP, we discovered that, after seeing the differentkinds of forage options available, farmers eitheridentified other problems they could solve withforages or changed their farming practices tocreate new opportunities for using forages. Oneexample was a site where farmers showed interestin growing grasses in small plots to be grazed bytheir cattle, but ended up expanding Stylosanthesguianensis ‘Stylo 184’ to feed their pigs. Offeringa narrow range of choices at the start would havestifled this type of innovation.

Figure 3. Regular visits by the development worker buildtrust with smallholder farmers.

59

Field Methods: Evaluating Forage Varieties and Technologies

Once farmers had been evaluating forage options forat least one season, we used a process of participatoryevaluation to understand:• which varieties/technologies they liked and why;• which varieties/technologies they did not like and

why;• what problems they had encountered; and• what they planned to do in the following season

(expand?, test new varieties?, test new ways ofintegrating varieties into their farming systems?).During Participatory Evaluation in the field,

individual farmers were asked to describe theirpreferences for the different varieties by either:1. Preference Ranking of the varieties they had

been evaluating (where ‘1’ = the best variety, ‘2’= the next best and so on) OR

2. Preference Rating of the varieties they had beenevaluating (on a scale of 0–10 where ‘0’ = verypoor and ‘10’ = excellent) OR

3. Preference Weighting of the varieties they hadbeen evaluating (allocating 50 ‘counters’ betweenthe varieties, where the more counters given toeach variety, the more highly it was preferred)‘Preference Ranking’ was not used very often

because it did not give any indication as to the extenta farmer liked one variety relative to the others, andvery often farmers preferred one or two varietiesmuch more than the others. The advantage of‘Preference Rating’ was that it told us how muchfarmers preferred each variety on both an absoluteand a relative scale. Sometimes, however, farmerswould give similar ratings to many or all of thevarieties (maybe to please the development worker),as shown by Farmer ‘C’ in the example in Table 2.In these cases, ‘Preference Weighting’ was useful tohelp separate the farmers’ preferences, but on itsown, ‘Preference Weighting’ did not tell us howmuch farmers liked each variety on an absolutescale.

Care is needed when interpreting the results ofparticipatory evaluations. In the example in Table 2:

• Not all farmers evaluated all varieties/technologies.As a result, you have to be careful when looking ataverage ranks and ratings for each variety. In theexample, only two farmers evaluated variety/technology ‘T’ but this resulted in the highestaverage rank.

• You also have to be careful in looking at averageranks and ratings if there are different groups offarmers in the group. For example, if some of thefarmers keep only pigs and some keep only cattle,they are likely to prefer different varieties forlogical reasons. In the example, Farmer ‘D’ hasvery different preferences to the other farmers,which may be because she has very different con-straints and opportunities in her farming system.In cases like this, it may be necessary to ‘dis-aggregate’ the evaluation data (that is, analyse itseparately for the different groups of farmers).

• Preference weighting helped to separate the similarratings given by Farmer ‘C’

Once farmers had rated (or weighted) each variety/technology (which usually took no more than fiveminutes), we would ask them to explain the reasonsfor their choices by describing the positive andnegative attributes of each of the varieties/tech-nologies. This helped us to understand their criteriafor accepting, rejecting or modifying different foragevarieties/technologies.

From field experience with participatory evalua-tion we learned that:

• Not all evaluations had to be ‘participatory’. Insome cases, we were interested in technical aspectsof the performance of the varieties/technologies inthe field (for example, yield in relation to soilfertility). It was important, however, if we were todo a technical evaluation, that we did the partici-patory evaluations first, so as not to influence thefarmers’ responses.

1 = Figures in brackets are this farmer’s weightings using 50 counters

Table 2. Example of ‘preference rating’ and ‘preference weighting’ with four farmers.

Variety/technology Farmers’ ratings Averagerating

Averagerank

A B C1 D

P 8 9 7 (17) 4 7 3Q 7 9 7 (11) 7 7.5 2R 4 4 7 (9) 4 4.8 4S 0 – 6 (3) 3 2.2 5T – – 7 (10) 9 8 1

Total varieties tested 4 3 5 5

60

• The amount of data collection and the value of thedata collected are dependent on the amount ofexperience the farmers have had with the foragevarieties/technologies. Formal evaluations werenot necessary during the first six months of tech-nology development when farmers had not yetlearnt much about the varieties/technologies

Field Methods: Local Expansion

Soon after commencing on-farm forage technologydevelopment, we were faced with issues of localexpansion. This was where farmers started expandingforages on their own land and neighbouring farmersstarting to plant forages. Local expansion raisedquestions for the FSP about local availability ofplanting material and the role of the developmentworker in the local expansion process.

From field experience with local expansion welearned that:• Local expansion often occurs through the energetic

efforts of a local ‘champion’, often an innovativefarmer. We needed to engage closely with thesefarmers to help them in their efforts and also toensure that new farmers continued to have accessto a broad range of choices and not just one optionfavoured by the local ‘champion’.

• Sometimes farmers launched into local expansiontoo early and development workers needed toencourage them to evaluate the broad range ofvarieties throughout a full year. In some of thedistinctly wet/dry sites where we worked, forexample, farmers started expanding varietiesduring the first wet season that subsequently diedin the following dry season (e.g. Brachiaria ruzi-ziensis ‘Ruzi’).

• Local expansion often occurs spontaneously,without the need for extra seed. We had thoughtwe would need to foster local seed supply systems,but if farmers were sufficiently impressed by thevarieties/technologies, they would either expandthem using vegetative planting material or collectsmall amounts of seed locally. In some instances,local expansion was limited by the lack of plantingmaterial and the development worker had to eitherassist with small amounts of seed or simplyfacilitate farmer-to-farmer exchange of locally-available planting material.

• In some countries (especially the Philippines andIndonesia), we worked through farmers’ groupswhich helped give momentum to the local expan-sion. Such farmers’ groups will be especiallyimportant once local expansion of promisingforage technologies gets to a scale that is beyondthe capacity of the individual developmentworkers to support (see Braun and Hocdé 2000).

Field Methods: Monitoring and Evaluation

As local expansion continued and became morecomplex, we realised that the different stakeholders(national partners, the project) needed informationabout the forage technology development, both tounderstand how the process was developing and toplan for future development. What was beingadopted and why?

Late in the project, we implemented a monitoringprocess called the ‘Adoption Tree’. Details of theprocess and field experiences are presented in Horneet al. (2000b).

The changing role of development workers

Perhaps the biggest challenge of the differentapproaches used by the FSP (Figure 1) was that theyrequired a complete change in the roles of develop-ment workers and farmers. Previously, developmentworkers were required to implement governmentprograms, collect data and promote technologies,and the farmers also expected this. With the partici-patory approaches, the role of the developmentworkers evolved as the activities in the field evolved(Figure 4) moving towards an increasing partnershipbetween farmers and development workers. This wasa significant change for both and a change thatresulted in the rapid expansion of on-farm foragetechnology development at most sites. Although theFSP has not yet reached a stage of scaling-out(expanding to completely new areas), it is expectedthat the momentum will continue and that some ofthe more innovative and active farmers will them-selves become extension workers in the process.

Some Lessons Learned …

The process of implementing the participatoryapproaches has been very challenging for all thedevelopment workers involved. Through this processwe have learned some valuable lessons:• In the early stages of the fieldwork, careful farmer

selection can make a huge difference to the sub-sequent success or failure of the participatoryprocess. With time, however, as local expansionstarts to happen, this is not as important, asfarmers ‘select themselves’ (by spontaneouslyplanting forage or asking to join the FSP program).

• There are many documented methods or ‘tools’ ofPRA, for example, village and resource mapping,seasonal calendars and matrices, that can be usedin the field to facilitate diagnosis, evaluation andmonitoring. Like a carpenter’s tools, however,these methods are a waste of time without askilled person to use them. While these methods

61

are generally well known (many developmentworkers can faithfully describe ‘matrix ranking’!),the essential skills of facilitation and communica-tion with farmers have been under-emphasised.These skills take a lot of time and commitment todevelop and consolidate.

• The participatory approaches usually made thedevelopment workers feel a lot more confidentabout visiting and working with resource-poorfarmers. In their new role, they could admit tofarmers that they did not have all the answers, buthad some ideas and technology options toevaluate. In this role, development workers andfarmers were able to learn new skills and ideasfrom each other. Although these approachesrequired a lot of time and commitment fromdevelopment workers, the results were more sus-tainable and better directed towards farmers’needs.

• The ‘tools’ of participatory diagnosis can be a‘two-edged sword’, lulling development workersinto a false sense of achievement. We must notlose sight of the fact that the goal of our work is toengage actively with farmers in working towardssolution of their problems, not simply trying tobetter understand those problems.

• The changing role of development workers(Figure 4) involves them encouraging farmers tobecome more actively involved in developmentprocesses that affect their livelihoods. This

process can be threatening to officers in central-ised research and development institutions, whomay feel they are losing control. As technologiesstart to expand in the field, it is essential that thedevelopment workers’ institutions become moreactively partners in the process.

• There are often institutional pressures that pushdevelopment workers to put most of their effortinto helping a few farmers or to work in areaswhere there is little potential for forage or to comeup with quick results by encouraging farmers toplant large areas in the first year. It is our experi-ence that success comes from first encouragingfarmers to evaluate and innovate on a small scale.It is better to have a small success than a bigfailure.

The participatory methods being developed,applied and adapted by the FSP are far from perfect.At some sites, it was not possible to conduct diagnosisbefore commencing the technology developmentprocess. On several occasions, poor site selection ledto diagnosis being conducted in communities whereit turned out there was little potential for impact fromforage technologies.

Despite these problems, however, both farmersand development workers responded enthusiasticallyto their new roles and the result was on-farm foragedevelopment which had substantial momentum atmany sites.

Figure 4. The changing role of development workers and farmers involved in participatory technology development.

Development worker facilitatesfarmer-to-farmer exchange,

providing feedback andsuggestions; farmers innovate,

expand and initiate change

Developmentworker facilitates,and farmers make

the decisions

Developmentworker decides

Development worker provides,varieties and information ontechnology options; farmerschoose, test and evaluate

Incr

easi

ng p

artn

ersh

ipbe

twee

n fa

rmer

s an

dde

velo

pmen

t w

orke

rs

Siteselection

ProblemDiagnosis

Testing andevaluation

Localexpansion

ScalingOut

Time

62

Through these experiences, development workerscollaborating with the FSP are moving away fromdependence on a ‘manual of methods’ towardshaving a ‘toolbox of skills and approaches’ that theycan use to respond to new situations and make thedecisions required to nurture the participatoryapproach.

At present, there is only a limited number ofdevelopment workers in each country who have this‘toolbox of skills and approaches’ and a majorchallenge now is how to provide these skilleddevelopment workers with opportunities to mentorand help other development workers to learn aboutparticipatory approaches.

ReferencesBraun, A.R. and Hocdé, H. 2000. Farmer Participatory

Research in Latin America: Four Cases. In: Stür, W.W.,Horne, P.M., Hacker, J.B. and Kerridge, P.C. ed.Working with Farmers: the Key to Adoption of ForageTechnologies. These Proceedings, see Contents.

Gabunada, F. Jr., Heriyanto, Phengsavanh, P., Phimphach-anhvongsod, V, Truong Thanh Khanh, Nacalaban, W.,

Asis, P., Vu Thi Hai Yen, Tugiman, Ibrahim and Stür,W. 2000. Integration of adapted forages on farms inSoutheast Asia – experiences from the Forages forSmallholder Project. In: Stür, W.W., Horne, P.M.,Hacker, J.B. and Kerridge, P.C. ed. Working withFarmers: the Key to Adoption of Forage Technologies.These Proceedings, see Contents.

Horne, P.M., Magboo, E.C., Kerridge, P.C., Tuhulele, M.,Phimphachanhvongsod, V., Gabunada, F. Jr., Le HoaBinh and Stür, W.W. 2000(a). Participatory approachesto forage technology development with smallholders inSoutheast Asia. In: Stür, W.W., Horne, P.M., Hacker,J.B. and Kerridge, P.C. ed. Working with Farmers: theKey to Adoption of Forage Technologies. These Pro-ceedings, see Contents.

Horne, P.M., Stür, W.W., Orencia, E.L., Gabunada F. Jr.,Phengsavanh, P., Bui Xuan An and Tuhulele, M. 2000(b).Monitoring forage technology development – the adop-tion tree. In: Stür, W.W., Horne, P.M., Hacker, J.B. andKerridge, P.C. ed. Working with Farmers: the Key toAdoption of Forage Technologies. These Proceedings,see Contents.

Oparaocha, S. 1997. Gender roles and relations in livestockmanagement: the Hmong communities in Nong Het, LaoPDR. AIT, Bangkok, unpublished MA Thesis.

63

Selection and Targeting of Forages in Central America linking Participatory Approaches and Geographic Information Systems — Concept and Preliminary Results

M. Peters1, P. Argel1, C. Burgos2, G.G. Hyman1, H. Cruz1, J. Klass1, A. Braun1, A. Franco1 and M.I. Posas3

THE challenge is to match the potential of forages toimprove sustainability of tropical agro-ecosystemswith wide scale utilisation by smallholder clients.The potential of forages in improving the social,economic and environmental sustainability of small-holder production systems in the tropics is well rec-ognised. Potential benefits of forages include theincrease of livestock production through improvedfeed. Positive effects of forages on crop productioninclude the reduced dependency on external inputswhile maintaining or improving soil fertility; incor-poration of forages in rotations have a positive effecton breaking pest and disease cycles. Forages also canreduce competition of weeds and lead to recupera-tion and reclamation of land. Synergistic effectsbetween crop and livestock production can increaseefficiency of land and labour inputs, in addition toutilisation of land not suitable for crop production.

However, adoption of forage-based technologies,in particular legumes, has so far been limited.Besides, an unfavourable policy environment givingpreference to external inputs, the limited acceptanceby smallholders can be attributed to lack of farmerparticipation in the development of forage germ-plasm and the lack of co-ordination of research onfeed improvement, soil fertility and community par-ticipation. Moreover, methods for extrapolation andup scaling will need to be improved.

The Approach

Based on the limitations to adoption described abovewe utilise an integrated approach for multipurposeforage germplasm development emphasising thefollowing key components:• Farmer participation.• Integration of on-farm with on-station work.• Synchronising demand and (artesenal) seed pro-

duction (i.e. integrated community-based seed-supply systems).

• Increasing the capacity of stakeholders.• Involvement of local, national, regional and inter-

national partners.• Extrapolation of results using advanced

technologies.

Developing Forage Germplasm with Farmers, NGO’s and NARS

In 1998, we commenced in Honduras an initiative toselect forage germplasm with farmers using partici-patory methods. We started evaluation with a refer-ence site approach, with extension to satellite sitesplanned for the future. The collaboration with SERT-EDESO (Servicios Técnicos para el DesarolloSostenible), a NGO residing in the reference site,facilitates the communication with farmers, while theinteraction with DICTA (Dirección de Ciencia yTecnología Agropecuaria) working at the nationallevel, is expected to enhance the up-scaling process.For using forages as feed we interact closely with theCIAT-led Consortium TROPILECHE.

In Tables 1 and 2 preliminary results from theselection of grass and legume germplasm are pre-sented. Based on these evaluations all farmersinvolved in the initial evaluation have requested seedfor planting larger plots. We are currently developingwith the farmer’s possibilities for artesenal seed

1Centro Internacional de Agricultura Tropical (CIAT), AA6713, Cali, Colombia. Email: b.hincapie@cgiar.org2Dirección de Ciencia y Tecnología Agropecuaria (DICTA),Apartado Postal 5550, Tegucigalpa, Honduras3Servicos Técnicos para el Desarollo Sostenible (SERT-EDESO), Yorita, Yoro, Honduras

64

production. DICTA has agreed, in collaboration withCIAT, to back-up this process by capacitation andbasic seed production. We are also in the process ofevaluating results from trials using shrub legumesand grass and legume species for soil reclamationpurposes.

In 1999 the approach was extended to Nicaraguaand in 2000 we intend to commence work in CostaRica. Experiences gained from this initiative andother work with farmers is expected to focus futurecharacterization and collection demand of foragegermplasm.

Develop Expert Systems linking Biological and Socio-economic Data with Geographical

Information

Developing a forage database

In an effort to make information gained from thisand other work available to a wider community, weintegrate experimental data into a forage databasewith a graphical interface. Figure 1 shows a screen-shot from an early version of the tool.

In contrast to many other forage databases, thetool in development is deriving information from

Table 1. Summary of evaluations for the participatory selection of grasses by farmers. Grasses were scored on a scale of 1(least preferred) to 5 (most preferred). San Jerónimo, Honduras.

Accession Evaluation Points Total

Ranking

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Brachiaria brizantha CIAT 26110 18 20 18 18 18 20 18 18 18 20 20 15 6 6 3 5 241 1

Brachiaria – hybrid FM 9201/1873 18 20 16 18 20 20 20 16 14 20 18 13 6 8 3 3 233 3

Brachiaria humidicola cv. Llanero CIAT 6133

12 14 16 14 16 16 14 10 10 8 14 7 6 6 5 5 173 5

Panicum maximum CIAT 16028 16 14 16 20 16 16 20 16 16 20 16 13 10 6 5 5 225 4

Panicum maximum CIAT 16051 16 12 12 12 8 10 12 12 16 14 14 9 4 4 5 3 163 6

Panicum maximum cv. Tanzania CIAT 16031

16 18 18 20 18 20 18 16 14 18 18 13 10 10 5 3 235 2

Table 2. Initial absolute evaluation for the participatory selection of herbaceous legumes by farmers. Legumes were scoredon a scale of 1 (least preferred) to 5 (most preferred species), San Jerónimo, Honduras.

Accession Evaluation PointsTotal

Ranking

1 2 3 4 5 6 7 8 9 10 11

Arachis pintoi CIAT 17434 cv. Pico Bonito in Honduras 5 5 5 5 5 5 5 5 5 5 5 55 1

Arachis pintoi CIAT 22160 5 5 1 1 1 5 5 5 5 5 5 43 3

Centrosema brasilianum CIAT 15387 5 5 1 1 1 1 1 1 1 3 3 23 8

Centrosema macrocarpum CIAT 25522 3 3 1 1 1 3 3 3 3 3 1 25 7

Centrosema plumieri DICTA 5 5 1 1 1 5 5 5 5 5 3 41 4

Centrosema pubescens CIAT 434 5 5 3 3 3 1 1 3 3 1 1 29 5

Desmodium heterocarpon var. ovalifolium CIAT 23762 3 3 1 1 1 5 3 3 3 3 3 29 5

Stylosanthes guianensis cv. Pucallpa CIAT 184 5 5 5 1 3 3 3 5 5 5 5 34 2

65

Figure 1. Screen Shot of CIAT Forage Database (under development).

Figure 2. Initial maps showing the distribution of forage germplasm evaluation sites according to altitude level and life-zones, after Holdridge (1967).

66

actual experimental data — down to accession level— over a wide range of environments across LatinAmerica and Africa. The incorporation of data fromAsia is planned.

We expect to have a first version available forshipment to key collaborators in 2000. We intendcontinuous updating of information and extension toincorporate further information as it becomesavailable.

Developing GIS-based decision support tools

Based on the forage database, we are developing aGIS-based Decision Support Tool usable for mapping

and extrapolating forage adaptation to different socio-economic and biophysical environments. A version totarget forage germplasm to biophysical environmentsis scheduled for 2001. Initial maps developed areshown in Figure 2.

We are also developing models to incorporatesocio-economic information such as different pro-duction systems, market access, social preferencesetc. into the GIS-based tool.

ReferenceHoldridge, L.R. 1967. Life Zone Ecology. Tropical Science

Center, Costa Rica.

69

Scaling-Up: The Roles of Participatory Technology Development and Participatory Extension Approaches

J.G. Connell1

Abstract

While Farmer Participatory Research (FPR) and Participatory Technology Development (PTD)are becoming better recognised and accepted within mainstream research, similar acceptance ofparticipatory approaches has not occurred within mainstream extension. Two cases are examinedin which participatory extension approaches have been applied with success. The introduction of anew crop in Thailand, wheat, was achieved by initiating a PTD process within the extension,which resulted in a range of farmer developed technologies, that replaced the Officially Recom-mended Technology. The Pilot Extension Project in Laos attempted to initiate a broader basedparticipatory extension approach, for the development of the National Extension System. The keyto this was to have extension methodologies, and training programs which would be robust enoughto be applied on a National basis. Apart from these approaches for participatory extension, the‘diversity of farmers’ production environments’ is proposed as a compelling pragmatic rational forthe inclusion of participatory approaches in mainstream extension.

FARMER Participatory Research (FPR) and Partici-patory Technology Development (PTD) are becomingbetter recognised and accepted within mainstreamresearch as effective in developing technologiesappropriate to farmers’ needs and conditions. Theproblem remains as to how to scale-up from the fewfarmers who have been engaged in the PTD, to otherfarmers in the village, and in other villages.

In areas with diverse production environments, asthis scaling-up occurs, and a new technology movesto farmers who have different production conditions,it would seem that the PTD process should be main-tained. If this is seen as a problem of ‘technologydevelopment’, it is natural for scientists to be con-cerned about scaling-up. But once one takes on thetask of working across villages, districts andprovinces, this is clearly beyond the resources of theresearch sector. If scaling-up is to be achieved at thislevel, it must begin to involve some sort of NationalExtension Service (NES).

Extension, however, will also need strategies todeal with diverse production environments, of whichPTD is one. Attempts to develop Participatory

Extension approaches do not date far back (Connell1992). There has always been a grey area whereresearch ends and where extension should take over.But in the past decade, as we have begun to seeresearchers concerned with scaling-up, and exten-sionists attempting to engage farmers in adaptingtechnologies to their own conditions, this grey areaand the overlap between research and extensionseems to becoming even wider.

This paper aims to examine the need for partici-patory approaches, in particular PTD, in the contextof extension, and then to look at the issues andstrategies which would be needed to establish theseapproaches within NESs. It will do this by looking attwo cases which illustrate these issues and providesome concrete guidelines as to how they could beaddressed.

The Role of PTD in ‘Extension’

When extension was dominated by the ‘transfer oftechnology’ paradigm, extensionists felt their rolewas to introduce and train small groups of farmers inthe use of a new improved technology, which wouldthen spread to other farmers in the area. The scaling-up, or spread of technology from the small group of

1Pilot Extension Project, Agriculture Extension Agency,Ministry of Agriculture and Forestry Laos. Email: AFD@carelaos.org

70

farmers, was variously called ‘diffusion of tech-nology’ or the ‘adoption process’ (Van den Ban andHawkins 1988). This approach to extension has hadmixed success, yet the role of Extension Workers(EWs) as ‘technology messengers’ and the use of the‘transfer of technology’ approach to extension, isstill very much alive today.

While the process of ‘diffusion of technology’ canstill be used to describe the spread of technologyamong farmers, the diversity of farmers’ productionareas greatly inhibits and complicates this process.Particularly in upland and highland environments,the physical characteristics of the farmers’ fields;soil fertility, texture, weed population, incident lightand rain, etc., can vary significantly within a fewhundred metres. In addition the socio-economic con-ditions of the farm-families, such as availability oflabour, funds, equipment, technical knowledge etc.can vary from one family to another. All of thismeans that while one technology may suite onefarmer in a particular plot, it may not suit anotherfarmer, or another plot of land.

In this situation, no matter how effective a PTDactivity has been to develop a new technology appro-priate to one group of farmers, this technology mightnot be appropriate on the other side of the village,much less in the next village. Clearly EWs workingin an area with this sort of diversity will need anextension approach which engages farmers in aPTD-type process, so that the farmers will adapt thestarting technologies to their production conditionsthemselves. The case of the introduction of wheat torice farmers in Northern Thailand serves to demon-strate the crucial role that PTD had in the context ofextension.

CASE 1:Introduction of Wheat to Northern Thailand

During the 1970s, SE Asia had the world’s highestrate of increase in wheat consumption (Byerlee1984). CIMMYT’s SE Asian Wheat Program (1981–93)1 aimed to support the NARS of a number ofcountries in the region in the development of wheatproduction for import substitution. With wheat’sexcellent tolerance of drought, the two productiondomains identified for wheat were: as a rainfed cropin upland areas, established at the end of the wet-season; and, as a second crop, in irrigated paddy,following the main rice harvest.

While there were many initial concerns aboutwheat’s ability to stand up to increased disease andpest pressure in the warmer climate of the region, no

problems were expected with the planting tech-nology for such an extensively grown crop. As aresult, when the Department of Agricultural Exten-sion (DOAE) began extension activities in 1983, the‘officially recommended technology’ (Of-Rec-Tech)was straight off the experiment stations, without anymodification for farmers’ conditions. This consistedof: • Full soil preparation (with formation of seedbeds

in paddy areas);• Seeding in rows;• Rates for seed and fertiliser application, and

planting dates. Immediately, there were problems across the

board with this technology. Thai farmers wereunfamiliar with seeding in rows, and so tended toover-seed, resulting in inter-plant competition andpoor plant development. In the paddy areas, farmersover-irrigated to the point of soil-saturation, causingdamping-off and root-rot diseases. Extension effortsfrom 1983 to 1986 resulted in consistently poorstands with frequent complete crop failures. A fewsites did produce well, lending some hope that wheatproduction was feasible. But the overwhelming con-clusion was that while the crop could be grown inwarmer environments, it was just too sensitive to beviable as a crop in farmers’ fields in Thailand.

In the 1986 planting season, somewhat by chance,the Program suggested to a group of farmers at onelowland site that they try two alternative plantingmethods which were less intensive than the Of-Rec-Tech.

Alternative Tech. 1 – minimum tillage, with rowseeding, and Alternative Tech. 2 – full soil preparation, broad-cast seeding + harrowing. While these technologies gave a slightly lower

yield than where the Of-Rec-Tech was appliedcorrectly, the ‘alternative technologies’ appeared tohelp farmers avoid the common errors they had beenmaking. Farmers who broadcast seeded were able tojudge their seed-rates; and minimum tillage (whichlacked raised seed-beds and irrigation channels) leadfarmers to irrigate by flash-flooding, thus eliminatingthe incidence of over-irrigation.

What was more interesting, was the range ofadapted and innovated technologies which appearedin farmers’ fields. Of the 26 farmers at the site, 11used the alternative technologies, with six of thefarmers using more than one technology in theirfield. What was more, two entirely new technologiesemerged!

This sort of innovation by the farmers was anexciting development. It appeared that it could offerthe Thai Wheat Program a shortcut to identifyingappropriate and robust technologies, by stimulating

1Thailand, Philippines, Indonesia and Sri Lanka. Of these,Thailand developed the most vigorous program.

71

farmer experimentation. In the following season, thisextension approach was tried in a variety of con-ditions; in rainfed and irrigated production areas,with lowland and ethnic minority farmers, andimplemented by District EWs and NGO develop-ment workers. In all cases it had the result of stimu-lating farmer experimentation.

On-Farm Research and Joint Monitoring Tours, inwhich scientists and extension staff together sur-veyed results in farmers’ fields, ensured a constantcross-fertilisation of ideas between researchers,extensionists and farmers. While the Of-Rec-Techstill stood, an informal extension strategy developed,with many extension workers offering farmers anumber of alternative technologies. By the 1988planting season, a survey conducted by the DOAErevealed that only 60% of the 400+ farmers surveyedwere using ‘alternative technologies’ (Connell1999).

Farmers’ informal ‘trials’ often appeared as smallplots alongside their main field. On other occasions,they committed larger areas to compare differenttechnologies. Some of the variations in the pro-duction technologies were made in response toproblems or issues felt by the farmers, whereas inother cases farmers seemed to be just trying some-thing different. A dozen farmers in a village, alltrying various technologies (not all sensible!), didnot always give the impression of being constructive.However a longitudinal study conducted at one siteshowed that there was indeed an evolutionary path inthe direction these informal trials took over a numberof seasons.

This sort of evolution of farmers’ production tech-nologies was seen in Pai District, where farmersbegan growing irrigated wheat using the Of-Rec-Tech in small plots (<0.1 ha). In the followingseason, they began to increase their area of wheatand changed to minimum tillage + row seeding toeliminate the cost of soil-preparation. As theyincreased their area of wheat further, the time fordigging furrows for row-seeding became a limitingfactor. So they switched back to broadcast-seeding,being willing to pay the cost of full soil preparationto save the time and labour for seeding. Finally a fewfarmers tried to reduce both the cost for soil prepara-tion and the time for seeding together by exper-imenting with zero-tillage and broadcast-seeding. Atthis point the technology development seemed toreach a steady state, with this becoming the mainenduring technology applied in Pai (Table 1).

This sort of trial and evolution of technologies byfarmers was not structured or guided in any way.The key factor which helped to stimulate it on such awide scale, appeared to be the extension strategy of

simply offering farmers a selection of technologiesand inviting them to identify which suited them best.

This role of offering farmers a choice was con-firmed statistically during the IDRC ParticipatoryExtension Project (1992–1994). The project aimed toexamine whether participatory extension approachescould be introduced and applied by EWs of a largeNES, such as Thailand’s DOAE. Training in exten-sion methodologies, which included offering farmersa choice of technologies, was provided to EWs fromeight Districts. At the sites where EWs had providedfarmers with a choice of technologies, farmersadapted some component of the technologies pro-vided (Department of Agricultural Extension 1995).The level of farmer adaptation was much less at siteswhere farmers had been provided only the Of-Rec-Tech

To those committed to the ideals of participation,simply offering farmers a choice of technologies,might seem a poor brand of participation. And theymay be right; that without other changes in attitudeand behaviour of EWs towards farmers, this will notbe sufficient to realise farmers’ full potential, orempower them. However there are some importantimplicit changes which take place when EWs beginto offer farmers a choice of technologies. When theydo this, the EWs are admitting that: (a) They are no longer the bearers of the ‘best’ tech-

nology, or the ‘right way’; and that(b) Farmers do have the role and capacity to evaluate

and select technologies themselves. If such anapproach could be applied as a general extensionstrategy, it would be a significant and worthwhilestep in the institutionalisation of participatoryextension approaches within mainstreamextension.

Table 1. Farmers’ adjustment of production technologiesin response to new constraints as they expand productionarea. (Pai District, Mae Hongson Province, North Thailand).

Season/Farmers’ objective

Soil preparation +Seedling method

1988 seasonOf-Rec-Tech.

1989, 1990 seasonsreduce soil-prep. costs

1991 seasonreduce time/labour

1992 seasonreduce costs + time/labour

full tillage + row seeding↓

minimum + row seedingtillage

↓ ↓full tillage + broadcast seeding

↓ ↓zero tillage + broadcast seeding

with straw mulch

72

Implications and role for PTD within extension

PTD and national extension services.

The provision of a choice of technologies to the Thaiwheat farmers was pivotal in the development ofmore robust technologies. Without this, wheat wouldnot have gained a foothold as a new dry-season cropin Northern Thailand2. Thus it represents a signifi-cant demonstration of the role and impact of PTD,not as part of a FPR Program, but within the contextof a broadly based extension program.

While the development of new technologies isexciting, this aspect should not be overemphasised,as genuinely new technologies are going to be anoccasional event only. The full range of functions ofproviding farmers a ‘choice of technologies’, forextension in diverse production environments, can besummarised as the following:• Farmers have alternatives as ‘starting-points’,

from which to identify the technology most suit-able for their particular plot of land;

• From time to time, it can stimulate a PTDprocess which may result in worthwhile ‘spin-off’ technologies;

• Ensures a relationship of mutual respect betweenEW and farmers, where the needs of farmers arerecognised, and farmers are engaged in decisionmaking. Any attempt to alter the procedures of a large

organisation such as an NES can be a daunting taskthat many senior staff will shy away from. However,simply offering farmers a choice, or menu of tech-nologies, is not a difficult adjustment for an EW andwould not require massive amounts of re-training. Inareas with ‘diverse production environments’, this issimply a pragmatic response to the need for farmersto have alternatives to choose from. Put in this con-text, it should receive ready support from adminis-trators to place into mainstream extension.

Interface between FPR and Participatory Extension

Given the opportunity for PTD to be stimulatedwithin the context of extension, the implications thishas for research need to be examined.

Results of ‘adaptive research’ will always be site-specific where farmers have diverse productionenvironments. If the research paradigm were followedfor ‘scaling-up’, additional adaptive research wouldbe needed at each new site/village. In terms of

researcher time and funds, this is simply not possible.Yet the reality is that such technology adaptation is,in fact, needed for diverse production environments.Instead of scientist-managed adaptive research, somelevel of technology adaptation could be achievedthrough farmer-lead ‘adaptive research’, by initiatinga PTD process through Participatory Extension(Figure 1).

Figure 1. Potential adjustment of the demarcation betweenresearch and extension effected by general application ofParticipatory Exensions.

If widespread technology adaptation by farmerscan be initiated, then it should be possible to re-direct scarce research resources to issues whichfarmers are not able to address. This then poses twonew questions: (a) What are the types of issues which farmers are

well able to deal with, and which researchersneed not focus on, and visa versa?

2By 1993, there were over 2000 farmers growing about1000 ha of wheat in the 7 Provinces of Northern Thailand.Farmers’ yields in villages where farmers had developed afew years experience were averaging 1.6 t/ha, with max-imum yields in farmers fields in excess of 3 t/ha (Connell1999).

Research and ‘transfer of technology’ Extension,linkages

Research andParticipatoryExtension,linkages

Pure Research

↓

Applied Research

↓

Adaptive Research

Pure Research

↓

Applied Research

↓

Adaptive Research

Multi-location trials

↓

Demonstration plotsParticipatory

Extension

farmers farmers farmers

Widespreadextension

farmers farmers farmers

73

(b) How far does research need to go in refining newtechnology before it is handed over to extension,if the researcher knows that it will go through aprocess of adaptation under ParticipatoryExtension?

Participatory Extension approaches are just begin-ning to find a foothold in mainstream extension.Thus the opportunity to examine these questions isjust emerging.

Participatory Extension Approaches

NESs must work with not just one village, but withmany villages. An extension office at a District leveltypically might be responsible for working in 60–100villages. While considerable diversity can beexpected of farmers’ production environments over awhole District, this can still be managed at the farm-level by EWs providing farmers with choices oftechnologies to stimulate PTD. However PTDfocuses only on the technology aspect of extension,and should not be confused with the broader objec-tives of Participatory Extension approaches.

The real-life issues that farmers face are broaderthan just applying a new technology. Each seasonfarmers need to decide how best to use theirresources of labour, land, cash, etc. In the past,farmers had a relatively fixed set of framework con-ditions, and could rely on their Indigenous TechnicalKnowledge to guide them in making these decisions.As farmers have become irrevocably involved inmarket systems and require outside inputs, they mustnow take into account new factors in their decisionmaking, such as: which cash-crop to grow, marketprices, whether to purchase new equipment such as asprayer or a pump; and so on. The introduction andtransfer of new production technologies to farmers isan important aspect of improving productivity, butonly one aspect of the every-day challenges theymust face.

Extension therefore needs to have objectivesbeyond simply improving productivity. It shouldhave Human Resource Development (HRD) objec-tives also, which would include enabling farmers toidentify issues affecting their production; and toevaluate and to decide on the potential benefits ofimproved practices.

Participatory approaches applied by projects andnational organisations usually include these HRDobjectives. While they may be effective in enablingand empowering farmers, such projects also havespecial attributes, such as selected staff with a highdegree of commitment and special training, andexcellent resources and back-up. These conditionscannot be replicated within most NESs with theirexisting rank and file EWs and limited funds3.

For NESs to apply participatory extensionapproaches on a broad basis, they will need to faceoperational challenges in several areas:

Extension Methodology: to identify extensionmethodologies which will be participatory, yet havesimple enough procedures, and be sufficiently robustthat rank and file EWs can accept and apply them ona regular basis.

Training EWs: the training strategies will need toaddress not just knowledge and skills development,but also changes in behaviour and attitude. Thisneeds to be achieved not just with a small selectedgroup, but across the board with rank and file EWs.

Administration of Participatory Extension: admin-istrative systems for participatory extensionapproaches must include planning at the local level,and will need to accept that progress is indetermi-nate. This may conflict with the traditional expecta-tions of administration to exert direction over staffand extension activities, achieve set goals and ensurefunds are used effectively and accounted for.

There will always have to be some play offbetween quality and quantity when mobilising anational program. NESs also need to balance nationalgoals, budgets and policy. Apart from these issues,the practical issues of operationalising ParticipatoryExtension approaches remain. One attempt to do thishas been the Pilot Extension Project (PEP) recentlyconducted in Laos.

CASE 2:Participatory Extension Project, Laos

Background

Lao PDR has a network of Provincial Agricultureand Forestry Offices and District Agriculture andForestry Offices (DAFO) spread throughout the 17provinces and 132 districts of the country. Within theDAFO, staff are allocated as specialist staff toSections for Crop Production, Livestock, Irrigationand Forestry.

The DAFO is the key institution interacting withfarmers, but most of its work has been focused oneither administrative type activities (collecting dataon production, crop losses etc.), or improving theinfrastructure to expand irrigated paddy area. Very

3Two notable exceptions to this are the FAO’s IntegratedPest Management (IPM) Program and Agritex’s Participa-tory Extension Approaches in Zimbabwe. While IPM mightnot be considered by all as mainstream extension, it cer-tainly has been applied by large numbers of rank and fileextension staff. The process of Agritex’s ParticipatoryExtension bears great similarity to PEP’s four ‘Steps ofImplementation’, but elaborates these in far greater detail(Hagmann et al. 1999).

74

little of the work would be considered ‘extension’,such as working with farmers to introduce new pro-duction technologies etc. The technical expertise ofthe DAFO staff, even in their own subject areas, isquite limited, and their communication skills arepoor. The official extension approach (introduced in1993) is that of the establishment of Model Farmers.DAFO staff would rarely have been able to workconsistently over a number of seasons on any issue,so their experience and concept of extension is non-existent.

The objective of the Pilot Extension Project(PEP)4 was to develop strategies which could beused to establish a NES.5 With effective extension inLaos being more or less a blank slate, this appearedto be an opportunity to build a participatory exten-sion system from the ground up. While this was true,there were still many pre- and mis-conceptions thatstill needed addressing and clarification.

The project was based in the Agriculture ExtensionAgency of the Department of Agriculture with a smallgroup of technical staff comprising an ExtensionTraining Unit (ETU). The ETU then worked at thelocal level with staff of the PAFO and DAFO. Thestrategy was to implement extension on a pilot scalein four DAFOs in two Provinces. In each of theDAFOs, six staff (approximately 30% of technicalstaff) were trained and supported to work as generalistEWs in a number of pilot villages (10–12 villages perDAFO, making a total of 46 pilot villages).

With the understanding of extension at such a lowlevel, PEP had to make a number of WorkingAssumptions:

• The initial focus of a young NES should bedirected at farmers’ ‘basic-needs’ (e.g. small-holder production of rice, livestock raising etc.)rather than introducing new cash-crops, orattempting dramatic changes in the existingfarming system. This level of work is applicable tothe majority of villages nation-wide, and the skillsneeded for this level of work can be developedfairly quickly with DAFO staff.

• DAFO staff should become generalist EWsassigned to a number of villages where theywould assist farmers in whatever basic production

activities were required, including rice, livestock,horticulture, etc. Over the three years of the project (1996–1999),

the Project developed three Working Models for theimplementation of extension:• A Community-Based Extension Methodology;• A functional model for the DAFO structure (with

emphasis on extension over administrative duties);• An Extension Management System (EMS).

The Project also developed two HRD Programs,necessary to provide staff with the capacity to applythe Working Models, in case they would be acceptedby the Ministry to be applied nation-wide: • Capacity-building of the DAFO technical staff;• Leadership development for administrators of

extension.For the purposes of this paper it is worth focusing

on three of these: (A) the Community-Based Exten-sion Methodology; (B) capacity building for DAFOtechnical staff; and (C) the Extension ManagementSystem (EMS).

A) Community-based extension methodology

Various ‘tools’, such as PRA, have allowed a famili-arity with participatory approaches to become wide-spread. However a PRA can only help to initiate workin a village and there is a danger that, once the PRAhas been conducted, extensionists or developmentworkers will be at a loss as to how to continue to main-tain a participatory operating approach. The otherextreme of participatory approaches is that they canbecome so complex that only staff with a high degreeof experience and commitment can apply them.

When working with rank and file EWs, theparticipatory approach needs to be readily described,and easy to follow. This will lose the detail that somepractitioners might want to see. But this is inevitablewhen processes are to be applied by large numbersof individuals. If the process developed is more com-plex than this, the bulk of staff will become frus-trated in their attempts to apply them and begin totake short-cuts. At the same time, it is worth remem-bering that, while the bulk of staff may be operatingat a medium level of ‘participation’, there willalways be those few staff who will operate to a muchhigher degree of participation, due to their own skilland commitment.

Four ‘Steps of Implementation’

The Extension Methodology introduced by PEP hadfour steps, which could be used to structure activitieseach season.

4The Pilot Extension Project was funded by the NovartisFoundation for Sustainable Development. 5PEP was implemented by the Agriculture ExtensionAgency, under the Dept. of Crop Production. While therewere other projects, in other Departments of the Ministry,aimed at improving extension in other sectors such as for-estry, livestock, irrigation, PEP was the only project whichtook the DAFO as the target unit, and dealt with the imple-mentation and management of all activities by the DAFO.

75

The function of each step6, as it would be appliedwhen starting to work in a new village which has notpreviously had regular extension, is as follows (seeFigure 2):

Figure 2. Four ‘Steps of Implementation’ for PEP’sCommunity-based Extension Methodology.

Step 1: ‘Preparation’

Farmers need to identify their problems and needs.DAFO staff then respond with a selection of tech-nical options, from which the village can decidewhat they would like to try. (In this way the tech-nical options are not presented as recommendations,but as answers to the farmers’ expressed needs). Thevillage should then choose a few ‘selected farmers’to try it first.

This initial step is carried out with the wholevillage involved, so that the issues have becomevillage-issues. All villagers are then engaged andwaiting for the results of the ‘selected farmers’ trial.

Step 2: ‘Training and Implementation’

Support is then focused on the ‘selected’ farmers toprovide them with the technical information theyneed to try the new technology. (In the first year thiscan often take the form of working directly withfarmers on a small trial plot, rather than formaltraining).

Step 3: ‘Follow-up’

Farmers implementing a new technology the first timeusually need to be shown where mistakes have arisenand to be provided with additional technical advice.DAFO staff should also encourage farmers at thispoint to compare their new activity with their old wayof doing things, to begin the process of ‘evaluation’.

Step 4: ‘Farmers’ Evaluation Planning’

Towards the end of the season the ‘selected farmers’should evaluate the results of their trials, and thenreport their results and experiences back to thevillage. Other farmers can then ‘evaluate’ these interms of their own conditions, and the village as awhole can begin to make provisional plans for thefollowing season.

The Community-Based Extension methodology isparticipatory in that each of the steps engages thefarmers in some degree of analysis and decisionmaking (this will be true if the training approach inStep 2 uses adult education principles!). The partici-patory nature of this process is further supported by:• Staff providing farmers with a choice of technolo-

gies so that it is clear farmers will make a decisionbased on an evaluation of their experiences.

• Inputs provided only in sufficient quantities for atrial to be conducted. This ensures farmers interestsare focused on the trial and not on obtaining freeinputs.

• The selection of farmer representatives shouldalways be carried out by the villagers. EWs pro-vide guidelines for the type of person, type of areasuitable for the trial, etc., but leave the selectionitself to the village.

• Extension activities planned to target ‘villageclusters’, with Farmer Exchange Meetings heldbetween villages. This allows farmers to seefellow farmers as sources of information andstimulation, as well as the EW. If conducted on aregular basis, a network can develop within thecluster.

6 These four steps were derived from what is commonlycalled the ‘project cycle’ which has the steps of a) datacollection; b) assessment; c) planning; d) implementation;e) evaluation. For the four steps in PEPs Community BasedExtension, the first three steps of the project cycle havebeen lumped into a single step called ‘preparation’.

Farmers’ Evaluation+ Planning

Follow-up

Training +Implementation

Preparation

1# Preparation 2# Training +Implementation

3# Follow-up

4# Farmers’ Evaluation

1# Preparation

1# Preparation inthe next season!

76

The process just described for the four Steps ofImplementation applies to the first year of initiatinga particular activity, when issues are still being iden-tified and new technologies are introduced and triedfor the first time by farmers.

But this is only the first stage in scaling-up. As anew technology is introduced and then adoptedwithin a village, two factors will be changing.Firstly, the number of farmers using the new tech-nology will gradually increase, from a few farmersusing it on a trial basis, to larger numbers using itextensively. Secondly, farmers’ interest, and skills inthe new technology, will also increase, from initiallya healthy skepticism, to a commitment to applying itrigorously, as well as greater competence in itsimplementation. Thus, while working directly with afew farmers on small trial plots is an appropriateextension activity in the first year, this would not bea constructive extension activity in later years.

Figure 3. Three ‘Stages of Development’ for typical villagesin adoption of a new technology.

To provide a framework for DAFO staff to identifythe ‘extension objectives’ for each season and planthe ‘extension activities’ accordingly, the Projectdescribed three Stages of Development (Figure 3).These are:

1. Engaging farmers’ interestIn the initial stage of working with a new village, theextension objectives should be to: • identify farmers’ problems; and • to work with a few farmers, to demonstrate tech-

nologies that farmers are interested in. (The description of the four Steps of Implementation,applied to this Stage of Development).

2. Intensifying Technology SupportAs the interest and numbers of farmers increases,DAFO staff need to find ways to meet the greaterdemand for technical assistance. The extensionobjectives should be to: • support the expansion of the use of the new tech-

nology (i.e. assist in the supply of inputs); and, • establish and provide formal training to Village

Extension Workers (VEW) who can provide onthe spot advice to new farmers.

3. ConsolidationAs the number of farmers continues to grow, therewill be a need for farmers to organise themselveswithout relying on DAFO staff for supply of inputsetc. Technical support is still needed for thosefarmers just beginning to take up the new tech-nology, with an additional need to stimulate furtheradaptation of the original technology (i.e. the PTDprocess) in an on-going manner. DAFO staff shouldaim, where appropriate, to: • form Farmer Production Groups to: (a) manage

supply of inputs and the sale of excess; and (b) con-tinue technology development within the village;

• develop the Farmers Exchange Meetings into a‘network’ within the village cluster to continue tostimulate technology development and adaptation. The four Steps of Implementation still provide a

framework for EWs to structure extension activitiesfor each Stage. For instance, as the numbers offarmers begin to increase in the second stage (Intensi-fying Technology Support), the step of ‘Preparation’should focus on (a) planning of inputs needed to scaleup, and (b) the role and selection of VEW. And so on.

The time each village spends moving througheach of these ‘stages of development’ depends on thecapacity of the villagers themselves, and the com-plexity of the issues they address. Many villages maynot see the third Stage fully established.

What started out sounding fairly simple with thefour Steps of Implementation, now begins to soundoverly structured. While three Stages of Develop-ment do reflect processes commonly encountered inthe establishment and scaling-up of a new tech-nology, they are not intended as a blue-print for theDAFO to follow. Their main function is to alertDAFO staff to the fact that conditions do change

Bank/Credit

Markets

Input suppliers

Committee+Farmers’ProductionGroup

1# DEVELOPING FARMERS’ INTEREST

2# INTENSIFYING TECHNOLOGY SUPPORT

3# CONSOLIDATION

77

from year to year. As a result, they will need toassess the status of development each season to beable to set appropriate extension objectives and toplan suitable activities to meet these.

B) Program of capacity-building for DAFO technical staff

For the application of participatory approaches tobecome a reality within an NES, it will depend to alarge degree on how these staff can be trained. Theaim was that the capacity-building program devel-oped by PEP be applied nationally for all DAFOstaff. To do this, the program was designed to beconducted in-service, with the staff carrying outextension in pilot or training villages, so that realimpact would be achieved in the course of thetraining.

The capacity-building program had four maincomponents:

On-the-Job Training:– to instill the processes of extension.

Formal training:– to provide concepts and technical skills.Workshops:– for review and reflection.Short Technical Inputs:– for specific technical knowledge.Many aspects of the program are particular to

Laos, but the On-the-Job Training component of theprogram is of general relevance to training EWs inparticipatory extension approaches. The full TrainingProgram is provided in Annex 1.

On-the-job training

Formal training can provide staff with knowledge,but skills, and to some degree the necessary attitudes,will come more effectively from experience. Staffwere initially provided with concepts and a generaloutline of the Community-Based Extension method-ology in a formal workshop. From then on work wasat the DAFO level.

Figure 4. On-the-job training process.

1# Guidelines for thenext step of extension

2# Back-up during field workin the village

3# Review of workin the village

4# DAFO staff continue working andapply the lessons in other villages

Trainer returns to base!

78

For each step of the extension process throughoutthe season, trainers travelled to the DAFO andworked directly with the staff there. The On-the-JobTraining process used was as follows (Figure 4):

Provide objectives + guidelines for implementationTrainers met with DAFO staff prior to each exten-sion activity in the villages, to discuss DAFOobjectives and to provide guidelines for theirimplementation. Back-up during field workTrainers accompanied DAFO to the village, wherethey provided back-up to the staff in implementingthe activities.Review experiences and summarise lessonsFollowing this, the experiences were reviewed andsuggestions made for improvement.Independent implementationTrainers returned to base leaving the DAFO staffto continue to work independently in the remainingpilot villages. Leaving the DAFO staff to work independently

was an important part of this process. It ensured thatthe ETU staff were clear that their job was trainingonly, and not to implement extension. It also ensuredthe DAFO staff knew that it was they who had tocarry the load and to make the real achievements.The first step of On-the-Job Training on the next visitof the trainers included the DAFO staff reporting onthe work they had continued independently. Thismentoring process was consolidated with strategicworkshops where the DAFO staff could reflect on theoverall process and their growing skills.

In the first year, the main emphasis of trainingwas on On-the-Job Training. It was only after havingworked through one cycle, and observed farmers’responses, that DAFO staff could appreciate thefunction of extension and how it could be carriedout. They could then recognise their own weaknessesand became receptive to formal training. Thus, in thefollowing two years, the formal training componentwas increased. The On-the-Job Training componentis gradually reduced to provide back-up for newtypes of extension activities, as staff meet as theywork through the three Stages of Development(Figure 5).

The whole capacity-building program coveredthree years. This period was necessary for thetraining to guide staff through key extension activi-ties they would require for the three Stages of Devel-opment. This is a relatively long period for trainingto cover, but as the training was in-service, staffwere active in the field and actually achievinggreater impact under training, than they had everachieved previously.

C) Extension Management System

This is an easily forgotten area of extension as mostefforts are focused on the technical aspects ofdeveloping extension methodologies and trainingstrategies. But unless the administrative and manage-ment’s systems support Participatory Extensionapproaches, staff will not be able to work in thisway. The Project identified three main issues whichwould need adjustment:

• Planning for extension is centrally driven with settargets. If DAFO staff are to work according tofarmers’ needs and to respond as the situationchanges, then the planning of extension activitieswill need to be locally based.

• Monitoring of extension activities in the past hasnot been difficult as there was little activity tomonitor and any reporting was simply madeagainst set targets. However, if funding is raisedto adequate levels and DAFO staff can workeffectively, the level of activity in larger numbersof villages will increase, with the results achievedbeing quite variable according to the situation ofeach village and the stage of development it hasreached. Reporting will need to provide infor-mation on the progress and impact of widespreadextension activity. Furthermore, administrativestaff at higher levels need to learn how to use suchreports in a functional and constructive manner.

• Funding of extension in the past has beenextremely limited. The funds that were availablewere ‘handed-over’ piece-meal with no accountingof actual use. For extension to be functional, anadequate level of funding will be needed. For thislevel of funding, a transparent system ofaccounting against actual use will be needed.

These three issues are deeply embedded in theexisting policy and administrative environment, andare not likely to change in any fundamental manner.However, PEP tried to reduce them to a technicallevel by encapsulating them in what was dubbed‘Extension Management Systems’ (EMS). Theseincluded:

• Planning procedure for staff on an annual, monthlyand personal basis.

• Reporting of activities to both inform and toengage the DAFO Heads

• Accounting procedures which provided quarterlyfunds managed and acquitted by PAFO andDAFO staff.

All of the EMS implied a decentralisation of theadministration of extension to the level of the PAFOand DAFO.

79

Implementation and impact of PEP’s models and programs

These models and programs were implementedduring the course of the Project with a fair degree ofconsistency. Where there were lapses, the reducedimpact that resulted only served to confirm themodels further. The main impact of the three modelsand program described in this paper were as follows:

Extension methodology

DAFO staff were able to engage the majority offarmers in all the pilot villages in a consistentmanner. The main extension activity had focused onimprovement of rice production. By the end of three

years 42% of all farmers had begun to use improvedtechnologies over 20% of the total paddy area in the46 pilot villages (Figure 6). The typical yieldincrease was over 50%. This is not difficult toachieve with a few new seed varieties and use ofchemical fertiliser. But the fact that this wasachieved in such a consistent manner by the DAFOstaff, across four DAFO in 46 villages, did indicatethat the Community-Based Extension methodologywas effective and robust enough to be applied by theDAFO staff.

This improvement in productivity could simplyhave resulted from the fact that some form of exten-sion had begun to reach farmers. But there wereaspects of farmers’ behaviour which indicated the

* training is preceded by data collection by DAFO staff in the field** elective training input*** late in Training Year 2#, or, early Training Year 3#→ On-the-Job Training: guidelines/back-up in the field/review

Figure 5. The 3-year training program.

TRAINING PROGRAM

YEAR 1 YEAR 2 YEAR 3

On-the-Job Training On-the-Job Training On-the-Job Training

→ → → → → → → → →→ → → → → → → → →→ → → → → → → → →

(back-up in the field)

→ → → → → → → → →→ → → → → → → → →

(back-up in the field)

→ → → → → → → → →(back-up in the field)

Formal Training Formal Training Formal Training

Extension: Concepts+ Methodology 5 days

Facilitation Skills+ Development(inc. PRA) 5 days

Extension Planning 1# 3 days

Communication Skills 5 days

Basic Technical Skills – Rice* 10 days

Extension Planning 2#+ Reporting 5 days

Group Development* 5 days

Basic Technical Skills –Irrigation Management 5 days

Basic Technical Skills –Livestock/Fish* 5 days

Workshops Workshops Workshops

Review of PRA and Roleof Village Extension Worker 5 days

End of Year Review 3 days

End of Year Review 3 days Evaluation of ExtensionMethodology + Vision 5 days

Short-technical Input Short-technical Input Short-technical Input

Rice Production 8 days Dry season crops 1 day

Seed selection (rice) 1 day

Compost making 1 day

Fruit trees 2 days

Total 26 days Total 23 days Total 28 days

80

‘participatory’ element of the Community-BasedExtension Methodology had had some effect onfarmers’ capabilities:• Farmers who had begun to use the improved tech-

nologies were articulate in what effect the inputshad had, and were adjusting their own practicesaccording to their conditions.

• In the Farmer Exchange meetings, representativesfrom different villages were making coherentstatements of the status and issues affecting riceproduction in their villages, and discussing theseamong themselves.

• Farmers who had been trained as Village ExtensionWorkers, were beginning to assist other farmers. But perhaps one farmer expressed the changes to

their approach to change, when he said: ‘In the pastwe had just tried new things. If they worked we con-tinued, and if they didn’t we stopped. Since DAFOstaff have been working with us, we now think aboutwhy something new works, and if it didn’t, why itdidn’t’. If this sort of thinking can be stimulatedacross the Districts and Provinces, this would be asignificant step to enabling and empowering largenumbers of farmers.

Staff training and capacity-building program

This had improved the knowledge and skills of theDAFO staff in a real and obvious way. Staff alreadystrong gained additional skills, particularly in areas

of facilitation and communication. Staff who initiallyhad been considered ‘weak’ began to function effec-tively. This was evident in their ability to conductvillage meetings, advise farmers in the use ofimproved technologies and to conduct short trainingcourses for Village Extension Workers.

In addition to the technical skills, staff who hadworked through the full three years of the Capacity-Building Program were then able to identify issuesand develop their own programs. As such they arenow able to continue to implement extension on aprofessional basis within the District. And, finally, areal commitment and self-pride in their work haddeveloped.

Extension management systems

During the course of the project, the annual andmonthly planning procedures became a regular fea-ture of the staff operation, not just for themselves,but within their respective DAFO. Operating fundswere provided at a minimal level, but on a regularbasis. In the final year of the project, these were pro-vided to the PAFO and DAFO on a quarterly basisand then managed and accounted for by them with ahigh degree of efficiency. Only reporting could notbe instilled. It was not possible for this to be projectdriven, in the absence of any demand for this type ofindicative reporting from the top.

Figure 6. Impact of extension: rate of adoption of improved rice technologies (46 pilot villages).

1996 1997 1998

Year

45

40

35

30

25

20

15

10

5

0

% H

ouse

hold

s an

d%

Are

a

% Households

% Area

81

Nonetheless, the EMS was able to give a profile tothe difficult area of management, and to demonstratesufficiently that it is an area which needs to beaddressed.

PEP was implemented from within the existingsystem, and was subjected to the various day to daydifficulties that affect the system as a whole. Thusthe impact described above should not be seen asresults achieved within a special situation, but in factrepresents what could be achieved on a regular basis.Indeed, even better results should be expected if themodels and programs were refined further and thetrainers more experienced.

Despite this optimism, there are two serious con-straints to the application of these on a national level.These are: (a) the need for adequate operationalfunds to provided to DAFO on a regular basis and(b) the need for a team or task force of trainers to beestablished who would be committed to the veryconsiderable but necessary task of training andmentoring DAFO staff on a national scale.

While these two factors are not going to fall intoplace quickly, already two of the working assump-tions and models applied by PEP have appeared inthe recently promulgated ‘Vision for the AgricultureSector’ published in October 1999. These were:(a) DAFO staff should work as generalist extensionworkers; and, (b) that the operational administrationand management of extension should be devolved tothe PAFO and DAFO levels. Meanwhile the appli-cation of PEP’s Working Models and HRD Pro-grams remains under review. Nonetheless PEP hasalready contributed to discussion on the establish-ment of extension in Laos which is likely to continueto evolve over the next months.

Integrating Participatory Extension Approaches into Mainstream Extension

From the two cases discussed here, it should be clearthat there are opportunities for the institutionalisationof participatory approaches in mainstream extension.

Firstly, PTD processes can be applied by EWsthrough a fairly simple mechanism of ‘offeringfarmers a choice of technologies’. Being a relativelysmall shift in the behaviour of EWs, this can beachieved without expensive and time-consumingretraining programs. As was the case with the intro-duction of wheat in Thailand, this can stimulate ageneral process of experimentation by farmers,which can then be harnessed by conducting JointMonitoring Tours by the researchers and EWs. Inthis way the issue of ‘linkage’ between research andextension is achieved directly in farmers’ fields,rather than across conference or meeting tables.

The case of PTD for the introduction of wheat inThailand was concerned just with the ‘transfer oftechnology’ aspect of extension of a single com-modity. Participatory Extension needs to include farmore than this and should include HRD objectives ofstrengthening farmers’ ability to analyse and makedecisions. To achieve this will require substantialchanges in the extension approach. For the case ofPEP in Laos, this was achieved through the appli-cation of a process (Community-Based ExtensionMethodology), rather than introduction of a fewtools or techniques (such as PRA).

To achieve such changes in extension proceduresand EW behaviour, considerable training inputs arerequired. In the case of PEP, the key element of thetraining strategy was a mentoring process called On-the-Job Training. It is difficult to see how changes inan extension process and EW behaviour can beachieved without this sort of mentoring over one ormore seasons. For this to become effective, a groupof trainers would need to be established as an Exten-sion Training Unit. This sort of input can only beachieved if there is going to be significant politicalcommitment at the top to the development of aneffective NES.

Often the call for increasing the ‘participatory’nature of extension (and research) has been lead byidealism to ensure greater equity for farmers. Tomany administrators, this does not automaticallyequate with improved production. If ParticipatoryExtension approaches are to be accepted by NESs,they will have to be able to demonstrate concreteachievements (not just for administrators, but for thefarmers too!). This was possible for both of the casesdescribed above. Such results, however, take time todemonstrate and in the complex environment ofextension, can easily be attributed to many otherfactors. Thus new rationale is needed to justifyfarmers’ involvement in decision making. One of themost compelling arguments for this is the ‘diversityof farmers’ production environments’. Once this isemphasised, it makes good sense that farmers be‘offered a choice of technologies, so that they canidentify the best one for their own conditions, or thatthey be consulted on their needs and opportunities.‘Diversity of farmers’ production environment’could therefore provide a simple pragmatic justifi-cation for higher degrees of participation in main-stream extension.

ReferencesByerlee, B. 1984. Wheat in the tropics: Economics and

policy issues. In: Villareal, R.L. and Klatt, A.R. ed.Wheats for More Tropical Enviroments. Proceedings ofthe International Symposium Sept. 24–28 1984, Mexico,D.F., 303 – 315. Sponsored by UNDP and CIMMYT.

82

Connell, J.G. 1992. A minimalist approach to participatoryextension. Journal of the Asian Farming Systems Associ-ation, 1 (3): 361–369.

Connell, J.G. 1999. Developing production of a non-traditional crop in S.E. Asia: Wheat in Thailand. WheatProgram Special Report, No. 44, Mexico, CIMMYT.

Department of Agricultural Extension 1995. ParticipatoryExtension Project: An action-research project to examinepotential for the development of participatory extensionprocedures for a national extension institution. Final

Technical Report for IDRC Project file: 91–0231,Bangkok, Thailand, DOAE.

Hagmann, J., Chuma, E., Murwira, K. and Connolly, M.1999. Putting process into practice: OperationalisingParticipatory Extension. Network Paper No. 94, London,UK, The Agricultural Research and Extension Networkof the Overseas Development Institute.

Van den Ban, A.W. and Hawkins, H.S. 1988. AgriculturalExtension, London, Longman Scientific and Technical,100–123.

83

Participatory Approaches and Scaling-up

E. van de Fliert1, R. Asmunati1 and W. Tantowijoyo2

Abstract

Participatory research and extension approaches have been accredited as being effective inachieving impact, although success has seldom been demonstrated on a larger scale. Participatoryactivities are often characterised by intensive guidance processes, which would not always allowfor scaling-up. However, to achieve sustainable impact, particularly when dealing with complexenvironmental issues where farmers have to work collectively, appropriate mechanisms forscaling-up need to be sought. This paper presents a framework for integrative research anddevelopment as a tool for planning and evaluation of participatory projects aimed at impact. It isargued that the objectives of an activity, i.e. research versus extension, should be clearly defined,and methods planned accordingly. Basic and applied research carries risk, and should therefore notnecessarily be scaled up. Adaptive research, however, can be incorporated in participatoryextension programs, in which farmers test and adapt innovation collectively and enhance theirexperimental skills for further application on their own farms. Appropriate mechanisms forscaling-up as well as key factors in the institutional and/or community setting will have to beidentified on a case-by-case basis to enhance the chances of success. Anticipating scaling-up at anearly stage of project implementation through the establishment of linkages with potentialmechanisms is likely to facilitate accomplishment later.

THE IMPORTANT role farmers can play in agriculturalresearch, development and extension, if only given achance, has become widely accepted (e.g. Chamberset al. 1990; Haverkort et al. 1991; Jiggins and deZeeuw 1992). Participatory research and develop-ment advocates the involvement of farmers ascollaborators at all stages of the process, particularlyat the early stages of problem definition and settingof research objectives. Benefits of such an approachinclude early definition of concepts of what tech-nology users are likely to adopt, and adaptation ofprototype technology to meet implementers’ needsand preferences, hence, the likelihood of greateradoption (System-wide Program on ParticipatoryResearch and Gender Analysis 1997).

Despite the growing recognition for participatoryapproaches, several criticisms are often fanned,among others relating to replicability and, hence, thepotential for scaling-up. So far, there is little docu-mentation of large scale replication of pilot successes.

Participatory research networks do not relate effec-tively to established, conventional extension mech-anisms, which are in many countries primarilydesigned to transfer standard technology packages ina linear mode from formal research centres to a selectgroup of contact farmers. These extension systemshave no way of dealing with location-specific tech-nologies and the ‘transfer of processes’ (experientiallearning, experimentation), as required for sustainableagriculture (Van de Fliert 1993).

This paper presents a methodological frameworkfor participatory research and development aimed atachieving impact on a larger scale, and some hands-on experiences of participatory research and exten-sion approaches, and of anticipating scaling-up. Theframework has been developed and implementedwithin the context of sweet potato Integrated CropManagement (ICM) and ICM training developmentin Indonesia by the International Potato Center andits local partners, as described in the followingsection, but should be applicable to the developmentof sustainable agricultural systems in general.

The sweet potato ICM project in Indonesia wasdesigned according to an initial version of the frame-work, and the framework developed as the project

1Integrated Pest Management Specialist, International PotatoCenter (CIP), ESEAP Regional Office, Bogor, Indonesia.Email: Van-de-Fliert@cgiar.org2Project Coordinator, Mitra Tani, Yogyakarta, Indonesia

84

advanced and expanded to other countries and topotato Integrated Pest Management (IPM) research.

Our experiences in these projects are specific to thecontext of sweet potato and potato IPM/ICM systemsin Southeast Asia, where we are dealing with highlydiverse smallholder farming systems. IPM and ICMare complex concepts requiring location-specific,informed decision-making and, under smallholderconditions, collective action. A predisposition of thisframework is that, for achieving the overall objectivesof enhanced problem-solving and decision-makingcapacity, and next, impact at a larger scale, intensivefarmer training is needed.

It is emphasised here that this framework is pro-vided, not as a cookbook containing recipes to befollowed rigidly, but rather as a systematic map fornavigating farmer participatory research. This paperwill specifically focus on how the use of a frame-work like this can help anticipate scaling-up of par-ticipatory research and extension approaches.

The Context of the Sweet Potato ICM FFS Project

During 1995–1997, the International Potato Center(CIP) in collaboration with Mitra Tani (a localNGO), the national Research Institute for Legumeand Tuber Crops (RILET) and the Duta WacanaChristian University (UKDW), implemented aproject to develop a protocol for a sweet potatoIntegrated Crop Management (ICM) Farmer FieldSchool in Indonesia. The project strategy consistedof participatory approaches and methods at allstages: • planning, implementation and analysis of a needs

assessment; • development of ICM components and integrated

approach; • development of a Farmer Field School protocol

for extension purposes; • training-of-trainers; and • pilot program planning, implementation and

evaluation. Project activities were implemented in major

sweet potato growing areas in East and Central Java,where crops are grown intensively throughout theyear with fairly reliable water supply.

The Farmer Field School (FFS) model applied inthis project is a training approach that, over the pastdecade, has been developed and applied as the pre-dominant methodology for Integrated Pest Manage-ment (IPM) training in the majority of Asia’s rice-growing countries. Having its foundation in non-formal education principles and initially beingdesigned for rice IPM training, it emphasiseslearning by doing, and empowering farmers actively

to identify and solve their own problems (Van deFliert 1998). Participation, self-confidence, and col-lective action and decision-making are fosteredduring the experiential learning process. Thisapproach seems highly consistent with the require-ments of active ecosystem management by farmersimplied by IPM and ICM. An IPM farmer fieldschool lasts for a whole growing season, involving agroup of at most 25 farmers in weekly sessions of,on average, four hours. The trainer is a facilitator ofthe experiential learning process, not an instructor.Each training session contains a set of activities thatfoster farmers’ analysis, decision-making andproblem-solving, including:• Field monitoring of observation plots in small

groups, considering environmental factors, cropdevelopment, pest and natural enemy occurrenceand interaction, and damage symptoms on theplants.

• Agroecosystem analysis, in which drawings ofobservations are made, and conclusions aboutcrop status and possible measures are drawntogether.

• Presentation of the agroecosystem analyses, anddiscussion aiming at a collective agreement onwhat crop management measures should be takenduring the coming week.

• Special topics dealing with locally occurring fieldproblems, or providing opportunities to discoverprocesses, causes and effects of phenomena occur-ring in the field; the ‘insect zoo’ (an enclosure) isa tool often used by farmers in the field school tostudy plant-herbivore-predator relations.

• Group dynamics exercises to enliven the school,strengthen the coherence of the group, and makethe members better aware of the importance anddynamics of group processes.The field school model is designed to allow

farmers to learn to take informed pest managementdecisions based on their own observations andanalyses. This ability often gives them more con-fidence in their own farm management skills, which,in turn, is reflected in improved, overall farmmanagement. A second important purpose of thefield school is that farmer groups are stimulated totake collective action, which in certain cases may beindispensable for effective IPM implementation, par-ticularly for pests such as rodents.

A successful IPM field school often results infollow-up activities, spontaneously organised, andfunded, by the field school graduates, such as IPMclubs in Vietnam (Eveleens et al. 1996). Duringthese follow-up activities, the farmers may studynewly occurring cultivation problems, organisecollective control measures, and even get into wideraspects of community development, such as rice-fish

85

culture and collective marketing of produce (Van deFliert and Wiyanto 1996).

In order to institutionalise the sweet potato ICMFFS model and scale-up training and implemen-tation, 82 farmer trainers and staff of the NationalIPM Program and 30 local NGOs were trained astrainers. Work plans were prepared for follow-upactivities to be conducted on a self-supporting base.Mitra Tani initiated a second phase of the sweetpotato ICM project to monitor and evaluate thesefollow-up activities during a two-year period (1998–1999). Documentation of the monitoring and evalua-tion outcomes is expected to provide an interestingcase from which other projects may learn, and tocontribute to the general appreciation of partici-patory research and development.

A Framework for ParticipatoryResearch and Development:

Cycling from Problem to Impact

The framework in Figure 1 presents a possible routefrom problem definition to impact within the contextof sustainable agriculture development (E. Van deFliert and A.R. Braun, unpublished). Anticipatingthe various stages of this framework is consideredimportant when aiming at large-scale impact. The

framework emphasises iterative phasing or cyclingof activities and a division of major responsibilitiesamong the various stakeholders, distinguishing threemain realms of activity: (1) research and develop-ment; (2) extension and implementation; and (3)monitoring and evaluation.

These three realms are strongly interconnected,and their respective activities will partly overlap intime and space. Additionally, the process is notlimited to a linear set of sequential activities, butallows for cycling within and between the activityrealms.

Research and development

The research and development realm comprises co-creative processes to identify the problems, generatenew information and innovations, consolidate themwith adequate existing farming practice, and thentranslate them into learning objectives and activitiesfor enhanced farmer performance. These processesare likely to be highly iterative and synergistic. Par-ticipatory research targeting the needs of poorfarmers should begin with collaborative identifi-cation and analysis of problems, needs and oppor-tunities, in an attempt to gain an understanding of thebroad agroecological and socioeconomic context.

Figure 1. Framework for integrative, farmer participatory research aimed at impact.

Research & Development

problem identification

innovation development:→ basic research→ applied research

definition of implicationsfor the implementationof the innovations

training development:→ farmer training

curriculum development

→ training-of-trainerscurriculum development

Extension & Implementation

impact

farm-level effects

practices

farmer-to-farmerdissemination

farmer training/learning

training-of-trainers

Monitoring

&

Evaluation

86

This includes the identification of already existingalternatives to solve the problem(s), which may needto be tested under different conditions, and shouldeventually be consolidated with innovations. Theproblem identification phase should lead to the (par-ticipatory) priority setting and formulation of theoverall project goals and specific research objectives.The final output is a prioritised research agenda.

Once the research agenda is set, innovationdevelopment follows. This phase is likely to includeboth a basic and an applied research component.Farmers’ involvement in innovation development isparticularly desirable at the level of applied research.Their role may vary from ‘analysts and evaluators’(Fano et al. 1996) validating existing technologies to‘research collaborators’ determining and testingtreatments in their own fields (Ashby et al. 1995;Braun and Van de Fliert 1997).

It is emphasised here that participatory technologydevelopment serves an essentially different purposefrom extension. Research carries risk that we do notwant to extend to larger groups of farmers attendingextension activities with the expectation that theywill learn something. Therefore, we involve onlysmall groups of farmers in participatory researchactivities and clearly define the objectives andexpected outputs of the exercise together with thefarmers. On the other hand, experimentation can beused as a learning method in extension, but mostlyserving learning and/or adaptation rather than tech-nology generation purposes.

The development component in the research anddevelopment realm emphasises the translation andvalidation of innovation development outputs inrelation to the agroecological, socioeconomic andcultural conditions in target areas. The developmentprocess should not end with applied research, whichis often considered the final step of researchmandates. Applied research should be followed upby deliberate attention to training development.

Experience has shown that linear, top-downresearch and extension, as practiced in conventionaltechnology transfer models, often failed because ofinappropriate technology and/or inadequate ‘pack-aging’ of the messages (Röling 1988). Moreover,consistency is needed between the nature of theinnovation and that of the extension approach andmethods applied to convey the innovation to farmers(Röling and Van de Fliert 1997). Therefore, toensure consistency, we should not only look at theinnovations per se, but also define the capacities thatpractitioners need to implement them as well as therequirements for the support system (input supply,markets, etc.).

This leads to an analysis and definitions of what achange in agricultural practice effected by the

developed innovations implies for the farmers. Whatknowledge, attitudes and skills do they need toimplement the new practices and ideas? Answeringthis question is central to the development of theapplied technology, and a prerequisite to thedevelopment of training strategy. The process ofdefining the implications of the implementation ofthe innovations may provide new insights forproblem identification and/or raise issues that needto be fed back to the phases of applied or basicresearch, or even problem identification.

Training curriculum development is the next com-ponent of research and development and thereforewithin the responsibility of scientists. Preferably,technical and social or extension scientists wouldshare responsibility, and farmers and extensionofficers would be involved in field-testing and vali-dation. Training development implies designingactivities, modules, and media for farmer training,carrying out pilot studies, and revising them accord-ingly. Once the curriculum for farmer training is set,a curriculum development for training the trainerscan begin, preferably applying the same methods asthose used for farmer training.

Extension and implementation

Extension and implementation encompass the phaseswhen efforts are made – either in formal or a non-formal settings – to share the innovation with largergroups of farmers who then test, evaluate and incor-porate (or reject) them in their farming practices.Changing farming practices should ultimately lead tosubstantive impact.

Extension—defined here as a function of dissemi-nating an innovation to a wider audience—is notusually considered part of the mandate of researchinstitutions (Fano et al. 1996). Therefore, suitablemechanisms and partners must be found to performthis function. To ensure that potential partners cancarry out extension work efficiently, scientists canplay an important role, contributing both technicaland methodological skills. These skills may be com-plemented by those of GO or NGO extensionworkers, who have a comparative advantage as com-municators at the village level. However, potentialtrainers must be trained themselves before they canbe expected to run a curriculum according to thetraining model specifications. The participation ofaccomplished trainers is critical to success in thefield.

In many developing countries, extension serviceslack the human resource capacity, in terms of bothquantity and quality of staff, to reach a critical massof their target audience effectively (e.g. Röling 1988).Much of the information obtained by farmers is

87

disseminated by other farmers, either directly bysharing experiences, or indirectly through demon-strations of sample field practices and the resultingeffects. Recent experience with IPM training inseveral Asian countries has shown the positive impactof involving farmers as trainers, and of enhancingfarmer networks in order to support farmer-to-farmerdissemination deliberately (Eveleens et al. 1996;Braun 1997). Farmer facilitators must be selectedwith care and given additional training on facilitationmethods. Training programs must also address farmerinteraction and horizontal communication require-ments from the start. i.e. during the planning stage.

The major actors in the implementation realm are,of course, the farmers. Farmers decide whether ornot to implement, adapt or reject an innovation.Enhanced knowledge and skills, obtained throughtraining, contact with fellow farmers or any otherform of learning, are catalysts for change in farmingpractices. While much research has been devoted tostudying the process of adopting innovation (Rogers1995), in terms of sustainable agriculture, adaptinginnovation, to farm-specific conditions, is considereda more valuable output, particularly under themarginal conditions of poor farmers. The ability toadapt guidelines rather than follow a standard recom-mendation is evidence of farmers’ enhanced capacityto experiment, analyse, evaluate and, finally, solvemany of their own problems without having todepend upon external advice.

Response mechanisms, however, are critical inthis realm, because farmers often receive contradic-tory messages from other sources (e.g. promotionalcampaigns by commercial companies aiming to sellalternative inputs) that could lead to confusion.Questions arising during implementation need to beaddressed by trainers, whose role includes sup-porting the adjustment process and helping bridgecommunications between farmers and researchers.

When farmers’ capacities and practices change,tangible effects at the farm level can be expected.These may include yield increase, reduction ofexpenditures, or a more balanced ratio of pests tonatural enemies in the field. When such changesoccur on a larger scale, an even broader impact canbe expected, such as the improvement of ruralpeople’s livelihoods and/or a healthier environment.If initial outputs prove beneficial to farm families,they will most likely be disseminated further, con-tributing to a general increase in the knowledge baseof the farming community.

Monitoring and evaluation

The monitoring and evaluation realm forms a maze,overlapping with and collating the other two realms.

Researchers must observe and measure what happensduring training and implementation, and must relateand/or recycle the information back to the researchand development realm for further adjustment orimpact assessment. Systematic monitoring andevaluation of projects assures the capacity to makeadjustments before it is too late, to learn from experi-ences and to justify the research investment. Rapidfeedback is critical when farmers are presented withnew variables (for example, a new variety, a newtechnology, or a more complex, integrated innovativeapproach). In participatory projects, monitoring andevaluation should be planned and implemented inconjunction with the farmers. Farmers should particu-larly be involved in defining indicators for evaluation,and in analysing evaluation results. In the case of sus-tainable agriculture, evaluation indicators shouldalways relate to the objectives and expected outputsof each phase. Within this context, well-defined indi-cators usually focus as much or more on people andthe environment as on technology and economics(Van de Fliert 1998). Monitoring and evaluation ofclearly defined indicators should generate valuablefeedback for adjusting current project methodology,improving future research and development, and pro-viding examples for other projects.

In order to be able to justify the research anddevelopment investment, the monitoring and evalua-tion system should be designed to analyse the out-puts in relation to the objectives set for each specificphase. This is depicted by the horizontal links inFigure 1, where the expected outputs of the activitiesand elements in the extension and implementationrealm relate directly to the objectives of the activitiesin the research and extension realm at the same hori-zontal level. After the evaluation exercise, we shouldbe able to answer the following questions: • Is the impact of the activities consistent with the

overall goal? • Do the farm-level effects concur with the intended

objective of the innovation? (for instance, wasthere a reduction of pesticide load on the farmecosystem as a result of IPM practices?)

• After training, have farmers’ capacities and prac-tices reached the levels required for implemen-tation of the innovation?

• Do dissemination mechanisms result in effectivefarmer-to-farmer communication?

• Are the processes of farmer education and training-of-trainers compatible with the curriculum design? These horizontal links clarify the idea that, in order

to achieve positive impact, research and developmentteams should seek mechanisms for incorporatingextension and implementation requirements whensetting their objectives for research and development.

88

Scaling-up for impact

Impact as depicted in the framework in Figure 1 isthe desired change relating directly back to the initialproblem definition and, hence, overall goals of aproject. Impact occurs as a result of effects at thefarm level, which in turn are induced by a change infarmers’ management practices. Our understandingof actual impact of a project implies that the processof change occurs at a fairly large scale with regard tonumbers of farmers, as opposed to potential impactas may have been demonstrated only on a pilot scale.This means that targeting for impact requiresscaling-up. Particularly with sustainable agriculturalapproaches such as IPM, where a more naturalbalance of the larger agroecosystems needs to beestablished for optimal effectuation of technologycomponents, collective action of farmers within thatagroecosystem and hence large-scale implemen-tation, is required.

Change leading to impact consists of two importantcomponents. One is the internalisation of (adapted)technology components in the farm managementsystem – which is more than adoption of a technologyonly – and the other, often neglected, component isthe change in knowledge and skills leading toenhanced problem-solving and decision-makingcapacity of farmers. As predisposed in the frameworkabove, achieving impact for sustainable agriculturalapproaches requiring enhanced problem-solvingcapacities calls for extensive farmer learning, whichis mostly achieved through intensive training andeffectuation of farmer-to-farmer dissemination. Asexperienced in IPM training in Asia, training usingparticipatory, experiential learning methods, such asthe FFS described above, have proven to be the mosteffective.

In order to provide quality training to reasonablylarge groups of farmers, a mechanism responding tothe needs of a certain project needs to be installed.The ‘hardware’ of such a mechanism consists of thechannels or institutions to be identified andmobilised, taking into consideration the desiredcoverage of training activities, expertise availablewith regard to approaches applied, and funds avail-able. The ‘software’ includes the training modulesand curriculum that should contain elements guaran-teeing replicability, and inducing motivation andcapacity building of trainers for self-supportedexpansion, including farmer-to-farmer dissemi-nation. Our experience is that deliberate attentionshould be given in training of trainers and training offarmers to importance and possible means and con-tent of farmer-to-farmer dissemination to ensure it tohappen in a satisfactory way. IPM training, forinstance, is field-based and cannot be disseminated

with the same level of effectiveness by talk only. Asa result, farmers participating in IPM training shouldbe provided with methods and ideas to effectivelyinform and teach their fellow farmers about IPM.

Scaling-up Participatory Processes

Farmer participatory research activities often resultin good output with regard both to technologiesdeveloped, since they are more adapted to farmersneeds and opportunities, and to enhanced problem-solving capacities of the farmers who have beeninvolved in the process. The latter is a very valuabletrait for farmers, allowing for adapted technologyimplementation and internalisation within the farmmanagement system. However, extending this traitamong more farmers should in most cases not bedone by involving more farmers in participatoryresearch activities, but rather by designing training/learning activities that enhance farmers’ exper-imental and analytic skills. As the two-pier structureof the above framework suggests (research-develop-ment versus extension-implementation), partici-patory activities can occur in either realm but serve adifferent purpose.

Concretely, participatory technology developmentdoes not serve the same purpose as training toenhance farmers’ problem solving capacities forfarm-level adjustment of technological guidelines.Any participatory activity should have clearlydefined objectives – research versus extension – tobegin with, not to raise false expectation among thepartners in the process.

Our experience is that research activities,especially those carrying risk, should be limited to asmall group of highly committed partners. Onlywhen proven technology components have beendeveloped can we think of scaling-up by sharingboth guidelines and methodologies with largergroups of farmers through some sort of learningmechanism. Nevertheless, socialising participatorytechnology development activities and processes, forinstance through community-level analysis work-shops or field days, is considered important forinducing wider interest among communities for bothparticipatory research and extension.

A main characteristic of the Farmer Field School(FFS) model is experiential learning about ecologicalprocesses by providing opportunities to farmers todiscover and experiment. Experimentation, however,serves primarily as a learning purpose in the initialmodel, although the FFS setting also allows forexperiments to test and adapt technology com-ponents. In the Sweet potato ICM FFS developed byCIP and its partners in Indonesia, the FFS model wasextended by incorporating activities to familiarise

89

experimental methodologies to the farmers andencouraging them to design, conduct and analysetheir own experiments on the FFS plot. This hasproven very effective and several alumni groups con-tinued to do collective and individual experimentsfurther to adjust the ICM guidelines to their farmconditions, and convince themselves of compatibilityof the ICM components in their farming systems.Again, the purpose here is (large scale) adaptationfor final implementation, and not (participatory)technology development which had long before beenpreceded by the FFS implementation in order to pro-vide its technical content.

Anticipating Participatory Scaling-up

Aiming at wider scale impact calls for anticipation ata very early stage of project management. In additionto thorough project planning, an important factor inthis is the early establishment of linkages with organ-isations or individuals that can provide mechanismsfor future scaling-up. Identifying such mechanismsshould preferably be part of the needs and oppor-tunity assessment. Linkages can be established andstrengthened by involving key persons in criticalevents during the various stages of project implemen-tation, for instance by inviting representatives ofcertain organisations to needs assessment analysismeetings, or an end-of-season evaluation andplanning workshop of a participatory technologydevelopment team. When project developmentreaches the stage of institutionalisation, roles, respon-sibilities and mandates of the various institutionsinvolved should be made very clear. An analysis ofexpertise and capacities may help to determinewhether there is a need for additional input, such astraining or provision of materials.

In the case of the sweet potato ICM FFS program,we originally intended to ‘scale-up’ through the eightfarmer researchers who had been involvedthroughout all stages of the project. The effort totrain them as master trainers, however, failed,because they lacked the experience of having learnedthemselves through FFS methodologies, and there-fore did not pick up the facilitation skills needed.

As a result, the National IPM Program wasselected as the major mechanism for scaling-up,since this program had an impressive, nation-widecadre of very capable and well-trained FFS facili-tators (Asmunati et al. 1999). The selected trainersfrom major sweet potato growing areas only neededupgrading with respect to (1) sweet potato culti-vation aspects, since they only had experience in riceand soybeans, and (2) the particulars of ICM asopposed to IPM, including a range of crop manage-ment and methodological activities.

It was also realised that the National IPM Programonly operated in irrigated areas, indeed coveringseveral major sweet potato growing areas. However,in order also to reach farmers in rainfed areas whoprobably needed the ICM technology more, NGOnetworks working in the field of sustainable agricul-ture were involved as a second mechanism forscaling-up. Since NGO programs generally have acommunity rather than a commodity focus, thetraining-of-trainers for NGOs had to be adaptedaccordingly, by emphasising the principles of ICMand FFS approaches instead of focusing on ICM andFFS for sweet potato, as was the case for theNational IPM Program training-of-trainers.

Due to the wider scope of ICM, the project hadinvolved the Directorate of Food Crops Production,in addition to the National IPM Program which fallsunder the Directorate of Plant Protection. Althoughthe role of this Directorate was very minor throughoutthe project, the presence of the head of the Sub-directorate for Rootcrops during an evaluation work-shop resulted in the preparation of a proposal for sub-mission to the Directorate’s Planning Bureau for athree-year Sweet potato ICM FFS Program. Thistargetted 490 FFSs, potentially involving 12 250farmers.

At the time this paper was written, the decision forprogram approval was still pending. The experience,however, shows that through long-time acquaintanceand involvement, possibilities, albeit minor ones, forscaling-up opened up that would not have emergedwithout long-lasting, and initially seemingly fruit-less, efforts.

Conclusions

When planning for scaling-up (participatory) projectefforts, it should first be clearly defined what wewant to scale up and what our objectives are, whichglobally could be categorised as either research orextension objectives. Do we need participatory tech-nology development aimed at generating andadapting technological guidelines to be replicated inmore, diverse places for a more widely adaptedresearch output? Or do we want to extend researchoutput, which may include both technological guide-lines and experimental methodologies for location-specific adaptation of such guidelines, to largergroups of farmers? These two objectives call for adifferent scale of scaling-up, the second being muchlarger than the first, and requiring a differentmechanism to be installed for scaling-up.

In terms of the framework presented above,activities with research objectives take place in theresearch and development realm, whereas scaling-upfor extension purposes comes under the extension

90

and implementation realm. The latter implies thedevelopment of learning protocols, training oftrainers and organising wide-scale implementation offarmer learning activities.

Attention to experimental methodology in farmertraining to enhance farmers’ experimental skills,allowing them to verify and adapt technologicalguidelines under their specific farm conditions, isperceived to contribute greatly to the successfulimplementation of sustainable agricultural practices.

The farmer field school model provides a goodenvironment for teaching such skills, althoughmodules adapted to the agroecological and social-cultural setting of a FFS program need to be devel-oped. Teaching such skills to farmers requires strongtrainer capacities, hence solid training of trainers,especially in countries with top-down extensionsystems primarily delivering preset recommendationsto farmers.

Selecting and preparing the right mechanism toachieve both quantity and quality in scaling-up istherefore of utmost importance. Setting clear objec-tives and anticipating implications for achievingthese, in a participatory way, are necessary to buildstrong programs at the desired scale.

ReferencesAshby, J.A., Gracia, A.T., del Pilar Guerrero, M., Quiros,

C.A., Roa, J.I. and Beltran, J.A. 1995. Institutionalisingfarmer participation in adaptive technology testing withthe ‘CIAL.’ Network Paper 57, ODI AgriculturalResearch and Extension Network, 43 p.

Asmunati, R., Wiyanto and Van de Fliert, E. 1999.Integrating Sweet potato ICM and FFS Approaches inGovernment and NGO Initiatives in Indonesia. Paperpresented at UPWARD Planning Meeting, 1999–2003Program Cycle, ‘Enhancing Sustainable Livelihood withRootcrops: Potentials of Local R&D’, Bogor, Indonesia,29–30 April 1999.

Braun, A.R. 1997. An Analysis of Quality in the Indo-nesian Integrated Pest Management Training Project.Report of a Technical Audit Conducted for the WorldBank of the Indonesia Integrated Pest ManagementTraining Project, Jakarta, National IPM Program, 71 p.

Braun, A.R. and Van de Fliert, E. 1997. The Farmer FieldSchool Approach to IPM and ICM in Indonesia: User par-ticipation. In: Local R&D: Institutionalising Innovationsin Rootcrop Agriculture, 44–64 (UPWARD), Los Baños,Laguna, Philippines, UPWARD.

Chambers, R., Pacey, A. and Thrupp, L.A. Ed. 1990.Farmer First: Farmer Innovation and AgriculturalResearch, London, IT Publications.

Eveleens, K.G., Chisholm, R., Van de Fliert, E., Kato, M.,Nhat, P.T. and Schmidt, P. 1996. Mid-term Review ofPhase III Report. FAO Inter-country Programme for theDevelopment and Application of Integrated Pest Controlin Rice in South and Southeast Asia, P.O. Box 3700MCPO, 1277 Makati, Metro Manila, Philippines.

Fano, H., Ortiz, O. and Walker, T. 1996. Peru: Inter-Institutional Cooperation for IPM. In: L.A. Thrupp,ed. New Partnerships for Sustainable Agriculture,Washington: World Resources Institute, 64–68.

Haverkort, B., Van der Kamp, J. and Waters-Bayer, A. Ed.1991. Joining farmers’ experiments. Experiences inParticipatory Technology Development, London, ITPublications, 269 p.

Jiggins, J. and de Zeeuw, H. 1992. Participatory Tech-nology Development in practice: process and methods.In: Reijntjes, C., Haverkort, B. and Waters-Bayer, A. ed.Farming for the future: An introduction to Low-External-Input and Sustainable Agriculture, London, MacMillan,135–162.

Rogers, E.M. 1995. Diffusion of Innovations. 2nd edn.,New York, Free Press.

Röling, N. 1988. Extension Science: Information Systemsin Agricultural Development. Cambridge, CambridgeUniversity Press.

Röling, N. and Van de Fliert, E. 1998. IntroducingIntegrated Pest Management In: Röling, N.G. and Wage-makers, M.A.E. ed. Rice in Indonesia: A PioneeringAttempt to Facilitate Large Scale Change. SustainableAgriculture: Participatory Learning and Action,Cambridge, Cambridge University Press, 153–171.

System-wide Program on Participatory Research andGender Analysis 1997. A Global Programme on Partici-patory Research and Gender Analysis for TechnologyDevelopment and Organisational Innovation. Agren Net-work Paper No. 72, London, ODI Agricultural Researchand Extension Network.

Van de Fliert, E. 1993. Integrated Pest Management:Farmer Field Schools Generate Sustainable Practices. ACase Study in Central Java Evaluating IPM Training.Doctoral Dissertation. Wageningen Agric. Univ. Papers93/3, Wageningen, PUDOC.

Van de Fliert, E. 1998. Integrated Pest Management: Spring-board to Sustainable Agriculture. In: Dhaliwal, G.S. andHeinrichs, E.A. ed. Critical Issues in Insect Pest Manage-ment, New Delhi, India, Commonwealth Publishers,250–256.

Van de Fliert, E. and Wiyanto. 1996. A road to sustaina-bility. ILEIA Newsletter, 12(2): 6–8.

91

Farmer-Based Extension in the Philippines: The World Neighbours – Mag-uugmad Foundation

Experience

L.A. Moneva1, J.B. Cadao1 and J. Jackson2

Abstract

Some experiences of the World Neighbours Mag-uumad Foundation, in the uplands of Cebu, thePhilippines, are discussed. The Foundation works on the premise that the sustainable developmentof the region is in the hands of the local farmers themselves and that extension of appropriate tech-nologies is best led by the farmers. A Soil and Water Conservation Package was demonstrated ona model farm and was readily adopted by the farmers. Farmers conducted their own experimentsand technology adoption was promoted through the ‘alayon’ (traditional co-operation groups)system. The farmers soon came to realise the importance of livestock on their farms. In collaborationwith the Forages for Smallholders Project, forages have been promoted for use as hedgerows andfor feeding livestock. It is concluded that ‘alayons’ are the best venue for learning, but that there isa need for more training to keep up with the rapid rate of adoption of new technologies.

THIS paper contains stories and lessons drawn fromthe experiences of Mag-uugmad Foundation, Inc.(MFI), or Mag-uugmad as it is popularly known.Mag-uugmad is an NGO formed in June 1988 by theimplementing staff and farmer leaders to carry outthe soil and water conservation program initiated byWorld Neighbours (WN) in some watershed areas ofCebu in 1981.

The Province of Cebu, the case study setting, is amountainous island situated in the central part of thePhilippine archipelago. The upland areas of Cebu,including its critical watersheds, are severelydenuded. This situation was precipitated by theexpansion of farming communities within the water-shed areas, the encroachment of inappropriate low-land farming practices into ecologically fragile sitesand the excessive extraction of resources from theremaining forest.

Mag-uugmad is working towards a farmer-centeredand process-oriented development pathway. Thisdirection is anchored on the belief that the rehabili-tation and sustainable development of the uplands are

the moral obligations of the very people who dependon its resources for a living, the farmers themselves.The promotion of forages integrated with soil andwater conservation through farmer-led extension isthe main thrust of this paper.

Participatory rural appraisal (PRA) was used toexamine the impact of forages integrated with soiland water conservation (SWC) in villages within thewatershed area. Field surveys and farmers’ work-shops were conducted to evaluate the processes ofFarmer-Led Extension.

MFI’s Approach to a Farmer-BasedExtension System

A farmer-based extension system (FBES) is adevelopment approach wherein farmer extensionistsshare with other farmers lessons in farming systemsdrawn from extensionists’ own experiences. Farmerextensionists also assist fellow farmers in identifyingkey farm problems, implementing appropriate tech-nologies and facilitating the formation of ‘alayon’ orfarmers’ workshop.

The farmer-based extension system evolved fromthe early successes of the soil and water conservationprogram. The SWC package was readily adopted by

1Mag-uugmad Foundation, Inc., Cebu City, Philippines.Email: mfi-cebu@mozcom.com2Cornell University, New York, USA. Email: jrj10@cornell.edu

92

farmers because their need for a steady supply offodder and woodfuel was immediately addressed bysuch components as hedgerows and forages. Themomentum of technology transfer was triggered bythe remarkable success of the early adoptors. WorldNeighbours, the project initiator, saw the need todevelop farmer extensionists to assist in the imple-mentation of a rapidly expanding program. Thosefarmer extensionists, then called farmer instructors(FI), were selected by WN from among the out-standing farmer adoptors in the community.

Farmer-based extension strategies

The farmer-based extension strategies proven to beeffective are model farm development, alayon for-mation, participatory farm planning, small-scaleexperimentation, and training/cross-visits. The inter-play of these strategies is the key to the success ofthe farmer-led extension.

Model farm development

The model farm displays the various SWC tech-nologies, cropping systems and land-use patternsappropriate to the locality. It differs from a demon-stration plot in the sense that it is not just a showcasebut a product of a long process of technology adapta-tion by the farmer. The model farm serves as a livingexample of the technologies promoted by the exten-sionists, without which trainings and cross-visitswould not be successful.

Alayon formation

The ‘Alayon’ is a traditional form of cooperation inthe village, wherein farmers group themselves andwork on each others’ farms on a rotation basis. Thismutual sharing of labour hastens the pace of thetechnology adoption, especially the ones that arelabour intensive. The alayon also serves as venue forgroup learning, problem solving and the promotionof equitability among farmers.

On-farm experimentation

Farmer extensionists conduct their own experimentsto ascertain the appropriateness of new technologiesbefore these are disseminated to other farmers.Farmers learn of new technologies mostly fromcross-visits to successful farmers and technologyresource centres but seldom do they adopt the tech-nologies on a wide scale without testing them first.

Experimental plots aid the farmer in extensionwork. Neighbours may take an interest in the exper-iment, keep track of its progress and readily adoptthe technology if it turns out to be successful. Therole of women in the conduct of the trials is veryimportant as they are more keen on monitoring.

Trainings and cross-visits

Trainings and cross-visits are especially helpful tofarmers from distant villages, where SWC modelfarms do not exist and the alayon is not practiced yet.The training methodologies proven to be effectiveare farm tours, farm practicals, sharing of experi-ences and other participatory techniques.

FBES management

The Farmer-based Extension System is managed bythe farmers themselves. Mag-uugmad deliberatelylimited its role to the facilitation of complementarysupport and the upgrading of farmers’ skills in exten-sion. In this way, the leadership and managementcapabilities of farmers are enhanced and their senseof ownership of the program is gradually reinforced.

The Creation of Sustainable Upland Agriculture Resource Centre (SUARC):

The Convergence of Farmer-led Extension Strategies

The soil and water conservation program waseventually turned over to peoples organisations(POs) by Mag-uugmad in the mid-1990s. A formalmanagement structure was created by the POs inpartnership with Mag-uugmad to consolidate thegains of the extension and the direct pace of the pro-gram expansion. This structure was called the Sus-tainable Upland Agriculture Resource Centre.

The Sustainable Upland Agriculture ResourceCentre (SUARC) is now a learning institution whichis co-managed by local farmers. It is the well springof lessons in upland development and serves as themain vehicle of Mag-uugmad in reaching out toupland communities. SUARC offers technical assist-ance, consultancy research and development toupland communities in Cebu and in other provincesin the Philippines. It is also offers extension tomember people’s organisations.

MFI-FSP forage project

Mag-uugmad, in partnership with the Forages forSmallholders Project (FSP), collaborated in a projecton ‘On-Farm Testing and Production of Forages infour upland barangays of Cebu City’. The projectintended to strengthen the technical capabilities andintensify forage production of farmers who areinvolved in the MFI-assisted Livestock DispersalProject.

Methodology

In answering the above question, Mag-uugmad con-ducted a survey and a focus group discussion (FGD)

93

in one of the sitios (sub-village or hamlet) where weworked with farmers on livestock production withforage production components. To address exten-sion, during the FGD we asked farmers who theywould approach if they needed technical advicerelated to forage production. We also used the Adop-tion Tree (introduced by FSP) to determine theextent of farmers planting forages within the sitioand in adjacent farming communities.

Mag-uugmad Learnings from Surveys and Interviews

Before soil and water conservation was introduced,farmers practiced tethering. After World Neighboursintroduced SWC, farmers planted live erosionbarriers which had multiple uses. Farmers alsostarted integrating livestock in the farming system.An average family in a sitio owns a carabao, a cowor a goat. The farming system is integrated withhedgerow species being chosen for soil erosioncontrol and its palatability to livestock. Among thespecies which have gained widespread acceptanceare: Setaria sphacelata (setaria), Pennisetum pur-pureum (napier grass) and Paspalum atratum.Farmers are using these forages to feed all their live-stock, propagating them through cuttings and tillers.

As for extension, farmers learned about forageproduction from their neighbours. They are gettingtechnical advice from PO leaders who were trainedby Mag-uugmad.

Analysis of MFI’s Learnings

Farmers will always look at immediate economicgains, low cost of inputs, little labour requirement,compatibility with skills and farm resources before

they decide to adopt a technology. In Mag-uugmad’sfarmers’ experience, adoption of forage production isremarkably fast.

After the farms were relatively stabilised throughsoil and water conservation structures, farmersfocused on soil fertility management technologies(use of animal manure, composting, etc.). It wasduring this stage that farmers realised the necessityof integrating as many livestock in the farmingsystem as they could.

From the extension point of view, the project isreadily accepted by farmers because of their need fora steady supply of forages.

Implications for Extension

Mag-uugmad’s learning from PRA activities indi-cated that, in the sitio where the project has beenadopted, there was at least one model farm thatshowcased the integration of forage production in thefarming system that served as an eye-opener to otherfarmers before spontaneous adoption. An exposuretrip from other villages to this sitio expedites thewidespread expansion of the project.

Alayons have proven to be the best venue forlearning. The alayon, being a community-basedinitiative, is highly replicable in other upland com-munities in the Philippines.

Since forage production has addressed the feltneeds of farmers, the technology will spread by itselfin the upland farming communities. But as adoptionis advancing faster from farmer to farmer and fromvillage to village, there is a need for more trainingand seminars on forage technologies. Farmer exten-sionists are also needed to sustain the momentum ofextension and the pace of the project expansion.

97

Feed Resources for Ruminants in Smallholder Farming Systems in Southeast Asia

D.A. Pezo1, E.F. Lanting2, Wong Choi Chee3 andP.C. Kerridge4

Abstract

The demand for meat and milk in Southeast Asia has almost doubled in the past two decades,and this trend is expected to continue in following years. There is a need to enhance productivityin the prevalent smallholder crop-animal systems, in order to match these increases in demandwithout jeopardising the natural resource base. Some estimates suggest that available roughagesmay be enough to sustain the actual livestock population, but better quality and high yieldingimproved forages, and other nutritional options are needed to ensure more efficient use of availableroughages. Current options for improving the efficiency of feeding systems in smallholder livestockfarms include: a) appropriate budgeting of available feed resources; b) greater use of improvedforages; c) supplementation of low quality feed resources; and d) treatment of crop residues.Research on genetic improvement of crop residues and the manipulation of rumen microbialpopulations should provide new options for the medium to longer term.

The Southeast Asian Setting

Trends in livestock production

THERE HAS been a significant increase in the percapita consumption of livestock products in South-east Asia during the past two decades. This has beenassociated with increases in family incomes, urbanis-ation and modification to diet patterns, along withthe higher elasticity of demand that characterise live-stock products. Between 1961 and 1995, total meatconsumption per capita increased from 9.4 to 21.0kg/year, but pigs and poultry remain the main meatsources (FAO 1999). The changes in per capita con-sumption, and human population increase, suggestthat there will need to be a 3.5 to 4.0-fold increase inlivestock products by the year 2020 (Delgado et al.1999).

The increase in demand for livestock products aswell as in population density will put more pressureon the natural resource base, as further expansionwill be on more fragile lands. Also there is likely tobe more competition between the use of resourcesfor livestock and crop production.

Overview of Smallholder Farming Systems

Almost all farming systems have a livestock com-ponent which includes both ruminants and non-ruminants. Most livestock are raised in traditionalsystems, except for large-scale intensive systems ofpig and poultry production practiced in somecountries; however, the latter are not addressed inthis review.

Most pigs and poultry are raised in traditional low-input systems. Local breeds are commonly used andanimals are allowed to scavenge outdoors, and arefed some household residues and crop by-products assupplements (Edwards and Little 1995). In suchsystems, animal productivity is low (ca. 100 g/dayLWG in pigs, and 50 eggs/bird/year for poultry), butalso production costs are low (Vercoe et al. 1997).Where the main purpose of pigs and poultry is togenerate income, farmers use improved breeds and

1ILRI, Los Baños, Philippines. Email: D.Pezo@cgiar.org2Livestock Research Division, PCARRD, PO Box 425, LosBaños, Laguna, the Philippines. Email. pcarrd@phil.gn.ape.org3MARDI, GPO Box 12301, Kuala Lumpur, Malaysia.Email: CCWong@mardi.my4CIAT, AA 6713, Cali, Colombia. Email. P.Kerridge@cgiar.org

98

concentrates, and under those circumstances pro-ductivity is much higher. For example, in NorthernVietnam, where cassava and other crops are grownfor feeding pigs, productivity is 250–500 g/day(D. Peters, unpublished data; Nguyen 1994).

In lowland rice-growing systems and intensiveupland systems, buffalo and cattle are used for landpreparation and transport, with manure, meat and/ormilk forming useful by-products (Vercoe et al.1997).

In extensive upland systems, cattle and goats arelargely kept for building wealth and generating cashwhen needed. The increased demand for meat hasresulted in a steady increase in the cattle and goatpopulation (Figure 1). This has allowed smallfarmers to increase their asset base. Buffalo numbershave remained constant or decreased slightly, due toincreasing mechanisation in lowland rice areas (Parisand Sevilla 1995). However, in some cases buffalonumbers remain larger than expected. This is prob-ably due to their ability to utilise crop residues andweeds or forage in waste areas and thus continue tohave a role as an asset and in contributing manure forrice-based systems.

In both lowland and upland systems, livestocktend to occupy niches that do not compete with otherforms of land use. This applies to both use of theland itself and the utilisation of farm labour. Thesystems in which they tend to be absent, and couldbe increased (Chee and Faiz 1991), are specialisedperennial tree crop systems such as rubber, oil palm,fruit, tea and coffee. These are often relatively largeenterprises where owners look on livestock as asource of damage or as a problem in management ofsuch estates. Large waste areas available for freegrazing of cattle, such as Imperata grasslands, areunderutilised compared to equivalent areas in Africa.

The livestock population remains greatest in thedensest cropping areas. Livestock are associated withintensive rather than extensive land use (which con-trasts with the situation in South America). This isbecause livestock are closely integrated with crop-ping systems; the crop providing sustenance for thehousehold and the livestock contributing draughtpower, manure for crops and cash for the family orto finance crop activities. In spite of these inter-actions, livestock compete with crops for land,labour and capital resources, and the competition isbecoming greater as human population density anddemand for land increases. However, where theseresources are used for livestock production, it isbecause the products derived from livestock areequal to or more valuable than crop products.

The integration of crops and animals in small-holder upland systems is the main avenue forincreasing animal production in Southeast Asia(Devendra et al. 1997).

Ruminant Production Systems

The ruminant production systems practiced in South-east Asia can be grouped in three major categories: (a) Systems integrating arable cropping and live-

stock;(b) Extensive grazing systems;(c) Systems integrating animals and perennial tree

crops.Combinations of cropping and livestock are the

most common ruminant systems. In these systems,livestock are controlled in intensive cropping areasby stalling or tethering and shepherded grazing ofroadsides, idle land and crop stubble (Devendra et al.1997). It is common to ‘cut and carry’ grasses,

Figure 1. Changes in livestock population in East and Southeast Asia, between 1961 and 1998 (FAO, 1999).

45 000

40 000

35 000

30 000

25 000

20 000

15 000

10 000

5000

0

Buffaloes

Cattle

Goats

Sheep

1961 1970 1980 1990 1998

Hea

ds, x

100

0

99

weeds, and/or crop residues to animals maintained inpens or even to those tethered. Supplementation withconcentrates is seldom practiced, and it occursmainly during the dry season.

Extensive management of native or naturalisedgrassland systems is not as common in the region asit is in Africa and tropical America. Extensivesystems tend to be low-output systems, in whichanimals graze native pastures or the understoreyvegetation of upland primary or secondary forests.Poor control of cattle in extensive systems oftencauses problems to crop producers who have to useprecious labour and capital resources to build fences.Livestock in extensive systems are more likely to beowned by absentee owners. In those intensivesystems where animals are always stalled, tetheredor shepherded by farmers, there is greater equity inownership and more efficient livestock production(Piggin 1997).

Integrating ruminants is common in coconutplantations but has often been discouraged in otherplantation crops such as palm oil, rubber and fruittrees. There is considerable opportunity for expan-sion. Considering that stocking rates of 3.0 sheep or0.25 cattle/ha are achievable under tree crops, it hasbeen estimated for Malaysia that if only 50% of theoil palm plantations are grazed, an extra 3.3 millionsheep and 275 000 cattle could be reared. Also, afurther 1.1 million sheep could be reared if only 20%of the rubber plantations were to be managed undergrazing (Chin et al. 1998). However, the adoption ofthese systems by a large number of farmers willdepend on resource endowment, supporting infra-structure, market potential and policies favouringintensification, diversification and specialisation(Devendra et al. 1997).

Feed Resources in Smallholder Farming Systems

Feeding systems for large ruminants in indigenoussystems in Southeast Asia are characterised by theuse of a wide diversity of feed resources. These arebased largely on the use of crop residues, forages,and other resources not directly used for human con-sumption. Competition is more likely to occur indairy systems because of the higher value for milkand milk products. However, the dairy cattle andmilk buffalo population is relatively small in theregion (FAO 1999). Rice straw and other cropresidues, naturally occurring grasses and broadleafweeds and shrubs are the dominant feedstuffs. Nitiset al. (1982) reported that a total of 56 differentgreen feeds were fed to goats in Bali. These includedtree leaves (35), grasses (15), legumes (3) and others(3). The distribution of the species of green feed fed

varied according to climatic zone, topography, landutilisation and soil condition. Agro-industrial by-products and non-conventional feed resources mayalso be important under local situations (Devendra etal. 1997).

Rice straw is dried and stock-piled as hay for usewhen there are no other feed sources, in the dryseason when other feedstuffs are in short supply or inthe wet season when land is occupied by crops. Inthese cases, farmers either feed the animals with ricestraw taken from the piles or let animals eat directlyfrom the piles. In other cases, fields from which riceand other crops have been harvested are grazed, andcattle benefit by having a mixed diet of crop residuesand green weeds. Farmers fully utilise these ‘waste’resources before using other options such as pur-chased feedstuffs and committing labour to produceimproved forages.

With a greater demand for meat products, farmersare modifying their feeding systems. In a few cases,there has been stratification of the industry wherecattle are bred on extensive systems and then movedto other areas for fattening. As an example, cattlebred on Masbate in the Philippines are moved toBatangas for use as draught and for fattening. Inother cases, improved grasses and legume trees andshrubs may be planted for cutting and feeding tostalled cattle.

Feeding systems for pigs, small ruminants, andfish make greater use of concentrates. These may bederived from alternative food crops such as cassavaand from food crop by-products such as rice bran.Nevertheless, high quality roughages such as leaf ofcassava and sweet potato, grasses and legumes arealso fed to small ruminants, pigs and fish.

Permanent pastures represent only 3.9% of thetotal land in Southeast Asia (Table 1). However,native and naturalised forages, as well as ‘weeds’ arenormally found in crop fields, roadsides, river banks,rice bunds and areas under fallow, and even asunderstorey vegetation in the forest, as well as underplantation crops (coconut, oil palm, rubber),although the latter are often underutilised. The esti-mated availability of these native forages from non-arable land is 110.4 million tons DM/year (Table 3),which would allow the maintenance of almost 30million ruminant livestock units (RLU1).

Crops provide a range of residues, agro-industrialby-products and non-conventional feedstuffs, whichcan be utilised by ruminants and non-ruminants,either in crop-animal or specialised systems. Most areproduced at the farm level (e.g. rice straw, sugarcanetops, sweet potato vines, cassava leaves, grain legume

1RLU Equivalencies: buffalo = 1.0, cattle = 0.8, goats = 0.1and sheep = 0.1

100

Source: FAO (1999).

aBased on 1997 crop area and production (FAO 1998); bAt 1:1 grain-straw ratio; cAt 1:1.2 grain-straw and cob ratio; dAt 1:3 grain-straw ratio; eAt 2 t DM/ha;fAt 0.5 t DM/ha; jAt 1 t DM/ha; hAt 5 t DM/ha; iAt 15% of cane produced; jAt 40% of fruit produced; 1At 3.7 t DM per Ruminant Livestock Unit (RLU)/year.

Table 1. Land use (× 106 ha) in Southeast Asia, 1994.

Land use type Brunei Indonesia Cambodia Laos Malaysia Myanmar Philippines Thailand Vietnam Total

Arable land 3 17 126 3 700 800 1 822 9 534 4 970 17 085 5 511 60 551Permanent crops 4 13 045 105 50 5 782 542 4 400 3 360 1 247 28 535Permanent pastures 6 11 800 1 500 800 281 345 1 280 800 328 17 140Forest and Woodland 450 111 774 12 200 12 550 22 248 32 400 13 600 14 800 9 650 229 672Other land uses 64 27 412 133 8 880 2 722 22 934 5 567 15 044 15 813 98 569

Total 527 181 157 17 638 23 080 32 855 65 755 29 817 51 089 32 549 434 467

Table 2. Availability of crop residues and agro-industrial by-products (1000 tons), from the most common crops in Southeast Asia.a

Crop residues/by-products Brunei Indonesia Cambodia Laos Malaysia Myanmar Philippines Thailand Vietnam Total Carrying capacity

(1000 R.L.U.)1

Rice strawb 1 50 632 1 414 1 970 18 900 11 269 11 269 21 280 26 397 135 253 36 554.86Corn stover & cobsc – 11 190 72 94 58 300 5 198 5 460 1 844 24 261 6 544.86Wheat straw4 – – – – – 333 – 3 – 336 90.81Sorghum stoverd – – – – – – – 690 18 708 191.35Camote vinese – 496 16 28 8 10 274 18 606 1 456 393.51Cassava leavesf 1 650 7 2 20 4 108 600 138 1 530 413.51Peanut hayg – 645 9 9 1 512 46 100 263 1 585 428.28Sugarcane topsh – 2 015 40 15 120 365 1 835 5 275 1 200 10 865 2 936.48Sugarcane bagassei – 4 291 31 13 240 576 3 900 8 846 1 350 19 247 5 201.89Pineapple pulpi .4 215 6 14 65 – 581 800 74 1 755 474.32

Total 2.4 70 134 3 571 1 589 2 482 21 000 23 211 43 072 31 890 196 951

Carrying capacity .6 18 955.1 965.1 429.5 670.8 5 675.7 6 273.2 11 641.1 8 618.9 53 230.0

Ruminants population (1000 RLU)

6.1 14 999.0 3 010.0 2 176.0 783.5 10 702.6 5 433.8 10 221.0 6 045.3 53 368.3

101

haulms and straw), whereas others are by-products ofcrop processing activities, frequently performed off-farm (e.g. rice bran, lower grades of rice grain, coprameal, oil-palm cake, sugarcane molasses, pineappleand citrus pulps, fruit rinds and peelings).

The fibrous crop residues represent by far themajority of the total volume of feeds produced inSoutheast Asia, and constitute the basic feedresources for most crop-animal systems. The avail-ability of cereal straws in Southeast Asia is 158million tons, 86% from rice. Considering theruminant population, the average availability perRLU is 2.9 tons/year, this being in excess of thepotential intake of this resource, which is 1.28 and0.73 tons/year, for buffalo and cattle, respectively(Devendra 1997). However, there is a wide variationamong countries, for example, in Vietnam and Indo-nesia, there are large excesses (4.7 and 4.1 tons/RLU/year), whereas the availability in Cambodiaand Laos (0.5 and 0.9 tons/RLU/year) is below thepotential intake of a RLU. The available cropresidues and the agro-industrial by-products theoreti-cally could provide all the roughage demanded bythe present ruminant population (Table 2); however,these estimates are based on the assumption that allresidues are exclusively used for feeding, are readilyaccessible to animals, and all the above-ground bio-mass is consumed. There is a need to quantify howmuch of these crop residues are actually used, and tostudy what proportion could be consumed understubble grazing, considering different crop andanimal management conditions.

There is a wide variation in the nutritive value ofcrop residues and agro-industrial by-products(Table 4). There is also variability within a given feeddue to factors such as genotype, temperature, rainfall,soil fertility, fertiliser management, harvestingmethod, post-harvest handling and storage. Ingeneral, fibrous crop residues are poor sources offermentable nitrogen, as their crude protein (CP)content is below the level required by rumen micro-organisms (6–7% CP). Most are low in easily degra-dable carbohydrates, and in mineral nutrients, as wellas in nutrients required to balance the products of

digestion to requirements. All these result in limitedintake, poor rumen function, increased methane emis-sions, and low animal productivity (Leng 1993),unless strategies are applied to overcome the nutri-tional constraints of these fibrous crop residues.

Options for Improving the Efficiency of Current Feeding Systems

Appropriate budgeting of available feed resources

There is a strong relationship between feedingpractices and cropping calendar in the crop-animalsystems commonly practiced in Southeast Asia(Figure 2). Seasonal changes in the availability ofcrop residues and other forages determine importantvariations in diet composition (Figure 3), resulting incyclical periods of feed scarcity and under-nutrition,alternating with periods of nutritional adequacy, andeven feed surplus, with consequences on nutrientintake and animal performance (Figure 4). Theimpacts on rumen function and nutritional efficiencydue to these changes through the year, as well as ofthe diversity of feeds used, is not fully understood.

Any attempt at balancing the offer and demand ofnutrients year-round should be built on a clear under-standing of the seasonal changes in the animal/herdnutritional requirements, as well as the availabilityand access to different feeds at the farm level, and inthe local markets. For example, in the case of rice-based farming systems, the most critical feedingperiods for ruminants are the dry season and the timewhen rice is harvested. In the former, lack of mois-ture limits forage availability and quality, while inthe latter, competition for labour between rice har-vesting and forage gathering, as well as the reducedarea available for tethering or growing cut and carryforages, limits the amount of feeds on offer (Sevillaand Carangal 1995). In some areas, for example, therainfed lowlands of Laos and Northeast Thailand, thedeficit extends to the entire rice-growing season, asanimals cannot access the rice fields. Theseproblems are becoming more critical with the inten-sification of crop cultivation (Roxas et al. 1997).

1Average yields proposed by Roxas et al. (1997).

Table 3. Availability of forages for ruminants from potentially grazing areas in Southeast Asia.

Land use type Area(million ha)

Yield1

(ton DM/ha/year)Availability

(million tons DM)

Permanent pastures 17.1 0.8 13.68Forest and woodland 229.6 0.2 45.93Permanent crops 28.5 0.4 11.40Other land uses 98.6 0.4 39.43

Total 373.8 110.44

102

Table 4. Proximate analysis, calcium and phosphorus values of crop residues, agro-industrial by-products, non-conventionalfeedstuffs and food waste materials.1

Feed resource

Dry matter

(%)

Crude protein

(%)

Ether extract

(%)

Crude fibre (%)

Ash(%)

N-Free extract

(%)Ca(%)

P(%)

IVDMD(%)

Atis (Anona squamosa)peelings 96.2 6.3 2.6 35.0 4.0 48.3 1.72 0.17 27.0seeds, roasted 96.7 14.7 22.4 44.1 1.9 13.6 1.14 0.21 9.0

Banana (Musa spp.)peelings 89.4 5.1 9.1 13.5 10.4 51.4 1.58 0.14rejects (pulp + peelings) 89.3 3.7 3.0 3.0 4.0 75.5 – –

Brewer’s bagasse/Brewer’s spent grains 91.8 20.6 13.1 13.1 2.6 42.3 0.36 0.39Cacao (Theobroma cacao)

pods, boiled 85.8 7.6 2.6 36.0 7.4 32.3 0.60 0.17pods, raw 90.7 9.5 2.3 31.3 10.1 37.5 0.90 0.24

Caimito (Chrysophylum caimito) rejects 93.4 4.5 20.3 19.5 4.0 45.2 1.06 0.11 34.0

Cashew (Anacardium occidentale) pseudocarp 87.8 7.3 6.1 8.9 3.7 61.9 0.67 0.18 –

Cassava (Manihot esculenta)leaves, sun-cured 90.0 20.1 4.6 16.1 7.5 41.4 1.43 –petiole – leaf meal 88.4 18.6 6.4 13.1 7.2 43.2 0.28 0.39

Castor (Ricinus cummunis)oil meal 98.9 36.8 6.2 28.2 6.7 21.0 1.88 0.70

Chico (Manilkara zapota)meal 97.8 2.4 4.2 16.9 2.7 71.6 1.02 0.12 46.0

Coconut (Cocos nucifera)cocopress cake 96.8 7.6 11.2 17.87 2.9 57.4 – –

Coffee (Coffea spp.)pulp 80.5 5.8 2.7 21.5 6.7 43.8 0.57 0.23hulls from roasted beans 88.3 5.12 1.0 51.9 2.6 27.8 0.33 0.07

Corn (Zea mays)cobs 89.7 3.5 0.8 29.7 4.5 51.2 – –cobs, fresh 27.0 0.8 0.3 10.2 0.6 14.9 0.06 0.02cob meal 92.2 3.0 0.9 34.6 2.4 51.3 0.19 0.04husks, dried 88.4 7.1 – 27.1 16.9 – 0.16 0.13leaves, fresh, wilted 45.0 2.8 1.0 14.14 4.1 22.5 – –leaves, sun-cured 90.0 5.5 1.6 29.5 8.7 45.0 0.35 0.13stems, fresh, wilted, milk stage 24.0 1.7 0.8 8.6 2.3 10.5 – –stems, sun-cured 89.0 4.00 1.3 34.8 5.6 43.6 0.25 0.11stover 89.0 5.7 0.4 28.3 8.9 45.7 0.41 0.16 48.3

Cowpea (Vigna unguiculata)empty pods 87.4 12.2 1.3 25.6 6.0 42.4 0.74 0.25seed 92.1 20.4 1.5 4.5 3.3 62.4 0.06 0.35vegetable pod meal 94.6 20.1 2.5 16.0 4.6 51.4 0.60 0.48

Durian (Durio zibethnus)rind 92.0 4.9 1.5 18.0 7.8 59.8 0.62 0.15

Guayabano (Anona muricata)peelings 94.8 6.4 3.7 32.2 3.0 49.6 1.18 0.16 32.0seeds, roasted 98.7 13.4 24.7 54.6 1.8 4.3 0.94 0.20

Jackbean (Canavalia ensiformis)fresh, full bloom 19.0 4.4 0.6 4.8 2.5 7.3 0.56 0.07leaves, sun-cured 89.0 19.9 2.8 21.9 11.3 33.1 2.55 0.33seeds 94.9 21.2 2.7 10.7 3.1 57.1 0.94 0.72seeds, roasted 92.0 25.8 1.8 9.8 3.7 50.9 0.31 0.34

Jackfruit (Artocarpus heterophyllus)leaves, fresh 34.0 4.7 2.2 7.0 5.1 15.0 – –leaves, sun-cured 89.0 10.9 3.9 17.7 22.6 34.1 – –peelings with seeds, sun-cured 89.0 9.5 3.30 11.5 10.3 54.2 – – 72.0seeds 97.0 12.3 1.1 5.1 4.6 73.8 – –seeds, boiled 92.7 13.5 2.2 9.5 2.9 64.7 1.12 0.20 72.0

Lanzones (Lansium domesticum)peelings 93.0 9.9 8.7 17.2 5.6 51.4 1.28 0.16 42.0

Mango (Mangifera indica)peelings 94.2 4.6 4.8 15.6 5.8 63.5 1.20 0.15 49.0

103

1Adapted from: Gerpacio and Castillo (1988) and Zamora and Baguio (1984).

Mungbean (Vigna radiata)empty pods 89.5 7.6 0.7 33.7 3.2 44.2 1.42 0.13 46.0sprout coat 95.2 11.4 0.4 35.5 5.8 42.0 0.94 0.14 35.0

Peanut (Arachis hypogaea)germ meal 96.7 26.9 42.7 17.4 3.0 6.6 0.20 0.61hay 54.4 8.7 0.8 16.6 8.3 20.1 2.72 0.21 56.6seed coat 94.9 14.5 11.6 21.9 2.4 44.4 0.50 0.13shells 92.5 4.8 0.6 66.5 6.6 13.6 1.38 0.07 20.0straw – 12.4 – – – – 1.10 0.15tops 90.5 15.8 2.8 21.4 7.4 43.1 1.63 0.22 52.0

Pigeon pea (Cajanus cajan)leaves 88.8 19.6 1.7 7.0 4.0 56.5 0.84 0.56

Pili (Canarium hypogaea)nut coat 93.4 9.9 15.6 27.6 3.2 37.0 0.78 0.33

Pineapple (Ananas comosus)bran or pulp 88.0 3.2 0.9 16.3 3.9 63.8 – –wastes (skin, pith) 11.7 0.4 0.1 1.4 0.6 9.1 0.05 0.01

Rambutan (Nephelium lappaceum)peelings 98.9 5.9 1.5 17.0 2.2 72.2 1.74 0.10 45.0

Ramie (Boehmeria nivea)Fresh, midbloom 15.0 2.3 0.4 4.9 2.0 5.0 0.63 0.04hay, sun-cured, ground 91.0 21.1 2.9 10.9 19.2 36.6 – –hay, sun-cured, ground, midbloom 86.0 19.3 2.4 13.2 15.9 35.2 3.70 0.24leaves, fresh 21.0 4.0 0.9 2.0 7.2 6.9 – –leaves, sun-cured 88.0 17.7 4.2 15.8 19.5 30.4 – –stems, sun-cured, midbloom 88.0 7.4 – 45.2 7.7 – – –

Rice (Oryza sativa)hulls 90.0 6.5 5.1 20.9 12.4 45.2 0.12 0.18straw 92.6 3.7 1.5 31.4 21.2 34.8 0.43 0.12 31.2

Sincamas (Pachyrrhizus erosus)hay, sun-dried 89.2 15.5 3.7 26.9 7.8 35.3 1.11 0.32

Sorghum (Sorghum bicolor)hay, sun-cured 91.0 9.4 3.2 32.1 8.2 37.8 0.35 0.14stover 81.7 5.2 1.3 22.5 9.1 43.6 0.36 0.16 54.2

Soybean (Glycine max)hulls 89.6 17.8 3.4 25.4 4.5 38.5 0.75 0.23leaves, fresh 18.0 3.5 0.5 5.8 1.7 6.7 0.25 0.07leaves, sun-cured 87.0 16.9 2.6 25.4 8.7 33.5 1.27 0.40pulp 91.3 27.5 11.0 19.0 3.1 30.7 1.20 0.38 70.0

Sugarcane (Saccharum officinarum)bagasse, dehydrated or sun-cured 93.0 2.5 0.8 39.70 4.50 45.45 0.48 0.27bagasse, wet 78.0 2.8 0.6 8.00 5.60 61.47 – –filter cake press (sugar mud) 92.2 6.6 6.6 12.56 19.86 46.59 2.69 1.36leaves, fresh 24.0 1.7 0.5 8.80 2.10 11.11 0.08 0.06top, aerial part with leaves, fresh 31.0 1.6 0.5 10.40 2.70 15.62 0.06 0.05top, aerial part, sun-cured 91.3 3.6 1.4 28.2 7.9 50.2 – –

Sweet potato (Ipomoea batatas)hay, sun-cured 85.0 10.2 2.4 17.7 14.2 40.6 1.17 0.72leaves, fresh 27.0 7.8 0.3 2.9 2.5 13.2 – –leaves, sun-cured 86.0 13.6 3.8 8.5 10.0 49.9 – –

Tamarind (Tamarindus indica)leaves, dehydrated 91.0 13.3 1.9 22.0 6.9 46.5 2.54 0.23leaves, fresh 36.0 5.2 0.8 8.6 2.7 18.3 1.00 0.07

Tapilan (Phaseolus calcaratus)hay 84.5 13.5 – 31.1 9.8 – – –

Tomato (Lycopersicon esculentum)pomace 90.8 16.0 8.2 30.4 5.7 30.6 0.34 0.37

Winged bean (Psophocarpus tetragonolobus)seed meal 92.6 31.6 15.8 10.0 3.6 31.6 0.76 0.42

Table 4. Proximate analysis, calcium and phosphorus values of crop residues, agro-industrial by-products, non-conventionalfeedstuffs and food waste materials.1

Feed resource

Dry matter

(%)

Crude protein

(%)

Ether extract

(%)

Crude fibre (%)

Ash(%)

N-Free extract

(%)Ca(%)

P(%)

IVDMD(%)

104

Figure 2. Cropping pattern and feed resources flow model for a crop/animal system practised in a sub-humid agro-ecozonein the Philippines (Palacpac 1994).

Figure 3. Diversity of feed resources utilized in a rainfed rice-based cattle system practised in the sub-humid agro-ecozoneof the Philippines (Palacpac 1994).

Figure 4. Seasonal variations in the proportion of nutrient requirements covered and BW changes in a rainfed rice-basedcattle system in a sub-humid agro-ecozone of the Philippines (Palacpac 1994).

11 1 2 1 2 3 4

Grazing Grazing

“Cut and Carry”

Dried/Preserved rice straw

Rice

Straw

Stubble

Mungbean

Hay/pods

Stover

Siratro Siratro

Weeds &Pastures

Rice Hay

FoodCrops

Legumes

CattleFeeding

Corn

5 6 7 8 9 10Months

12

9

6

3

0

kg D

M

Rice bran

Legumes

Mungo pods

Corn stover

Rice straw

Grass/Weed

May Jun

Jul

Aug

Sep

t

Oct

Nov

Dec Jan

Feb

Mar

Apr

CP

TDN

BW

200

175

150

125

100

75

50

25

0

%

May Jun

Jul

Aug

Sep

t

Oct

Nov

Dec Jan

Feb

Mar

Apr

105

Use of improved forages

In most crop-livestock systems, where poor qualitycrop residues constitute the basal diet, action couldbe taken to increase the availability of leguminousshrubs and trees, as well as herbaceous legumesduring the year (Steinbach 1997), in order to provideat least the fermentable nitrogen that is needed. Thisrole is more likely to be played by legume trees andshrubs, since many of them can carry green leavesinto the dry season, and they can be more readilygrown in borders of crop fields and spare areas thanherbaceous legumes. Leucaena leucocephala hadbeen widely promoted as a popular fodder legume inSoutheast Asia until the psyllid (Heteropsylla cubanaCrawford) infestation became a major problem (Chinet al. 1998). However, the advent of psyllid resistantand acid tolerant genotypes, as well as natural preda-tors, will ensure that Leucaena remains important(Shelton et al. 2000). Gliricidia, Calliandra, Ses-bania and Erythrina, as well as some of the newlyintroduced legume shrubs and trees such as Desman-thus spp. and Desmodium cinerea (= ‘D. rensonii’),Flemingia spp., are also playing a major role in someSoutheast Asian indigenous livestock systems.

Improved grasses should function as buffers incrop-animal systems, complementing the existingfeed resources, especially during the most critical

periods of the year. Improved grasses have beenshown to be not only more productive, but also moredrought resistant than native grasses (Stür et al.2000).

A range of improved forage technology optionsbased on forage species with broad adaptation tosoils and climate is now available for differentfarming systems (Gabunada et al. 2000). Farmers areadopting different options to meet their own needs atthe farm level. For example, those farmers thatidentify lack of labour as limiting, see intensivelymanaged forage plots as an attractive way ofreducing the demand for labour for livestockmanagement. In many cases, farmers intend to usethese intensively managed plots only at specifictimes, such as days when they have to go the market,when some family members are sick, or duringperiods of peak labour demand for other agriculturalactivities. Also, the availability of improved foragesfor ‘cut and carry’ has increased the capability tocollect manure for crop production.

The range of technologies tested by farmers hasincreased as they become familiar with forages andthey see more opportunities on their farms. The entrypoint is usually farmers’ interest in growing foragesin ‘cut and carry’ plots for feeding their animals.However, some technologies aimed at better

1FM = Flemingia macrophylla.

Table 5. Feed intake, digestibility and live-weight gain of ruminants fed roughages, supplemented with legumes.

Type of basal diet, legume and animal species used

Level of legume% DM

Straw/grassDM intake

% BW

Total DM intake% BW

Rationdigestibility

%

Live-weightchangeg/day

Gliricidia maculata as supplement (Doyle et al. 1986)

Untreated rice straw/young bulls 0 2.7 2.7 47 –11310 2.8 3.0 46 –5420 2.5 3.0 49 –9433 2.2 3.3 55 10

Urea-treated rice straw/young bulls 0 3.2 3.2 41 –289 3.1 3.4 45 63

13 3.4 3.9 50 13426 2.8 3.8 52 130

Gliricidia sepium as supplement for urea-treated rice straw/goats (Trung et al. 1989)

23 1.70 2.21 44 –3046 1.78 2.79 47 –10

Calliandra callothyrsus as supplement for Napier grass/sheep (Wina et al. 1997)

0 3.46 3.46 66Wilted Calliandra 30 2.21 3.80 54Fresh Calliandra 30 2.30 3.91 78

Stylosanthes guianensis (SG) as supplement for untreated rice straw/sheep (Lanting and Sevilla 1998)

58 (SG) 0.60 3.05 58 1840 (SG) + 7 (FM)1 0.58 3.00 55 22

106

resource management are emerging in the moreintensive systems. These include grasses grown incontour hedgerows or for stabilising gully erosion,legumes as ground covers and grasses and legumesfor feeding fish (Gabunada et al. 2000). In all cases,improved forages are used to complement and notreplace existing feed resources such as nativegrasses, weeds and residues on old cropland andstored crop residues.

Supplementation

Animals fed low quality forages, such as fibrouscrop residues or mature grasses, frequently requirefermentable nitrogen and minerals for efficientrumen function. A sub-optimal level of any of thesewill result not only in poor rumen function, but alsoin a low protein to energy (P/E) ratio in the nutrientsabsorbed by the host (Leng 1993). To achieve higherproduction levels, there is a need to balance the ratioof protein (absorbed amino acids) to the energy(volatile fatty acids) derived from rumen fermen-tation. Supplementation with by-pass protein, starchand even long chain fatty acids can do this.

There are many opportunities to produce, at thefarm level, the supplements needed for moreefficient utilisation of the basal diet. Herbaceous ortree legumes (Doyle et al. 1986; Trung et al. 1989;Lanting and Sevilla 1998), as well as protein-richcrop residues, such as cassava leaves (Wanapat et al.1997) and sweet potato vines, could be sources offermentable nitrogen and minerals. On the otherhand, non-marketable roots, grains, and fruits couldprovide starch and/or easily fermentable carbo-hydrates (Chimliang 1989). Table 5 shows the bene-ficial effects of the inclusion of legumes in the diet,to supplement a basal diet of rice straw.

Another strategy is the use of small quantities ofconcentrates (0.3 to 0.6%BW or 12 to 20% of thetotal ration DM) to supplement the basal fibrouscomponent of the diet (Leng 1985), improving notonly rumen function, but also animal performancethrough the provision of by-pass nutrients. However,emphasis should be put on locally available agro-industrial by-products, such as rice bran, copra mealand oil cake, or non-conventional feedstuffs, such aspoultry litter, that can be produced on the farm ornearby, since smallholders usually have limitedaccess to commercially available feeds, due to logis-tical and financial reasons (Steinbach 1997). Someexamples of the responses of ruminants to the use oflocally available concentrate feeds as supplements toa rice straw basal diet are shown in Table 6.

Another option to supplement low quality foragesis the use of urea-molasses or multi-nutrient blocks,the latter being a source of by-pass nutrients, and

even a vehicle for anthelmintic dosage. The adventof the cold process for the preparation of the blocksmade this technology more accessible to small-holders. However, the degree of adoption varies withlocal conditions: characteristics of the basal diet,availability and cost of ingredients, end-productprices and marketing capacity (Sansoucy 1995).

Supplementation with urea-molasses blocks andmulti-nutrient blocks improves rumen functionefficiency, but to get increases in animal produc-tivity, there needs to be sufficient roughage available(Ricca and Combellas 1993). The beneficial effectsof these blocks are not only due to urea, but also toother microbial growth factors, such as sulphur andtrace elements, contained in molasses (Sansoucy1995). Positive effects of the use of blocks have beenobserved in Southeast Asia with different ruminantspecies and production systems. For example,Hendratno et al. (1991) found that supplementationwith urea-molasses blocks increased the birth weightand live-weight of kids by 15% and decreased thekidding interval by 5.3%. Thu et al. (1993) foundthat working buffalo fed rice and urea-molassesblocks, at the beginning of the working periodploughed 20% more land, but after a month the dif-ference was 40%, as supplemented buffalo worked ata higher speed, recovered faster from work and lostless weight. Wanapat et al. (1999a) observed thatdairy cows having access to multi-nutrient blockshad a higher roughage intake, and improved milkyield by 2 kg/day, as well as higher milk fat content,resulting in higher economic returns.

Crop residues treatment

Alkali treatment of fibrous crop residues to improvetheir digestibility in ruminants has received a lot ofattention from researchers (Devendra 1997). Itspotential of being managed by small farmers indeveloping countries was demonstrated when urea-ammonification technology was studied in villagesin Bangladesh (Dolberg and Finlayson 1995). Eventhough the benefits of urea treatment of crop resi-dues on digestibility (Aryal 1989), intake and animalperformance are well known (Table 7), its adoptionby smallholder farmers has been limited.

Studies in Bangladesh and Sri Lanka demon-strated that smallholders apply this technology onlywhen they have plenty of straw, lack access tograsses, and when milk production and marketreturns cover treatment expenses (Saadulah and Siri-wardene 1993). Similar reasons have been expressedby farmers in India, but they also indicate they couldapply this practice if water, urea, plastics and othercovering materials are easily available (Singh et al.1993). The extraordinary uptake of this technology

107

by farmers of the Central Plains of China is anexception, since, in only six years, the amount oftreated straw rose sharply from a negligible amountto 11 million tons. If used for beef production, thisamount of treated straw supplemented with by-passprotein could produce over 500 000 tons of carcassmeat. The adoption of the straw treatment tech-nology in China has been facilitated by a stronggovernment support in technical assistance at thevillage level, ready availability of subsidised credit(0.7% interest rate per month) and assured supply ofurea at low prices (Dolberg and Finlayson 1995).

New Opportunities in Tropical Feed Resources Research and their Implications

for Small Farmers in Southeast Asia

Genetic improvement of crop residues

Wide variation among genotypes has been detectedin the ratio of leaf blade: leaf sheath: stem, as well asin chemical composition, in vitro digestibility andintake of rice straw (Table 8) (Pearce et al. 1988),

but the effects appear to be strongly influenced byenvironmental factors (Roxas et al. 1997). Kush etal. (1988) suggest that only after environmentalinfluences on quality attributes are understood andminimised, could an effective breeding programtargeted to enhance rice straw quality be justified.However, an important goal in such a programshould be to maintain or even improve grain yieldand quality, parallel to the enhancement of theresidue’s nutritional quality (Kush et al. 1988).

Results obtained by Vadiveloo (1996) suggest thatbreeding rice for enhancing straw quality is notincompatible with grain yield improvement, butattention need to be given to other traits. For example,a reduction in lignin concentration could result inhigher digestibility, but also may increase the lodgingsusceptibility (Ookawa and Kuni 1992), and insectresistance would also be at risk, unless high levels ofantibiosis to pests could be developed in those lowlignin genotypes (Pathak and Khan 1994).

In sorghum, Saini et al. (1977) suggested thatdigestibility, neutral detergent fibre and tannin con-tents are under genetic control, and consequently

a,bMeans with the same superscript in a given column do not differ statistically (p < 0.05)1DM intake of urea-treated rice straw only

Table 6. Effect of supplements on total intake, daily grain/milk production and feed efficiency in ruminants fed roughagesas a basal diet.

Animal species and ratio DM intake/% BW

Live-weight gain(g/day)

or milk yield(kg/day)

Feed efficiencykg feed/kg

animal product

Native sheep, fattening (Wina et al. 1995)

Napier grass (NG) 2.55 44.4 11.560% NG + 40% Calliandra calothyrsus (CC)L 3.83 61.8 12.460% NG + 40% CC + Urea 3.24 64.1 10.160% NG + 40% CC + Urea + Ammonium Sulfate 2.82 60.9 9.260% NG + 40% CC + Cassava flour 3.23 73.5 8.8

Young bills, fattening, 126 days (Sevilla et al. 1976)

50% RS + 50% Commerical Concentrate (C) 2.74a 540b 11.5b

35% RS + 30% C + 35% L 2.84a 710a 9.4a

Growing heifers, yearling to calving (Trung et al. 1987)

35% RS + 65% C 2.19a 450a 12.0a

35% RS + 20% C + 45% Poultry Litter (PL) 3.10b 450a 18.0b

Milking cows, first lactation (Trung et al. 1987)

35% RS + 65% C 2.70a 8.2a 1.1a

35% RS + 20% C + 45% Poultry Litter (PL) 3.50b 7.8a 1.4a

Milking cows, late lactation (Wanapat et al. 1999b)

Urea-treated RS (TRS) ad lib + Concentrate: Milk (C:M) 1:2 2.01 6.3a

TRS ad lib + C:M 1:2 + 0.56 kg Cassava Hay (CH) 1.8 6.1a

TRS ad lib + C:M 1:4 + 1.70 kg CH 2.3 6.1a

108

improvements through breeding should be possible.Moreover, Rattunde (1998) detected large geneticvariation for grain and stover yields in sorghum, withthese attributes not negatively correlated. Hence,there are considerable opportunities for selectionfavouring both traits. An impact assessment studyfor sorghum and millet (Kristjanson et al. 1998)estimated that a 1% increase in digestibility of thestover would result in increases in milk, meat anddraught power outputs ranging from 3.2 to 10.7%.

These results suggest that genetic manipulation ofchemical and/or morphological attributes is anotherapproach to enhance the nutritive value of cropresidues, and could eventually contribute to theimprovement of animal productivity in smallholderfarms in Southeast Asia. There are opportunities todo so through traditional breeding programs, andthrough genetic engineering. This is an area ofresearch that deserves joint efforts of agronomistsand animal scientists.

Manipulation of rumen microbial populations

Some of the legume tree foliages commonly foundcontain anti-nutritional factors, for example, tannins,mimosine, that are toxic either to the animals or torumen microorganisms (Kumar 1992), thus limitingtheir use as animal feeds. Previous exposure to somefeeds containing those compounds can promote the

proliferation of rumen microorganisms capable oftolerating or even detoxifying some of them (Odenyoet al. 1997). Jones and Megarrity (1986) demon-strated not only the presence of bacteria capable ofdegrading mimosine in animals previously exposedto Leucaena leucocephala, but also that it waspossible to inoculate non-adapted animals, pre-venting toxicity problems. Odenyo and Osuji (1998)isolated Selemononas species capable of detoxifyingcondensed tannins. These bacteria were isolatedfrom the rumen of East African native sheep, goatsand antelopes adapted to Acacia angustissima. As inthe case of the mimosine-tolerant bacteria, it is likelythat these microorganisms could also be successfullytransferred to non-adapted animals, to improve theefficiency of utilisation of high tannin forages, aswell as decreasing the detrimental effects of tanninson rumen cellulolytic activity.

There are also some possibilities to manipulate theprotozoa and fungal population in the rumen. Imai etal. (1995) demonstrated that Malaysian cattle andbuffalo have a higher population of protozoa species,considered to be fermenters of cellulosic materials,than do cattle from temperate areas. Orpin and Ho(1991) showed that the population of rumen fungi ishigher when animals are fed highly fibrous rations,but comparisons have not been made with animalsacross genotypes, feeding systems and environments.

a,b,cValues followed by the same letter are not significantly different (p < 0.05).

1IVOMD = In Vitro Organic Matter Digestibility; 2IVDMD = In Vitro Dry Matter Digestibility.Adapted from: Pearce et al. (1988) and Kush et al. (1988).

Table 7. Effect of animal genotype and straw treatment on feed intake and live-weight gain in heifers receiving urea-treatedor urea-supplemented straw (Schiere and Wieringa 1988).

Parameter Sahiwal Sahiwal × Native Jersey × Native

Upgraded Supplemented Upgraded Supplemented Upgraded Supplemented

Live-weight gain (g/day) 282a 105bc 185b 70c – 39c

DM intake (% BW)Straw 2.33a 1.89b 2.49a 1.83b – 1.70b

Grass 0.13 0.14 0.25 0.26 – 0.25Rice bran 0.29 0.31 0.54 0.56 – 0.54Total 2.75 2.34 3.27 2.65 – 2.49

Table 8. Variability in morphological and quality attributes in rice straw from different genotypes.

Morphological attributes Quality attributes

Weight,% of total straw

Range Nutritive value Range IVOMD1 Range

Blades 36–52 Nitrogen, % 0.38–1.52 Stem Internodes 42–77Sheaths 29–43 Sulphur, % 0.01–0.13 Leaf Sheath 38–56Stem 16–27 IVDMD2, % 30–62 Leaf Blade 45–60

Voluntary Intake, % BW 1.0–2.7

109

Conclusions

• The increases in demand for meat and milkobserved in Southeast Asia during the past fewdecades are creating opportunities for farmers notonly to hold livestock as an asset, but also toincrease productivity through improving calvingrate and turnover (fattening). There are also somedevelopments in the dairy sector, which requirethe availability of better quality forages, to reducethe dependence on external inputs.

• In Southeast Asia, there is scope for increasing theavailability and utilisation of feedstuffs by live-stock, and consequently animal productivity,within those systems integrating ruminants withannual crops and also a greater integration of live-stock with plantation crops.

• The livestock feeding systems practiced in small-holder farms in Southeast Asia are complex,because of the number, as well as the diversity offeed resources utilised. These systems tend to beclosely dependent on cropping patterns, and inmost farms, crop residues are the major elementsof the basal diet. Therefore, technological inter-ventions designed to improve these systemsshould be built on the characteristics of the indig-enous or existing systems, incorporating scientificknowledge in tropical ruminant nutrition, andlooking for opportunities to favour synergisticeffects to improve the utilisation of crop residues.

• The use of improved forages constitutes an optionfor improving the efficiency of current feedingsystems in smallholder farming systems. Culti-vated grasses should function as buffers com-plementing existing feed resources, whereasherbaceous and woody legumes can provide thelimiting nutrients for efficient rumen fermentationin diets based on crop residues, mature grassesand weeds. There are also other opportunities forimproved forages to function as multi-purposespecies in these systems.

ReferencesAryal, I.K. 1989. The effect of urea treatment and/or sup-

plementation on the feeding value of rice straw forsheep. MS Thesis. UPLB, Los Baños, Philippines, 80 p.

Chee, Y.K. and Faiz, A. 1991. Sheep grazing reduceschemical weed control in rubber. In: Shelton, H.M. andStür, W.W., ed. Forages for Plantation Crops. ACIARProceedings No. 32, ACIAR, Canberra, 120–123.

Chimliang, T. 1989. Cassava as energy source for on-farmcattle fattening. M.S. Thesis, University of the Philip-pines, Los Baños, 57 p.

Chin, F.Y., Wong, C.C. and Halim, R.A. 1998. Com-mercialisation/exploitation of forage resources for theanimal industry in Malaysia. In: Ong, H.K., Ho, Y.W.,Wong, C.C., Idrus, Z. and Bejo, M.H., ed. Local Feed-stuff Utilisation: Potential and Realities, Proceedings,20th Annual Conference of the Malaysian Society ofAnimal Production (MSAP), 27–28 July 1998. Putrajaya,Malaysia, 1–12.

Delgado, C., Rosegrant, M., Steinfeld, H., Ehui, S. andCourbois, C. 1999. Livestock to 2020: The Next FoodRevolution, Food, Agriculture, and the Environment Dis-cussion Paper No. 28. IFPRI – FAO – ILRI. WashingtonD.C., U.S.A, ILFRI, 72 p.

Devendra, C. 1997. Crop residues for feeding animals inAsia: Technology development and adoption in crop/livestock systems. In: Renard, C., ed. Crop Residues inSustainable Mixed Crop/Livestock Farming Systems.CAB International, Wallingford, UK, 241–268.

Devendra, C., Thomas, D., Jabbar M. and Kudo, H. 1997.Improvement of Livestock Production in Crop-AnimalSystems in Rainfed Agro-ecological Zones of SoutheastAsia, Nairobi, Kenya, ILRI (International LivestockResearch Institute), 116 p.

Dolberg, F. and Finlayson, P. 1995. Treated straw for beefproduction in China. World Animal Review (FAO)82(1). [ILRI/FAO. 1999. Farmers, their Animals and theEnvironment. Multi-document CD-ROM]

Doyle, P.T., Devendra, C. and Pearce, G.R. 1986. Someeffects of chemical treatment or supplementation on thekinetics of fibre digestion in the rumen. In: Doyle, P.T.,ed. Utilisation of Fibrous Agricultural Residues asAnimal Feeds, Melbourne, Australia,. School of Agricul-ture and Forestry, University of Melbourne, 28–37.

Edwards, P. and Little, D.C. 1995. Integrated crop-fish-livestock improvements in Southeast Asia. In: Devendra,C. and Gardiner, P., ed. Global Agenda for LivestockResearch. Proceedings of the Consultation for the South-east Asia Region, IRRI, Los Baños, the Philippines, 10–13 May 1995, Nairobi, Kenya, ILRI (International Live-stock Research Institute), 71–84.

FAO 1988. FAO Production Yearbook, 1998, Rome, FAO.FAO 1999. FAOSTAT Database. Electronic Version.

Rome, FAO.Gabunada Jr. F., Heriyanto, Phengsavanh, P., Phimphach-

anhvongsod, V., Truong Thanh Khanh, Nacalaban, W.,Asis, P., Vu Thi Hai Yen, Tugiman, Ibrahim and Stür.,W.W. 2000. Integration of adapted forages on farms inSoutheast Asia: Experiences form the Forages for Small-holders Project. In: Stür, W.W., Horne, P.M., Hacker,J.B. and Kerridge, P.C. ed. Working with Farmers: TheKey to Adoption of Forage Technologies. Proceedings,Workshop held in Cagayan de Oro, Philippines, 12–15October 1999. These Proceedings, see Contents.

Gerpacio, A.L. and Castillo, L.S. 1988. Nutrient com-position of some Philippine feedstuffs. Technical BulletinNo. 21, University of the Philippines at Los Baños, 108 p.

Hendratno, C.; Nolan, J.V. and Leng, R.A. 1991. Theimportance of urea-molasses multi-nutrient blocks forruminant production in Indonesia. In: Proceedings Inter-national Symposium on Nuclear Related Techniques inAnimal Production and Health. Vienna. 15–19 April1991, Vienna, Austria, IAEA, 115–169.

110

Imai, S., Abdullah, N., Ho, Y.W., Jalaluddin, S., Hussain,H.Y., Onodera, R. and Kudo, H. 1995. Comparativestudy on the rumen ciliate populations in small exper-imental herds of water buffalo and Kedah Kelantan cattlein Malaysia. Animal Feed Science and Technology, 52:345–351.

Jones, R.J. and Megarrity, R.G. 1986. Successful transferof DHP-degrading bacteria from Hawaiian goats to Aus-tralian ruminants to overcome the toxicity of leucaena.Australian Veterinary Journal, 63: 259–262.

Kristjanson, P., Zerbini, E., Rao, K.P.C., Kiresur, V., Hofs,P. and Ravindra, K.V. 1998. Genetic enhancement ofsorghum and millet residues fed to ruminants: An ex-ante assessment of returns to research, Nairobi, Kenya,ILRI, 67 p.

Kumar, R. 1992. Anti-nutritional factors, the potential risksof toxicity and method to alleviate them. In: Speedy, A.and Pugliese, P.L., ed. Legume Trees and other FodderTrees as Protein Sources for Livestock. FAO AnimalProduction and Health Paper No. 102. FAO, Rome, Italy,145–160.

Kush, G.S., Juliano, B.O. and Roxas, D.B. 1988. Geneticselection for improved nutritional quality of rice straw –a plant breeder’s view point. In: Reed, J.D., Capper, B.S.and Neate, P.J.H., ed. Plant Breeding and the NutritiveValue of Crop Residues. Proceedings, Workshop. ILCA,Addis Ababa 7–10 December 1987. [ILRI/FAO. 1999.Farmers, their Animals and the Environment. Multi-document CD-ROM]

Lanting, E.L.F. and Sevilla, C.C. 1998. Improving the utili-sation of Stylosanthes guianensis CIAT 184 in ricestraw-based ration with the inclusion of tannin-containing legume for sheep. In: Sarabia, A.S. and Lihat,T.S., ed. Proceedings of the 35th Annual Convention ofthe Philippine Society of Animal Science, 15–16October 1998, Manila, Philippines, 263–277.

Leng, R.A. 1985. Determining the nutritive value offorages. In: Blair, G.J., Ivory, D.A. and Evans, T.R., ed.Forages in Southeast Asian and South Pacific Agricul-ture. Proceedings. International Workshop. Cisarua,Indonesia. 19–23 August 1985. Proceedings No. 12,ACIAR, Canberra, Australia, 111–123.

Leng, R.A. 1993. Practical technologies to optimise feedresource utilisation in reference to the needs of animalagriculture in developing countries. In: Mack, S., ed.Strategies for Sustainable Animal Agriculture in Devel-oping Countries. FAO Production and Health Paper No.107, Rome, Italy, FAO, 107–120.

Nguyen, T.T. 1994. Role of women in the promotion oftechnology transfer in sustainable livestock productionsystems in Tuyen Quan Province. In: Preston, T.R.,Ogle, B., Le Viet Ly and Lui Trong Hien, ed. Sustain-able Livestock Production and Local Feed Resources,Ho Chi Minh City, Vietnam, Agriculture PublishingHouse, 110–114.

Nitis, I.M., Lana, K., Sutji, N. and Sudana I.B. 1982. Goatfeeds during the wet season in Bali. In: Jainudeen, M.R.and Omar, A.R., ed. Proceedings on Animal Productionand Health in the Tropics, Malaysia, University PutraMalaysia, 241–244.

Odenyo, A. and Osuji, P.O. 1998. Tannin-tolerant ruminalbacteria from East African ruminants. Canadian Journalof Microbiology, 44: 905–909.

Odenyo, A., Osuji, P.O., Karanfil, O. and Adinew, K. 1997.Microbiological evaluation of Acacia angustissima as aprotein supplement for sheep. Animal Feed Science andTechnology, 65: 99–112.

Ookawa, T. and Kuni, I. 1992. Varietal differences in thecell wall components affecting the bending stress of theculm in relation to lodging resistance in paddy rice.Japanese Journal of Crop Science, 62: 378–384.

Orpin, C.G. and Ho, Y.W. 1991. Ecology and function ofthe anaerobic rumen fungi. In: Ho, Y.W., Wong, H.K.,Abdullah, N. and Tajuddin, Z.A., ed. Recent Advancesin the Nutrition of Herbivores, Kuala Lumpur,Malaysia,Vinlin Press, 163–170.

Palacpac, E.P. 1994. Optimising utilisation of feedresources in rainfed lowland and rainfed upland rice–based farms in Pangasinan and Cavite. MS Thesis.University of the Philippines – Los Baños, Laguna,Philippines. 128 p.

Paris, T.R. and Sevilla, C. 1995. Priorities for socio-economic research in farming systems in Southeast Asia.In: Devendra, C. and Gardiner, P., ed. Global Agenda forLivestock Research. Proceedings of the Consultation forthe Southeast Asia Region, IRRI, Los Baños, the Philip-pines, 10–13 May 1995, Nairobi, Kenya, ILRI (Inter-national Livestock Research Institute), 107–116.

Pathak, M.D. and Khan, Z.R. 1994. Insect Pests of Rice.Los Baños, the Philippines, IRRI, 89 p.

Pearce, G.R., Lee, J.A., Simpsom, R.J. and Doyle, P.T.1988. Sources of variation in the nutritive value of wheatand rice straws. In: Reed, J.D., Capper, B.S. and Neate,P.J.H. ed. Plant Breeding and the Nutritive Value ofCrop Residues. Proceedings, Workshop. ILCA, AddisAbaba 7–10 December 1987. [ILRI/FAO. 1999.Farmers, their Animals and the Environment. Multi-document CD-ROM]

Piggin, C. 1997. The role of leucaena in swidden croppingand livestock production. In: Conference on IndigenousStrategies for Intensification of Shifting Cultivation inSoutheast Asia. Bogor, Indonesia, 25–27 June 1997.ICRAF (mimeo).

Rattunde, H.F.W. 1998. Early-maturing dual-purposesorghums: Agronomic trait variation and covaraitionamong landraces. Plant Breeding, 117: 33–36.

Ricca, R. and Combellas, J. 1993. Influence of multi-nutrient blocks on live-weight gain of young bullsgrazing sorghum stubble. Livestock Research and RuralDevelopment, 5(3): 31–38.

Roxas, D.B., Wanapat, M. and Winugroho, Md. 1997.Dynamics of feed resources in mixed farming systems inSoutheast Asia. In: Renard, C., ed. Crop residues in sus-tainable mixed crop/livestock farming systems. Proceed-ings, International Workshop. ICRISAT, Patancheru(India). 22–26 April 1996, Wallingford, UK, CAB Inter-national, 101–112.

Saadulah, M. and Siriwardene, J.A. de S. 1993. Reasons forsuccess and failure of straw treatment and straw feedingin Bangladesh and Sri Lanka. In: Singh, K. and Schiere,J.B., ed. Feeding of Ruminants on Fibrous Crop Resi-dues. Proceedings, International Workshop. National

111

Dairy Research Institute. Karnal (India). 4–8 February1991, Delhi, India, IARC, 277–288.

Saini, M.L., Paroda, R.S. and Arora, S.K. 1977. Inter-specific hybridisation in sorghums for improvement inforage yield and quality. Indian Journal of Genetics andPlant Breeding, 37, 218–224.

Sansoucy, R. 1995. New developments in the manufactureand utilisation of multi-nutrient blocks. World AnimalReview (FAO) 82(1). [ILRI/FAO. 1999. Farmers, theirAnimals and the Environment. Multi-document CD-ROM]

Sevilla, C.C. and Carangal, V.R. 1995. Research method-ology appropriate for rainfed lowland crop-animalsystems. In: Devendra, C. and Sevilla, C.C., ed. Crop-Animal Interaction. IRRI Discussion Paper Series No. 6.IRRI, Los Baños, Philippines, 251–263.

Sevilla, C.C., Perez, C.B. Jr. and Gatmaitan, O.M. 1976.The effect of ipil-ipil (Leucaena leucocephala Lam deWit) leaves in rice straw-based rations on growth per-formance and carcass characteristics and yield of bulls.Philippines Journal of Veterinary and Animal Sciences,2: 21–32.

Schiere, J.B. and Wieringa, J. 1988. Overcoming nutritionallimitations of rice straw for ruminants: Response ofgrowing Sahiwal and local cross heifers to urea-upgradedand urea-supplemented straw. Asian-Australasian Journalof Animal Science, 1: 209–212.

Shelton, M., Piggin, C., Acasio, R., Castillo, A., Mullen,B.F., Rika, I.K., Nulik, J. and Gutteridge, R.C. 2000.Case studies of locally-successful forage tree systems.In: Stür, W.W., Horne, P.M., Hacker, J.B. and Kerridge,P.C., Working with Farmers: The Key to Adoption ofForage Technologies. Workshop held in Cagayan deOro, Philippines, 12–15 October 1999. These Proceed-ings, see Contents.

Singh, M., Amrith Kumar, M.N., Rai, S.N. and Pradhan,P.K. 1993. Urea-ammonia treatment of straw undervillage conditions: Reasons for success and failure. In:Singh, K. and Schiere, J.B., ed. Feeding of Ruminants onFibrous Crop Residues. Proceedings, International Work-shop. National Dairy Research Institute, Karnal (India),4–8 February 1991. IARC, Delhi (India), 289–296.

Steinbach, J. 1997. Alternatives to crop residues as feedresources in mixed farming systems. In: Renard, C., ed.Crop residues in sustainable mixed crop/livestockfarming systems. Proceedings, International Workshop.ICRISAT, Patancheru (India), 22–26 April 1996. CABInternational, Wallingford, UK, 197–213.

Stür, W.W., Ibrahim, T., Tuhulele, M., Le Hoa Binh,Gabunada, F., Ibrahim, Nakamanee, G., Phimphachanh-vongsod, V., Liu Guodao and Horne, P.M. 2000.Adaptation of forages to climate, soils and use in small-holder farming systems in Southeast Asia. In: Stür,W.W., Horne, P.M., Hacker, J.B. and Kerridge, P.C.,Working with Farmers: The Key to Adoption of ForageTechnologies. Workshop held in Cagayan de Oro,

Philippines, 12–15 October 1999. These Proceedings,see Contents.

Thu, N., Dong, N., Hon, H. and Quac, V. 1993. Effect ofurea-molasses cake on performance of growing andworking local buffalo and cattle. Livestock Research andRural Development, 5(1): 46–53.

Trung, L.T., Palo, L.P., Matias, J.M., Abenir, E.E., Lapinid,R.R. and Atega, T.A. 1987. Dried poultry manure andleucaena in rice-straw based blended diets for dairycattle. In: Dixon, R.M., ed. Ruminant Feeding SystemsUtilising Fibrous Agricultural Residues. InternationalDevelopment Program of Australian Universities,Canberra, Australia, 119–210.

Trung, L.T., Antonio, F.S., Palo, L.P., Atega, T.A.,Lapinid, R.R. and Panaligan, O.G. 1989. Urea-molasses-mineral block as a supplement to rice straw andGliricidia vs. copra meal + rice bran as supplements tourea-treated straw for goats. Terminal Report submittedto the International Development Research Centre(IDRC). UPLB, Los Baños, the Philippines, 20 p.

Vadiveloo, J. 1996. The relative importance of leaf andstem fractions on the nutritive value of whole rice straw.Proceedings, Silver Jubilee Malaysian Society of AnimalProduction Conference held in Kuching, Sarawak,Malaysia on 28–31 May, 1996. Malaysian Society ofAnimal Production, Kuala Lumpur, Malaysia, 210–211.

Vercoe, J., Coffey, S., Farrell, D.J., Rutherford, A. andWinter, W.H. 1997. ILRI in Asia: An assessment ofpriorities for Asian livestock research and development,Nairobi, Kenya, ILRI (International Livestock ResearchInstitute), 54 p.

Wanapat, M., Petlum, A. and Pimpa, O. 1999a. Strategicsupplementation of high-quality feed block on roughageintake, milk yield and composition, economical return inlactating dairy cows. Asian-Australasian Journal ofAnimal Sciences, 12(6): 901–903.

Wanapat, M., Pimpa, O., Petlum, A. and Boontao, U. 1997.Cassava hay: A new strategic feed for ruminants duringthe dry season. Livestock Research for Rural Develop-ment, 9: 2 (electronic version).

Wanapat, M., Puramongkon, T. and Siphuak, W. 1999b.Feeding of cassava hay for lactating dairy cows duringthe dry season. Asian-Australasian Journal of AnimalSciences, 12: in press.

Wina, E., Kayadu, M. and Tangendjaja, B. 1995. Effect ofurea, ammonium sulfate or cassava flour on the perform-ance of sheep fed a mixture of elephant grass and freshCalliandra. In: Prosiding Sem. Nas. Sains dan TeknologiPeternakan, 176–181.

Wina, E., Tangendjaja, B and Gunawan. 1997. Wilting ofCalliandra calothyrsus: Its effect on sheep performance.In: Prosiding Sem. Nas. II. Ilmu Nutrisi dan MakananTernak. 15–16 July 1997.

Zamora, R.G. and Baguio, S.S. 1984. Feed compositiontables for the Philippines. Los Baños, Laguna, PhilippineCouncil for Agriculture and Resources Research andDevelopment (PCARRD), Book Ser. No. 13, 264 p.