Participatory Technology Development With Farmer Research Groups: Lessons From Zanzibar

MINISTRY OF AGRICULTURE, LT\IESTOCKAI{D NATI.IRAL RESOURCES ZA}I{ZIBAR CASH CROPFARMING SYSTEMSPROJECT

CONFERENCEPAPER HARARE, OCTOBER1995

PARTICIPATORYTECHNOLOGYDEVELOPMENT WITH FARMER RESEARCHGROT]PS Lessonsfrom Zannbar

By; fsmdl Mgeni Martin Walsh Rupert Woods

zccFsP POBox 2283 Zarnbar Tel./fax:(05a)nU

CONTENTS Pageno. I

Introduction Local social,economicandinstitutionalfactorswhichled to the adoptionof the FRG approach

I

Why form FarmerResearchGroups?

2 ,'

Initial orperiences

2

Currentmethd for selectingandforming $oups

4

Methodsfor workingwith groups

6

Problemareas

9

Other approaches beingusedby ZCCFSPandMALNR

l0

The institutionalconstraintsto furtherdwelopmentofthe FRG approachin Zanabarand prospeotsfor the future

t2

Lessons

t2

Summaryof methodology

t4

ABBREVIATIONS

FRG MALNR PRA

zccFsP

FarmerResearchgroup Mnistry of Agriculture,Livestockandnaturalresources ParticipatoryRural appraisal ZannbarCashCrop FarmingSystemsproject

e

:

INTRODUCTION Project(ZCCFSP)startedin l99l with the The ZanzibarCashCrop FarmingSystems principleobjec'tivesof developingsustainable cashcrop farmingsystems,increasing foreign orchangeearnings,reducingexpenditureon importsandraisingthe incomesof rural householdsonZawibar (PembaandUngujaislands). It is fundedby Overseas DevelopmentAdministration(ODA) ofthe UK andmanagedby the Natural Resources Institute (NRI). It was originallyfoundedon the principlethat new export cropscould be devel0fedto replace,or at leastto supplement,the existingexport crop of cloves,for which the world marketdemandhasgreatlyreduced. At the outset,therefore,the approachtakenwasto selectindividualcropswhich were thoughtto havea good market,which would grow well andbe acceptableto Zanzrbari farmers. A processofPRAs, marketsurveysand agronomicinvestigationresultedinthe compilationof a list of candidatecropsfor furtherresearchanddevelopment. andthe decision This essentially top-downcrop-orientedapproachmetwith little success was soonmadeto adopta moreparticipatoryapproactr,to concentratemore on the farmingsystemspart of the projeottitle. with the developmentof FarmerResearchGroups This paperdescribesour experiences (FRGs)asa key methodfor enhancingfarmer-participatoryresearchand extension. LOCAL SOCIAI4 ECONOMIC AND INSTITUTIONAL FACTORS WHICH LED TO THE ADOPTION OF T'RG APPROACH The choiceof most suitablemethodsfor working with farmersdependsto a largeextenton institutional,socialandeconornicfactors. The conditionsthat havebeenexistingwithin asfollows: Zaru,rbarcanbe summarised Wthin Mnistry of Agriculture,Livestock andNatural Resources(MALNR) e A top-downapproachhasbeenused,usingthe principlethat the researchsectionsdo the researchandthenthe resultinginformationis 'taught' to farmers. o Researchhasmainlybeenfocusedon theoreticaltechnicalfixes:fertilizers,spacing,time of planting. It is seldomproblemorientatedor takesinto accountfarmerknowledgeor socio-economicconditions. On-stationtrials havebeenusedandsomeon-farmtrials with scatteredindMduals. o Lack of coherentpoliry. The policyis relatedto countryobjectives(eg productionof rice) and seldomrelateto farmerproblons or conditions. o Projectisationandunderfirnding.Donor-supportedprojectsettractthe bestofthe staff andare often autonomousin decisionmaking. Very little co-ordinationbetween sestions. Fundingfrom the govemm€ntis essentiallylimited to payrngstaffcosts and little is givenfor runningcosts. o Long historyof free or subsidisedinputs(fertilizers,seeds,agro-chemicals, tractor senvices).Thisis now reduceddueto lack of funds.

r Organisationby crop or discipline:makesit difficult to respondto moregeneral problemsfacedby farmers(eg declineof soil fertility). Rural areas e High populationdensity,smallfarm size,considerablelandborrowingin manyareas. . Significantvariationbetween&re&s:agro-ecological;cropsgolvn; importaoceof oF farm income;acc€ssto markets;involvementof womenin agriculture. r History of governmentcontrol hasresultedin a reductionin the senseof responsibility of farmersfor solvingthek own problemsanda lack of communitystructurewithin villages,outsideof family groups. Also an expectationof direct benefitswhenworking with government. o Low input-low output agriculturepractisedin most rueas.

WEY FORM FARMER RESEARCH GRO[]PS? The reasonsfor usingthe FRG approachcanbe summarised asfollows. o Relevantresearch.With farmerresearchgroupsthe researchpriorities arisefrom the constraintsandopportunitiesfacedby the farmeruwithin the group. If the group is well chosenthe new ideasor technologieswhichare developedby the group will be relevant to a wide sectionof farmingfamilieswithin that farmingsystemzone. o Orgonistion. From an organisationalpoint of view it is easierto work with 15 farmers in a group, ratherthan l0 individuals o Contimrity. lf farrnersdrop out otherscanjoiq the group will havea'memory' ofwhat hasbeendoneandthereis more chanceof activitiescontinuingifthe direct project assistance ands. Long-termcontact. o Exchangeof informationwithin thegroup. The learningproc€ssis muchquicker as farmerscanlearnfrom eachothers' experiences. o Responsibility. Farmersaremorelikely to take responsibilityfor their own problemsby talking throughthemand searchingfor possiblesolutions. o Collabwation betweenformers.Thereareadditionalbenefitsfor the frrmers from cooperationin a group: sharingresouroes;marketing;andattractingattentionfrom other organisations. o Collaborationwithin MAINR lt was felt that the FRG approachwould be a good methodfor encouragingco-operationwithin the MAINR Different sestionscan collaborateon differentproblems. The collabor&tioncanbe dernand-ledby the farmers, which reducesthe problemsof 'ownership' (the problemof one sestionbeingaskedto helpwith anothersection'strials). o Training.Good trainingexperiencefor ministrystaff

INTIIAL EXPERIENCES ZCCFSPis a so-called'processproject'. This meansthat the preciseend-pointandthe methodsfor gettingtherecannotbe clearlydefinedfrom the outse{. To this enda flerdble approachhasbeenusedto try to developmethodsbestsuitedto Zanabu conditions.

3 The farmerresearchgroup programmehasbeencautiouslybuilt up. Formingone group at a time and developingthe approachover tirne. In this way it is hopedto avoid makingtoo manymistakes.This is particulartyimportantwith farmerre.search groupsbecause considerableinvestmentis madeat the outsetto start a group. Cunently thereare a total of six FRGs,tlueeon eachisland(PembaandUnguja). A rangeof differentmethodshavebeenusedfor startingandworking with thesegroups. gainedduringthe formationofthe first few groupsare describedbelow, The experiences beforediscussion of the currentmethods. , '

Farming SystemZones As a first step,beforeforming anyFRG, the islands were sub-dividedinto FarmingSystem Zones, The objectivesofthis exercisewereto describewhat was happeningin agriculture within the differentzonesandthen,througha seriesof meetingswith other sectionsof MALN& to identify the mainagicultural constraintsandopportunitiesfor eachzone. Issues

locotion

It was decidedthat formationof FRGswould relateto the FarmingSystemZones. The key issuesregardingthe locationwerewhetherto target areas:which havethe most cashcrop opportunities;or which havethe mostproblerns,or wherechangeis happeningfastest;or to try andcover all areas? Also whetherto selecta villagerepresentativeof a certainzone or a villagewherethey haveaccessto two zones. Iswesregardinggroup selection Therearetlree mainfactorsto consider r Size o Composition r Method of forming a gxoup In the pastfarmerswere selectedmainlyon the basisof willingnessto co-operatein doing research. Whenforming a researchgroupthereareadditionalrequiranents,suchas wealttqfamily groupings,race,gender,ageandwillingnessto do research.Thesefactors haveto be considered for selectionof a groupwhichis representative of a wider population andwhichwill be ableto work together. For our purposeswe wanteda groupwhich was most representativeof the peoplemost involvedin agriculturewithin that area. It is easyto get homogenousgroups(eg womenor family-basedgroups),which are the sort of groupswhich the extensionserviceusuallyworks with. A mixedSoup, however,is more difficult to achierrebecauseit tnaynot representa naturalgrouping. The first goup wasformedby inviting a contactfarmerto selecta group. The resulting gfoup hasa reasonable mix of ages,but too manyof the group membersarerelated. There is only onewomanin the group, but womenarenot muchengagedin agriculturein this afea.

For the secondgroupit wasdecidedto ta(getwomenonly. This decisionwasbasedon the observationthat womendo most of the agriculturalactivitiesin the selectedfarmingsystem zone,whereasmenaremostlyinvolvedin earningoFfarm income. Following a more detailedstudy,howwer, it was discoveredthat althoughwomendo most of the agriculturalworh they were not muchirrvolvedin farmingon tho coral rag, an important farmingareafor the village,becausethey arenot ableto do the necessarybushclearance. A discussionwas heldwith the womenaboutthis issueandthey decidedto includefour menin the group. This group now consistsof eight womenandfour men.

CURRENT METHOD FOR SELECTING AI\ID F'ORMING GROUPS. Village selection Therewas muchdebateaboutwhereto form the mostrecentgroup. Therewere somethat favouredstartingin a FS zonewherethe projecthad not hithertobeen'involved,but where thereare perhapsfew cashcrop opportunities;othersfavouredhavinga secondgpoupin the largestzone,wheretherearethe most constraintsandopportunitiesfor agricultural development. The compromisethat was reachedwasto selecta villagewithin the largest zone,which was far from the existinggroup within that zoneandwhich alsohad accessto the adjoiningzone. It was felt that anytechnologiesdevelopedwould therebyhave relevance to two zones. Severalvillageswere chosenwhich met this generalselectioncriteria. Thesevillageswere visited over a period of two daysandinformationwas collectedon agriculturein each village. Finally onevillagewas chosen,whereit was consideredthat the most new developments were happening(independentof outsideassistance). Group selection Having decidedon the village,a teamof two peoplewere s€ntto the villagewith the task of identiSing possiblegroup members. Tkee separateinformantswere identifiedwithin the village,who wereselectedon their abilityto be ableto identify possiblegoup members, Theseinformantswere told aboutthe aimsof forming a researchgroup and were askedto zuggesta mix of people-differentages,gorder, wealth,family grouping. The importantcharacteristicfor all group membersbeinga willingnessto participateand try newtechniques. Following thesediscussionsandbasedon a comparisonof the threeseparatelists a final list of 15 nameswasdrawnup. Our ocperience is that the idealgoup sizeis from 12to 15 people,any largerthanthis andthe group becomestoo difficult to managewith no extra benefits. Creatinga researchpragramme We havefoundthat the importantfirst stepwith farmerresearchgroupsis to gain a thoroughunderstandingof their farmingpracticesand socio-economicconditionsbefore

makinganydecisionsabouta possibleresearchprografirme.Thereforewe conducta type of mini-PRA The methodologyusedis asfollows. l. Soonafterthe selectionof a researchgroup,we hold a discussion with the groupabout the conceptof researchandthe aimsofforming sucha group. A dateis thensetfor doing the PRA. 2. ThePRAteamare selected.The aimis for 8-10teammembers,drawnfrom the sestionswithin MALNR that havean interestin collaboratingwith the researchgroup. The teammeeta few daysbeforethe PRA anddrawup a checklist for datacollection.Topics that arecovered include:Land useandownership;labourissues;soil fertility and practices;croppingsystems(food andcashcrops);forestry;livestock; associated marketing;incomesources;householddecision-m#ng;andgenderissues. 3. Informationcollectionis baseduponsemi-structured intervieqrsandvisitsto the individualplots of eachFRG member.A teamof two or threeinterviewersspendone day with a singlefarmer,the morningvisiting the farmer'splots andthe aft6rnooncompleting the interviewat the farmer'shome. 4. Eachday's findingsarediscussedcollectivelyeach€veningandanynecessary modificationsar€madeto the checklistof the following day. 5. A meetingwith all the membersofthe FRG is heldon the final morningto providean opporhrnityto presentanddiscusspreliminaryresultsof the surveyandoutline subsequent stagesof the researchprocesswith farmersbeforethankingthemfor their co-operation. 6. Back in the office,a dayis givento the collectiveanalysisofthe data. Minutesof the analysisaretakento form the basisof a report summarisingthe findingsof the study. Also notesare compiledon eachfarmer,which generallytakesabouta week. Responsibilifyfor drawingup the reportis givento two teammembers. 7. Theteamthenget togetherto drawup a list of possibleareasfor researchandany furtheractionneeded,basedon the collectiveanalysis. 8. Finallythe researchproposalsandcommentsof the tearnare discussedwith the FRG to modi$ andrefinethe proposals,togetherwith anyadditionalsuggestion. The advantages anddisadvantages of this mini-PRAprocessare asfollows: Advantages e importanttrainingfunctionfor project andMALNR staff. Improvesinterview,report writing andanalyticalskills. o More rapidthan conventiondPRA which involvesmatrix andwealthranking. o The informationgatheredcanbe directly used. Disadvantages o Sometypesof informationare difficult to obtainwith zucha rapid proc€ss,for example informationaboutincome.

6 o Farmersexpectquickresponsefrom the researchteam. o Time consumingfor farmer.Takesroughlylrl2 daysfor eachfarmerand 4 - 5 daysfor the wholegroup.

METIIODS FOR WORKING WITH GROUPS The methodsusedfor working with groupshasdevelopedover time. The history of researchwith the oldestgroup (DayaFRG) is givenaBan exampleof this developlnent. Initiallythe ernphasis wasput on doingsimpletrialswith new cashcrops(chillies,ginger, furmeric). Simpletreatmentswereused,sushas* or - shade,* or - mulch. Monitoring was donebi-monthly. This approachwasnot very successfirlasthe farmersexpecteda lot from the new cropsandtwo of the threecropsprovedunsuitablefor the local conditions. Subsequently, the groupwas giventhe choiceof a wide rangeof tree crop seedlings, availablein the project andforestrynurseries.The seedlingswere sold dt a subsidisedrate, choiceof species,numbersandwhereto plant wasleft to the farmer. This was still not very satisfactoryasit was difficult to dealwith major iszues,suchas soil fertility andredevelopmentof the cloveplantationarea. A detailedstudywas then undertakenof eachfarmer(asdescribedfor PRA above). From this a muchwider rangeof researchthemesaros€. The themesareactuallyquite similarfor all the FRGs,although thereare somedifferences.The currentthernesfor DayaFRG are: Soilfenility maintenance.This coversinvestigationssuchaserosioncontrol, compost making a$oforestry, mixedcropping,economicuseof fertilizers,mulchapplicatiorl useof legumes,rotatio4 greenmanure,cover cropplngandfallowing. AIso the adviceto concentrateefforts on a smallerarea. Farm oatptt diversification. This is particularlyaimedat developmentofthe clove areas, whoreonly cloves,citssavaandbananasaregrown. Oneobjectiveis to ertend the home gardenideainto the cloveplantation,to havemulti-storeycroppingof a rangeof crops. Main emphasisis on trylng new species,extendingcultivationof someexistingspeciesand trytng improvedfood crop vuieties, Otheractivitiesin this themeincludenursery establishment, monkeycontrol andbreadfruitdry,ng. Developmentof rice areds. It is felt that the rice valleysareunder-utilised,despitethe fact that manyhavewater all the year. The objectiveis to improveproductivitythroughcrop diversificatioq diversificationinto new uses(eg sugarcaneor vegetablesinsteadof rice) anduseofbundsto improvewaterholdingcapacity. Following the PRA in Daya,the finding$were discussedwith the group and the individual farmersdecidedwhat sort of investigationtheywould do. Now thereis a mix of : t Semi-formaltrials. Examples:comparisonof artificial fertilizer, compostandcattle msnureon bananas.Thesetrials could be categorisedasresearcherdesigned- farmer implemented. t Informal iwestigations.Examples:compostmaking,monkeycontrol, trying new crops/varieties.Couldbe describedasfarmerdesigned- farmerimplemented.

7 o Discassionon topicol iswes. Examples:developmentof clove are:!s,managernent of forestareas,discussion mayarisefrom the discussions, aboutcrops. New investigations otherwiseanydevelopmentsarenoted. Often the point is madein thesediscussionsthat the farmersmusttakeresponsibility themselves andthat they shouldco-operatebecause if only onepersonactson their own they are likely to fail. c Training. Studytours, visits Measarementsand recording All trials are kept very simple. The majorobjectiveis to get the farmers'opinioris-about the viability of a technology(technicalandsocio-economic).In sometrials the farmer aretaken. measuresthe yield, in othersno measurements aretakenis an erosioncontrol trial. This is a An exampleof a trial whereno measurements rangeof specieshavebeenplantedon for the farmers and a wide new concept completely the contour,togetherwith a comparisonofvegetativeand mechanicalmethodson a single farm. The ideais to get a first responseofthe farmersaboutwhat type cif rnethodsmay havea chanceto work. This will also serveasa demonstrationplot. Thereis not a strongtradition of recordinginformationor writing paperswithin the MALNR andwhereaseflorts aremadeto improvethe analyticalandreport writing skills, paperworkis alsokept to a minimum. The currentthinking is to havea short annualreport on the progressofeachgroup. Inputs Material inputs shouldbe kept to the minimumpossible. Thereare cases,however,when the outcomeof a trial is completelyuncertainandthe costsof materialsor labourare high; or wherethe benefitsto the project,in termsof the knowledgegained,maybe potentially greaterthan thosefor the farmer;or whena largeplot hasto be plantedto get realistic results. In thesecaseswe fe€l it is justifiableto give a gteaterlevel of zupport. In a recentexample,a farmerwas paidfor someof his labourfor implementinga large erosioncontroltrial on his farm.Someofthe plantingmaterialwasalsoprovided. Before agreeingthe rate to be paid an independentlabourerwas askedto give a quotefor carrying out the work andthe paymentof the farmerwasbasedon this. Erosioncontrol techniques are new to the islandsandthe aim wasto comparea largenumberof techniqueswithirr a singlefarm. Farmer training and studytours The useoftraining or studytours is an importantcomponentofthe work. Sometimes farmersaretakento other areas(eg from one islandto the other)to get first-hand experience.This maybe to seea specifictechnology(eg compostmaking,vanilla cultivation,erosioncontrol), or to seenew farmingsystems(eg cattlecrop integration). Trainingin the villageis alsogivenfor somenewtechnologiesfor which the farmershave had little experience.

8 Meetings Meetingswith the grouptakeplaceat inegularintervals,depending on the time ofyear and the researchactivitiestaking place,the intervalcanrangefrom 2 to 10 weeks.In between the groupmeetingsindividualfarmersarefollowedup. Group Ieader ship/organisation Someof the groupshaveappointeda chairmanandsecretary,othershavenot. Decisions of this kind are left to the groupsto decide. For the group$which haveseveralrtr6mbers from a singlefamily group it is perhapsbetternot to haveanychairmanasthis may cause rifts. Topical studies Sometimesissuesor researchpossibilitiesarisefrom the work with the FRGsfor which moredetailedstudyis required.An example(plannedfor October1995)is a studyon the agroforestry possibilitiesfor the coral rag (soil type). The objectiveof this studyis to identify indigenousandimportedtree specieswhich are currentlygrowing on the coral rag, to investigatethe perfonnanceof eachspeciesandthe way in which they areused. From this studywe expectto be ableto designagroforestryinterventionsfor adaptationby the KangaganiFRG. Remlts It takesat leasta yearbeforeanyrezultsof co-operationwith FRGscanbe seenandmay take longer- It needstime to foster ganuineparticipation,for the farmersto understandthe conceptof researchandto start to overcomethe problemsofunrealistic expectationsfrom the farmersgroups. Therearetwo areasto considerwhentrylng to evaluatethe effectivenessof a group: the changeswithin the group andthe effecton farmersoutsidethe group. The changesthat haveoccurredwithin the groupsare significant. To takethe example againof the first group,Daya: r All farmersare startingto useorganicmatterrnanagement methodswhich they were not usingbefore-useofmulches,compostandcattlemanure. r Farmersaretaking responsibilttyfor reducingthe monkeypopulatiorqwhich is allowing themto plantlargerrangeof cropsoutsidethe immediatelocalityof the village. o A largenumberof new cropshavebeenassessed, somecontinueto be grown others wererejected. . All farmersare startingto diversifytheir cloveplantationareas,plantingsomecropsthat they hadformerlylaughedat the ideaof (eg timber species,cinnamon,other tree crops). . Somefarmersare startingto raisetheir own tree plantingmaterial. o Somefarmersareconcentratingtheir efforts in a singlearea(beforethey were using scatteredplots). r Startingto try sornenew cropsin rice valleys(eg sugarcane). o The farmersarenow'researchminded'.

9

The Dayafarmersgroup is now startingto repaythe effort requiredto start it. None ofthe farmershavedroppedout. Someadditionalfarmerswant to joi4 but the group havenot wantedto expand, The farmersreport that other peoplein the villageusedto makefirn of the farmerswhenthey havetried new cropsor techniques,but the group farmersthink that the last laughwill be on them. We haveyet to evaluatethe impactofthe farmerresearchgroup work on farmersoutside of the groups. For most of the groupsthis would be premature,but a studywill be done within the local areaof the frst group later this yearto determinethe extentof tlie dispersal ofnew ideasandadoptionofnew technologies,

PROBLEM AREAS Expectation The dominantproblemwhentry to work with groups,or individualsfarmersfor that matte,r,is that of expectation.The expectationof what the farmerwill get out of participatingwith the project and expectationofthe relativerole of farmerandproject in problemsolving. This issuemustbe tackledbeforeanyrealparticipationcanbe realised. The history of strongstatecontrol, subsidiesandtop-down approachhasreinforced farmersunderstandingthat: . farmerswill be grveninputsor other incentiveswhenco-operatingwith the government. e all problemsoutsidethe immediatecontrol of an individualwilt be tackledby the governmentandnot by the community. o anytrials are seenasbelongingor beingfor the benefitof the Ministry andrrot for the farmer. Onefarmersaidthat he thoughtof the governmentasbeingmotherandfather. Therehasbeenlittle history of real participatoryschemesinZarn:ibu. Although therehas beena moveby variousdepartmentswithin the governmenttowardshelpinggroups,this is often counterproductiveasthe groupshavebeenforming solelyto attracttangible resources(tools, inputs)andnot througha desireto try andalleviateproblemsttuough joint action.Thetypesofgroupsthat havebeenencouraged arethoseforjoint production, which often fail. Thereareno local NGOs working within the agriculturalsector. Conceptof research It takessometime for the farmersgroupsto really appreciatethe conceptof research.The Dayagroup,for example,expectedthat theywould be ableto quickly plant large areasof introducedcashcropsandtherebymakemoney. Whenthey were givenplantingmaterials of new cropstheywantedto plantthem in a largeareabeforethey knew how thesecrops would perform. The project objectiveswere to experimentwith a rangeof options, plantingsmallareasonlS and comeup with improvedsolutions.

l0 Workingin multidisciplinary mode The project'sremit is developtnent of cashcrops. The idealwhenworkingwith FRGsis to havea multi-disciplinary whereeachmemberof the team teamof researchers/extensionists, cancontributedifFerentareasof expertise.Currentlythereis no suchgroup within Zanabar,so thereare a numberof alternatives: r involveothersections; o coverall areas; r limit the group exclusivelyto cashcrop problems. ,In practicewe areusinga mixtureof the three. We aredoingsomejoint work with other sections(panicularlyforestry),we aredealingwith someproblemsoutsideof our direct remit andwe are alsorejectingsuggestionsfrom the FRGswhich arecompletelyout of our (or suggestotheragencies capabilities whichmigtrtbe ableto helpthem). Working outsidethe remit of the project is not idealas it mayresrlt in duplicationof effort, but on the otherhandthe divisionsbetweensectionsare somewhatartificial and are not a logical dMde but relateto donor activity. Most MALNR staffreceivegeneraltrainingin agriculture.Someg* specialised trainingbut thenmaybe postd to anothersectionon return from training. The structureof the Ministry haschangedmanytimeswithin the last l0 yearsandwill probablychangeagainin the future. Someproblemsaredifficultto classify,anexamplebeing the problemof damageby monkeys. We felt that this problemis onethat relatesto the PlantProtectionDivisior\ th€y saidthat monkeysare underthejurisdiction of the distrist governmentoffice, who in turn saythat it is an iszuefor the Departmentof Environment,and so on. In reality, monkeys are a major constraintin the area,becausethe farmersareunwilling to plant anycrop in the areasawayfrom the village apartfrom cassava,rice andthe existingclovetreesor wild trees. Thereusedto be a governmentprogrammefor shootingmonkeys,but this programmeceasedmanyyearssgo. Our input hasbeento: give trainingon trappingandpoisoningmethods;to persuadethe villagersthat it is their problemand only they cansolveit throughcommunityaction;andto advisethat if thereis largescaleplantingof cropswhich are susceptibleto monkeydamage, then damagesufferedby eachindividualwill be reducedandalsomonkeyhabitatwill be reduced.

OTHER APPROACHNS BEING USED BY ZCCFSP AND MALNR The FRG progranmeis the mainmethodusedby the project for doing long term indepth researchanddwelopmentof new technologiesin a participativemanner,but it is by no meansthe only approachbeingusedby the projector the widerMnistry. Other ZCCFSP havebeentried, manyofwhich were subsequently Different approaches modifiedor rejected.Onemethodwhichhasbeendroppedis theuseof so-called'pilot trials'. This approachwasbasedon commercialdevelopment of singlecrop. Within a givenarea

ll farmerswereidentified(opento anyfarmer),potentialtradersfoundandplantingmaterials were distributed. The mainproblemswith this approachwsre: the difficulty to correctlyidentify a crop whichhadreal scopefor expansion to a commercialscale,the usualproblemsof expectationof the farmers,aswell asthe physicaldifficulty ofproviding sufficientplanting materialof goodqualityat the right time. TheForestrysub-Commission hadmorezuccess with this typeof approachwith developmentof Casuarinaproduction.This speciesfilled a real nichefor buildingpolesand couldalsobe grownon'free land', Plantingmaterialwasgivur freeto farmersfor many years,but is now soldbecausea real dernandhasdevelopedfor this species. The currentproject programmeis betterintegratedandlesstop-down. c Individual crops. Somework continueswith individualcrops,for which thereis considered to be realcommercialpotentialandsomeresearchquestionsremain.'" Examplesarethe agronomicassessment of vanilla;top-working andvariety selectionfor mangoe9. . Networking, Of individualsin differentareasfor agroforestryinvestigations. o Topicalstudies. Whenmote detailedinformationis needed.An exampleis the rec€nt of blackpepperproduction. assessment c Policy andplanning. Promotionof farmingsystemsapproachwithin MALNR, Developmentof CropFact Sheets o Germplasmdevelopment Introductionof new propagationpractices;identificationand selectionof superiorgermplasm.Dorelopmentof capacityfor village plantingmaterial production. o Detailed socio-economicstudies. On rural households,their economicstrategiesand relatedmatters. o Mmketing. Particularlyresearchanddevelopmentof regionaltrading opportunities; facilitationof export(reductionof bureaucracy). Approacheswithinother sectionsof M4LNR In the pastthe emphasis within the Ministryhasbeento do traditionalon-stationandonfarm researchusingresearcherdesigned- researchermanagedtrials. More recentlyit has beenrecognisedthat this methodwas achievinglittle. Subsequently manyofthe different sectionswithin the Ministry are movingtowardsa group approach. Unfortunately,the problemsof poor co-ordinationwithin MALNR hasresultedin different approaches beingusedby differentsections. Participatorygroupscanwork well if the contactpurposeor objectiveis clearandsimple (eg tree plantinggroups)or is left openandflexible(eg FRGs). The problemoccurswhen groupsareformedwith objectivesthat don't respondto a clearlyidentifiableneed. Thus the PlantProtectionSectionencounteredproblemswhenthey formedgroupsfor pest control in certainfood crops. It soonbecameclearthat the mainproblemsfacedby the farmerswere problens of marketinganddecliningsoil fertility andnot pestsper se.

t2 Potentialsolutionsto theseproblemswereto diversifycrops,agroforestryandother practices,ratherthandirectpestcontrolmethods. organicmattermanagement The Extensionsectionhasalsostartedworking with groups. The problemsthey havefaced include: the groupsthey areusingare often formedfor reasonsdiffFerent from the (eg objectives for attractingeconomicassistance, or for socialreasons);the field staffhave little experienceof working with groupsin a flexibleparticipativemode; andthereis a Iack of ready-madetechnologieswhich canbe extendedto farmers.

THE INSTITUTIONAL CONSTRAINTS TO F{IRTHER DE\IELOPMENT OF THE TR'G APPROACS IN ZANZIBAR AND PROSPECTSFOR THE FUTURE. Therearemajor institutional,economicandpolitical changesin the offing andit is very difficult to predictexactlywhat future this FRG approachmay enjoywithin Zandbar. Thereis a continuedproblemof projectisation, despitethe changes to th€ structureof rnu Mnistry designedto reducethis problem. Eachproject importsdifferentideasand approaches.It difficult for the govemmentto controltheseprojectsbecausetheir is minimal. Evenlessresourcesaredirestedto contributionto the runningexpenses sectionswhicharenot donor-supported,Also incentivesgivenwithin projectsin the form of allowancesandtrainingexacerbates the projectisation.Allied to this is the lack of a realisticoverallpolicy on agriculture. On the positiveside: r The FRG approachreducesthe dependency of the group on the institution,through minimisingthe useof inputsandttrough participation, o The MALNR staffbenefitfrom prolongedcontactwith farmersandthe flexibletwo-way approachneededfor working with groups. r Chanetrng institutionsis a slow process,oneof thebestinfluenceson positivechangeis to demonstrateeffectivemethodsfor agriculturaldwelopment. o FRGsare an effectiveway of ensuringthat frrmers' constraintsandoppornrnitiesform the basisfor the work progranrme. r FRGsareprobablyoneofthe cheapest methodsfor dwelopingappropriateagricultural technologies. Therearea numberof methodological issuesfor the future: . Will the learningcurve drop ofl Is it most appropriateto continuewith the same groupsor would it be betterto changeto newgroupsaftera certainperiod. o Whennewideasor technologies aredweloped,do you alterthe approachanddo more networkingor start extensiongroups?

LESSONS The mainlessonswhich canbe drawnfrom the FRG programme: r It is importantnot to importmethodswholesale.Methodsneedto be adaptedto suit the very differentcircumstances existingwithin differentcountries.

l3 Whendeveloping& new approachit is bestto startslowly,so that lessonscanbe learnt built uponbeforetoo manyresourcesor farmersare involved. This is andsuccesses often difficult whenprojectshaveshortlife spansandunrealistictargets. o FarmerResearchGroupsrequirea lot of time, effort andthereforemoneyto setup. It is essentialto selectthe group andlocationvery carefully. r FarmerResearchGroupsare an excellentmethodfor developingrelevant,useable technologies.

t4

SUMMARY OF METIIODOLOGY

Locationof FRG

r Basedon farmingsystemzones. r Village which is representative. e Createshortlistof villagesmeetingcriteria,do rapid surv€y, selectone.

Group selection andformation

Mix of age,gender,family, wealth, 10to l5 members. Representative oftype of peopleusingland. Must bewilling to participateandexperiment. of possible Threeinformantsin villagemakesuggestions given goup members, criteriafor mix of members, basedon makea combinationof the suggestions. Discussconceptsof researchandaimsof researchgroup Mni-PRA. Mixed group of MALNR staff(6-8), makecheck andvisit plots), list, visit eachfarmerindividually(discussion give initial feedback to meetingat endof field work to farmers,write report on mainpoints andon individual farmers. Drawup areasfor research(MALNR statr) Discussresearchproposalswith farmers,modifyplan.

Methodsfor working with groups

Identifyresearch themes Mix of : designed- Farmerimplemented) Formaltrials (Researcher Informal investigations(Farmerdesigpedandmanaged) Discussionon topicalissues. Trainingandstudytours. Mnimise paperwork. Topicalstudieswhenmoreinformationneeded. Minimalinputs,exceptwhenrisk to farmersis high.