The Problems Of Agricultural Development In The Clove Areas Of Pemba

MINIST-IT-YOF AGRICULTURF. LTYASTOCK AND NATURAL RESOURCES

ZANZTBAR CASH CROpS FAflMINc

SYSTEMS fROJf,CT {ZCCFSP)

WORKING PAPER No-: WP 95/27

AUGUST 1995

THE PROBLEMS OF AGRICULTURAL DEYELOPMf,NT IN THA CLOYE AREAS OF PEMBA Principal Findingsof a Study of Ilaya Farners'Rsearch Group

Salim.Abbas Abdulratman Ali Abdulrahman 'rnji Fafm*Khalid Musa Khamis Ali lftrtib Aweina Omar Suleirilan Shehe illarsud Suleim*n MartinWilS{€d) RuBert liloods

ZCCFSP FO SOX 2283 Z-anubar Telepilone/ fax: (054)33121

CONTNNTS Pcge

Preface

iii

INTRODUCTION

I

ovERvInw

1

Appendix A: ResearchMethods

)

[,oc{ttionaf theStudy

5

Ohjectivesof the Stufu

5

ThePrCIcess of'Research anclDevelopment

6

Appendix B: Principal Findings

I

I,anil Use

8

Land Awnership

9

Labour

l3

Cr,spsand their lufanagement

l5

Pestsand Diseases

18

ivfarketlngaf Crops

')n

I"ivestock

22

Forestry

24

HouseholdIncome

26

LU

PREFACE This paperpresentsthe principal findingsof, a nnultidisciplinarysurveyconductedwith Lhe m.embersof Daya Farmers' ResearchGroup (FRG) in westem Pernba (Farming SystemsZone 3). The main aim of this surveywas to providea detaileddescriptionof the farming systemsin Daya,with specialemphasison the constraintsand apportunities availableto farmersin this areaand in partieularthe rnembersof Daya FRG itself. The results of this study are norv being usedto plan a researchstrategyand prograrnmeof researchactivities together with the FRG whioh will addressthese sonstraints and opportunities. To fhe extent that the farming systernin Daya is representativeof others in Pemba's Zone 3, the study will also be used to help fo,rmulatewider potiey recomrnendations for this aad otherplantationareasin the theeastof Pemba. Daya FRG is one of three FRGs in Pembawhich ZCCFSPis now working with in collatrorationwith otherssectionsand projects in the Ministry of Agriculturg Livestock and Natural Resourees(MALNR). Similar work is also being.undertakenwith three FRGson Unguja. ZCCFSPis in the processof developinga rnethodologyfor working with tlresegroupswhich puts farrningsystemsresearchinto practieein a way which will be of maximum benefirtto the farnrers af Zannbar as well as to MALNR and the governmentand peopleof Zanzibaras a wbole. The ultimafe goalsof this approaehare to ensurethe active participationof farmers,tradersand other Zanzibarisin agricultural researchand planning;to help build a Ministry which is fully res,pnsiveto farmers' needsandcan designinterventionsap,propriateto these,and to therebyfoster sustainable agriculturaland economicdevelopmentin the most effective andefficient wayspossible. F{erewe would like to take the opportunityto thank the membersof Daya FR.Gand all the other villagers who helpedus in our researchand madeour staysuch a pleasantone. We are also gratefulto our colleaguesin ZCCFSPPeinbawho helpedwith the practical affangementsfor fieldwork. Finally, other membersof the researchteam would like to extendtheir specialthanksto Salim Abbas for his tireless wor,k in organisingour food and apcomodation SalimAbbas(ZC{FSP} AMulrahmanAli (ZCCFSP) Ab4*hahrrleqj{an(ZCCFSP) FumaKhalid (ZCCFSP) MusaKhamis(FoodCrops) Ali Khatib (ZCCFSP) AweinaOmar(ZCffSP) SuleimanSbehe{ZCCFSP) MasoudSuleimel (Forestry) Martin Walsh(ZCCFSP) RupertWoods(ZCCfSP)

Wete.Pernba Ausust1995

TIIN PROBLEMS OF AGRICULTI]RAL I}EVELOPMtrNT IN THN CLOYE AREAS OF PEMBA Frincipal Findings of a Study of Daya Farmers' ResearchGroup

Iatroduction 1. This reportpresentsthe principal findingsof a studyof Daya Farmsrs'Research Group(FRG),which is locatedin the clove-growingareaof Pemba{Farming Systems Zone3). The studywasconductedin 1995by the staff of the ZanzibarCashCrops FarmingSystemsProject(ZCCFSP),togetherwith colleaguesfrsm other sectionsof Zanzibar'sMinistry of Agriculture,Livestockand NaturalResources(MALNR). Its immediateobjectilr wasto find out more aboutthe farnring systsms{FS) in Daya,in particular to gain a deeperunderstandingof famers' agricultural knowledge and practice than earlier studies had revealed. This understanding based upon partieipatoryresearchd'pa,rtieipatoryobservation'in anthropologists'parlance),was then taken as the starting-point for planning further participatcry research and extensionactivitieswith the FRG. In combinationwith the resultsof other,simitar, FRG studies,it also fed into a more generalanalysisof trendsand opportunitiesin this aad other FS zones,therebyhelping ZCCFSPto plan its wider programmeand makeappropriaterecornmendations to MALNR and other institutionsconcemedwith agriculturaldevelopmentin Zwzibar. 2. The reportis dividedinto threemain sections: an overviewand two appendices. The overview highlightsthe study'smain conclusions,focusingupon the situation facingfarmersin Dayaandelsewherein Pemba'sFS Zone 3" The principalmessage of the study is contained in this overview. The first appendix outlines the methodrrlog"v of the study,and is primarily intendedfor readerswho have a practical interest in applying such methodsthemselves. The secondappendixpresentsmore detail on the stu.dy'sfindings toplc-by{opic, and is in effeot a summary of the preliminary analysis undertaken by the research team shortly after completing fieldwork. The interview notes which formed the basis for this summary can be consultedin ZCCFSP'sfiles. Readerswho are interestedin the practical outcornesof the research,includingthe ongoingprogr{unme of activitieswith fhe FRG,areadvised to refer to ZCCFSP'squarterly and other reportsrelevantto this work, as well as to consultdirectlywith the projectstaffinvolved.

Overview 3. Until recently Zawibar was f,amousfor the productionof cloves. For more than a hunderedyears its economywas heavily dependenton clove monoculture and the incsmesderivedby the state,traders,landownersand agricultr.nallabourersfrom the clove crop. The rapid expansi.,on of clove productionby the rvorld's main consumer, Indonesia,and the consequentoversupplyand drastic fall in world market prioesover the pastdecadehavehad a devastatingirnpact uponZanzibar'sclove industry and the

national€conomyas a whole. Accordingto c{rrrentmarketpredictionsthereis little hopethat priceswill rise to anythinglike their former levels. The f'utureprospectsfor large-scaleclove productionin Zar;ibar are therefbregrim, and while clove t'arrning hasnot suffereda suddendeath,it is sertainlyin a stateofaceeleratingdeoline. 4. The vast rnajority of Zanzibar'sclovesare gro\4rnon Pemba. Clove farms, many of them originally establishedas plantationsand share-plantationsin the nineteenth century,dominatethe hilly landscapeof the wes{emside of the island,spreadingover approximatelytwo-thirds of its toial surfacearea. Farmersin this areahave beenthe mosthard hit by the fell in clove prices. While farmerselsewhereon th€ islandhave also suffered considerablyfrom the fall in cLoveincomesand its wider economic impact$,in generalit hasbeen easierfrr them to adjustto the change. There are a variety of reasonsfor this, the mnst obviousbeing that clovesneverplayedquite the samedominantrole in their livelihoods: relativelyfew farmersown clove farms"and seasonalclove-pickingt'as only one of a rangeof income sourceswhich hotrseholds exploited in addition to their own subsistence-oriented agriculture. ln the west, howevcr, the vast majorify of farmers own cleves and until recsntly relied heavily upon their incomesfrom the elove harvestto buy food and meettheir other domestic needs. 5. The historicaldsminanceof cl,ovesin the local econornyhasmadeit extremely difficult for farrnersto reorient ttrems€lvesto their recent demise. The majonry of clove treesstitl standin the f,arms.It is illegal to fell productivetreeswithout official sanction,and the formal proceduresare slow and cumbersome-In any event many farrnon blame the low priee paid for harvestedcloves upon the government(rather than the world market) and believe that prices will recover if thc governme*t or its policies change. As a rssult there is at presentlittle inclination among farners to €ngagein wholesalefc'llingandreplacement of,clovetrees. Insteada largeproportion of the clove fbrms are neglected,and the land between the trees is reverting to secondaryforest. This provides an ideal habitat for venret monkeys, whose depredationsupon other food and cash crops act as a further disincentivs to the expansionof cultivation. 6. The inherited patterns of land-holding and labour use also place formidable obstaclesin the way of agrioulturaldevelopr*ent.Largeacreage$ sttcloves are owned by peoplewho live outsideof the local community(someof themoutsideof Pemba), and therear€ consequent limitationsupontheir utilisationfor other crops,especially tree crops. Meanwhilethe division of labour and inputs to agricultureremain geared to the econornicsystemwhich they evolved under. Wornen's role in agriculture is confined primarily to subsistencerice cultivation and the processingof this and other food crops. Men's inputsto agricultureare more extensive,but in mostcasescould not be describedas intensive,especiallywhencornparedwith the work nndertakenby farmerselsewhere(includiog womenoutsideof the clove plantati"onarea). This can be seen as a legacy of their historical reliance upon a tree crop which required minimal maintenance,much of which was underlakenby labourersfrom outside of the clove-growing area who were paid from the proceedsof the hawest. The transition from a low-input to a higher-input systemis by no means an easy one,

particulady t{ren the financial resourcesto employ outside labour are no longer so readilyavailable. 7" The typieal respo$seof farmersto this siaratisrumore details of which will be describedbelorv, has beento expandsubsistenceproduction in a piecemealfashion by planting cassavaand bananasin scatteredplots in the clove farms. Many of these plots havebeenopenedup in patcheswherecloveshavedied frsm disease.This is SuddenDeath,a diseasewhich is knovsnto attack poorly managedplontationsmore readily fhan those which are well earedfor, and which appearsto be thriving in proportionto the increasingneglectof clove farrnr. For farmerscassavaandbananas haveobviousadvantages They are familiar food cropswhic.hcanhelp to fill the gap formerly fiIled by food purchases,while any surplusescan be sold. There is, in particular, an increasingmarket for bananas,which are being traded in increasing quantities to Unguja and especiallyZanzibar town^ Otherwiss farmers have been slow to expandproductionof alternativecashcropq partly becauseof their lack of confidence in the markets for thern, as well as the operation of other eonstrail$s alreadyalludedto above8. The cloveareasof Pembathereforepresentthe agriculturaleconomyof Zarulbar with what it is arguablyits most pressingproblem. This problem is compoundedby the environmentalrisks inherentin developingahernativesto cloveswithout eareful planningand appropriateadvice from policy makers. Througlrouthistory fhe westefl side of ttle islaad has been perceivedas Zanzibar's largest area of aaturally fertile land, full of potentialfor agriculturaldevelopment.However,while the potentialis there (asovera centuryof suceessfulclove productionproved),it is a rnistaketo think that the soils in this areaare any more fertile than elsewhere.In fact the subsoilon the hills of Pembais notably poor and deficient in minerals: the deepsoil pocketsin sornepartsof the coral rag are much rnore fertile. The apparentfertility of the clove plantationsdoesnot derivefrom the soil and ia underlyinggeolory, but from the thin la}rerof organicmatterdepositedandmaintainedover thousandsof,ye*rsby the fbrest vegetationwhich grew'uponit- The naturalbalancewas a fragile onq and become even rnoreprecariousw{ en most of Pemba'sindigenousforestswere felled in the nineteenthand earlytwentiethcenturies.By happyaccidentthey wererepl.aced by a tree crop which playeda similar, if less effoctive, biological role in maintaining soil fbrtility. Ironically the periodic and current negleetof ctroveplantationshas firther helpedto rnaintainfhe balance, 9. The implication of this is that without somekind of treecover or other measuresto maintainsoil fertility the greenhills of Pembawill give wayrto a barrenand relatively unproductive landscape,sirnilar to that which has already developed in other deforestedparts of the tropics,ineludingsomeplacesalong the East African coast. Short-termeconomicgainshaveto be measuredagainstsuchrnedium-and long-term consequances. The danger is that these consequenceswill ba ignored or not recogniseduntil it is too late. This has alreadyhappenedin other areasof Pemba: where rool crops havebefl continuonslycultivated for more than two deesdes,soil fefiility and productivity have declined markedly, much to the concernof farmers. Given the recentchangern the economicfortunesof theeclove plantations,there is an

urgentneedto tackle this issue,to avertthe risks of irnplanneddevelopmentand to rvork towardsa sustainableagrieulture. 10. This is a task which requiresthe active collaborationof planners,researchers, extensionistsand farmers. This is evident from our own work together with the farmersof Daya, as weil as the resultsof researchwe have conductedelsewherein this area. Without farmers' participation in this researeh,or the multidisciplinary collaborationof differerlt deperfmentsin MALNR, rve could not havs arrived at the eoaclusionsoutlined in the precedingparagraphs.Likewise, we would not be able to exploresolutionsto the probternswe haveotrtlinedwitlrcnt ongoingparticipaticnand collabsrationof this kind. We hopethat readersof this report will comeaway with a better understendingof the problems of agricultural developnr.entin the clove plantation areasand how tbge might be tackled \l all concerned. The sectiom which f.otlou' outline the methods we have used in reaching our conclusionsand expandat greaterlengthuponthe analysispresentedin theseintroductoryparagraphs.

Appendix A,r ResearehMethods Lacation af theStudy tl. The principalfocus of studywas the villageof Daya, which lies in the middleof the Mtambwe pe.ninusla on the westerncoastof Pcrnba. The smallvillagessttlementis locatedon a high ridgq where the road fi'om Bahanaseforks north to Uwondwe and south to Mkanjuni. Daya is in the heart of clove-growing country end Pemba's FS Zone 3, whieh cov€rs most of the west of the island. Administrativelythe village is pa* of lvftanfiwe Kusini .r&e&icin the Wete Distrig of Pen&a's North Reglon. h 199?it comprisedsome80 householdswith a reportedtotal populationof more th.an360 people,approximatelyhalf of whornwereadults. 12. Researchwas undertakenin participationwith the menrbersof llaya FRG. This gfoup lvffi fonned in 1993 et th€ instigationofZCCFSP, follewing a PRA in Daya and Mkanjuni villageswhich was conducledat the end of 1992by a multi-disciplinaryteam from different sectionsof MALN\ led by ZCCFSP. Daya FRG was the first to be formed on Penrba(where thore ms now three FRGs! and indeedthe first in Zanzibaras a whole. Prior to 1995 ths main work undertakenby Daya FRG in collaborationwith ZCCFSPwas a seriesof on-farmrials with difbrent spiceuops. At the time of the presentstudythe FRG had a total of 12members.only oneof whom wasa wornirl Abjectiwsolthe Stady 13. The 1995 srudyhad a numter of airns. First amongthesewas the need to gain a mor€ detailed understandingof the farmingsy$t€min Daya,andthe role of FRG membersandtheir householdsin this. Althoughthis coveredsomeof the samegroundasthe 1992PRA its odectiveswere rnore specificand its outputs rnore clearly ffined. One objective was to provide the information rrccessaryfor FRG membersand ZCCFSPto plan fu*her researchand developmentactivitiesin the village. and to €nsure that these activities would be directed towards overcomingthe key consraims and eleloiting the prinicpalopportrmitie*which the study hadrevealed. 14. d secondobjectivewas to test the viability of the researchmethoditself as part of the processof startingrvork with FRGs. AlthoughDayaandother FRCrshad beenin existoncefor sometinre, a lol of researchactivities were being carried out in a comparativev&ccum,without a clear idea of their relevanceto farmersand the FS or af what impactsthey might have. At leastin Dayathe earlier PRA had prwided a generaldescriptionof the losal FS, but this kind of informationwas not availab{ein the 'rnini-PRA' and, following its caseof other FRGs. The 1995siudy in Dayawar the first FRG-foc.used satisfactoryoutcom€,becamethe model for subsequenf*udies, including ones undertakenwith new FRCsimmediatelyafter their formation. 15. A third objectivewasto promot€andgive freshirnpetusto multi-disciplinaryplanningand research within MALN& as well as within ZCCFSPitrdfl This had beenan important feature of the original seriesof PRAs,but had nol subsequentlybeenfollowed ttrouglL largely becausethesestudiesdid not Iead directly to axy joint programnresof action in the villages concemed. The researchteam for the Daya PRA thereforeincludeda forester and a food crops officer aswell as nine staff membersfrom differentseetisnsofECCFSP Pemba,andthe Forertry-Departmentin particularhaseollaboratedclosely in subsequent researchanddeveloprnentactivilies. 16. Within ZCCFSPitself the work of the agronorny,post-harvest,and socio-economicssectionshad hitherto beenpmrly integrated,and a lot of researchconductedindependently.This cften meantthat there was little or no socio-economicinput to the plannrngof agronomic research,and that the sosio-economics sectionwould rnerelybe calledin at a lat.erslageto assistin monitoringthe impactsof researchwhich had alreadybeenundertaken. In contrestto this approach,the Daya study revolved aroundthe aetiveparticipaticnof memberssf all sectionsfrorn the start of the researchpro€esstkough to its conclusion. Indeedthe workwas not 'sectionalised'at all but an attemptmadeto draw upon the different skiltrsand interestsof staff as individuals, regardlessof their sectionalaffiliation. Socioeconomicinputshave sincecontinuedto be integated into the work of the Pembateam asa whole, not just in rheFRG smdies"but in all a$pefisofthe tearn'swork. 5

T\p Pracessof Researchand I)ewloprnent 17. The Dayastudyandthe actionsstemmingfiom it cornprisedthe following steps; 18. Productionof checklist. Collectivebrainstormingin the office to producea checklistof research questions. The checklistproducedf,orthe study covereda rangeofoverlapping topics: Iand use and oramership, labour,soil fertility and associatedpractices,crop varidies and plantingmaferial,diseaseand pests,forestry livestock, rnarketingof produce,sourcesof income,householddecisisn-making,and genderissues{runningtluough all of thesecategories).Particularemphasiswas also given to the need to understandpatternsof changeover time' rr,ilharriew to assessing cllrr€nt andfuture opportunitesfor agriculturaldevel,oprnent. 19. Fieldwork by the whole team in Daya. Data collectiog fullpwing the topics outlinpd in tle checklist,was basedupon semi:structuredinterviewsandplot visits with FRG mernbers. The method was for a teamcf two or tkee interviewersto $pendone day with a singlei.nterviewee,the morning visiting plots and the affernooncompli*ing the interview at the intErviewee'shome. This wsrk was done in as informal a styl€ as possiHe, and no problemswere encounteredin securing farmers' participation io the exercise. This also had an important training furraion for project sta$ whose interview skillswere $urpmed in iuteractionwith their colleagues{hencethe useof smalllearnsinstead of single interviervers). In the eveningeachteam of interviewsrspresenteda stuftnary of the day's findings,andthese',rysre discuss€dcollectively*nd improvementssuggested,alongwith importanttopics to follow up the next day. On the final morningin the field a meetingwas held with all the membersof the FRG, together with other local resourcepersons. This provided an oppnrtunity to presentaad discusspreliminary researchresults, as well as to thank the FRG for its cooperaticn and outline process. rubsequent stagesoftlre actionresearch 20. @ Back in the ofiFceRoteswere cornpiledin the form of a report on each farrner iaterviewed. Completion of theseinternal reports generallytook up most of the week fbllowing fieldw-ork. This provedto be the weakestlink in the processbecausein somecasessigrrificant facis and observationsresordedin the interviewswere not reproduccdin the reports, de*pite a prior ernphasison inclrrdingas nruchdetail as possible. Somevery good individual reports,however,were written. and it is hopedthat this aspectof the researchwill improvewith further practice. 71. Cp.llegttyrunskq$. Following the comfllation of individual f*rmer reports, the whole research teammst to undertak€eollectiveanalysisof the datathey had gathered.Eachofthe m4|or topiesin the original checklistwas reviewed in a roundlablediscussionand the principal findings relevant to each were listed. Edit€d minutesof the analysiswere subsequentlyproduced and distributed rvithin the project, and it is thesewhjch form the basisf,or the presentreport, written up somewhatlater by tlre ZCCFSPsooialanthropologrst 2?. Action plannin&and-imp-lementation.An action plan was then produced an the basis of the colle,ctiveanalysisand the views of FRG memb€r$.This step had the following stages: {a) collective pr,oductionby the project team of a possibleplanfor further researchand actions; (b) dircussionof this with the FRG and rnodification and refinesrent of the pmposals, togeth€r with any additional slggestions: (e) implementation(curer,rtlyongoingin Daya) fther MALNR dcpart$ents have also beeninvolved in this processin cases*rhere prr:posedactionsfall within their remit. The Daya action plan andits implementationaredescribedin detail in other ZCCFSPreports,andwill not be considered further here- thoughit shouldnot be forgottenthat this was andis one ofthe key stagesin the researeh proc€ssandoneof ihe most importantourputsof the study. 23. Input to wider policy aad planning. The idorrnfrtioa gatheredduring the Daya study was also subcequentlyabsorbedints the wider aspectsof ZCCFSP's work. This has included the refined description of Pemba's FS zones {a process describedin ZCCFSP's quarterly reports and other pub'licationsl and more generally, the fir*her developmentof the researeh,policy and planning framwork of the pmject as a rvhole. The findingsof the Daya study and further participatoryresearch baseduponit - togetherwith the findingsfrom parallelstudieson other FRGs. havethereforeplayedan

inryortant and coutimringrole in informing the direction of the projeet and its interactionsrdth other sectionsof MALNR, including recommendationsand proposalsfor the future developmentof the ministryandthe role of farmerparticipatoryresearchandextensionwithin it.

Appendix B: Principal Findings 24. Th€ following is an uneditedsummaryof the main poiats raisedduring collectiveanalysisof the Daya study. More detailedinformationon particularpoints and different asp€ctsof farmers' practice canbe found in the unpublishedreportson indMdual farmerswhich were compiledby the researchteam andare avaitablefur consultationin ZCCFSP'sofrices. trn+d Use 25. Therearethreemaintypesof farmlandin Daya Theseare: A. llorne Gardens,aroundrhe housesin the village,which lies on a high ridge. B. ClgW llantati-o-ns,on the hill slopes crOp.

the aree,where cloves are the dominant

C. RieeValleys,betweenthe hills, and including(a) rice oultivationin the valley bottoms,and (b) cultivationofbananasandother cropsat the edgeofthe valleys.

TheHoreGerdetrs 26. Crops cuhivated in the multi-slorey home gardensindude the following: hananas,breadfruiq ramtrutan,cleve$,coconuts,cinnamon,black pepper,pigeonpeas,p,ineapples, cassal4 andcocoyams. 27. Cassavais most prevalentin young homegardens. Whm bananasand other higher storey arops havematured,lessspacerernainsfor cassavqwhich cannottoleratetoo muchshade. ?8. Smallarnountsof blaekpepperare plantedaroundthe houseson treeswhich are alteadygrowing. Someare planteda :little distanceaway fiom the houseg and if thereis no spaceat homefarners will plant elsewhere.More black pepperhasbeenplantedat Daya in the yearssinceZCCFSPbeganwork there. 29. Pigoonpeasgrow well in the Daya area,but becauseof the depredationsof monkeysand birds fhey canonly be plantednearthe houses. ThoClove-Plantationq 30: Despitethe drasticfall in the value of clovesasa cashcrop, no onp wasotsarvedto haver;r,rtdown his or her treescompletely. Onefarmer saidthat he hadgot soilleoneto arf back fhe lower branchesin order to prcvide more light to the lower storey. It is possiblethat there is sometheft of branchesat night. Usua$ only deadtreesaf;esold firr firewood and eharcoalproduction. lt is difficult to cut trees whenthey are still greenandtheywould haveto be concealedfrom the authorities-The possessionof a certificate(clletil permittirrycutting of deadtreesis not very importantin Daya. 3l Somefannersstill believein the viability of clovas. It is r seasonalcrop with a reliablemarket (if not a high price) and thereforethey still want to keep their cficves. Many farmersare waiting for the October 1995electionsto $€eif the econorniesof elove production irnproves. They think that the governmentis stealingfram them{by ofiering low pricosand downgradingtheir cloves)andbelievethat the price of cloveswill rise if the governmentchanges.Farmerssaythat clovesfetch the equivalentof Tshs3,000per kg in Kerrya,wheretlrey are sometimessmuggkdto. Nonetheless,no new plantingwas observed,althoughsomenewly plailted trees were seenin 1994. To this e)dent,farmersseemto b$ hedgrngtheir bets. 32. Sorneclove trees are eged and it is very difficult to climb them and pick the crop. Othersare (the proportion of poorly-rnanaged overgrow{lwith bus} throughthe neglectof management plantations has risen as the price of cloves has fallen). SuddenDeath, a diseasewhose incidencereflects poor management, is a problem. Farmersgenerallyreplacedeadcloveswith other crops. The rangeof crops

I

found in the eloveplantationsis similarto that found in the hornegardensand includescassava,bananas, iackfruiq cinnamon,black pepper,turmericandcardarnom. 33, Altheugh nrany fhrnrersreplacedead cloves prirnarily with bananasand cassava,there are also eacouragingsigns of widu diversifieation,especiallya{nong membersof Daya FRG. Oae farmer replacesdeadtrees every year with orangesand bananas,or sometimescinnarnonand black pepper. Another replacesthem w-hhother fruit trees, cinnamonand black pepp€r. A fhird farmer replaced cloveswith cassavaon one plot three yearsago,but now grows gingerandtunneric there. On another plot, four yearsago. he replacedclsveswith cassava,bananas"rambutanandjackfruit, and also plalrted someclove se,edlings.Other replacerner*cropsmentisnedinclude$ugarcane.Onefarmerwas planning to reducethe clsve bran€hesin an overgrorvnareaso that he could plal* black pepper. He did not plant cassavain one arqaafflicted by SuddenDeath becausehe thought that the trandr*'asno longer fertile: insteadhe plantedforestrytree species. 34. Someof the areesof governmenflandin Daya do not haveany cl,oves,but betel andbanrboocanbe found on them. One hrnrer plantedbananason governmentland but replacedthem with cassavaafter the past outbreakofBlack Sigatoka. 35. To some€xtentcultivationhasshiftedcloserto people'shomes. Theclove plantationsusedto be dominatedby pure standsof cloves,but now diversificationis underwayto increaseyieldsand income, reducelabour inputs (for examplein weeding),and ircure againstthe risk of one crop failing. These strategieswere rnadeexplicit by rnorethanonefarmerwe interviewed. The Rice Vallels 36. Farmerscharacterised agriculturein the past asbaaedprimarily upon clovesand rice. Rice wasnot cultivatedvery seriously,and young peopledid not take muchpart in agrieulture. There was a lot of importedrice in the shopsandrery few peoplecould dependon their own rice cultivation. This began to changcwith the faminei* 197l-72 when a lot of peapleexpandedfood produetion. A lot of food also cann liom Mombara at this time on the blacl nrark*. Peoplehad enoughmoneyto buy food, t.herewat just none to buy. The causeof this farninewaa the abruBtcancetrlation of food imports in favourof constructionmaterialsto buildthe,rovemnrent flatson Pemba. 3?. Now other fosd cropsare planted,opl"r"t, nearthe housesandnear the valleys. When farmers are guardingthe rice fields cessavais often plantednearby. Flswever,manycrops cannotbe plant€dat tlle edge of the valle.ysbeeuse of the risk sf theft and the ravagesof vervet monkeys. This has prcmptedafurther concenaationof effrrt on the homegardensaroundthe villagehornes. 38. Ssmefarmershaveplantedcinnarnonfar away from the village. Monkeysdo not eat them and if a lot of treesare plantedby diffrrenf farmersthen thei' can sharethe lossesto humanthieves. Farmers usuallyonly have l0-15 cinnarrontreeseach. Somehadplantedcinnamonat hornebeforeworking with ZCCFSF,but beganto plant tr€es somedistanceawayafter the pmject c+nne.They still do no have a market lor cinnamon,or do nol knolv aboutit. 39. There are m&nyrice valleyn,large and snrall,aroundDaya. h4avuikais the largest. In the valley bpf,tomsit is unrrsualbranyhing to be plantedafler the mainrice harvest, Rirrgarice can resprout,but only a smallsecondcrop is obtained. Onefarmerhad plantedsugarcane, but fhe crop was destroyedby monkeys. Land *wnership 40. Outsideof the villagethereis generallyno separatetree ownership(i.e. separateftom ownershipof th€ lald on which they grow), asthereis elsewhre on Pemba.Thehomegardensprcvide the ercception to thig and trees may be owned by former occupentsof the land, who planted thern prior to the existenceof a houseple,

41. Farrnerssaythat all ofthe non-goveramnTt landis ful$ occupiedandthat rhereis no land for young peopleto cultivate{i.e. nonewhich they canacquirewilhout inheriting,borrowing,renting or buyiag: in resl termsthereis a lot of underutilisedland in Daya). 42. Land holdngs are quitefragmented,and onefarmerwe interviewd counted24 sepnateplots: the normis l0 or less" Inberitancs 43. Distinctionscan be madebetweenfunily (i.e. jointly owned) and private (i.e. individually owned) land, and betwe€nfrmily land which has beendivided fbr use by individua,lfamily membersand that which hasnot. Wives can own separatefarms, and theseare often outsideof Daya (includingthe east coast),especiallywhenthe wife's family hailsfrom elsewhere. 44. One ftrnrer useshis mother's clove pftot: he is her only child so his useis nominallythe sameas ownership(the expectationbeing that he will inherit this farm). He has plantedcardamorn,cinnamon and turrneric there, and plansto plant bananasand sorneblack pepperin tlre more open areasof the farm. 45. A youngunrnarriedrnanliving with hisparentsuseshis father's land: he doesnot haveany land of his own. He also has a plot borowed from a nr{ghbour ard relative. This is a probtrembecausebe cannottake anydecisionsaboutthe family trandand cannotplant 'permanent'oropson any ofthe laild that he hasaceess to. 46. The wife of one farmerhasfbur plots: a clove plot inheritedfrom her par€ntsandtwo ca*savaand bananaplots nearhome,plusa rice plot at Mawika. 41. Land can also be rcquired by gift. A clove fann at Sengenyar+.asgiven to a farrnerby a friend. Whenthe latter died his sunrivingchildrendid not eleimit es pert of their inheritancebut confirmedthe transferwhich had beenrnadeby tlreir fgher. In reecgnitionof this the neu*owrer sharessomeof the incamefrom the farm with thorn. It is consideredto be his privateland: in additionto clovesit alsohas cinnamon,cassavaandindigenoustree species. 48. Before the 1964Revclutionland inheritanceoperatedin the normalwey in the rice valleys. After the Revolutionthe larger valle'ys(of i0 or more acres)like Mawika were declaredto be government landa*d colleetivelyowned. This was enactedby forea Eachpersonin the village was giveu one ,!sre to crrltivate: a large meetingwas held c. 1968-69in which it was announcedthat thosewho rvantedit could get on€ acre. Some of the original owners lost out and in cons€queRce rnoved elsewhere. Everyone paid fixed surns to purclrase subsidisedseed, fe.rtiliser,and tractor servicesfrorn the gov€rnment. Howerer, it is allegedthat the produceof thesefields would largely be consumedby the village ehairman.In areaslike Daya this systemwas nst maintainedfor long and the valley plots have long sincerevertedto privateownership. Shareplanting 49:' Shareplanting,by which a psrson who clearsand plants land is zubsequentlygrv'ena half of ttrat land, was once widely practised,and was particrdarlyimportant in expardirg tlre clove plantatiorn. It was usually done when cleariagthick farest, but was *lso possibleon 1*rid *4rich had already been planted,and could be undertakenby &ther and son(i,e. ar a mea$sfcrrthe sonto acquirea portion of his father's land). The planteralsooften too& the first harvestfrom the whole of the plot which he had planted. 50. Thereis no sh*replantingsystemnow in the overgrorm clovebush. lt was formerly practisedjust for the plantrngof cloves at a tirne when cloveswere v€ry valu*ble. Farrnersopined that there is no prospectofreviving the 50:5Oshareplanting systemwith respectto othertreescrop$.

t0

GovernrnentLand 5l. Some farmersplant in the government-ownedforest, fbr exampleat Kinyasini near Sengenya. Someareasof governrrentland (rnost of it nationalisedafter the Revolution) have not beendivided. Nonethelesssomefarmershaveevenplantedclovesand occasionallyother tree cropson this land. They arejust trying their luck. One farrnersold governmentland that he had cleared"albeit for a low price and with no documentation.Otherwisefrrmers tend to think of soverm€ntland which thev are usins, without authorisationasborr,owedland. 52. There are also casesof hornegardensbeing developedca governmentland. Two f,arrners,for haveplantedbaraoaqcassayaanda little blackpepperon nationalisedland clcseto the village. erxample,

Erotsg 53. It is relativd earyto borrow land in DxyE *both the plantationand rice valley areas,and lending andborrowingarewidespreadpraetices,engagedin by mostfarmers. 54. Thereare a numberof reasonstbr borrowing(andlending)land: (a) to obtain accessto land which the borr.owerdoesnot possess,especiallyvalley land f,or growingricg an essentialpart oflocal far*ing practice; (b) to gain accessto land which is more accessible to the bonower, for example*earer to homethanthe fieldswhich he or sherrny haveinhuited; {c) to obtainlandwhicir is moreopenand lessshadedthan the borrswerso*1, thus permitting him or her to plaatsass:Iva,essentialfor householdsubsistence; (d) to get mcre fertile land. for examplewhere fertility on the borrowers own fields has declined; (e) to get landfree lrorn pe$ts,wherethe b'orrowersown landis not; (f to obtain land near which is closeto that being warked by otherg in arder to er{oy the advant*gesof company,labourgooling, andincreasedseeurityagainsttheft andverrnin; (g) to gain accessto land in caseswherethe borrower o\ilnsnone at all, especiallyin tlre case of mmarried menwho havenot beenallscatedfamilyplots. 55. Onc farmer,lbr example,was formedy cultirratingalonein a smallriee valley. However.he had a lot of problemswith birds and monkeys,and his children were afraid to go there unaccompaniodto guardthe field. He thereforeabandonedthis solitary plot and borrowed a rice field in Mavuika where many people were cultivating. Another farmer had danted black pepper, cinnamoaand turmeric on borrowed land. This was very fertile becauseit had not been planted for many years and therefore coutaineda lot of *rade anddecompnsingvegetation. 56. No payrnentsare madefor borrowing, apart from smsll gifts of the harvestedcrop to the land owner(s). One farnrerh*d boncwed half an acrecf,rice land at Mavuika from an owner who lives at Kjsiwani"andhad doneso for tha pastfive years. He borrowsfree of chargq but gives 3-5 pi*hi ot-rice to the owner wheneverhe gets a good yield. Another farmer had bo,rrowedrice land for the past 7 years. The owners,however,rnaytake this landbackwheneverthey wi*h. 57. Peoph tend to stay on the samepieceof borror.vedland, often bacauseof its proximity to their homeandbe,cause theydo not own better-locatedplots. 58. The reasonsllorlendinslandincludethefollowins:

11

(a) the owner is an absenteeand has left sorneoneelseto ruuragethe plot, usuallya relative {who maythereforeuseit hfunselfor lendit to someoneelse); {b) the plot is dista:rtfiom the homeof the or.vner(this is the obverseof (b) above,andresults in a rnorerationaldistribtionof landutilisation); {c) the owneris unabletrr clearovergrownland: a borrower will clearthe burh. which is good of eloves. for management 59. Somepeoplewrite an agreernentwhenthey lend/ barrow land andagreeon the termsof its use. 'Perrnanent'treecropscannotgenerallybe planted removed or by the borrow'er. It is, however, possiblefor a borrower to sell cropswhich heor shehasplantedbackto the owner of the land.thorrgh the latter may refuseto buy them. 60. One farmermanagesa half acrgundividedfamily cloveplot on behalfof his elderbrother who lives irl Dar es Salaam. He rentsout the harvestfrom c.3O0chve trees. He has plantedjackfruit (though only one survired frorn tea) but does not wanl"to plant more treas becausehe might lose the whole investmsrtif *omethinghappenedto the olderbrother{if itbelonged to ore personand that poson sold or took the land back tle would probablyget somecompensationlor the treesthat fuehadplanted). He has also planted bananasand cassava. He plaas to plant more bananasbecausea farrner is also cultivating a neigtSouring plot and the risk of theft has therefore decreased. He may aiso plant cardarnomandblack pepp€r. He will not plant cinnamonbecauseit is rnorevaluableand easyto steal andis consider€d to be a'permment' crop. 61. Anotherfarnrerhasborroweda quarteracreclcve plot whichwas originallyplantedundera 50:50 shareplantingerrangement(prezurnablythe reasonfor its smallsize). Theowner died aud left it to two sonsr*ho live ar Kikunguni ($humbaVianrboni)with their msther. They cometo Daya at harvesttirne antl also at other times to cleanand maintaintheir trees. The borrower has plantedcassav&and will plant a few bananaqespeciallyat the baseof the slope. He would plaat pineapphsif there w€r€ no monkeysto spoil thern. 62, Clovsscannotbe plantedon borrowedland,but betrcngto the owner. Sornebonowers, however, prcvert the owBer'saew clovesfrom growing well becausethey will spoil their cassa\a. This also happetsin other areas(outsideofDaya), whereborrow€rshavebeenknown to pull up clove seedlingg andothersto put dieselfuel on the roots. Thereis thereforesomepotentialfor confliet betweenowners andborrowers. 63. Someownersdo not allow borrowersto f'armwith ridges(rnatuta) and fuffows becausethey will cut the clove roots and can also cause soil erosion (if poorly aligned). The use of small rnounds {m*knngo\ is thercforemoreusual.exeeptat the bottom of slopes. Thesestrictures,howeveqact asa eonstrai*ton the develoBment of erosioncontrsl methodsusingcontour ridees, Renting 64. Thereis not a lot of renting of clove trees now becausethere is no value in doing it. Treeswere usuallyonly rentedaftsr the crop tad beenseen. A similarpracticeobtainsfor mangoesand cocgnuts. Thereare two typesof rentirryfor coconuts Kdi ya tm+'alqr" rent for a whole year at a fixed price, fior exampleTshs50 for everytree, andkdi ya jangua, just for a singleseason.

T2

Mort&asing 65. Thereis reportedlyno mortgaging(relwni) of treesor land at present,thoughthis usedto occur in the past drrringthe clove era. The deb4was reducedeveryyearand interestpaid. In the past gold and watcheswere alsopawned. After the Revolutionin 1964f'armswere returnedto peoplein debt- It is saidthat a licenceis neededto mortgageFoperty. Purchaseand Sale 66. The saleof farmsis saidts havedrqppedin frequencybecausecloveshavemuchlessvaluethan they usedto have. One ftrmer baugtrta clovefarm solelyin order to ob'taincassavaplots nearto his home. 67. The vaftreof plantatioulandis usuatlJcalsulatedon the nrrmberof ireesin it. Treesca-nnotbe sold and owned separatelyfiom land- They can, however,be sold lbr cutting, Forfirewood and charcoal production(thoughin the caseof clovesthis is usuallydcnewhenthe treesaredead). t,ohour TheDivisionof Labour 68. Women's work focusesespeciallyupon fice weedingand harvesting,food processing(including ,pounding,grinding andcooking),andlooking after childrenat home. Many wornenhavesmallchildren, andthereforework most at homeandin the homegardens.It is claimedthat in the pasl womenhardly performedany agriculturallabour at all, but that householdsrelied primarily on cloves and the income ob'tainedfrom them (to purchasefoodstuffsand the labcur inputs of mainlandernand others). Older women(and men) are saidto do mqre agriculturalworh than their juniors. One fanner opinedthat the problemwith family labouris that rdves do relativelylittl€ ftndare lazy. 69. One farmerrswife with four valley plots does all tile rice culiivation herself with her children. Childrenalsohelp her in fetchingwatsr, cuttingfirewood. cooking and weedingca$savaandbananasa$ well as rice. The advantcgeof this division of labour is said to be that the woman can r€st and familiarisegirls with their domesticdutiesreadyfor when they get rnarried. Ano'therrvife performsatl of the rice tasks herse[ and also helpsto dig holes ficr bananaplantingand weedscassava. Women frequently participatein the cultivation of bananasand cassavathe two most important subsistence cropsapartfrom ric*. 70. In rice cultivationthe typical pettem is fbr the husbandto clear the land at the start of the season, fur husbardandwife to cultivatetogetheqandfor the wife to weedandharvestthe rice, helpedby their childrenwho nrayalso underfakemost of the crop grarding prior to harvest. Weedingusually begins two rnonthsafter plantingandcrop protectionbecomesinlensivefrom the fifth rnonthonwards,earlierif there are a lot of monkeysaround. February-Marchis the busiestperiod for riee sowing,April-May for weedingand July.{ugust for harvesting. 71. One couplehavesevenchildrenbut someare too youngto work in the fieldsand the older onesare rnerrist women. The hrusband andwife therefore$hareout agriculturaltasks, Again the wife doesmost of the rir;ework and sornecassavacultivation. Sametinreshe hireslabourto work on their rice plots at Mavuika,but usuallyhe cannotafford to do so. 72, ln anotherhouseholdthe wife is employedasa schoolteacher"Shehelpsher husbandin the fields at weekendsand is otherwiseresponsiblefur all of their rice cultivation. ?3. One unmarriedmsn works aloneandthis yearplansto plantmtemauplandrice. In the pasthe has cultivatedvalley rice but found it to be rnuchharderwork. Another unmarriedmale farmer lives with his mother and three sisters. His mother and sis€rs do all the rice operationsand sometimescassava mrdbananaplantingand weeding. They alsodo the clove-pickrng He is involved in everythingexcept

l3

rice weeding,bird-scaringand harrresting.He cannotafford to pay for a tr&ctor and so has neverused one. 74. Labouravailabilityis nory a big problemin Daya. A lot ot'young mm haveleff rhe village, someof them to do businessof differentkinds and accurnulatemoneywhich they can then comeback with to build houses.Youths startedto travel out of Dayaespeciallyafter*rc 1964Revolution. 75. A lst of fannshaveweedsandarp not aswell managsdasthey wErein the past. Peopledo not likp to spendmofieyon thei.rcassavaplots. Labour Pooling 76. {liina,labour pooling and exchange,is practisedbut is not the norm. It is maiilly confinedto cooperationbetl"eenwomsnin weedingand hawestingrice. Two or three\r'eruenmay take dar.bturns to help one anotherin specificoperations,for examplerice broadcastingweedingandharvesting. This is only done in sowing and weedingriee amcngclose relativescr friends. Peoplewithout rice plots might be invited alongduringharvesting,firenor wo$ren,and obtaina portion of thahnrvestin return. Hired Labour ?7. Mainlanders,especiallyNyamweei.weretraditionallyhired for perfiormingheary agriculturaltaskq includingland preparation. This is still the casebut there is only one Nyarnweziman now availablero perfiormmost of this work. He is thereforevery busy and can only take larger contracts. A woman famneruseshim for othertasksincludingthe weedingof cassavaandbananas.Shealsc sometimeshires labour for sowing rice. One or two local men do occasionaljobs but they are not consideredvery reliable. 78. Locals are saidto dislike working for athersbeceuseof its historicalassociationswith slaveryand the role of the Nyamwezi(this is not true of all Pemba). One farmer saidthet it is not difificult to get casuallabourersif you hnvEthe moneyto p&y.Another confirmedthat hired labour is availahlebut that is it very experlsive,indicatingthat aost andthe lack of incorne(espociallysincethe declineof cloves)is oneof the reasonsfor the declinein labourhirine. 79. Castnl labonris sometinresemployedin cassavacultivation. It costsTshs5-6 to heveone planting mound,koongo la muhego,prepared,up to Tshs600 for 100 mounds. Sometimeslabourersare also hired lo harvestaassava. 8O. Hired labour is also used especiallyon clove plots. To cut overgrownbushcosts Tshs 300 per singleclove tree. Clove-pickersusually approachthe tree ownersthernselvesand pick for Tshs60 per pisAi. Pickemusedto comefrsrn Kiuyu (Minungwioi) ofl easternPembato pick clovesbut they do not come eny rnore. One unmarriedl,ocalmanpicks cloves ss olr€ of his main sourc€sof income,but in generalthis work is no longerlucrativeenoughto attractmanypickers,local or o{herwise. 81. One farmerreportedthat hejusl usesfamily labourin agriculturebecausehe cannotafford to hire. As a res+rlthe hasreducedthe numberof his plots to five. Anotherfarmerhires peopleto harvestblack peppff, at Tshs80-l00perpishi. A third farmer$t&tedthat he hir*s labourersto helpin rice cultivation if he is sick at the tifte. Nfechanisation 82. In somevalleysa tractor e€xnatbe usedand so farmerrcultivateearlyby hand. Otherwisathereare two govemmenttractors at Wete to servicethe whole district in which Daya lies. The tractor driver usuallyclaimsan sxtra 'fes' for his work. During the last sea$onfarmersgwp a Tshs5,500sontribution to buy fuel for the tractor to plough each acre at Maluika, and the sameagainfcr harrowing (though rnanydo this themselvesby haud). Theywere then askedto pay anotherTshs3,000each. The fiactor driver prornisedto comebut nwer did. In hifavuikathere was thereforea delayin planting because farmerswerewaiting for the tractor" andin the.endthe tractor only cultivateda little on somefarms.

T4

83. One problem is that everyonewants the tractor at the sametirne. Another problem lies in the economicsof this subsidisedservice.The official govemmentprice for ploughingor harrowing is Tshs 2,000per acre. Ckreaeretakes20 litres of fuel, currentlycostingTshs5,400. Eventhe Tshs 5,500 daimed by the tractor driver vvouldnot be sufficientto cover costs,including wages,oil and repairs. Nonetlelessfarmerscontinueto believethat the governmentshouldpay for tractor fue1and sub*idise the service(insteadit appearsthat farmersare supplernenting the salaryof the tractor driver). 84. There are scmeprivate tractors avsilablein the district bw they *re usuallyemployedin carrying heavy loads like stones and sand. They carurot €orrpete in price with th€ governrnentsubsidised tractors. Somepeoplecould afford to pay Tshs7-8,000psr acer. Howwer, thetopograptryof the area is not idealfor tractor$,andthe iand is too hard at tirnes,andonly soft ercugh during the rains. (lrops and tfutr MunnEyement PlantingMaterial 85. Obtainingplantingrnaterialis nct a problemin the caseof rnost commoncrops. Farmersuzually either obtain planting material frorn their own fields or exchangewith ot.herlocal farmer to get the varietiesof rice, bananasand cassavawhich they desire. Becauseof their associatiouwith ZCCFSP, FRG rnenrbersalso experiencefittle difficulty in obtainingthe seadsand seedlingsof less commonly grown crops. 86. Rice seedis not normallydif;Ecultto find unlessthe crcp hasfriled. Mtema *pland ricg however,is difficult to obtain, largdy becauseit is not aswidely grown asthe rain-fedvarieties. 87. Farmerssaytlrat they often haveto buy bananasuclcersif they do not possesstheir own. Puhrs* suclcerscufrently cost Tshs2A, andmkun wa temboTshs50 or more. Thereis a shortageof banane suckersduring planting time and farrnerssametimeshaveto travel sornedistanceto bu1'for them at higherthan normalprices. 88. One farmernotedthat gl{rg€randturmeric canbe difificultto obtain. More probloms,however,are experiencedwith cropshke coconutswhich are only sporadicaltyplantedin the Dayaarea.

Menaes$s[t*hac1itrs 89. Onef,armerhad l0-I5 yearoJdblack pepperplated by nbu andmjengatm suppcrttrees. Another lbrmer, however, had plantedhis black p€pperon jackfruit, cloves and coffee, as a result of which harvestingthe pepperwasa problem,becausethe plantshadgrown high op into their supports. 9O. furother farmerlad planted55-60pine*pples2.5 metresaparton op€nlandwith sandysoil which is not very fe*ile. 91. Bananasare rnanagedby reducingthe numbersof shootsper stool. Dead leavesare ctit off and placedat the basealong with stemssfter cutti,ngoffbunches. One farmer $tatedthat it is best to plant bananasin the wli rains becausethey take 1.5 yea-rs10 mature and will tlrEreforeflower during the secondffiasikfi seasonand get enoughrain andmaturewell Somevarietiestake 12 monthsto mature. 92. CEssavawas saidto be bestplantedbebre the long rainsbecauseit will rat if there is too much rain before it is Established.Mstuta are used in rryretareasand makoongoelsewhere. Cassavais usually plantedduing the rainy s€asonsand weededtwice. Dependingon variety it takes 8-12 months to rnature. h is harves€dwith a hoe or by pulling up the stemif the soil is moist. 93. Rice seedis usuallvbrcadsast,a*d it is saidtc be best to do so beforethe rains. One farmer has rece$tly establisheda rursery plot and will transplantbecauseit usesless seed and the weeds are correspondinglyweaker. He will be the first to try planting in rows More water is neededto transplant. One farmer had seenbundingstructuresat Mangwenabut was worried that cther farmers

15

would thinlc he was stupidif he tried to build them himself. Land preparationis problematicbecauseof the late arrival of tractorsin Mavuika. Rice weedingis performedonceor rwice, mainlyby wontenand children{see*bove}. 94. Mterna upland rice is usually planted for a single seasonafter felled trees have be,enburned: thereaftercassavais usuallyplanted. Crop Varieties 95. The following varietiesof,bananasare planted: m*tna vs terxba,befundi {- kir,ukwi?),pukusa, bakobako{which did nst grow well on a clove plal}, mhuifu (susceptihleto Black Sigatnka),kigvruwe, ki.jakazi,karolmi, rnzigo,msinTvri, ad macharn mengi. Mzuzu. mkonowa tembo andbakobokaarc known asdiseaseresistant;wlulem*anowa tembo,mil)ifu andmzuzudo well aroundhouses. 96. The frllowing varietiesof cas-sava arc planted: mjawa (or jtma = leibega?),mal*is dwhichhss the (: chane?),mc$aniwellsrw, kibirtti sn+vasi,kibiriti bestyield), msgerczsrmesanja,mqjani memb*wtb
16

lo 7 pishi per tree on averag€.No fedilisersare used,md Justuprootedweedsare spreadon the plot. On anotherplot, of flat land with clay soil overtyingsandyloam, the sarnefrrmer plafttedcassavaand bananas(alsorambutenandjackfruit) four yearsago after SuddenDeath had affectedthe cloves. He hadgood harvestsin the firs yearbut yieldshavedeclinedprogressivelysince. 106. Farrnersrecognisethat overgrownclcve areasarefertile: this is a kind of unintentionalfallow. 107. In mauyc&$escassavaean only be plantedfor about 5 yearsafter clcaringland, One fanner had plantedc&ssavaon top of a hill until it would not grow again. Now there is nothing grewing on this land, iust grass,and it has beome very infertile. fupther casewas cited, however, of the l4-year continuouscultivationof cassavaandbananas:fertility haddeclinedon this plot but was still good.

gerl-Elassq 108. In thepast;hegovernrnent usedto promateerosioncontrslalotmore andfine peopleforplantrng on st€€pslopes. Onefarmerthoughtthat therewas lesserosicnth*n in the pastbecausethe slove areas are changiagto bush Anotherfarmerthoughtthat erosionis not a big problernat Daya becausethere ate fiany trees and grasses.Ilowever, he recognisedthat it deiesbecomeinfuile on the tops of hills becau'*ethe rain takeso$the topsoil, especiallyclay soil with gravel, He therefareleaves grassto grow-at the tops ofhills during the rains. Somecrops, for example,bananas,are thought to help in grosionconirol. Chemiqd.Fe{tilispa lO9. Wh€n farmerscan affirrd them, chernicalfertilisersare generallyonly usedon rice, althoughone fannerrecalledusingureaand TSPonboth rice and camava.until the priceswent up. Fertiliserusewas widespreaduntil 4-5 yearsago andsould resultin doublethe eurr.entyieldsof rice. 110. One farmer, who had used fertiliser extensivelyin the past (obtained fsr free when he wrs a MALNRemployee),tlrought that fertiliserwasespeciallyne€dedonpoor s
t7

in the hornegardensin order to savelabourratherthan better usespace. i 15. Intercropplngis prervalent Cassavais planted betweenbananasurtil there is too nnrch shade,after which pineapplesmay be planted,with the bananasbenefittingfrom pineappleweeding. 116. One farmer intercrcpped 5 varieties of cassavabecauseof a shortageof planting material. Anotherintercroppedfor the following statedreasons:(a) to get mor€yield frorn the samearea {.b)to saveon labouroperations"and (c) to increasesoil fertility. Anc'therfarmer saidthat intercroppingwas pfactisedto reducethe ri* ofcrop failure. 117. The firllowing tree speciesare saidto grow well with crops: youngcloves,jackfruit, str4tle artd mlondaa, Mchenru msils andmburs are found on poor soils and are thougll to havea bad effea. A shrubcalledtnwivuwiwt,issaidto makefhe soil fertile by addingrnanyleaves. Cassavaprovidesshade fur tree cropswhenthey are fir* planted. Pestsanl Dissdnss Bananas 118. Overallthere are no big problemswith bananas.Whole fields were uprootedin the past because of Black Sigatokabut they have been planted again. Farmerssaid that Black Sigatokaused to be prevalentbEcausethey did not haveany control measures.Howwer, now there is better manageftent and farmerstake cff th€ affectedleavesand sterns. Scmefarmerscut and burn the leavesin the dry $eason. Not all farmers, hcwever, burn the atrected leaves. Sigatoka especiallyaffeets mhvi**, kigunwe andmkonaws tembo. farmers do not reducethe stemsas recommendedby PPD, although the spacingaf bananasis not bad. The exceptionis hijakazi, which puts out so maay suckersthat farmersdo not bother to try to reducethem, hencethey grow in thick clumps, especiallynear the h.ome-sieads. 119. Bananaweevil (funza) especially*ffects puhusa andmkarp wa ternb
18

{a) tango (alss calledscufi), Mannikins,Lanclnru spp.,whi.charevery commoq numberingup to 100in a flock. {b) l*vche (plural makweche) arld.mr.wwr. the tslack-winged Red Bishop, Euptectus hordeaceas.Kweeheis the nameof the malein breedingphimage;femalesand out of season malesarecalledmnqw. l2a. Tbe problemof birds lastsfor 2-3 monthsbeforeharvestingand the seedsbeeometoo hard for the birds to eat. Womengo to the vall,eysbefcredawn, around 5 am, and staythere until 7 pm. They are helpedby neighboursand alsotheir childrenwhen tlrey conretorn school. To scarebirds sornepeople string tapeliom sld musiccassettesacrosstheir fields,and thes,eare saidto makea noisein the wind. Othersuse a tin caa{d
l9

Marketing of Crops Geperal 132. There used to be Intlians with big shopsar Uwandwg including rvholesaleand clove-buying. Therewas no needfor farmersin Dayato go to Wete. Farmersusedto sell tireir clovesbit by bit, not all in one go. for example10kg at a time to providean all year round income. Peoplekept the clovesin sacks- up to 100in their hompstores(dtsla). This 'free market' disappearedafter tbe Revolution. 133. There is nc rrarket centre in Daya and it is diff&eultto sell produce there (except for cloves). Middlemeaeithercr:meto :Dayaor farmerstaðeir commoditiesdirectto town. A lo,rryccmesonce or twice a wesk to collect both passengersand goodl. The lack of transport is a problem, though oxc&rtsareavailable. Somegoodsare transportedby seafrom Mtambweto Wete: for exampleit costs Tsho lO0-200per bananabunchaccordingto the negotiation(the fare for a personis Tshs 30). Some Dayaloctls own shopsin Wete whichbuy bananas,black pepperandcardamom. 134. Peoplefiom Mic*reweni,Wingwi, ShumhaN{jini and MaeiwaNg'ombe cometo the village to buy cassav,Abananas,breadfruit and orher products. To lhis extant mar.ketingthese products is not a problem in Daya. There is considerabkseasonalvaliatiort in the prices for agriculturalecmrnidities, which are high in Ramadhanand (traditionally) &rring clove-picking. During the holy month trade in spicesincreasesin particular,andblackpepperard csrdamompricesincreaseda lot this pastRamadhan. Ban4nag 115. The bananatradeis increasingbecauseof the increasedpriceof bananasinZwzibar town and improvedsealinks to Unguja. Mkorn wa lembahasr good marketin Mombasaar well as elsewhere:it is not gmwn a lot on Unguja, Palwsa, mzuzuandespeciallymtwikeare taken ta Zat:aibartown market **m Daya, though not by 1.oca1 fermers,but outside buycrs. Most people now rvftnt to plar$ more bananas.The price ofbananasusedto be highestduring Ramadhanand the elove-pickingseasonand drop thereaffer. Now pu*z.sahasa lower priceduringRarnadhan. 136" There are thr€e peoplein Daya who regularlybuy aud take bananasto W€e, where they have shops. At leastthreeFRG membersclaim to earnmost of their incomefrom bananas.One farmer gets most money kom mlwrn wa tewb(, andpukusa (he consumesotler varieties"for examplemzflzu, bokoboko*ndmtwike,athome). Anothermakesmost onmkanowettembo,pukusa,przuza,asdmtwi*e. l3?. .{n estimated50-100bunchesare sold from Daya every-week. The price variesaccordingto the variety and the size crf the bunch. The following (selling) price rangeswere quoted per bunch of differentbananavarieties: rnkowswqtembo: Tshs2,00&.3,5$0 for alargebunch;Tshs80&1,500fbr a smallbunch wtwike. Tshs1,000-3,00O puhtx: Tshs500-800 bakaboko.Tshs600-1.000 mzuztt.Tshs800-2,000 mzigo'. Tshs300-600 Spices 138. Ginger is the principle crop of one y.oungfarmer, followed bJ eardarnonl black pepper and cinrrarnon.F{eis a spicespocialist.Cardamamand black pepperareoften *old irr rhe village shop and boughtby peoplewho cometo Daya. 139. Farmersget a bette{ price for black pepper frorn traders who corne to Daya than ftom shop ownersin Wete. The ndddlernen*ormally pay betweenTshs50G.800per kg, while the price ros€up to I,100 per kg earlierthis ymr (presumablyduringRantadhan).In the past black pepperwas mld at Tshs 300.500perkC

?a

14O Cardamomis one farmer'smain sourceof income: he hasc.50 plants. He cafl get 3 kg of dry cardarnomat han€st andthenreturn 10 dayslater to harvestagain. Anotherfarmer wantedto increase cardamomproductionnext yearbecauseil hasa big market: at presentit is his secondmost important crop. There are t'wo main harvestings€asons,mwaks and radr, but farmersharvest a little all year round. Cardamomis saidto fetch Tshs I,000 per kg in the village,Tshs 1,5O0-2,000 per kg in Wete" and Tshs2,000-3,000per kgduring (the past)Rarnadhan. I41. Cfiawsstill producea significantincomefor somefarrners.Onefarmerwith 1,0OO or moretrees gets Tshs 400,000=500,000 per seasoRif they producewell. Another farmer manages30 clove trees belongingto his mdher: the family harvestsup to 250 kg of dry cloves per seasonwhich gives an incomeof aboutTshs45;000flabour oostsarehiddenbecausethey harvestthemselves).Togetherwith his tlner sistershe prefersclovesas a sourceof cashbecauseof the ready market,but they complain about'tht tow incsme wtriph +loves provide comparedto their daly needs. Cloves arp their most important cash erop" fiollowedby b**anas (they can sell up to 30 brmchesper y€ar for a total of approximatetry Tshs I 5,0ffi). Fruits 142. Inthe pastthere was no market for breadfruit,but duringthe past seasorlpeopleftom Micheweni eameto Dayaandboughtthe filits for Tshs10 each. They harvestedthe fruits themselves.Onefarmer andhis farnilvhave3 trees: it is possibleto-harvestup to 1"500fruits per treein a good season.They *re so abundantthat any villagercan harrcstthem ftee of chargefur do*resticconsumption. 143. One farmertakespineapples himself,totown and selisthem for Tshs200-300each. Another farmer reported making an income of Tshs 50,000 per year from the produce of his home garden: pineapples,banaaas,sassava,bhck pepperandcloves, Food Crops 144. Cassavahasa iow priee in Dayq especiaBy during the breadfruit season. It is mainly grcwn lor domesic co*surnptionand only the surplusessoldfor cash. Riceis not sold at all, but only consumedat horne. 145. One farnrertri€d to grow swe,etpotatoes,groundnutsand pigeonpeasfor cashin the last three yearsbut was stoppedby the monkeyproblern. This indicatesan attempt, albeit unsuccessfulin this case,to diversi$ cashcrop producticn. Livestock Cpttle 146. Thereare not a lot of cattle kept in Daya. Sornefrrmers hare no cattle at all and those who do havean averageof only oneor two animals.each.Two FRG memb€rshavecross4redcatttrewhich they bought: they tether them in the usual way (see below). Otherwisemost of the cattle in Daya are differenftypesof tho loeal short-hornz*u. tr47. Therese€mto havebeenmarecattle in Daya in the past: periodic economiccrises(the 1971-72 famine,the declineof clove incomes)haveresultedin manyof thsa being sold off One woillen fbrmer saidthal her fatherhad a 1otof cattlein the past. Shesaysthat the herd was reducedbecausethe areain which they could be herded diminshedwhen cultivation expanded. There is rorne conflict with cultivation: peoplewho find cattle in their fields may cut their tails or legs. Somecattle were sold becausethe ownerswere too old and had no childrenat home(being at school instead)ts look after them. 148. Cattle are tetheredin the valloysby the rice. T?risis a problemwhen there is rice in the valleSn. On the clove slopescaftle may slip and fall over" especiallyduring the rains when there is too much

2l

water to heep thern in the vallrys and the slopesare slippery. \'illryers used to give their cattle to people elsewhereto herd (in other Mtambwe villages like Makongeni and Matuuni, or outside of Mtambwealtogether). 149. Theretag and to someeritentstill iq a 50:50 systemof payrnentin operation. The first heifer born goesto the owner and the secondone to the herder. For keepingbulls the herderis paid by the year. The tetheringrope is providedby the owner and milk is taken by the herder. Sorneowners put their cattle out with different herders. On Pembathey used to herd in large herds without tethering, there$yrnanuringthe fields. This is still donein Michewenibut not elsewhere.A smallhut is/wasbuilt for cows andheifersto keepthemdry whentherewas a lot of rain a$d alsoto keepawaydogs. Thereis a specialareain oae of the valleysnear Daya fer keeprngcattle at night, but flot €veryoneusestiis: insteadtheyjust leavethe cow$to sleepwherethey are. 150. Onefarmerwiti onebull and threehgi6tr hasgivenihem to semeooeslseto herd and they divide the products: milk" calves,andcashwhenoneis so1d. 151, Cattle owllers leadthemto whereverthereare good feeds,not jus on their own farms. Cows are usuallywater€dandmo.vedthreetirnesa day. For examplea cow mightbe milked in the early morning, watered"andtetheredby somegood grass. Thenat I t-12 arnit is put in the shade,arldat l-2 pm it is givenwater, beforebefugailowedto grazeagainandwatereda third time in the late afternoon. 152. Both menandwofircncan milk cowsbut it is mostly doneby men. Milk is soid locally ftom horne andcasts Tshs70 per squashbottle, It is usedespecialtyfor ehildren,as rnilk or in porridge. Peopedo not really buy milk far tea. A milker usuallyhasthe sameregularcustomer. Theneis not a big d€rnand becausetlrereis not a iot ofmoney in the community. No milk goes outsideof the areaat the moment, though one fanner thinks that he can sell milk in Wete. Thereis not enoughmilk now in Wete, tfiough in the pasl thereusedto plenty. One crossbreedprovides5 bottles in tlre morningand 5 bottles in tlre evcrr-ing; only a little of which may be retdnedfor homeconzualption. 153. Onefarmerhasa little lzucenaplantedas feed at home. I{e alsoplantedsomeelephaatgrassin a valley but this was eatenby veryets- He got theseplantsfrom the lrish-aidedlivestock project, which usedto visit Daya. He has 5 cross-bredcattle and cuts leucenato take to thein. Peopledo not take wild plantsor grassto s.owqbutjust let them eatit in the field. 154. Cattlearemainlykept asa {brm of security. Cattleare slaughteredif they are sick, andthe meatis sold either locally or at $omedistaaceforn the village. Cattle are also slaughteredat weddingsand funaalq at at Lhek#enu for a rnemberof the family who died long ago (calledfulili\. If a child dies a goat is *laughtered{hnkiki): this is not done muchnow. Otherwisecattle arekept fcr securityand sold if there is an unexpectedproblern,for examplea funeralor other cerelnonyor urg€nt needlo purchase sorn€thing. 155. One farmerusedto havetwo cattle but sold them about six yearsago, oae to pay for a wedding, the otherwhenhe hada problem{unspecified).Ill,healthpr€ry€nts him from keepingcattle again. 156. Another farmer had 12 cattle iil the 1970sand sold them to buy a dove and coconut fzrnn(600 clovesand 100cocornrts), After harvesing he usedsomeof the moneyto buy a singlecow and now has 6 caftle, 5 of which are cross-bred. The calvesare rearedin a shedbut the bulls nnd cows are tetheredarldgrazedoutsideof the village. He doesnot usethe manure. 157. Onefarmersellshis calveswhen they reachmaturity so that he can buy necessitiesincludingfood and clothes. He saysthat the vet lives too far awayfor AI to be readily available. He does not usethe maurre providedby his ttree caftle becausethe arnountsarevery small. 158. One farrnetwith one cow andone bull is thirdcingof using thc bull and calvesfor ox-ploughing. fle doesnot usemanurebecauseof the difficulty of carryingit. One farmer we interviewed,however, hadused.manure on sorneof his crops. ?)

for trading in cattle. A licenceis neededro slaughter. A 159. Sornepeoplein Mtambweare kno,pvn cleanareahasto be preparedand the eow impectedbefore and after it is sleuglrtered.But fhis is nct done a lot. The headand skin haveto stay there to be checked. The nearestvet is at Kisiwani, he is also the local livestoekotficer. A certificatefrom the shehais neededto teke a oow to town. (The Daya-Mkaniuushehadied aboutfive monthsago,his seeretaryisaowat Mkarjuni). 150. The maindiseaseproblern,especiatlyfor young cattle,is East CoastFever(ECF, matukni). TIns occurs particularly duffig lhe mosiht rarns. Farmerswith affectedanimds either cpntaet lhe vet or apply a hot pnga to the cheeksof the animal. Trcks are also a problem: fannerswho can contact the vel who usesacaricidesto eontrol tieks and iniectscattlewhenthey are sick. One farmeruseskerosene to minimisethe numbs efticks.

es${y 16l. Many farmerskeep chickensbut not in large numbers. Ducks and guineafowl are also kept in Daya,but no doves. They are mostly kept free range. One farmer wants to keep turkeys which he plansto obtainfronr his brotherin Dar es Salaam.He has 15 chickens,2 ofthem hybrid. He sdls some eggs: the price fluctuates,at presentit is Tshs25-30 per egg. Shopown€rsin the village buy eggsfor Tshs30 each. 162. Chickensare kept to provide both firo-dand castr. They and the eggs are usually sold locally. Scme locals buy them and take thern to town. Ducks are kept for food and the eggs are also sold. Guineafowl eggsare sold ip particular. Eggsare mostfavouredasfood for children. 163. Onefarmerreportedputting chickenwasteon eggplantsandbananas. 164. Newsast{eDsease (nagwn} is the main diseaseprotlern for poultry. To treat tlxs mclwnga (a kind of leaf vegeteble)is po*ndedtog€therwitb other food and given to the afflectedchickens. Ducks are also killed by NewcastleDisease,but guineafowl are resistant. One farmer saysthat he slaughters the whole flockif he seessignsof NervcastleDisease. tr65. Poriltry are also attackedby raptors{especiallythe bird of prey calledkigtnga}. &herAnimals 165. There:rreno goats"sheepor donkeysin Daya. Dogs and eatsarepresent" Dogs are mainlykept asguarddogs,thoughonly a few are activdy kept, andother dogsjust roam aboui. Someare alsokept for hunting. Dogs are not normally bought and sold. Cats are kept to control domestic vermin, includingcockroachesandrats. Il.onoyGalhcring 167. Thereis no beekeepingin Day4 just the collectionofwild honey. This is donemostin the season rvhenthere is a lot of honey, epeciallyDecember-February.The first person to g€t to a wild hive o
23

areaof forest is on governmentlaod at Senguiya. Farmersare oot necessarilyfarniliar with the species on their land especiallyon plots that they do rot use. 169. Secondarygfowh in elor,replantationsincludesnrlnra, muirwivu attdmtuguu, It is oflen mixed with clovesthoughin someareastherearenone 170. Onefarmerdefinedthreetypesof forest: (a) pure natural forest with tall trees, on privately owned land: pe{manentcrops can bo plantedby 50:50agreementar annualcropswithout any agreernent (b) ctoveandforest mixturg also on privatei family land (c) governnentforest: this canbe 'borrowed' for cultivation 171. Anptherfarmer definedtwo typesaccordingto the domilaqt spgrics: (aj wehenza ssila f crest (b) zdnra forest t72. One farmer recognisedthat the forest helps to bring good raiils as well as providing wood products. However"he notedthat it also harboursmonkeysand other vermin. He was also awarethat the soil underlargetreesis fertile becauseof the shadeandleaf-fall. lirqty"oodCo-$ection 173. Thereis no real shortageof firewood in Daya. Peoplecan cut treesfor firewood without refening to the owner. Howevel clovesare sfolenfor firewood. It is not recessaryto plant indigenousor other speciesfor firewood becasrse there are fixanytreff avaiiable,includingmchenzamsita endcloves, aad thereis no scarcityof firewood: it would not be profitable. 174. Oneyoung,r.lnmarried manin the FRG cuts firewood for saleif he hasa customer. Anotlrer farmer doesthis for extra cash: ?0 piecesof firewood sel{for Tshs600-700. llowever" thereis not a lot of business because therearefew custorners^ CharcoaiProduction 175. Somechareoalproductio*rar€aswere observedin the plantationareas. Charcoalis not boughtand sold locally, but the producerstake it thenuelvesto Wete. Daya is not a particularlyimportantareafor charcoalproduction: Finyais a big sourcefor Wete. Charcoalis especiallymadefrom cloves. Clove treeshaveto be purchasedfor charcoalproduction. Producersare not allowedto makecharcoalamong the clor,'esby the plot owners,becausethis will aftct the clovesthernselves.Sometimestrees are cut dovsnwithcut permissionin order to makecharcsal,or provide pole* and firewood. One farmer had plantedCasuorinoin order to makecharcoal. 176. Charcoalproducerscr.rtbig fresh lc,4gs and arrangethernon the ground {not nece$sarilyin a hole) and put earthon top. Fuel and dry wood are placedat tire entrance. The whole production processis saidto talcethrce weeksfrom start to finish,includingup to one weEkto burn the uroed. The fire is put out with water. A bag of charcoalcurrentlysellsfor Tshs700. Local.Tin$er Cutting 177. If someonewants ts cut a mango troe for boat building they have to agree with the owRer: norrnaltythey are sold. Treesare more often cut down ts provide wood for carper$ryand furnituremaking. 118. There are three timber cutting groups in Daya. One FRG nnemberleadsone group. which has Thiswasbeguninl9S3,maintytacutmtowfunfbrmakingboats. threemembersinall. Therehasbeen

24

a steadydeciineover tle yearsin the availabilityof preferredspeciesfor timber cuuing. The fullowing speciesarenow all rare anddifficult to obddn: nrule (mostly plantedtrees, but some sprout frsm the parent tree), valued at Tshs 50,000100,000each mlonfuo,valuedat up to Tshs10,000p€rtree mbantbukofi(fioundon the coastand offshoreislands) mkatgaclnle mkw*ga(mostlywild), valuedat Tshs?,000-10,000 pertree wfenesi.iackfruit,valuedat Tshs5,000-10,000 each 179. The most comrnoniycut speciesRowar€the following: mchenzamsita mwavi,valuedat Tshs1,000-5,000 each mtulullt : ml<
25

member,had plantedilrany uees,includingmvinje (Casuarirn), mto'ndaa,msaji, ar,dmfutngu,given to him by the ForestrydepartmentHousehatdlrcawe Men 185. Many farrnersdo not know their income,or it is difficult to get the informationfrom thern. They areafraidofjealousy andfearthat they maybe bewitched. 186. Men have a variely of diferent sourcesof incomg and there is no singledoninant pattern, fcr exampledependence cn a partictrlarcrop, aswas the caseduringthe cloveera. 187. Onc farmer dcpendsmainly r:n agriculture(in gener-al)and timber sutting He also gets some incornefrom plasteringhousesand slaughteringcattle and sellingmeat. Anoth€r frrmer's main income comesfrom cardarnom,his secondprincipalsourc€beingbananas.In order to m€etlarge expenseshe bonows moneyor gets it from dling cloves. A third fanner, with a large income from, cloves,gets exfa moneyfrom his youngerbrothersaad tleir children. His secondanil third sourc€sof incorneare cardsrnornard black pspper. Yet anotherrnalcibrmergetsmost of his incomeftom tailoring. For some yearshe worked as a driver and tlren in a garage. Oneyoung uamarriedrnandependson clove-pichng and selling spiee crnps. A primary school toachernotd that agriculture is more profitable tlran a governrn€ntsalary: he earnsTshs7,O00per rmnth from hisjob. Wonen 188. Wornentendto havefewerand smallersourcesof incomethan men,often from petty businesslike occasionallysellinga chickento buy cloth. Other sourcesof incornefor women include handicra$ production (making rnatg bags and plate covers), tailodng selling bannnas,seasonalclove-picking makingmqanfuzi, andfrying &gaa. 189. The only woman in the FRC gets her inconrefrorn tailoring and setlirlgfbod crops (especially bananas)and cloves. Sheis alsogiven cashby her grown chitrdren. GenderRelations 190, One rnalefarmer said that he can buy anytling he wisheswithout consultinghis wife. He can borrow fioney &om his wife if she has it. As he noted, the father andor lrusbandis the principal decision-makerin tlre household. 'Decision-malring 191. One farmer'swifb seeksadvicefrom her husbandif shewants1o sell her crops. isjoint. Sheusesher incornefor buyingclothesfor her shildrenand herselfande$ssntielgoodsto help her husband(soup,sugar,rice). Shehaseight childrenanddoesrrot want to use farnilyplanning. In her view a major constraintfor householdsis the local water supply: the well they use is distant and over hills. Shesaysthat cookingmaandaziis not profrtablenow becausewheatflour hasincreasedin price. 192 Anoths farmer'swife gxowsher own prneappleswhich shesells. The moneyis hersand sheuses it for damesticpurpos€s.I{er husbandcuts andsellsthem fior her. Shecan cut her husbandisbnnaf,as and sellthera. Yet anctlrerwife hasher own banenaplot from which slrederivessomeincome. Shecan take money sr clothes from her husba.nd(whn is a tailor) without his knowledgeand without this resultingin anyproblem. 193. One wife who is a teaoherusesher wagesto buy clotlresand nther necessitiesfor her children Shegles meneyto her husbandif he.is Sort and'hasa problemof somekiad. Interestingly,while the wife saidthat shejust givesthe rnoneyto him, trer husbanddescribedtheseasloans.

?6

194. The responsibilityof the uran is to provide tlte main dxh(bitowee) and in somecasesall of the food which the family eats. Ahhough mmy villagesare conseffative and wonlen do not go to tlrc market,in Davathey maybuy food aswell asmen. It is especiallywsmenwho pic,kvegetables. 195. The usual kitowea is fish espe.ciallysmall fu" dagaa, which are the cheapest. The following leavesare regularlycollectedby wcmar: kisumw" cassar,aleavs tntariro. sweet.potstoleaves{a specialvariety plantednearthe housesemongthe bananas) tnslugwa, cocoyarn(malnrli) leaves mchicha,Aunnanllrusspp.(this is pianted) mchunga,anotherkind of spinach{this is not planted,andis not acommonvegetable) m:*irmbale mhwryakwaya 196. If a yaung rnaneetswith his parerts then tre usually helpsthem with money. Youths wittr their own paid work control their moneybut also tend to give someto their parents. However, manyyllog menhavemovedout of Dayain searc.h of betteroportunities eJrewhere. 197. The patterns of mealsif usually as follows. At I I am the fa,rnilyhas breakfust (chai) afrer cuhivating. Dinner is usuallyin tlre evanirg, argund6-? pm. Cassavais the main staplu rioe is only eatenoccasionally(becausethe averageyield of 3-4 sacksis usuallyall cons*medwithin threemonths). Cassavais alsoeatenby childrenafter school. Sometimesporridge is cooked in the morning. I* Daya peopleonly usualtyeat twice a daygroupsin the vitlage. l9-8. Thereaxeno cooperative

27