/ .¡ r I ,. ,~ ~ .. , < - --~.-~-- -- '"""" " " WtECClON HISTO!;f(A ~ 7 1 l' . • GUIDeUNES " MITHODOLOGIES ~o,.' 8oC~~(;M'IC " pOUCY ANALYSIS: -SITE CHARACTERIZATlOW IN THE "5 " B- PROJECT 'j' . g't:> ~ (' DRAFT OUTUN~~v D~Paohlco (CIAT). l:A:lhrupp (WR~. 800 S~Vos1l (IFPR~· December 1992 ThII cIOCument outllnu guldellnes. methodologles. and data needa regatCllng ..! te cI'IInIct8rIzal'lCr looIoeconomlo aOO polley cllmen,lont o, the Global "Altemallvet'to Slash and Bum" proJ8Ct. l. !mIad 00111 IDd QtüctIvu: In 811fab11sh/ng researoh methods anel e$ctIng data 'ICr the "AltemaUvet to Slash aOO Bum" projId, It la Impottant to WOJk towardt the fuIIIlllng of the overall goala whlch ate: • to reduco the rate of tropical d9forestatlon • to Improve the well-belng of ruOUtce-poot farmera. The ob!.euv •• of the social seIence resaatch component (u Identlllad In the propoSal) lhOuId aI80 be taken ln10 consldefatlon: • ...., .. and prapare diagnosis of the tooloeconomk: and cultural1actors and 'lhe poIk:y anvIronment Jeadlng te s1ash and bum agricultura: • Datlgn and avaluata poIIcy altamaUvu to ellmlnate or reduce S & B deforestatlon. promote lultllnable agricultura and protecl the environment: / • De\'alop methodologles and tool8 for pollcy (and socIoeconom!c] eVllluatlon to facllltllte deCiSIOrHnakIng procees" fer Implemenlatlon ot polk:y alternaUvu: • ldefrtIfy and adapt succ:essful poIk:y experlencet trom other counttles aOO locatlona: • AIISall poIIcy declalon-maklng proceesll and 1d8nllly crltlCal Interventlon pelnta an meana te promote lmpIementatIon of poIIcy altarnatlvet. 11. Ka QuldJllnt,. f?roctuu, .n~ planl Ior Btuarcb Dtalgn; L lntatdl!lC!pllnar!ty; lhe researoh should antalllnterdlsclpllnary taame ot people in eaeh slte. For lila pollcy/$Oc:lotconomlo reseatch. a preferable mlx would Inelude an ec:onomlst. soclologlst or atJIhfopologlst. an agrlcultural/resource polioy expert. and 8 demographer or nutrftlonlst; and "'e peopkI shouId also work dlrect!y wIth 'lhe biophyslcal te8m. lhere should be a balance between natural and social scIentIsIa. b. AggyIs!t!on of ex""'ne Data' The reseatch t88m ahould avold duplleatJon of prevlous reaeateh and te take advantage ot exi&ting data, They should aequlre e¡d9tlng document9 and data that has lIready been collectad In thls tIeId. Such InformatIon shoul6 be fully revlewed tlefoiirr Sartlng 'lhe . i u,}., .... ... 011 beIIavIour lI.l4r Madru d •• 1b_t • .o f outalden lA die teleltdl proceu WII rtCOII __ •. 5 OM Of flcUitIIor lIl4 ~ dIrec:IDr lid II1II,.. PItA prlíO 111 ....... lA pritlcIpIe ti) 4IMIYe die prodItctloD of bowIedp, ........, .... die paentloD of poAIIIlllfllOludoDleato ..w IIoa ItwllIIoodI fbrmed tIIe RbjeCt'" ........ Al • rauh,loc:II poopa. .. 80 1oDpt.- • .,., or ~.IIut. ptI1"t1JIIn 111. p!.QCIIl of nmrllb ud cIMIop ••• lit lato lIIOdoII by_PItA. .. . TIIe • IdIodI of lUtA, • ,_, of PIlA, _lid ID .... 19101, ad DOW COIII(IIIIIe • ridI _ of .lflsldoa. ~ lid poup wort ... (Boz 3) Iba 11M ~ vaIubIt ... IlllllliIllMdIq dieloc:ll percepdoaa ofdle tImcdoIIaI __ of~ ..p IOc •• of lpicIIIIurII lIIDIWIdoA _lile GOJDPla\tloa popf oIOrCIIIII'" p nlClIl. lIId -. PItA ..,.0ICIIIt• .....mI .... fll'• • _ pnalce aleo crffet lit _billalDa loc:II peopla lit joIIIt lIldoIl (MIICIMIIIs ..., 1.1; DIvMnm, 1991). • 1. I, , _"Md __ ,~; ...,f iow4l.p''II': .... - ". ..; ¡ ' .. ".'.: . • ! b 1 OD I PIa'. D!IIr,mm"lI lIIlbIes ccmtrol ove!' die c:núIOIl l11li 1IIIl,. of die .... modeI._~ ! 4_I1Ifa1t IIDI 10 bllbInd. PJOYIdiDII IOIJRII of iId'ormadoD wIúc:II CID bI ......e d. IDOdIfied - i dlld.1.ocII CIIIIIOfItt. erIterIa I1IIIIJIIlIoIt arelllld ID die el ............ '*""" .... 'fIIIIcIb ... , fmm lIIIIpJIiDa 11III1IlI.ldeUID¡. to eo~ _,.. of IocIlpelc.pdODa of..-lllll1lll WItodcaI 1 ....ss l11li ID ~ npr 1I&IIIIIIk'III of IIouMIIoId l11li Ilvtlillood .,..... ...... .1 \ .iMiD& qIItIdODs wIIldI .. cI\recbd by die VIIueI ofdl, rtt IIIC%7, locIl peopI. are ~ " ID cnat:iYtI~ apIore 1beIr OWII venIoIII of dIeIr warIda. . ! TIItJ_ 3. CoIDpariIIJ .. YWbII l11li visual. 1 1 v. ... YIIIIIJ I ..... - (iD!M ... ~. ....) ..... ..... 1 ~. ...... , iIIiIlp. ... RMdivtI Cr.IM..mt1llill 1 llIya Ibtnaive """"'_'1_ oí _ HlP r..-. r..- " . ~..taGt Hip 1'IIt •• 5_ ... 8IIIIriaIf ... ,.",..,. ¡..w., ... or: DeIIIl iIIft. ... lIJ: litio EIIIic .1a.".'-- 'ioD ftow o-IItiYt AltI.¡bill)'oí~1o TI.D,W.. "" ..., HIP lIIItiatiw b 0I111Í_ ......, OwIIerabip oí ialonMIioo Ap¡woplillwd by 0I111Í'" S..I.I.u.n.d.: ca .. CIWIOId by ~ lankIDa lUId scorioJ IlUl'clJlls provido. opportwIhy 10 lIDdentaDd lOIIItofdleeompler.. iavoIwd ID decllloIHDItiD¡. 'l1Iey .. puticlllll'ly YIIuabIe ID die le lIIoa of loI.:IIIy wr.", ....' ¡thd criteria Ibr teleai. lIIId evahIIdD¡ panIc:uI., VIrieIiII or tecIIaoloJIeI. ,.. 1IIIdl.1bey ilrDllIl ÜlYaIIIIbI. JIIIde Ibr l,IrlcuItIInI nallu:.ben 11 to .... ~ l11li pe-- ••• Uve 11 repnb vltieCll dIolce. MtIIboda 1UCIl1I c:rop bio,papblol. MwOtt lIIId Pllbway diaJt'ne'lD& (FADf..AIrIca 1991) lI!d, JII(lft reetI1IIy, ~ ~ c cboicc: bctwCCII varietic:s of maize rrom r:escarch stations is simply compared witb dIoice$abollllalld e1carin,l. lalld use. trec pIanti1Ic and the management of woodIINls. Farmers can 1.3 Local knowledge for agroforestry and native change maize wrietíes aftcr _ season, bu! decisioIIs abolll woody pIaD! plants resoulCCS aad the soil are not so rc:adily reYCnible (WiIson 1981). DecisIous iD lbe preseat may determine resourcc eooditioos lor ¡e. ... eratíons lo ClOIlIe. DIANNE ROCHELEAU, ICAMOjl WACHIRA, LUIS MALARET, Thc: second diftkulty witlt AF ilUlOVlltion ÍI !be ClIOI'IIIOU$ lIItl,IC of 8ERNARD MUCHlRI WANIOHI specics alld AF pnctiees. ScieDlisas !acle proveo pactap for 'Ibe w.erse environments and circum!i&aJlClCS of rw-aJ peGpIc iD tbe regioo (Rocheleao ~ "'-..... -m ,.ac:11I5 in l\friI;a and Rainaree 1_). One n:sponae by AF l'eIe8rCbcn has bca tochoole a ~ry heeomc: C1IViroIInIenta fr:w prac:tíces lUId a mort Iist of spc:cies and teat 'Ibem undcr a wrict:, of has popular iD dcveIopIMIIt and cicdc:s ~ !he wodd. 1t is _ Gtlen mwted 11 a _ soIadoa 10 c:ir<:umstaaccs, yct 'lbere 15 sboIta&c of tiroe and resow:ces ror sudl a tri. ... runII develapment oceda. But 'Ibe tciealific ~y and deYCIoptIIIIIlt and-erTOr apprlllCh 10 AF rescardI. '">Y_"''''~·'''''-'··'-',,",,".'~'-''-'''.~~---'''- For both practical and cthical re.uons, die rural pool shoukI pro­ proride toOls for stltdyiog exiMiDa 'aatural' eaMlystems, traditional AF domiIIate ia tbcSc complex proce_ of tecbllOlogy and laOO WIC dlanp. I)'llIeIIIa and r_1 iIIIICMIIioas by rural pooplc. 1bcir metlloda pro.cm Praetically, forma! experiments bave limited lICIClpC because trees require a ramy km& . lU!lpIe soopc b iDcotpOO!ÍIII iDdipnous tedIDicaI ~lcdp, ÍDIfi8IeOOIll JoI: of spIICC aud a time to pow. Only • f_ fon:ual expcriments eapadty for inrlovation alld iDdiJII- eapac:ity for gpeliDlentatiOA into QID tbetelore be earried out, wilh ií:w repetitioDs in Ipac:c ud time. Nor !he idcntification el spedcs for domes1Ícuion and thc desip and tesIina of can forma! tesiiDg be u.ndertateII 011 • ~ 10 &t the DUIIlCI'01ll. distinc:t new AF and woodIIlId manasemeol 1)'IItC1RI. eDVÜUIIIDIllCS w!IidI are ~ fouad. Oreal eare is COIISC'lIJ1IeDtty Rapiel Rural AppnisaI (RRA) teebniques ean COI1Ibine readily witb lICCMIad iD decidla& what spccieI and what AF practiccllO submit to formal e~ metbods. However,lt is thc stylc nuber Iban !he speed of apcrimcldS (Rainuce 1983; HuIcy and Wood 1984; TortoS 1984). RRA whick is moR critical. For cumple. reacan:hers can _ etImo­ Moreow:r, die oompIexity and seope of!he dlaoges ÍIWot-IIIR beyond ecoIogicaI dats lUId u.mpJc: ooUedion metbods witbin a series of Informal the capaaty of fonDaJ teSClIIdl pro¡rammes uacIu CDIl'ImIIed. oondítions. iDtervicws witb rural oollllDlJllity JIIOUPII of 15 10 30 pcopIe. foIIowed by Rural peopIe bave bete a oolllpll'Ative aduDIIIF: they ~ ud .. 'd!aiDs' of bouaeIIoId lcvcI ud iDdiYidwll intelYiews, mappiDs of farms wbole Iy&tcms in all their divcnity and varilbility; 88 dicats tIIcy know and ooIIection arcas and partjc:ipedoD in ptheriDg tripa. proceMiog and wbat wiD lDeet their aceda and tbey are weIl plaa:d 10 adapt and adjllSl AF otber aclivüies. DmiIt& subacquetlt \Ra8ICS oí n:seardI me &ame tiDds of OOmpoDCnts 0Yer time. From lO ethieal poiDt oí vicw,lÜO. it is ripl lbal búorrnation-ptbcriog activities can be lISC!d for monitotIng ud tNaluatioD lIJe pool nuaJ majority sboItkl direc'I Iny proccss whic:h will Iranafotm me of eqleriments. wIledler fonnaI or informal. Tbis cu apply wbetller tbe nuallllndscape and die bioIogieal basís of thcir liveliboods. me experimeldS are oo-statiotI, otHimD or iJl.llIo-forest, _ a wide range of AF reaeardl and devdopmcnl _km iD IIcld mUS1 lhercrore 'user'- en(f.'researdler parmersIIips wilh ~ 10 experimeDtal desiaD careIuIy mis exisñng Ioc:al prlll:tille witb tbe scic_ of deSignio. and ilMIMns IIId management. le!IiIing _ practi<:es. tbcmsd_ as c:oowItanlS and cataly. in a The possibililies finge from research-desisned eqleriments oll-stalioo ro prucess of teSClIIdl, exlel1lioa ud ewIuation _lialIy 'owned' by rural rural peopIe', OWJI olMite c:x.perimenll that lIJe _ply 'disco_ec!' ud peopIc (Roc:heleau aDd Weber 1981). doeu.mcmted by resean:II institubOllS. Mc6t proJi'am1llCS aJO based Oll a llIO!'lil dirc:ct c:>lIaboration between the two groups, whicb indudcs • variety ltIfIId r.ur. . ~ (IRA) ad _oIoP:.aI_* PIto b t'8i1I itw ..... of roles for Iand usen and fonnaI research insdtutioas in experimental Af~ de. and managemeot (FeldsteiD, Poats alld Rocheleau, 1987). MOIIIt of!be immediate _rt. in romIIIIlDity-bascd AF researdl wiII f_ Tbe mel~ for agroforestry in general ud commtUIltyobased AF re­ on ecoIogíeaI adaptations of RRA oombined wilh espetilneDtal sitnations searcb iD particular must oonstitute a radical departun: {mm 'IrIIdiIkm.al where tbe WICf is IÜO a resean:her. Hnwever, me exaet dIoioe of metbods ap-onnmy lID!l even from many oí the fanoing sy"ems researdl melhods and how ro oombíDe and appIy them is SIlO laIgely a matter of laste, style that havo become estnbHsbed in fonnal scientific circIc:s. Whether in formal· and available lCSOllAlCS. For lIJOIt professiooal researdlen, fint atlemplll or informal ft"'Carc:h protlJ1illlmes tbe approach shouId often be more' witb soch an approadl will be somewllllt of a personal experiment ro derive ~l than agronnmic, as betits tbe foeus 011 !he place of 1l'CCS, a cohereot mclhodology (rom an ecJcclíc: coIIectioo of metbods 10 aDswer 1I.IOOCIIands and savarntaS iD tbe habitat of fannen; ud herdels. Witbin researdl questiOllS framed in response lo loc:al CÚ'cumstances, ccotocY, both quaJilalive and quantitative sampling ud monitoriog Tbe [WO cases wbicb foDow are nol models, bul eumples of suc:h led:miques llave beca developed 10 SIUdy wIloIc systcms and me comptcx gperiments. Tbe empha.i~ is 00 Iessons leunec!, ud implicatioDS for relalionsbips betweeo orpctIsms. aud IlIeir environmeots (Odum, 1984; follow-up. Conway. 19S5). Moreover, lhe theory and!he metlwdoJo&y are _11 suited lo a sIiI:Iing scaIe of aRIIIJ$iS from trecHOIl iotcraetions to regiooalland use Syslc:ms (Odum. 1984: Rocheleau, 1983; Hart, 1985; Conway, 1985), An ~ from Keaya: triiiIJ, _ iIIIId hM' .,. : whcrcas agronomy is IImly moled in tbe plot. In explotatory __f arm ~ c:ondw:1c:d by ICRAF in Mbíuni Location, Tbe developmcnt oi AF ud woodIaad man8gcment systcmll ror rural MadIakos Disttict, Kenya. we tried 001 a combination of me metbods landIcapeI can bmefil pardcularly from tbe oonverp_ of metbodi described above. Aa carlier projc:c:t in the area _ based on 1he DiaporiI iD t_ sub-lields oí ec:oIoaY - etltnobolally and agroecology. While lUId DClisn (DotD) metbod (Vont, 1984: JCRAF. 1983 a, b, Raintree, etbllObolanJ, draws its mctbods froIII human 00(101)' and ctbllOllrapbic 1983) and invoI'led a famlcleYel survey, on-fann AF Iríais and monitoring tradiIioDs in anthropoIogy (Posey, 1981) and naroraJiR traditioos in plan! of Ioeal fuma. In ~ cydes of dia¡nollis lUId trials, farmers had tOO animal ,ecoIogy (Qbfor, 1981), agroecology derlvCl illl reaeardl identllied prinrity probIems for AF nllCl8l'dl to addresa: poor soil fcdiIity, metbods mOa: from environmentaJ -aa meat and sJIl'ClIIS oootogy inadoquale di moiature, dry IICUIlIl fodder lIIoItaae and Iack of buildina (Hart, 1981; AIIieri, t9&l: Coaway. ]986). EdmoboIany and aa-togy material and fuclwond, Tbe propclIM!Id respo_to lbese coru:ems ineluded aI(ey aopping wi(b UucdelfQ ¡'¡'_pIulfa fO! mulch aOO fuelw~ and b,o_ii and Comb,elutn "pp. leaves from large dispersed Ir«s, and bad rehabilítatioo of gm:zmg laods tbrougb plantmg scatlered mulllpwpose been doing so tor yeaJS. Witbin Ihe year mas! Carmers in Ihe 1IÍCÍnity had lodder treeS io microcak:kmeBU. Ten fanners (ded some oombination of tried lrus al Ieast ooee 'after trimming tbcir &phorbia bedges. Tbe neXI diese in informal tdals whieh were on-farm, rcseardIer-desigRed and logicat step seemed 10 be refinemenl 01 tbe leclmique in order 10 iDcrcase f.amler mana¡ged (Rocheleau, 1985). tite mttrlent oontent and lo íncrease buIk withotlt fouling lbe cattle pens. Lacer work al tbe c:ommunity level and foUow-up of tite original len farro ADOther woman wbo _ present al tite samc djs('!!ssjotl al the farm trial lrials prvelopraGt ud domesti­ reasons COI: practiccs ud prefereoces, problema widl pIants ud source c::a1ion of In=es. incorporated Umbcr speáes from carlicr riaJs ud sought arcas and ideal for improviJlg lhe silualion; eventuall)' tb.c ¡roup tactled dutions 10 dleit OWII speci1ic tree-plantiag problema. 'Ibi Iist of prioritíes deci.sioos aOOut whicb plants to domc:sticale ud where and in wbat j¡, distia.et from the lormal (or COD'IIOIItioaal) ~~ oí: c:om.binalioos. Intc:rviews ofIen endl>d wid! quc:stioos for participants 10 <:oosider, followcd in a few da1' by another session whídI ¡ave pcople time • Ipedes Ideetioo ud JetlCÚC implOU)lROftt of pIaDt maerial; 10 lbiol: and to eoafer with family and menda (J Kyengo. pmonlII • developmetlt al prototype tedmoIogy; lXlIDIII1IDie.ation; Vonl:. 1986; Rocbeleau, 1985). • adapGtion of ~ lO &itcs; ami Thc housebold and indí'llidual. inler'llÍeWs varicd in time ami io ionnat. • widespread mcasioo of a 6J:cd package. depending OD tbe disposition of lhe persons involved. Bcth formal aad By oontraat, this experieIlc:e arguer, for íntrodudion of nany varielica or infonnal approadIes wcre uacd. 0De informal in-depth survey _ based 1pCCies aad a few sampIe tedmoIogics widt emphasis m principies and 011 a cbain of itúonnants from 'average' 10 c:xpert; aoother _ • more !lcmonstraUolI of some prornisinJI:OIIIpIl'IICMS - as cftoctire approaches 10 formal raodomized sample of 63 bouseholds (5 per ce:nl of populalion), llel¡> buíJd sustainalJlc R&:D pnxe_ for ~ people. whicb asked fanhcn ti! answer spedfic qucslions about lile eDVÚOament, oollectioo, use and prefcreru:es of wild plants, elc. (Mutiso. 1986; Wanjohi. 1987; Munyao, 19117; Wadtira. 1986; Rocheleall el al. 19&5). 1'IIe fonual PIut domesIIcilllon: IocaIIInewIedp¡ and 'dIa!n 01 intet 'iews' survey took tbree times as long aOO reproduced lhe .ame maio rcs. ... ts We were parti<:nlarty interested iR women's use of off-&m 1aRd!:. ",hicl! as tbe ¡roup interviews ud cbaio ol intc:rviews, with less detall and iIIcluded tbe gathering oí indigenous plants and tb.c apparance of more roherence. ¡lid more 'in<:ipical' bome gardcllS. We also wanled o belp develop TIte surveys 00 _mco's use nf off-lano lands aod gatllercd plants altcrnalives that wouId serve womeo most depeodent 00 ¡roducts gathercd yielded a list of 6S indigenous species used for food aOO 9') used lor off·lana. 1'IIe project started witb identi&atioo of specicsand spaces most medicinal purposes. amoRJ tbem woodyspec;ie1¡, wild leafy vegetables aad importaD/. 10 WOIllC1l pdlcren and iDvestIslllÍn¡ tbelr inl~rest in domesti­ wild roots (RocheIeau el al. 198f._ Mas! of lbe fruit-bearíDg woody species cadng I'lrrourcd wild species on farms or in managÍDlI wlOdland sysI.ems. were also majar sources of wood oc fodder, lISCS whi<:b had receivcd more We focused primariIy 011 food and medicinal plants am secoodarily 011 atteRiÍDII in pre'llÍOUS slIJVeys of tbe farmíng system. In the fonnaI survey 90 -00 flIeI ud foddcr pIaDUI. per cent of !he S per cenl &le ¡roop reportelIclibe tbe exisling situaion wilh reapect vegelaba 10 sorne ment, 10 per CCIIt slid they use wiId g¡ccns year rouad 10 !he role of wiId indlgcaons pIants in IaOO use sysIem ud docnment md '10 per ceot reported that they or tbeír dlildren eal wiId fruiIs daily (OJ ttaditional pno¡riee md locallrnowled¡e, *ndfyil1l recetl irmovations in wheoever availablc). Masl oí the respondeats also used herl:>al remedies plaot 11WI8JClmCnt. 'Ibe erfort reIied heavDy on informal suveyr. of groups, made from iDcü¡eaoos phms. lKIuseIIOId and indl'llÍdllal!. _, bocb the anmuni)' al Iarge md In muy casca peopIe noted that witd planta play • particularty crltical act.aow1cd¡ed localexperts(Pope,l986; RodIeIca el aI,l98S;MaJaret &: role al some CÍIMS of the year. Some of lbe wild ¡reem. sa .. Commel!IfIt. NJOTII, 1986). In particu.Iar, we developld the 'chab ol inletViews' ~ (lCitowe), are particaIarIy impotlllut for late p1a11ters ("IC, poor 1 IMthod. peopIt:. wbo 'OOrmw' or rcat oxen) since lbese grecas tlII the ¡ap betw_ - K ,_, __ ~ ""-_",,,,~,.,..~, • ....,~,,, 1:> NI .,t,. ..~ .>""",= ..._ ___, ,~ • 0IlS!,!' of file rains and the first b~ of cowpea Iea_ from che • cropland. Likewisc, sorne vegetables (SolllllUln nigl'UlfI and Amaranthus ver,etables Ibat were more palatable and casier to prepare tban !be usual spp.) and fruilS are espedaUy importan! dlll'ing the dry season, with O\Ier lIIix of galhered grecos. 25 $peQes of fruies used by lile sample group during Ibat lime (Wanjohi, Fruil trees were ooosidered 10 be a lentati1le success if Illey eSlablisbcd 1987. Muñso, 1987; Wachira. 1987). weU withaut serious daalage by peses, diseases or droughl. Another Of aII tbe speQe$ Iisted, farmers MleDti6ed four spccies oí leafy vege­ measure of success ..a s !be degree of ¡oterest expressed by self- help groups tibies, nine fruit-bearins IlpeCics aDd seveo medicinal plants as good and individual farmen in the5e gardens. 1be.r priorily intereses included: c:andidates fOl domesticatioo oo-farm. 11Ie crlceria ciced fOl' choice of development of vegetable pdens in homesteads and group sítes; bome­ candidates and ¡he sllW$l.ed planaillg niehes and plan! oombinations also $load fruil !rees; mixed lree !lursctics; mullHtotey homc gardens. Tbey belped 10 define usefuI crlteria fOl subsequent !!creeDÍllg oí eltotics and foJ rcd many of the difficulties expen. 1ectioo arcas. Thcy _ n:c.eptive 10 the dornestlcation oí indigenous trees eaQ:d in eadler alley cropping and grazing land trials. However. given tbe (iocludingwild fruits) aOO wild leaf)' vegelab1es in ga.n:lens, small tree plots size and Iocation of gardcns, lanners found Ibese problems much easier to cear Che home and io-between spa<:eS sucli as boundaries, gullies and along address. Tbe home g¡wden aliSO presents a \oweT risk enll'Íronmenl for draioage aoo soi] conservanoo struáures (Roc:he1eau el al, 1986). MosI experimcms and a110ws fa:rmen to observe tbe entire system clase at hand. participanes werc: aJso eager to try eltotic spccies 10 SUpplemCDt iodigenous YeI, even in tbis limited and weft defined space, farmen must dea! witb fruits anel yeac:lables er 10 sell as cash crop; to urban oonsumcrs. There was many related inooVlltions (such as effective fencing, pes!. control, intensive m especialIy keen ¡nteres! in cltotic AmJJrQfllhu:I species for home use a, soíl managemcnl and rc:search and lraining in plan! propagatioo ledutiques) 1eaf spinach (Wanjobi, 1981). Atbougb t~ ack:oowledged Ibe productioo Ihat are necessary to support !le'" planto and practices. pOOmIial oí managed woodIands and gming land, Ihe people surveyed EventÚllUy scveraI principies aoC! componcnes from Ibcsc garden 'mal,' werre O\IefWbelmingIy pessimisdc about woodland mao.agemeat exceP! for wil! lilcely liad loor way loto cropland, gmzing Iand and in-between spaces !hose wbo actually own sízeable chunles of laOO witb bush and woodland in Ibe targer Iandscape. vegelation (Wanjolú, 1!l87; Wachira. 1987). TIIrougIIout tbiS cycle of s\1l'Ye)'s and plan! oollectioo lbe trce planling extension conunued, witk substantial informal feedbacle 10 Ihe rescarch '--md foIIow..up questIons effort Ibrougb this activity, as ..e ll as through participant observatioo in Severallessons were leamed Il:om. fbe sludies. 1be substandal inlerest in group ..o rle. ~. MOf~ver. ~e surveys sparle~ ~ inceres~ which in dome1ucation of índigenous ..U d planls (tncludin¡ treci, shrubs and 111m red back lOto exle1lSlOtl and Informa,t mals. Slmilarly eifect1ve ~up he~0U5 planls) warnmts vigorous researeh and extension to fol\ow.up melbods aOO processes have been woven mto CARE agrotorestry proJuts on !hose species idenlitied ter domestication or far managemenl in place. in olher part.~ oí Kenya (Vonk, 1986; Bucle. 1987). 'lbe experience points 10 !he practical value of: In Kalhama several fanMrs (aU _men) ..h a had participated in . • • •• intemews requesced a...istance with the design and establishment of small • ~OIce. by fa~rs. o! tndigenous aOO ~Xotic spectes for AF 5)'5tems, home gardens ror plan! domestiCóltioo. 1bey also requeSled bel!, in . CCOf:tntg. 10 entena Identi1ied by lbem, . roeurio" seed and cullings of indír,eoous lrees vegetables and vines a.~ • IdenlilicatlOD. uf sourc:e ateas and screenmg of gel1llplasm for farmer- " . ., .' . selected spea.es· ..- ell as exoltes. Each of lhe farmers planned her own garden (Wanjohl, .:. . . '1987' Wachir.. 1987) chose her sile and cieared tilted aOO fem:ed he r '""'1 • leslmg propagallOn lecbnlques far selected Indigenoos and exulte species • ' • • • I , Y"" for AF systems;. pnor 10 planunwgh iteUhl lien. ftoenced • ducti ' . l' . 'Ih diu - 'lbe factors se1eclioo of indigenous wild species foJ Ihe ron og SOCIa expcmnents WI .",real lenure 8rTangemenlS, and gardemi included abunclance, case o( access. and palatability ({or botb • :sting differenl lechllol~ designs ror 'incer\ocking' land uses by fruits and ve¡¡etables). Also importan! as selection criteria (01' vegetables ultlple usen al shared .,ces. _re preparation requirement. (re, ..h elber tbey need 10 be fermented in One methodological qucstian is how !he surveys should differ in timin¡. milk, fried in oíl or boiled) and wbether lbey lOre wsed alooe or mixed inlO COnDal and'or conteDt jf Ibey wen: done again. One need is a broadcr base .taple dlsbes as a reJiSh. of ecolo8ka1 iDfonnation and carefuI ideotification of Iopic$ for separate Of !he severo gardcn tria1s established during tbis season. ave 'suoceecIecr. lfeatmcnt and fer systema __n :h. AD impol1ant gap in ínformalion iS a I S_, as ddioed by !be farmers meanl tbat gardens produc:ed eDough ,ummary oí indipllOUS icnow1cdlle and enviroornental pen;eption. A i .... __ .. -L",O .". ftCCD ycr,a"'cs iv hmpC mmumO';OQ OLa. ,.. CM1"b lo u:¡¡hu;e ~,_'*"""'<, " •. " __' .'"'' !QQf;:rf!J, ~J, _t¡l'A.~ ..' Id> h.ln .tí ata. Fta,. tQ ",_uid. t.u_ _" '''o~~''''h_ i ." .. ! ! ( lo • - - J'~ ,..íc'fI'U0ry resHf'tlt tU 424. SAIW 'UJ;SAX4 rnearch in 1M muo.icipa.lity 4( a..wria. .......i .. Orien&al in (he Ptritippinrs. . Si.tt condúiou an: cha:ract.,.m;c of mnM ttf tfte aplands of SR Arai. wtñdIha.....e , stroftlly addic lOÜs witb rice .... C~"t cmp. 1lteit pIt haw ~ lo _ • .,.-.... of lo. .. ",,"uc.¡oity _ .. p". ..I i.y _ n.:..- .ha. can lit ..ro by ........0 1 ..,;.:..t. ...... _h~ •. Tht init.W ft::IIl!aR:h ..".. c:mil'Cf out on Ihr Ratter ftelU ... 'hlt.area Md ..d ¡~11 vo -.¡ ....r - "' ..... d> ....1 1w"ololli... ....... tlIan hall of Ih. Ji.ldo in aa.ma. _ ..... h_ aliope of..- Ú>OIt IK. _ "p ID 47,., ............/ y ""'¡ed <- oí SoiIo. 191!). ea .... _ ......y . .._ .......,h in CJavnIa .. , i i • j • ¡ i • , I ¡, Apt, ..., ,Jutt. JIIL Mi- $fp. _ Nov. OK. ..... ,.. .... _ """oidrrjng adaptÍl>ll .he .....r o....'Y ~ oí COI" ... heda<..,"" fIN DOIÍon contlol and i~d m.t'~ o:-.....m t. "'$* .,) ..... h?ffM# lelll,::!*sllliMl·,¡¡"¡¡ . ... .,." A¡rofOC'C'$Iry Jyltrmt annbine woody plmtlJ and fmM! Uf" fOl'agt' ciopt lo .... ",,"""1 • ..,,¡.,., (Yo_. 1-1 _ maínt_ ooiI fe:rtili.y (Sandla. 1! liT). • ~ .• --•• ~tt§i _ .... lZ_''''y' ~~-~.-·~·-·FIII\<:Jw~-~-~-- üoé.on control _ iel'tility mainteomtt &r(: iaurdepe-nDI (lAmdgrrn and .. ---.-~e 'Hui - t. = ? • __ ._~.n.~_._._~F'UOW,.w-~~-. N. ... 1985). T" qu.'" Yo. ....g (1!li7) 'app.."m. . a¡¡rol ......r y "Ytt. .... """,. ..1 .t::i ~ 4 +. ES? ya·*, .;, 's ha» cros:Jon.. III1iI.Íntain lOi. orga:nit mattt'l and pbyaieal propetbe~ and promott __ ~_ ..... __ ~ » .. ___ ~ ____ ~~ __ ~.F..,... ~. __ ~_~ __ ~-. _ ... ~_~t~ .. w_ t _ t-lfJcit.nt nutri~t q-din«'. TeduwfOKY ¡.,,'OMng th~ ... of .m A·(ramt lor a~" __ ~ ____ • --- _ .. ____f ' ....! 'IIIieI. ---..... ~--~-~ ~~ ~~--.. ~~~- cOllt<>UT lal"""' ...d bumled iIeiJ>g ...../ uperld No"''''''', ondoted). l. ía tw, t~""ty .....,._addoltC ..:.r.ib. l,l.l .lr.i.M...l .--.,.- •. ,....... to- - o) and .'. .....1 1;-.-•• ~,..--CIrwdII. j a_p.p.r:o1pr "1 ia •• lor """"', Jan mcn: ..i OIl" th< <."" ... ur for ...... .,... (h"" 1 ~ ...... -"ing in ",,¡. ....u onaI. rallo"land (FlK. 1), Aldu.wgl\ rkc. maí:e:e-. (:~.and ~rt'M¡als N1TUOl)S aJ"t',RrOWD throoghOUi IIv:Afe~ &htl"C" 8ft' du'CC' distinct .MInes ..mkh ('f~,rapund ruughly ttl incn:.ung allihute 'lbt intm:iisdplinaty tesearch lA\fII)tved icle'.w.tt (rum IR Rl md .:he DA. Efforts and n.infaIL Upbnd rÍCt'-tallow rotatjum and C;aH;iMI y ......,.,....... in .he ónitial ......y Wy 1" ' •. ' . ,\ .... __ .. _-~-_ .. _-- • • • 426 F4nft.tr parJlLPpdll)'l"'f res" SAJo{ FVJJSAttA 450 ·C..bI.I, lAa4fUec,,-.~CI.o.""'l!lól-1Wl5 off. r_ aIIcd a.le- r 10 ¡>arts .. thcit ro "'. .. . ¡ (from ~ mió) • ...,.;00 ano! ldenuOed ....... riP ...a PI, cm" 1oc:aI_}.boIIt 15~AÍd Ih" p~ .....1 r ¡m,. . ¡ r. I n.... -~ do po' ..,...,..... ~ ",""y ..;d ..,;J• ....., d _ l'" ".ath .n6 _ !he lIO\f laye .rotI me...-d in • t,,", crc~ 1'_ .." ,d ro-' "r dul r~o ... lMir "".,..,.".... llI'I"'" &c\ .. t1 ......... .,. ... peñod olexplon.. ... y ..--eh """ ........" ""' ioowe. úcinS 1__. ..- 1. y." .'; '. .. _ ~ In V...,...,. ... __ ..... _. bu> al'''''''''' ........ ..........sI _Id (1 _ ...... ir, ¡..,..¡. ....... nins ",0IJIWl .. de.'"""""' \'IOOt!i< Abo" .. (96l1.) 01 Ihe ínl. ........ saiu (a in« ""a ...s ían by pl. ...i nt! .ere ........ iR !he "I'!i ."-'.: - I nc:arl>y Won4 wlth a h.iJh popul'ation omI .........ty .,,,d"" upIaood.). Iw 1M,. ... duccd ca"'"", dltd>e• • b.mds. _ 1". ...... treo ttooio. bu dniJ>a&Mlozy utcd in Cebo ..< rOed '" ......, pof raid ~ cont. fTom Cebo "!V.nI .....a dt !he tec ........y lo ..l ec",. S_. tom and rice or c~ OD ~rer JOila. G do't'l'l' in Cl.t.mía wh" r.ad aIro.d~ atttmp. .... 'o re"""" pmIIkm. oí ..... ........., and «na. .... m.n, depIe.m ....d wh<> we. . k. .n IO....,;,r beiag r. ..c •• by !ud c1 ....l ied!anO' ha'" tm ~d ....... _ of .be r......"<. ...._ ,. hod u""¡ pe. ....... te'tOtlrces lO. taipc- agam h, anothcr #Ca• k~ bad noti«d $UII nttUieBt bent'rtu {ro1 aIk)'l Si" fa.rmers 4Irul tWo .¡ k re. .. arehen ~'" to CM' o( Ibe Cebu pro,fect ua'sfif1- ,u. 10li0-' ........e _· p".,:tíce ••; ,¡iI1 .wItch< F. duri:na: tite pre-c:ioppin. dry '-:UOn rOl fh~' tra.ining_ Thit i'D'\"OMd a.-po 1M";"""" 1M ben lb'.: A~{r.m.r fOl' ConluUf J.yout. bundiRg-ditdlilllll lar strip eMablishme:ru Ind 00"'1"'" pito. COm!""" pr_e,;"" is limllrd tbe • .....1 0. coolrol. ond .be pl_inl of hcd",.OWIJ or fodllor _ (...,Ie< ...... .,. ~"'''. lor bauIinS_. .r ) ...d of ",. ............. AloM plan, hft.JNHtltm J"lrpwr'rÍlm) and ~. .m e Ultra ~"'Mrr de c.taI). Glirieutt.r upiltrn). duA, in putuJe ar tlte t<,n, _ .... "". ." ,h i'ft' bid e_IN'"'' di_ti< ..,Iltc..,al had le" _Up mcIhods are ÍllCI"casinJly used in agricultural resean:h, clllension, can be effeclm foc temponuy aeú1lÍlÓCl; rol i.nstuce, a single meetiDg may mol III.bcr developmcmt acúviliC$ (Kumar 1987). In ParIS 1 a.ad 2 (lee be sufftc:ieat \O pass OD ideas fmm fumcn to researchers or 10 provide a '~OtIps' in irulex) we laW bow JIOOP meetinpc.an lIelp in eUc:iting farm"",' fUNID far ~bange of iBformation, as cIesc:ribed below in tbe cases of ideas and pnlblems, iD discovering IAd enbanc:ing lheir kllOwledge alld i_toe worksbops in Colombia and Baa¡ladesh. OtIIer groups _. ÍftnOVations a.ad in trials lOO lechnology development. This seetion aad vcoed al iotcf1lals can prOYide CODlinuity in, for cumple, monitoring triaIs, tbaie \O foJJow (3.4. 30S and 3.6) present furtber experience. discussin¡ problems aOO progrcss, or carrying 0111 self-sustainíng projecl 'Famer groups' refcJ:s \O groups COIIIIflOIiCd mainly of members of Ihe work. 1bese and other points are illustrated in tbe folJowing descriptions. rural eommunity, ilion¡ with 0De oe more agriculturaJ fC$Ca:rcben andlor e>:rensionislS. Tbere are maD}' (,pes. sizes aAd PUlJIOSCS of groups in agrialtural research actiWies and devdoping a Iypology is diffit:ult. Masl Gmups iR fieId bearings grrops ltave _Tal ñmclions. 'Worksbop' is uscd foe sorne groups usu.aIIy Balb. le. IA us ••, OIrt : 10 be taiIored 10 local pem:ptW1IS. Th.: best approach may be 10 asir. peopIe ,,,....,8 ...... : 10 draw tbcir viIIage and tllen build 00 tbe coovcolions for represellling its L • layoul whidl the)' adopto Obviously, lbe western COIIVCIlUon oC aJways 1Ioooin. ." , having 1'101111 at tbe top of !he map may be a llindraoce to Ullders1¡lnding in timIadt lit. gnuioo io_., I 1. ... '''. . r 6raDo. some cimunsta:DCell. Joinl map makillll with researebers aud ianners needs GrIZÍllll : 10 begin wim one o. lwo oomrnonIy agreed rel'erellCe poiDlS, the remaindcr , • DE the tnap being constnu:red 011 tbesc. (Il Patistan Gllr pr.ac:tice 1'R1sIo ... -P. .......¡ , ....h has bcen 10 s_.r- . I,P oor w..1-Ioiding ........ !, I loodin¡ ilOIIIIIJ:UI:t sketch maps from a hip vanfa&\: point, usiog ."".1' " tbis approach. Other uses of mapp.in¡ metbods by farmen are discltssed in seetioo 2.6. -.......... ...._ Ij - ,I , O¡:poUÓi. . ."", .._ OIIloJs.ili¡rioo T_ (see ftpte 2.1) have greater practica! tWlity titan maps, in our 1- Men IA l ... experience. They caa focus attenlioll 00 tbe difl'en:nt _ or miro>­ ... ...... .....-.... ,. .. : ....... emnromncnts in a watelllbed, viIIa,ge or (anu. In agr'OeCO$ystems _Iysis, I ....1 0 , I they are: draWD up by researehers ",ha walk from me bigbest 10 tbe 10_ I poiDt in lID enviroRment, atXlOlnpanied by toca! people, amsuIting people la eadI -. The malo purpose of tnInIIeCII is 10 ldelltify the major . problems ud oppo'luaitiea in !he aJlIble, tbey should bebasedoft the local caJendar. They 0.. .... J are WiCful 10 elUlblefanners to ideo.tify c:ritic:al times in, for cxample, lite aool.18l crop cyc:lc. JaleDdan bave long becn used iD Farmiog Systems , f 'M¡ AI V] " , I Al sllllll DI J I f ,-, A1 _1 JI J I ltc:seaJCh. ibcy canbe used 10 cover aD tbe major e..-en1S and chan," lbat _ witbin lile rulll year. The most obW:rus and important dimeusions I :; "I];_J I lIfC dill'late, crOflPÍ~ pattems, Ii\'e&tock <_=5 of forage and key eveots ~ I ~====== such as caMag, sales and migration), labour demand, die! AOd nutrition. díseaes and prices 'or crops, livestocl< and otbe! produce ud foc food. c..,. 1 ~,Bn>OtI B• •, . Climatic: data ma- often be avaiJable from ofñcial records, but famers' c-i>u r. ...... iiidliíI:-J e ! l'W!I perceptíons <:a:i be vaJid as well as indic:ating Ibe view of caDditioos on I I !he basis oí wblch !ley mate fann deQsions. In IIOrthem Pwstan, where , s¡,¡.- I I raimall data is larg:Iy a\)sent, we fouoo lbal scmi-structured ¡ntemews ¡:-- AMa'" 11. ....... can give relative ninfalls. Questioning goes approxiroately as follows: ~[ _ U] SIt."" [ &01"--e' Wbicb is the wctteslmoo!h? Whicll is tbe next _ttest? ud so 00, foIIowcd Clhi.. te"'. r. 1:0... by: Wbi<:h is tbe drest? Wbicb is tlle next driesl?, togetber witb campari­ ¡.¡,.. sons of montbs: Wliclt is Ibe weller (or drier) of these two? R.e!ative - 1 YoIcsI amoonts can be rou:bly gauged by asking camparisons or wetness. wbetber I .... 1 Ii__ ..... ... ooe month is tbrel-quarteB, a haJf or a quarter as wet as tbe _ttesl ~~I ... I ..,.. .. moolb. Relative anounlS are adequate for initiaJ dlagnoslic JlIIlJIOSCS, ..."... show:ing lile patten joto wbicb erops bavc 10 fit. It mar be enough: slmply - :r: - 10 constmct a seasooal calendar atoulld major events sucb as lite ~nning filA JI J J A S a N o J F JI A JI J J of tbe rains, periodl of drougbt. liBI frosts or lite levcl of inigation canal I - - IIow. Rural people's Inowtedge of climatic evenla can be detailed. In 3n J ~ Il{--;:--] [ interview in WoIb io Northem Etbiopia (ERe 1988), two !armen I I recallcd !be numbcr of days of rainfaD in ea monlh lar Ibe prevíans five yem:s. Their rec:allprobably pic:ked up raiIlI'aII wbicb was agriClllturaUy iignificaDI and IllAl therefore have had an agri<:ultural validity superior 10 tllal of normal raiKall records, la any case, their adtievemeal sbOW$ Ibe fLí-f---=-n-rl~-ll--r----- value 01 assuming ¡hat rvral pcople bavc detailed knowledge and astiog Ibem about it. _.... - 111- Agric:ultural Jab¡ur demand for wom:en alld for meo caD be I';liáted in a Il2- similar menner. asáng ftnt about lite busiest and the next bllsiest montbs [=, aod so 00, then tI:e least busy, and so oo. 10 West 8eDgal, lile resultiDg r--;;';¡';---J r------ bistograms far WOIIeD lIPd men have becn drawn 00 tbe groulld, pro~ -- a debate about ÚIl dift'erellt Iabour peab for _en alld mea and ttr.e cootinuous ~deroa:nd, pointed out by the _n, of thctr domestic: citares. fígw-e 2.2: Seasotwl adetUla, fo. 11 vj~ in M,them l'akisla,. Visually, 18 mmtb calendars are better !hao 12 montb calendars ror , .. . .. 80 81 • • • • re.ealillC scasonat pattems. CDnventional West~ c.aJe;ndars begin in January, but tbe Iooal cal:eodar may !>tan at some othcr time, or il may be WIlIIIIt . betlt:f lo start witb a By event b'ke tbe staft of tbe mios. ~.MIIIin. M• ••r e .. FIow ditlgr_ (see figure 2.3) can elicit and prcsent sequeoa:s sw:h as tbe Wour ~Ie Df prochlc1ioo and marlctíllC. Key aspccts can be DOt6d aJouc.side tlIe •• 1/oW. for ex;ampk!¡ labollr rcc¡uln!lJlmts and monelary COSú. 'I1Icse ÚIeIl • lobou. ' •• a_ beoome simple produe:UOD accounts. •• •• f~ ..! d .. Vt'1III Diqmnrs caD aloo be uscd, for eumple: for ItIIdenIaRdiIlC irlsci­ •• tuliooal relauonsllip' wilruo a village. Even m quite smaIl vilJa¡¡es tbe • ,•• Rwnber of different insútul:lons and actions invol~ ilI dedsioo'makiDg , caD be considerable. 1bese can be ideotified and diaplnmed at a meeting •• 25q.IIUQ. of vilJ.ts or of a partlalLu group. Veon diagrams, Imown in PBkistan .. •• 'chapan diagrams', use toudting ar averlapping drcles of various 1ÍltS. : ,• ,...fisD Each drcle represents an individual or iDstituoon and the si:!Ie of the cime índialtes importanc:e (wtIicb can be disaJSSCLI by 1M group undertaking tbe I ,I 3l1li0 dlya. lis 15 exercise). Tbe dreles can be usod lo indlaIte tbe degree of OOI\tact or ,I overIap in terms of lI!'IÍlIilI& at decisíous. Overlap oc:cun if Olle mslitution I 01 individual asfr;$ aootbet lo do sometbing, el if die, have ro oooperate in ,I I some way, -uing 10 1M followíAg COtIWfttion: •• •• • separate drcIes mean no contad; •• • lOucbing cirdes indicatc tllat infonnalÍoo Í$ cxclumged; •• • smal! overIaps pnínt 10 SOI1le oooperaúoo in decisioos; • • large over. mean considerable cooperation. 1= I DIqram$ 10 aid ana.Iysis IMptClÍolI BeyOlld their use 10 clicit informanoo, diagrams can be used by and with ~~ rural peopIe as an &id 10 analysis (see also pp 93-100). M05I of those de$cribed are general purposc tools tor ídentifying problems. constra:ínts, solutions and opportunities. Mast of them bave becn used io group EJ-~ ~-days.lIsl25 d~ssíon5 by teams oí reseatChers OI extensioo specialists or combi­ 10TAlCllSf; R$IiII n31ions o{ Ihese (Conway 1986) but bave r«enlly bun extended more to &id anaI)'Sis by rural pea"'" 1l>emliIclves, as tbe fuHowing eumples iIlustrale. "-lis 2 Seasonal diallrammiag can focus attention OD key seasonal oonsttainlS. In one viiI. in Pakistan, fOl example. systenllltic seasooal diagrammmg '1ie/1140mIs/t .... revealed that the perlad when dysentcry was rife was also the time of (larJkg1 barvestillC. posil\g a problem witb an agJWullural as well as a human aspect. Similarly, seasonal diagrammiag in Soutb WoIIo in Etbiopia foWld Jalfa, SALE • lis 3lDO Ihat tlle peale mootb fur malaria _ also tbe montb of higbeSl male aros labour Jeqoirernent, for land preparation. More positively, tAe analysis oí NETAETIJIIN lis 2740 seaaooal diagrlm5 can point lO opportullities, sucb » wIIcn new crops can be grown. ~ VeDn diagrams for village institutions can Ul1Illarly be used 10 gem:nte Figure 2.3: P,oduaitm cyck /01' sud JIOItlllXU and foeus analysis. We bave found tbat they can be 0IDISUllCIed very taSily .. by cutlÚlg out paper ardes of differellt m.:s, labellmg tlIcm with tlIe names 82 • í I • • . 83 ' • '''-''~ _.~,$, • Il 'Y)"~ __ ",~"",",,,,,.,,,,,,,,,,,,,_....,, ........ ",,,...,,._ of tbe institutioo or iadividuat aod ~ arraollÍO& IIlem on a tabIe in a pattem Ibal emerges from lile d""ssUm and ezperiencc qI: !he par­ r \kipaats. Once anan¡cd 10 everyone's saaisfattion, tbe ciJCles can be stapled inIO posiu- ud ~ 10 ideIIIify oee.ds roe improw:d linb. bctter oycrlap. 01" tbe positioniDc oí new institutions. The power ud utility of diagramming can, finaD)', be ¡,lustratcd frem a worbhop cmied out in lile PIIiIppincs whicb focused on a SIMIl cIam al tbe outlet of Latc &II¡ in BiooI province (Qmway aad Sajisc, 1986). FoJlowlog CODStnIcdon of tbe cIam a numbcr of 5eftR problCDIS arose, primanl)' affectiog tbe takesíde d_!len aod inhabituts of tbe mWlid­ paJi')' of BuIIi. 1'IIose ad\lenely affec:tcd bceamc IIDderstandably uJI")', ti;) lile poInl Ibal !be fature of lile projecl _ in jcopardy. In order ti;) ay ud taps, whicb varies and a guidc to lIWIy otber activities. and it is undoubled1y a YÍlal 1001 for witb altitude from 190 10 330 days. CaIculalions ~ malle from seoonda., _ny pw:poscs. However, ,.,hile reoognizing the importancc of mapping in &Ita. bul tbeII interYiews with farmers _re used 10 alter and re6ne tbe. its wnvendoaal forros, our discussioll instead _ Dl'W approaches. initl.al bouadaries. Farmers were wdI aware of the major dilfeR:nces we {UaiilY mIlPping Í$ a method we ba"e use: bave emerged. Women teod 10 draw ve., small forros, often centred on a lemple aod Ctnch.on rarely sItowing any means of IllIIlSpOrt. Men, by Cotltrllllt, rarel)' omit transport. Moreover. in the cate of a di}' village iD Mabara.slttra wbere a The potent:iaI of diagrams for cticiting Ibe knowtedge oí rural peopIe and studenl named Mandavlr.ar tried il, il was noticeable that whtle many poot for anaIysis by and witb lhem. is only j\lSt beginning 10 be realized. people drew (lB1)' their immediate neighhours and fields. tbe ridter people ProfessioDals ooncemed with rural development havc tended 10 suppose drew io far more dctail covering tbe whole village. tltat rural people. cspecially if they are iUiterate. will 1'101 be llble lO We are sliII arullysiog wbat may be pAed in undcrslanding people's understaJld or w;e diagrams. Our own experience has beco that their coglÚtive maps througb drawing, and _ do no! claim mw:b at presento capabilitiei almosl aJways exceed the expectatioDs of outsiders. 1be best Howe'lcr, we can recotUmend its use as an 'unfreczing devjce' al !be rule oC thumb seems 10 be 10 _ that tbey can uodeIStand and ose bepnning of famJers' Ot sclentists' workshops. 11 can generate tremendous diagrams uBli! proved o!herwise. creativity in group sessions, as bappened al a sessíon al Baogladesh For Ihe fulwe. tbere is scope for mucb inventiveness, by rural people Agricultural Research lnstitute (BARI). There paintiogs done by groups and by outsíders. in dellising and using th= aod other diagrams. expIoiting _re botd and satlrical. in contrasl lO tbe more reslraintd individual ones. tlle advaotages tbey bave o"er more conventional metbods oí iovestigalioo Wbat we bope 10 do. bowe\lCT, is lo refine the teclmique so as 10 beip WI and analysis. uodersland bew peopIe relate 10 their resource environment. Topographical maps by putoralists by Jo same places, 'liíllage people ba\'e beco cncouragcd 10 draw maps of a 2.6 Maps drawn farmers and extensionists more strit:tIy IOpOgrapbical klod. One agency in Mexico regulad)' asks people lo draw maps of their own villages. An instana in Wat Africa ANIL K. GUPTA ANO lOS WORKSHQP relates 10 Jcrerny Swift's woñ wilh Wodaabe Fularri pastoralislS ín Niger (Cbambers 1983:99). The reseatd!. team askedsome berders \0 draw mal"'. 0iffermI views ef reality wbím they did wúbout diflk:ully. The maps diffen:nlíaled liT'" aa:onIing 10 tbeit eallogical cbanlderiStics, as migbt baYe been e~. but also .. Mapping eIIn be viewed as OQC specific type of c:fia&ra:mming method. as indicaled several special 2JOIIC$. These were afCIIlI 'll'bere the herdcrs' eattle noted by Gordoll Conway in sectÍOD 2.5. It mipt appear surprisill8 10 gol nig/ll blindness from "ilamin A deficiellcy io the d" seasoo, for wbídl i t " AA • .. 87 .... '. TSBF TROPICAL SOIL BIOLOGY ·ANO FERTlLITV PROGRAMME • • 21 Dec 1992 TO: Participants in Slash and Dum Consortium FROM: Cheryl Palm While reading the following discusion paper on the biophysical aspects of site characterization please include any comments and addittions. 1 have tried lo emphasize measurements that are not as common. At the saIne time 1 have prbably shown the TSDF bias on biological measurements so 1 would appreciate help from those of your with other bÍaseS. Keep in mind severa! issues, coordination with other networks, national, regional and international and the measurements and methods they promote; the role of GIS and how lo interact from the beginning¡ current models of use lo the Slash and Dum Project; quality control and the need for any centrallabora~es. In particular, we need lo plan what site characterization and training activities the biophysical group coDSÍders most important for the .. workshop. In discussions with Dennis Garrity and Míke Swift with thought exercises in biomass measurements (above and belowground), meteorological measurements, and experiemtnal design and plot layout could be important. In your response please include ¡ .. other faclors or exercises you think: important. • , . Intemationat Unlon of Blological Sciences Unesco - Man and the Biosphere Programme 1 .. STANDARDIZED MEASURFMENTS AND METHODS FOR SITE CHARACTERIZATION AND MONITORING OF BIOPHYSICAL PARAMETERS • FOR THE SLASH AND BURN PROJECT Cheryl Palm, TSBF, Box 30677, Narrobi The specific objectives of fue biophysical research agenda for the Slash and Bum Project fall into three general activities; 1) Characterization of fue bíophysical environment of key deforestation areas, including regional GIS data bases which together will provide fue environmental framework: where research wíll be cooducted and have potential impacto 2) Evaluation of existing land-use systems and newly developed technologies wifu respect to fueir species composition and interactions, productivity, nutrients cyclíng and budgets and other aspects of sustainability. 3) Quantification of fue contribution of slash and bum agriculture and proposed altemative land-use systems to components of global change, particularly green honse gas emíssions and soil depletion and amelioration. Each of fuese .. three activities requires extensive measurem.ents, sorne of fuem overlapping between the activities but ofuers are quite specific to a particular objective. We must be certaín fuat fue • list of required measurements contains fuose fuat are essential to meeting the objectives but at fue same time fue list must not be overwhelmíng so as to discourage complete compliance by the participants. Mefuods used for the measurements must be standardized to provide compatible data for synfuesis and comparisons of relative productivity and sustainability of altemative land use systems and for extrapolation of fue more promísing altemative land use practices. Several questions must be addressed when selecting which variables to measure and what methods to use: 1. Where are fue measurements to be made, or in ofuer words, what is fue site? 2. Is the measurement for site characterization or for monitoring? 3. What are fue types of experiments to be cooducted at each site? 4. What is fue reason for makíng a particular measurement? Is fue measurement useful for understanding fue productivity and sustainability of fue systems? 5. What is fue cost and time effectiveness of fue measurements relative to fue information gaíned? Can the measurement be estimated other parameters? 6. What methods are fue majority of fue laboratories already using foe thé sPecified measurements? 2 • With regard to the first questíon, research proposed for fue Slash and Bum project occurs at several levels or scales, from !he experimental plot to a global framework. • Research will be focussed around fue 8 benchmark sites. Each benchmark site is comprised by a primary research statíon and possibily some related substatíons within the same country. The benchmark sites and substatíons will represent the basic agroecological-socioeconomic zones in which slash and bum agricuture is a pressing land use problem. These benchmark sites and stations will be characterized in a general sense with respect to climate, vegetatíon, soils, and land-use practice and this informatioo applied to georeferenced datasets. The actual experiments will be conducted both on on-station and on-farm plots. A more detalled site characterization will be required for fue on-statíoo experimental plots than the general description fur the benchmark sites but by the nature of oo-farm experimentation less detall will be required. For the proposed work to be relevant both at the small and larger scale ít is essential that the ateas or sites, at each level, are representative and that there are key links between the different levels. The data collected or required at each level is quite different but the data obtained at one level must be relevant and sufficient to feed necessary information into fue next level. • It is atso important lo know if the rneasurement is for site characterizatioo or foe long­ term ¡nonitoring of changes in experimental plots under various treatments. Site characterization usually includes those parameters which define fue general environment in terms of locatíon, climate, soils, vegetation, and even land-use history. These parameters set the conditions for fue potential prodnctívity of the slash and bum systems. Long-term monitorÍng involves repeated measurements of variables referred to as fue response variables, and includes the physical, chemical and biological variables of the vegetation and soils which change as a result of changes in land use. The response can be fast for some of the variables such as change in vegetation biomass but much slower in others such as soil organic matter; therefore !he frequency for measurÍng the variables wiIl vary according to fue systems under sludy and fue response times foe fue different variables. Sorne of fue parameters that define the site characteristics in a general sense are also response variables which must be measured in greater detall with time. Another type of measurement can be called the driving variables or those that cause changes in !he response variables. Driving variables are considered to be extemal to the system, causing changes within the system, and are often associated wifu land-use changes and management. For!he purposes of slash-and-bum agriculture examples 3 • would include clearing and buming the veg~tion, fertilization and tillage. No specific experiments have been proposed yet for any of the Slash and Bum siles • tberefore it is difficult to recommend specific measurements. The types of experiments envisioned cover a broad spectrum and would include crop variety and tree provenance trials; component interactions, such as crop-tree and crop-weed; nutrient cyclíng and nutrient budgets. The discussion paper includes a rather comprebensive üst of measurements tbat could apply lO many different types of experiments. It is nol expected tbat all measurements be made at all siles but tbat subsets of data be taken depending on tbe objectives of each particular experimento There must be sufficient reason for cboosing a particular measurement lo be included for site characterization or monitoring studies. The reason for site characterization are more straightforward and relate to tbe climate, soll, and vegetation features tbat are relevant lo tbe production potential of an afea. Variables to measure for monitoring studies are not so obvious because of tbe diversity and complexity of tbe studies. A guideline lO follow when choosing measurements could be biophysical aspects of sustainability, namely production • trends and environmental impact. Examples for tbe production side would be trends in crop production, soil fertiüty, weed biomass and composition, and nutrient budgets. From tbe envil:onmental context, examples would include emission of greenhouse gases, 10S8 of soil organic matter as C~ and other forms of nutrient loss from tbe system, including erosion and leaching. Given tbe aboye considerations tbis discussion paper presents a recommended list of measurements according to two critera: where tbey are lo be taken - benchmark areas, on­ station oc on-farm experimental plots and whether tbey are for site characterization or monitoring studies, or both. Not all measurements are relevant lo all experiments. Specific methods are proposed for tbese measurements. The details oC tbe metbods are nol included in tbis paper, tbey will be provided once tbe suite of measurements has been decided. Many of tbe methods are fiunlliar lO agronomic research and are already used by many of the nationallaboratories and research centers; detalls for these metbods can be Cound in tbe American Society of Agronomy publications - Methods oC Soil Analysis, Parts 1 and • 2 or tbe Tropical Soil Biology and Fertility Handbook ofMethods. Otber metbods, however, .. are derived more from ecoiOgical reSearch and may nol be as wldely Imown or utilized . More detall of tbese metbodolgies and a rationale for tbeir selection will be presented for 4 those that are not so widely used. The TSBp·Handbook has been used as a guideline for the • measurements and should be consulted for background information and the details of the • methods. Particular emphasis has been placed on the soil and vegetation varilables. It is important these variables fall into two categories: pools (standing stocks) and processes (fluxes in and out of the system and between the pools). Both site characterization and monitoring include both pool and process variables. Monitoring studies which follow the changes in pool sizes with time provide valuable information about changes in the systems. Studies which also measure the changes in processes provide a better understanding of why the system is changing. To illustrate the distinction, a monitoring study that measures the pool of soil organic matter will indicate if the content is changing or remaining the sarne. If this information is coupled with process variables such as decompostion rates and litterfall or residue applications we will be understand more about why it is changing. This process information provides the basis for the design of more efficient management of systems. Sorne of the fluxes that are important in slash and bum systems are presented in Table 1, methods for measuring sorne of them will be discussed in the following texto • It is our intent to present the measurements and methods as a starting point for discjlssion, we hope to reach on consensus in the following drafts and workshops. A comprehensive list of measurements is presented below in separate categories of location, climate, land-use, soils, and vegetation. Those variables which are considered requisites for site characterization at the benchmark area, on-station, or on-farm sites are highlighted. Additional mearsurements that are important for characterizing and monitoring specific components of experiments are also presented. LOCATION The exact location of a site, with ~pect to longitude, latitude and altitude, is important for providing information for geographically referenced data banks. Additional information such as landscape position (alluvial terrace, lowland, hillside or ridge), slope, aspect and geology or parent material are important for defining the site. The benchmark areas would ideally provide information of the percent of the area they represent that is in • the different landscape position, slope categories (0-5, 5-10, 10-20,20-40% and aboye), and .. geologicai zones. The actual area to be represented by the characterization of these 5 • benehmark sites must be decided and coordiriated with the georeferenced information. For both on-station and on-farro plots it is important to record all the information in Table l for • site eharaeterization purposes. Table 1. Variables required for site eharaeterization of the benehmark areas, on-station and on-farm plots Location Benehmark Area On-station and on-farro plots Variable Latitude and general area represented by the site Exaet location, (Global position longitude locators (GPL, ego Loran, Mage1lan) Altitude percent of area in various altitude Altimeter categories Landscape percent of area in different landscape Description position categories Slope percent of area in different slope Inelinometer (%) " categories Geologyl percent of area in different Description of geology/parent • Parent geological zones material of exaet site Material CLIMATE VARIABLES The elimatie data required for site eharaeterization and monitoring and the methods for obtaining them are provided in great detail in the TSBF Handbook, p. 8, and the mSRAM Technical Notes No. 1 on Site Selection and Charaeterization, p. 53-88, and are based on the WMO guidelines. The measurements are features of most agrometerological stations. Many experimental stations are already equipped with meterorlogical stations, for those that are not we would recommend recording stations with data loggers. The measurements, methods, purpose (site eharaeterization or monitoring) and location (benehmark area, on-station plots, on-farro plots) are presented in Table 2. For the benehmark areas longterm data on mean monthly rainfall, and maximum and mínimum daily air temperature should be compiled for the site eharaeterization. The number of years the • records cover should be reported and is preferrably more than 10 years. More detai1ed daily .. monitoring of rainfall, temperature (air and soil), evaporation, radiation, wind and humidity are necessary for on-station experiments. The measurements should be taken as near lo the 6 experimental plots as possible and if it suspected that rainfall is different at the plots then additional gauges should be instaI1ed near the plots. Most of the data obtained on-station can • be used for the on-farm experimental sires, but again it is recommended that daily rainfaIl be measured for the on-farm plots. Table 2. Method location ond f of a.easurement of cl1.tfc variables. Clilhatic Method ..........k Area on"statíon en'fana Variable Plots Ra¡nfaU coUeeted daUy, receptlve orea of longt .... recordsE daHy, 11ft dally, ... gauge ahould be 20() cia' ...., .....t hly ""'" D (""..,l Te ft1)ereture daHy .... ond ,.ln, ooeasured 1.2-2.0 11 lq:tertl reeords, daUy "'"" aboYe the grOU"ld ...., ooonthly me. end min -- - ond .in ("el Sou r..., dally ..." ,.,in at O. lO .. clepth .. .. .. .. .. for specifle studies in -- - relovant treatments Radiation C80Ilbell ......,Ine record. . .. - .. - .. daH" - - . Evaporotion elas. A DOn .. .. - .... daHv ... I/ind .up ..........t er .. .. - - .. daHv . -- • Hunídítv As""""", ""vcl". ...t •• ...... - - daHv - - . • LAND-USE A fairly detailed description of the land-use of the area under study is important for detetmining potential points of intervention for improving the productivity and sustainability of the current agricultural system. One must distinquisb between a general description Of the land use of the area being studied and the actualland-use of the experlemental site. Both are important but provide different types of information. The former is for characterization of land use pattems in the bench mark areas and provides information the percentage of the area under different land-use management. The latter, past land use of a site used for experimental purposes, is important for un~erstanding the immediate factors, or driving variables, that will had an impact on the current status of the vegetataion and soll. Current land use oould be oonsidered the experimental treatments that are installed. The type of information needed is presented in TabIe 3. Much of the data for site characterization of the benchmark area will be obtained from interview surveys and others seoondary information available and will overlap with the information requested by the • socioeoonomic survey. Quantification of as many of the driving variables, such as fertiJ.i7;er use, stocking cates and crop and fallow length, rather than a qualitative description, is most .. 7 ¡ ., useful and wiIl aid in understanding and predicting the magnitude of the response variables. Of particular importance for determining the productivity of ateaS under different land • use and the impact of ateaS in slash-and bum agrlculture to the emission of greenhouse gases ís an indication of the amount and age of the vegetation cleared and the frequency of clearing a single area. For example, is an area c1eared once and then abandoned to íorest or put to another use or is the land cropped or used until it is degraded? Other important factors influenciog the maintenance or decline of productivity of slash and bum practice in the area are the length of the cropping cyc1e and subsequent fallow period, and an indication if the fallow period is decreasing andlor cropping perlod increasing. rabIe 3. The dlfferent _ts of IIII"d uoe that are IlIf'Ortant for orea lII"d slte characterf.atlon. (adapted f rono theT SlIf H_ D. 20 ) LIII"d uoe INFORMATIOH REQUIREO ARfA (AC) OH SITE Variable CHARACTERIZATIOH (SC) .. Type of IIII"d-use: pastures, % arH in e&eh category or class (TSSf AC croDS. Dlantations H_ D.20) Clearing: and buming ·Area cleared pe" yea,., type of land AC eleared (undisturbed forest, tallow (age), grassland • ·Clearing and ilumino method Management, type 01 use (crops. *erop/fallow ratio (years) AC ancl SC pe.tures. plantations, etc) *crop and ttee varieties and plant $pecina, *rotations and intercropping . *,rOl.ft!l covers *weeds ancI weed .c:ontrol *tillage, irrigation *8"i ..b , stocking rates ·"".turea olant"ImlOS i ri on - Input. *types (feni l izers, ·'-nures. .. pr. .... h,.gs, AC ancI SC .... lches) *quanti ties *timing lII"d method of .rolle.tl"" Dutputs *harvest prock.acts (fruits, grains, AC ancI se straw, etc.) SOIL VARIABLES Site Characterization - There are severa! soil variables that are essential for site characterization for most types oí cxperiments envisioned. TIte soil chemical and physical properties, along with the climate, will dictate the approprlate crop varieties and tree provenances for the area by setting the fertility and hydrologic potential and constraints for • growth and production. We recommend that the 2 or 3 dominant soil types for the .. benchmark area be classified according to Soil Taxonomy, to the family leve! . Soil c1assification according to Soil Taxonomy is quite detailed and involves sorne 8 chemical and mineralogical procedures not available to most laboratories. The Soil • Management Support Services (SMSS) oC US AID can ofien assist in the anaIysis and • c1assification of soils. Sorne of the sites h¡lVe undoubtedly been anaIyzed and classified accordingly. lt is not necessary to classify the soils oC the experimental plots but it can usually be ascertained from the informatioo oC the soils in the benchmark arcas. There are, however, severa! variables that are important for site cbaracterization of the oo-statioo and 00-farm plots (Tables 4A, 4B). These include particle sire anaIysis (texture), soil available water, from the physical aspects; and pH, ECEC (effective cation exchange capacity), %base saturation, N mineralization potential as measured by anaerobic mineralization, and P­ fixation, measured vía P sorption isotherms, from the cbemical aspects. Another measurement that could be added to the list that is not normally considered part of site characterization could be a test for the limiting nutrient(s) to production. Many experiments willlook at various treatments for maintaining or improving crop yields and soil • fertility, knowing the limiting nutrient aids in selecting treatments and understanding results. Measurements for monitoring fertility. nutrient and catbon <1ynamics In addition to site characterization there are varibles that are important for longterm monitoring studies (Tables 4A, 4B). This section presents sorne of those measurements. The depth and fiequency of sampling depends on the measurement and experimento BuIk density must also be sampled for experiments concemed with changes in the total stocks of nutrients in the system. All of these factors should be discussed as they apply to different types of studies. Nitrogen, as well as, pbosphorus are the primary limiting nutrients in tropical soils. Whereas laboratory methods for evaluating fue available nutrient supply and fertility status for fue basic catioos are quite reliable, there is no widely accepted method for estimating nitrogen and pbosphorus availability. Total N and P are generally large pools that do not change significantly even Ove! a few years. Inorganic N levels are extremely variable spatially and temporally and are not ususally correlated with crop productuction except perhaps under fertilized conditions. Certain 'metbods fur measuring available P do províde a guide for critical levels for sorne crops, but does not work for tree productioo. • Obviously, sorne measure of the more readily available portions of soil N and P are needed. Even thougb there is still no CODeensus we recommend in situ N mineralization and the 9 Hedley et al. (1982) P fractionation scheme, for means of assessing changes in N and P availabilities . • N-mineraljzation Mineralization is tbe process by which soil organic matter and added organic materials decompose and release inorganic N. The rate depends on many factors including, rainfall and temperature, tbe total soil N content as well as its composition. The amount and type of organic materials added to tbe soil will also alter tbe net mineralization rate and pattem. Severa! metbods have been proposed for estimating mineralization and tbey all have particular problems (see TSBF Handbook p 162-165) but tbey can províde relative indices for nitrogen availability. The TSBF group adopted tbe in-situ undisturbed core method outlined by Raison et al (1987). Paired pvc tubes are driven inlO tbe topsoil; tbe soíl from one tube is extracted immediately fur inorganic N while tbe otber tube is left, covered, in tbe soil for 14 days lO one montb before removíng and extracting for inorganic N. The difference in inorganic N • between tbe two times is considered to be net mineralization. In tbe wetter sites, tbe method is not as reliable as in tbe dries areas, partially dne to tbe spatial variability of tbe process but also due to metbodologica1 problems of compaction and water saturation. Despite tbese problems we encourage furtber attempts at estimating in situ mineralization by use of tubes witb.a larger diameter to avoid compaction and by sampling intensively in a small area. Intensive sequential sampling for mineralization can provide an estimate for tbe amount oí N made available to plants during the coorse of crop growth or during the year. This however is extremely time consuming and not very practica1; ideally mineralization cates from soil organic matter could be modelled witb information on the effect of soíl temperature and moisture, such an exercise could be important for determining the inherent mineralization rates oí tbe vanous soíl·dimate zones representative of tbe slasb and bum areas. Simulataneous anearobic or aerobic incubations carried out in tbe lab wonld help lO establish a relationship íor the field and lab measurements at different sites. Another use of intensive sampling of mineralization is to determine tbe nitrogen availablity patterns following applications of organic materials of varying chemica1 characteristics and decompositions rates. Coupled witb iníormation on changes in tbe light • fraction (see below) could aid in tbe selection of materials and timing oC application to synchronize nument availability witb plant demand and minimizes losses oC numents from tbe system. 10 .. Phosphorus frnctionation - Estimating pbospborus availability in tropical soils is problematic partially because of Ibe high P-fixation capacity of many of the soils with high • Fe and Al oxide or allophane contents. Much of the P in weathered tropical soils exists in recalcitrant pools and this P is generally considered ro be unavailable. Recent studies have shown that much of that recalcitrant P is utilized and that most of the P utilized comes from organic pools, rather than inorganic pools as in less weathered soils. Soil test data for P can show little or no change with time despite crop harvest removals and continued tree and croP growth, this a1so indicates that the available P is being replenished from the other pools. Certain pools of P that are not measured in most soil tests are apparently a readily available source Of Pi for plant uptake. A scheme developed by Hedley et al (1982) separates P inro different fractions using sequential extractants: resin, bicarbonate, NaOH, sonication with NaOH, HCl, and a residual P extracted by H2SO. and H20z. The different fractions are also separated inro inorganic and organic P. While this fractionation method is laborious and time consuming we believe it merits further consideration. By looking at the changes in the different pools with time in different land use practices we may be able ro better understand P dynamics and the controls • on P availablity in different soils. It may be that onIy certain pools in the frnctionation schemc¡ will be crucial for assessing changes in P availability. The dependence on organic inputs in slash and bum agriculture stresses ~e importance of investigating this pool more thoroughly. Olber candidate methods for looking at changes in soil P would be separating total soil P inro inorganic and organic forms or more by a two step extraction, acid followed by base, detailed by Bowman (1989). Another method that has been developed in New Zea1and that appears promising for a soil test for trees is cumulative P extracted by consequetive extractions with bicarbonate. Soil orpnic matter: C. N. and P In recent years there has been renewed interest in soH organic matter research because 1) the importance of soil organic matter in maintaining the productivity of low-input agricultura! systems such as slash and bum agriculture, and 2) the contribution of the 1088 of soH organic matter ro carbon dioxide and other greenhouse gas emissions. Despite its • importance, there is no 'soil test' for looking at soil organíc matter as it affects productivity. Current efforts are being made ro measure more active or labile fractions of soíl organic C, • 11 N, and to some extent P, and to see if there is a correlation witlt produetivity. Mierobial biomass (C, N, and P), soluble C, and anaerobic N mineralization are • suggested as estimates for a fast tumover pool titat would be relevant to nument avaiJability. There is not yet conclusive evic!enee showing a general relationship witlt any of tltese measurements and productivity. Part of tite reason is tltat tite correet types of experiment have not been conducted; much of tite time has been spent on metltods development. TIte Slash and Buro project will provide tite types of experiments needed to test tite value of tite various measurements. Mjcrobjal biomass TIte method recommended for microbial biomass C and N is tite chloroform fumigation-direct exlr.lction method of Vanee el al. (1987). This method works better en aeid soils titan tite original fumigation ineubation procedure and it is not sensitive to recent organic additions. TIte method is best used when comparing diflerent treatments en tite same soil but is probably not good fcn comparing microbial biomass on drastically different soils, altltough tite trends may be similar. TIte fumigation time may vary with soils. sandy soils probably onIy require ene day of fumigation whereas clayey or wet soils may • require three to five days (this should be ehecked al each site). Fumigated and unfumigated soils are extIacted with KaSO. and tite extract analyzed for C and N (TSBF HANDBOOK). TIte lj.II1ount extraeted from tite onfumigated soil can be considered an estimate of soluble C and N and the diflerenee between tite amount extIacted from tite fumigated and unfumigated soils is reIated to the microbial C and N, factors have been caleulated to convert tite values to biomass. There is evic!enee that short-term changes in microbial biomass refl.ect long-term trends in soil organic matter contento Ligbt fraetíon TIte light fractien consists of tite soil organic matter tltat is not bound to day minerals and can makeup 5 to 50% of tite soo organic matter. It is basically undecomposed or partially decomposed organic material in tite soil and is tlterefore distinetívely diflerent from tite clay bound, numified soil organic matter. Knowing tite % of tite toJal soil organic matter titat is in this light fraction may serve as a means of breaking down tite toJal into funetíonally pools. Preliminary studies indicate that tite ligbt fraetion may represent a pool of slower turnover time titan microbial biomass. Measuring the ehanges in • light fraetíon over time witlt changes in land use and management may provide a sorne ., correJatien to production trends and ultirnately ¡ead to a soil test parameter . Several methods have been proposed for measuring tite light fraetion; these methods 12 cover a broad range in technical detall and equiupment, from simply swirling lhe soil in • water and collecting !he material that floats to more complicated sonication and density • fractionation procedures. The method we recommend involves sieving lhe soil through a 2mm sieve, and lhem dispersing lhe soil in djlute (O.5M) sodium hexametaphosphate for 16 hours and then filtering through a 0.53um sieve. The material trapped on !he sieve is oven- dried and can be analysed fur e and N, !he fraction cannot be analyzed for P because !he dispersant contains P. Perhaps by modifying !he method and dispersing with NaOH P can also be measured. The light fraction must be ashed to correet fur lhe amount of lhe mineral soil adhered. This method may not work for well aggregated soils, so sonication for a specified time may be reqnired. SOILFAUNA Soil fauna are important foí forming and maintaining soil structure and also in the decomposition and mineralization of organic inputs and soil organic matter. Despite these important functions there has been Iittle emphasis on characterizing soil fauna. Changes in • land use result in drastic reductions in lhe number and types of soil fauna which can have an impact on both soil structure and fertility. Methods for collection, separating, and classifying soil fauna are presented in lhe TSBF Handbook. Experiments proposed by !he slasq and bum should place empbasis on !he characterization and function of soil fauna. It may be necessary to assign a group oC experts to coordinate this task given the speciality of !he subject matter. VEGETATION VARIABLF.S Measurement oC lhe standing stocks and cates of accumulation of biomass and numents of the vegetation, both aboye and belowground; and the amount, chemical composition, and decomposition rates of organic inputs are important for understanding the initial stocks of lhe site and the vegetation dynamics and nument fluxes of !he system and fur following !he production and nument accumulation of lhe replacement systems. The methods described beloware used for bolh site characterization and monitoring of vegetarion, the measurements will not be required fur all type oC experiments (Table 5). BIOMASS ANO NVTRIENI CONIENT Quantification of!he biomass and nument content of the vegetation that is cleared for slash-and-bum agriculture is necessary for understanding • 14 • • son.VAlt.lA8l.É ~ U>III'$ NarES CLASStnCATtON' BoiIT~.""", __ I ·. ... Jllll'lQ:wb~ __ ~ ' M ~" wilhSMS$ PH-\:"SIC-AL l'ItOI'KJfI1I8 _ .-.... .. "" __ rrstIF_.", _ ....... .-...-..... ... _NtOH.~.-d ~ to:6.,..¡;th. bicJt Fe .. Al __ «&cal. (EMBRAPA} ....... ~ OIltoiltyp:., ~ (fSBF tIWboot. ".51- &1=1 Sl¡ -........-... - Doubk tille (TS8F, p5); """" ! _ hDW ~ &14 g¡pcity iIUi.!J:!'(TS8F ...'" H.aibooko p.J»'" ~ wikIIat f'QiMl (i~ ¡ ~_TS8Fp:l);¡ASAI~"""'b.n - ~ la * fOOCinc ""'" ............ CHEMIC'.AL PltOl'Bmu oH 1:2.5 ~ (I'UF p.l1} .. .. CARBON ·T«t.l ~ ~ (Wf W.J&y IllJl«witb ~ &es« (l'S8f' p. 3$) " _. I Li&I<"""" Soil CIpIJic.....ntl ~ ~. -.1 -.. w mo<:J, ¡ I ;1 .. ..,. ... wod: ~wilh4ihM! ¡ • l' h-U......,1Oilt EIlloa. .~- 1m¡ """ ~ ____ .... C Fl&I:liplb diflllu;l ~ CY_ el al. 1981) NIl'J.OGEN * T« al - """'..... """ ...... ..". ..... -. ...... .". N liatb poMlW ~ ..... ('IDF .. ~. ASAl p.1l1) -... Ni ..... 'irtE !u!:!e~ _ (TSBF. ,14) .,...... ' . .. _ ..... ~bDa ... _ ..... 1o .-.....*... i~_~~ PHOASI'H_Oti IS • TOIaI , """'....." .. - ~ CIICtnclIWc pUl.:! "". til~flll'lllNtbMfor. __ (fS8F p. :.; MAl. GI) ..... ot.,¡JpH F~ twk>r« .. (1912) ......... Na"" KMIpntIh {1m, Aa ~ oll4iutiaa. GIpdty;: ~ llD1y k dta)ooy, ..... or MSwiIh~~_pHia XaF >10, EXCHANGE..\Bt..E CAT'IOMS: ~ wiIh ................. ph 1 (fStW p.J6) ot ~~&tpH1F-1Il c.. ... " ea.J4.íNxa ... Kt.y~ -' .... • I:imdr. el , ,UH-.tAl Ú' J' KiIIl&y tAJ""fIIII H") ~ .1N Ka (t'SBF p. 31) -'. ... ._.._. M_ _____ ~ pH~7 "" __ aQt __ C'EC~. J:.fIodiw. CIlio6 ~ ~ ~+cdolt+-«+caAL+cxH {IDlC} -' .... ~tcIII'IIMI; .. pH1 $"-~~- ~_"""""""'OO; ~'" • " LlMmJ2mm¡ Iiw"'" .- -*t I11III lIIXIOIdiDc lo ... typo. .H.1a.ri,rd.. 0.0.II.I.aI.l .e.l.' "- ..... -_-__... ... .... bo CICIfIIOCIOII ror tbD miacta! lKIi..I i i "'-. ""'""""" -T __- ~- (TS8F p. (O), toCM --.s l1li......." 1!IpOItDII '" 10 ..... F« __ ....... pcrcaaiaJ. l)'IIIIIIIt. lOCJl coria& COIIIbiBaI wiIb lbb.otn:uI (fSBF p. fi6.67); for crcp or ...J. .,...... by lOd. ~ .. --. or DIlaÍl. • • 22 • VAIIlABU! "do la l'dIIlM .. ~ fIar.. 1IdIdwlI_ o/.~ 1-_ 1 -.-ibIa la...,. ..-w ~ 3 - dUIk:uk aad lbouId _ ... by poup EXTElUfAL -lNPU,l'S ........... ..... 1 ......... ~tID""'-"kd 1 Muleta, _= .. ...-. ot cad!aa ..... 1 (ftuID lUPIo pbI) SYSTEMS LOSSES _---.101 ..... 1 RID':Il!'. auioa dcp:ada ce .. m.t .... _ - 2 ...--. ...... -ay..u .. _ .. 1up if.1ot 01 ........ _ 2 lIIfdDd 10 lb. .,..... ""'"" , _._. .JlIIu.mI -d-a.l¡. aolt/o. --'''~'''''''' ~ if ____ ,.. • INI"EIlNAL CYCUNO .... ............. ---- 1 . J..itaR&B. -. _idUD. pnmiap ..... 1 • .. D . 1 ___ lo .... dc:paII al ... lo.,..... 2 ~ 2 .... - .......... ".. '- 1.2 • • A database of research activities for the Slash and Burn Project. • Richard Coe lCRAF, Nairobi ]6 December, 1992 1. The Need. The Glohal Alternatives to Slash and Burn initiative is a proposed global research • activity. Current plans are for suhstantial field research to he conducted at 1\ si tes in 3 cOl1tinents. Research will also take place at salellite sites, as well as al other centres concentrating on regional and global synthesis of results. Currently sume • 17 collahorating institutes have be en identified. If Ibis research effor! is really to be a coordinated global project and not a series of isolated research activities, there has lo be a well establisbed system for information exchange belween all interesled groups. Participants in the project worldwide need lo be kept informed of whal is happening. what datl¡ has bee n generated and is available and what the key results are. One way of achieving this is through a computer based datahase. There are many possible structures fm such a database. This paper outlines (lne strueture, summarising the OUlput needs it could satisfy, lhe input requiremel1ts ano sorne of the steps needed for implementation. Some alternative database structllr":S are also described. This paper is not a technical specification ror any datahase. 2. Inputs, Outputs and Constraints OIll¡J/I/.\ The ¡¡¡¡ture of outputs which must be available from the uatabase can be describcd by listing example requirements of typical users. Table L • • 1 Table 1 Typical User Example Output required • Field researcher New initiatives al other si¡es. Modificalions to design and methods al other si tes. Key results from all locatiOIlS. New Collaborator What is going on where in his field (e.g. nutrient cycling. economic sllrveys .. ). DOllor Lis! of research activitics and major achievements without tcchnical details. Reviewer Activities and results in lhe area being reviewed (geographícal ur suhject hased). • The outputs required will range from brief su mm aries (Where is improvl!d ¡all"" research going on? What fmil tree species are being s!lluied?) to the Iliglih lechnical (Are the same i,otope techniques being used for l' work eve[)\\ h,;re'l). • lnpu/s The outputs described aboye will be obtainable if the dala input consiste oí a lis: of research activities. with such details as: Project name Location (with biophysical and socio-econornic characterisation) Peopleflnslitutes involved. Reasons for Ihe research (where does il fit in lbe whole frarnework) Objectives Methods Data generated Results. COf!.\lmints The t"" major conslraints which make Ihe dataoase pOJentially difficllll to 'él Uf' and maintain are lhe geographical distrihution of hotlJ use!'s ami source:; of (h,,;\. and Ihe need for continllollS updating. as new ac¡ivitics are initiated ¡¡r:eI resl:ll' generaled. '. 'rhe dalabase will only be useflll if il provides a user fricndly way for infofmatlllJ: 10 be exchanged between Ihe 17 collaborating inSI'!llleS, Each inst:lul.: 1\ il. 2 r generate information thal mus! be incorporated into lhe database, and each institule will need lO aceess information originating al other centres. • The value of informalion in the database goes down quickly with its age. f it is not kept up 10 date ir soon becomes worthless. Om: of the functiol1s "f lhe database will be to keep collaborators informed of la[e:,[ developmen!s. F, ,1' ¡his reaSO(1 it is essential to structure the database so il call he continuously up :atcd. 3. A possible structure The structure outlined below is based on ICRAF's dalahase of research :lel: ,ities. which is still heing developed. Dala flow and updating Figure 1 shows the flow of data. Each ,ite at which information is .. generated (i.e. where research aclivities are planned and carried out) has a copy of the data capture software. Informarion on • research is entered, building up a local database whieh describes activities al that site. The formal ami structure of lhe information is controlled to ensure uniformity belween si tes. A copy of the local database is then sen! to the coordinating sÍle. ThE local datahases from all sites are compiled into a central dalabase, which wiil e ,ntair, information on all Slash and Burn project research activities. The COl11p iatioL process will ensure thar information fmm differen! ;;¡tes is compatible. A copy or image of the central database can lhen be dislrihuted to anyoIl ~ who needs acees:; 10 the information. This will inclllde all research sites th lt art' generating data as well as other interested parties, 'lIch as project coordi: ators. donors and potential new collaborators. The image database can be search, d and the necessary Olltput reporls produced byeach user. The database has to be kept up to date. This can no! be done by individua users modifying the data in their image database as thi;; would reslIlt in differen·. llscrs having differen! versions. The data in the image therefore can no¡ be m( difie(; by uscrs. Inslead users lIpdate their local database. The new local data 1 ase' ¡s • scm to the coordinating site, whcre it is compilcd alld a new image pro :uccd The llCW image is then distríbuled 10 all users. TiJe: frcljuency with whi. h t:li, • needs doing depends on how rapidly new initiatives, changes in desig 1 anc methods and important results are generated. Al ¡he start and end () eQcl 3 growlIlg season would he reasonahle. These transfers of data hetween the coordinating and remote si tes could in • principie be done by electronic data transfer (E-maíl) and even done automatícally. The prohlem of telecommunícations with some sites and the: iarge volumes of data involved are likely to make transfer hy diskette more jlr;c.ctícal. Data ítems Wíthin the datahase lhere are records of several types. One type is "Experiment". contaíning fields that describe such things as the ]ocation, ohjectives, mcthods, design, measurcments etc. Another record type is "Location". The fields of this record type give the characterisation of the researeh !oeations (clima te, soil" laml use elc.). This si te characterisatiQn component of lhe data base is valuablc ¡tselr, gívíng a description of aJl sites in which the Slash and Burn Project is operatíng. The power of this structure líes in the link betwecll the two types of reeorus. A user of the databa.c;e needs to know the character"tícs of the si te at which an aetivity ís going on. However a researcher entering informalion does not want lO have to enter lhe same characteristics data for ead) ,"xperiment at that localion. By having linked records of characteristics and experiments, soil and climalé data • is tied 10 details of each experíment, yet only has to be input once per ,íte. Other record types can be defined according lO rango On-line work on a remote computcr is therefore very expensive. The situat¡¡¡li i, nlll likely to change during the lifetime (lf this project. CO/lll'lcxity of ¡he database • There is a wide spectrum of possible contento of the database. Simpler defínitions would have less technical detaí1 011 each research activíty. More complex definitions could encompass a wider range of researeh types and include raw data. In general Ihe value or Ulilíly of ¡he database will increase with inereasing complexity, hut with diminishing returns. Tlle cost of development and rnaintenanee will go up exponentially. As an example, consider the problcm of including raw data in lhe database. The added value of ¡ha! informatíon arises beeause it / means researchers working on UI:lily ...... regional or global synthesis of , ~--" - results have easy aceess to it. Cosl While slIch data exchange is essential for synthesis, the numher of people who need aceess to raw Complexlty data is limíted. Most can gel requírcd results from surnmaries. Figure 2 CasI effeclíveness af database. Howcver the cost of including raw data in Ihe database is very high, because of the c