AN ECOREGIONAL APPROACH TO ENHANCING AGRICULTURAL RESEARCH IN TROPICAL AMERICA al te naco otdeAg e tt oTopcol 1 ta ato oiC te f T p oiAglc lt e --"'\ ~1 1 ) ' y May 1995 To Cent o lnte nacional de Ag lculfuro T oplcal tntematlonal Center for Tropical AgrlcuHu e Chatrman T AC Members of T AC CG Secretanat TACS~fuL From R~Pv~Ofr? Teléfono (57 2)445-0000 Fax !57 21445-0273 Co eo elect ónlco (E-mal/) CIATOCGNET COM 0Jeccl6 postol(Mai/lngoddess)"AA 6713 Call Colombkl SubJect Ecoregtonal Proposal for Troptcal Amenca Date May 30 1995 Attached for constderatton at T AC67 please fmd a revtsed proposal for ecoregtonal research m Troptcal Amenca that atms to respond to TAC s earher suggesttons and comments In order that the proposal remam a coherent document on tts own substanttal matenal from earher drafts has been conserved However there have been substanttal revtstons to whtch 1 w1sh to call your attentton A preface has been added that descnbes the consultat10n process that has been followed and the evolutton of the proposal m response to T AC s comments A new mtroductory sectton on the context of ecoregtonal research has been mcorporated The management and coordmatton structure of the proJect has been greatly stmphfted and the posltlon of full ttme coordmator has been deleted A new sectton on managmg ecoregtonal research has also been mtroduced The mterests and commttments of CGIAR centers to ecoregtonal research are detatled m an appendtx that has been revtewed by the respective center dtrectors Fmally a workplan and budget for 1995 are spectfied and an tndtcattve budget for 1996 97 ts also mcluded The tndicattve budget pomts out resources that are anttctpated based on advanced d1scusstons wtth donors Great efforts have been made to msure that the revtsed proposal reflects the valuable feedback recetved from T AC and our partners We contmue to look forward to your support m thts endeavor wh1ch we beheve can stgmficantly contnbute to tmproved ecoregtonal research m Troptcal Amenca We welcome your comments ce Dtrectors General CIMMYT CIP ICRAF IFPRI ILRI IPGRI PREFACE Evolubon ancl ConsultatJon Process d Propcstl to Enhance Ecoreg1011al Research m LatJn Amenca In arder to des1gn an ecoreg1onal program that responds to the needs of stakeholders 1n Latm Amenca an extens1ve consultat1on process has proVIded essent1al 1nput to the development of th1s proposal In response to the call for proposals for cons1derat1on at TAC65 CIAT convened at short notlce key reg1onal partners toa consultat1ve meeting 1n San Jase Costa R1ca 1n August 1994 Part1c1pants 1ncluded CA TIE CIAT CORPOICA (Colombia) 108 and IICA. Out of th1s meetmg carne a proposal focussed on the hllls1des of Central Amenca and env1s1omng outputs as summanzed m the m1nutes of TAC65 that 1ncluded (1) an updated map of target ecolog1cal zones (n) s1mulat1on rrodels embraang agnculture and natural resource prooesses (111) relevant GIS data bases for the ecoreg1on (1v) a small cadre of profess1onals tra1ned 1n natural resource management strateg1es (v) develop1ng a capaaty for conven1ng atan ecoreg1onallevel at CIAT and (VI) estabhshmg organ1Zat1onal and fundmg arrangements for long term ecoreg1onal proJects TAC 65 comments on the proposal of August 1994 were that 1t was an ongo1ng programme Wlth a narrow geograph1cal focus on the h1lls1des of Central Amenca TAC65 also 1nd1cated the commttee would welcome a rev1sed proposal wh1ch also spells out the process and t1m1ng for expandmg the scheme to other ecoreg1ons and 1n particular TAC65 noted 'TAC continuas to see potent1al synerg1es between CIA1's commtrnent to the Andean hllls1des and CIPs 1nvolvement m Andean mounta1n development The Commttee would expect th1s to covered m a rev1sed CIAT proposal Upon receiVIng these comments CIAT convened a second planmng meetmg that was held m Cah Colombia 1n December 1994 Wlth the part1apat1on of CIAT CIMMYT CIP ICRAF and IICA. Th1s meeting recommended that the ecoreg1onal proposal to TAC should reflect the need to support vanous 1nd1V1dual agroecosysterns research consort1a already operatlonal 1n the ecoreg1on -the Global Altematlves to Slash and Bum led by ICRAF -the CIP 1mt1at1Ve for h1gh mounta1n agnculture -the Central Amenca consort1um of CIMMYTIIFPRIICIAT the savanna consort1um of PROCITROPICOS/IFDC/CIAT In add1t1on 1t was agreed to Introduce an ecoreg1onal mechan~sm that would prov1de servJces to the agroecosystems research consort1a as well as conduct research on 1nteract1ons among agroecosystems and develop a reg1onal analyt1cal capaoty for pnontiZJng targetmg and extrapolatmg research and 1ts results lt was emphasJZed that the ecoreg1onal mechan1sm would not be placed 1n a coordmatmg role above the agroecosystems consort1a but rather rt would coordmate the prov1s1on of serv1ces for wh1ch there are central economes of scale ( eg sorne pohcy research contmental scale data bases and models etc ) lt would also coord1nate the development and 1mplementatJon of research funded by the ecoreg1onal program but 1mplemented by IndiVIdual consort1a Th1s research would be part of an agreed agenda deahng Wlth methods development 1nformatJon exchange cross agroecosystems research and stakeholder strengthen1ng These concepts were embod1ed 1n the proposal submtted to TAC66 1n February 1995 Th1s proposal was reVJewed and mduded changes 1ntroduced by CIMMYT CIP ICRAF IFPRI and IICA Speofically th1s proposal a1med to complement the consort1a by ass1stmg 1n nat1onal program strengthen1ng as well as 1n the dehvery of three research outputs 1) Enhanced capaoty for cross country pnontJzat1on and for des1gn of natural resources research 2) lmproved methods for s1te select1on and extrapolat1on among watersheds Wlthm agroecosystems 3) Methods for 1dent1fy¡ng and pnont1z1ng natural resource problems at the watershed level In Apnl1995 a consultatJon meet1ng Wlth NARS regional mstJtut1ons and IARCs active 1n the four agroecosystems consort1a was held m Cah Colombia Part1c1pants 1nduded representat1ves from Braz11 (EMBRAPA) Colombia (CORPOICA and the Un1vers1ty of the Andes) Honduras (Nat1onal Autonomous Un1vers1ty) Nicaragua (INTA) and Venezuela (FONIAP and PALMAVEN) Reg1onal1nst1tutlons represented mduded CATIE CONDESAN IICA, PROCIANDINO and PROCITROPICOS whlle 1ntematJonal centers mduded CIAT CIMMYT ICRAF and IFPRI M. th1s meeting there was full consensus on the ut1hty of an ecoreg1onal program that would support and act Wlth the four part1opatJng consort1a lt was completely clear that the four consort1a would operate under the1r ex1st1ng govemance and that an ecoreg1onal program INOuld be an effect1ve means of meet1ng common ends for methods development and 1nformatJon exchange espeoally at the watershed level but also at the agroecosystems and reg1onal levels In addrtJon 1ntens1Ve bilateral d1SCUss1ons have taken place between CIAT and other 1ntemabonal centers Wlth act1ve research programs m the ecoreg1on leadmg to danficatlon of the specrtic po1nts of 1nterest of d1fferent centers m the ecoreg1onal program and définrtJon of the1r expected mvolvement 1n 1t See Append1x 1 for the statements of 1nterest 1n the Labn Amenca hum1d trop1cs ecoreg1onal program as approved by the 01rectors General of several part1apat1ng centers An Ecoregooal A¡:poach to Enhanang Ag7cultural Research 1n Trapea/ Amenca PROGRAM GOAL To 1mprove management of natural resources devoted to agnculture 1n Trop1cal Amenca 1n order to permanently reduce poverty and hunger mamtam resource quahty and 1ncrease agncultural produdlv1ty PROJECT PURPOSE To enhance the effect1veness of agncultural and natural resources management research m Tropical Amenca by 1mprovmg the capaaty to define and understand produd1v1ty problems 1n agnculture to develop and adapt su1table solutlons to these problems and to extrapolate results among agroecosystems through the development of georeferenced mformatlon systems and analys1s BACKGROUND ANO JUSTIFICATION Context á EooregJOnal Research There are a number of ongo1ng CGIAR affihated research ad1V1t1es m Labn Amenca that hnk strateg1c global research to effectiVe and clear partnersh1ps Wlth nat1onal programs 1n order to address the techmcal and human d1mens1ons of sustamable produdiVIty m landscape un1ts dehneated as pnonty agroecosystems Among the most notable of these are the followmg the CONDESAN netiNork for the h1gh altltude Andes m wh1ch CIP has served as the lead center for the CG system and 1n wh1ch CIAT and potent1ally CIMMYT are mvolved - the Altemabve to Slash and Bum Agnculture Program focussmg on the marg1ns of the lowland tropical forests m wh1ch ICRAF serves as global lead center for the CG wh1le CIAT coordmates Latm Amencan ad1v1t1es wh1ch mclude the part1apat1on of CIFOR IFPRIIITA, and IRRI - the Central Arrenca Hllls1des Program 1n wh1ch CIAT 1s serv1ng as lead center for the CG and 1ncludes part1C1pat1on of CIMMYT and IFPRI and CATIE and IICA from the reg1on the Savannas Consort1um wh1ch operates under the umbrella of the IICA sponsored PROCITROPICOS consort1um and 1n wh1ch ICRISAT and IFDC part1e1pate and CIAT serves as lead center for the CG In add1t1on new SysteiTM/Ide Programs are be1ng launched 1n the CGIAR that Wlll have rnaJor act1v1t1es 1n the reg1on These 1nclude for example the SysteiTIWide Llvestock ln1t1at1ve be1ng convened by ILRI and the So1l Water Nutnent Management Program wh1ch has targeted the rnanagement of aCid so1ls 1n the tropical Arrencas as one of 1ts pnonty research themes lt 1s antiCipated that the latter would be hnked 1n particular to the Savannas consort1um v1a the Manag1ng ACid Solls(MAS) consort1um LlkeWise ICRISAl's global sorghum research 1s be1ng hnked to the Savannas research consort1um From Wlthln the reg1on the lnter Arrencan Development Bank (IDB) 1s carry1ng out extens1ve consultatlons Wlth countnes 1n the Arrencas on the 1n1t1at1on of a mechamsm to support agncultural research of 1ntemat1onal SIQnlficance 1n the reg1on Th1s proposed cap1tal fund for agncultural research 1n the Arrencas would have 1nterests 1n the reg1on that would extend beyond those of the CGIAR system but nonetheless close artlculatlon bet\Neen the fund and the CGIAR would be essentlal both to 1rnplement research related act1v1t1es of mutual 1nterest and also because the fund would serve as a rnaJor forum for the definrt1on of research pnont1es from a reg1onal perspect1ve In th1s context CIAT has been requested by the CGIAR to serve 1n an ecoreg1onal convemng role for the CGIAR Clearly th1s can only be ach1eved by prov1d1ng serv1ces that complement ongo1ng act1V1t1es and encourage hnkages among them G1ven the range of 1mportant research already be1ng camed out 1n the reg1on there are rnaJor opportun1t1es to explo1t splllovers among these act1v1t1es 1n terrns of the development of common methods and data bases as vvell as the shanng of research s1tes data and other 1nformatlon Such a convemng role enta1ls a respons1b1hty to 1nsure the facilrtat1on of hnkages among these d1verse act1v1t1es Wlthout 1n any sense presum1ng to 1nvolve an overall coord1nat1on of the1r execut1on Th1s proposal therefore must be understood not as the totahty of an ecoreg1onal program for La1n Arrenca and the Canbbean but rather as a program to capture and foster extemahtles among a vanety of ongo1ng research act1v1t1es that all have an ecoreg1onal d1mens1on 2 Prublet1s and Opporlumbes.Aaoss the d1verse agroecosystems of the Trop1cal Amencas the natural resource base 1s bemg degraded even though overall aggregate growth 1n food product1on 1n the reg1on has barely kept up Wlth demand growth s1nce 1980 and has lagged beh1nd growth ach1eved elsewhere 1n the world Land degradat1on assoaated Wlth agnculture has been rampant m the reg1on In the hllls1des wh1ch cover sorne 90 mlhon ha 1n Central Amenca and the Andes low mcomes and populat1on pressure are dnv1ng a process of over explo1tat1on of the resource base Small fann s1ze and land scarcrty 1s lead1ng to the rap1d destructlon of the small remam1ng reserves of the h1ghly dJVerse natural vegetatlon as land 1s converted to agnculture On agncultural land fallow penods are be1ng reduced and fert11Jty 1s dechn1ng wh1le extens1ve areas are also be1ng degraded by soll eros1on These processes not only narrow the scope for atta1mng susta1nable llvellhoods for the h1lls1de populat1on but also are lead1ng to watershed degradat1on Wlth senous consequences for downstream resource users Because resource degradat1on 1n the h1lls1des typ1cally occurs over landscape un1ts that exceed the holdings of an md1v1dual fann manager and frequently 1nvolve off srte costs or benefits pollaes for collect1ve act1on and to 1ntemallze costs and benefits are essent1al and necessary elernents 1n an effectlve resource management strategy The forest rnarg1ns of the trop1cal lowlands m the Amazon basm and Central Amenca currently cover 40 mlhon ha out of a total of 800 mlhon ha of lowland forest 1n the reg1on Rap1d clearance of land for agnculture 1s expandmg the area of the forest rnarg1ns whlle putt1ng at nsk a rnajor reservo1r of b1od1vers1ty as well as contnbutmg s1gn1ficantly to the accumulat1on of atrnosphenc carbon Susta1nable product1on systems for settled agnculture are not avallable for th1s agroecosystem due to rap1d declmes 1n so1l fertJhty and encroachrnent of weeds lnsecure land tenure and poor access to markets also constra1n adoptlon of resource conserv1ng practlces makmg pollcy a cruaal d1rnens1on 1n 1mprovmg resource management Consequently the vast bulk of the rural populat1on 1n th1s agroecosystem 1s trapped at low 1ncorne levels desp1te a mass1ve decap1tahzat1on of the ex1st1ng natural resource stock In the savannas wh1ch cover sorne 240 mlhon ha of the trop1callowlands of South Amenca the agncultural front1er has been v1gorously expandmg Although th1s agroecosystem has great potent1al for contnbut1ng to world food supplles sustamable product1on systems for these new lands are lack1ng Current product1on systems qUJckly degrade the phys1cal charactensbcs of the so11 wh1le fa1llng to manage the1r 1nherently low chemcal fert1llty Consequently degraded pastures has becorne a maJor land use lncreas1ng the b1omass by replaang natlve and degraded pasture Wlth more product1ve grass legurne pastures can ra1se overall systerns productlv1ty and also may have a potent1ally 1mportant role 1n sequestenng carbon thereby contnbut1ng to the reduct1on of greenhouse gases 1n the atmosphere and help1ng to counteract tendenaes towards global warmng 3 The Andean h1ghlands 1s a v1tally 1mportant ecoreg1on both 1n terms of the large populatlon d\Nelling there that 1s pnne~pally dependent on agnculture but also as a source of severa! hlghly productiva nver bas1ns prov1d1ng life support systerns for rural and urban populat1ons located 1n the lowlands The whole Amazon bas1n ong1nates and depends on water sources located 1n the Andean ecoreg1on Hence degradaban of natural resources 1n the h1gh Andes as 1s also the case 1n the md alt1tude h1lls1des can have drarnat1c effects downstream Economc cond1t1ons have detenorated senously 1n the Andean h1ghlands where over 80% of the currently avallable agnculturalland 1s steeply slop1ng wrth a bread range of so1ls and vanable pattems of clirnate and water ava1lab11ity Poverty 1s 1ntense and Wldespread 1n th1s agroecosystem lead1ng to rnass1ve mgratlon both to the c1t1es and also to the Amazon bas1n A1 the sarne time nat1onal and reg1onal capaCity for agncultural research 1n Tropical Amenca has suffered due to 1ntens1fy1ng resource constra1nts Through the decade of the 1980 s a heavy burden of extemal debt severely hampered nat1onal capaCity to mvest 1n agncultural research Moreover the h1gh degree of urbamzat1on rnay have lnruted dornest1c consbtuenCies to support expend1ture on agncultural research In more recent years the adoptlon across the reg1on of a new economc model based on opemng the econom1es and redue~ng the role of the public sector has further curta1led public 1nvestrnent 1n agncultural research 1n rnany countnes In sorne 1nstances th1s has resultad 1n conserv1ng the skllled human cap1tal bu1lt up for agncultural research over the last decades whlle redue~ng operat1ng budgets In other countnes hovvever there has been ma¡or downs1Z1ng 1n the nat1onal research system Consequently rnany nat1onal research systerns espeCially 1n the smaller countnes have a s1gnlficantly reduced capae~ty to undertake tradrtJonal crop 1mprovernent research let alone embark on the new more cornplex research agenda assoe~ated Wlth natural resource 1ssues An Ecoreg101181 Approach Because nat1onal research resources are so limtlng and because the agroecosysterns of Tropical Amenca are so d1verse developrnent of solut1ons to the numerous and vaned problerns wh1ch degrade the resource base and depress agncultural productiVIty 1s beyond the 1nd1v1dual capaCity of all but the largest natlonal prograrns Agroecosysterns d1vers1ty prov1des a clear need for research cooperat1on among nat1ons 1n the reg1on LlkeWise d1vers1ty makes h1ghly attract1ve the capaCity to link research and technology transfer for s1mlar but often d1stant agroecosysterns A1 the sarne time there 1s a premum on knOWing the extent of d1fferent agroecosysterns and the1r aSSOCiated constra1nts so that hm1ted research resources can be concentrated where they can rnake the b1ggest d1fference Thus there 1s a need for a reg1onal effort to 1dentlfy common problerns and generalizable solut1ons as vvell as to develop technology and policy opt1ons to address pnontlzed Wldespread problerns through ¡o1nt act1ons that lead to an eqUitable en¡oyrnent of benefits 4 Ecoreg1onal agncultural research addresses 1ssues of enwonment-agncultural hnkages through a landscape based reg1onal perspectiVa lt 1nvolves the 1mprovernent of the productiVIty of agncultural systems wh1le ensunng the1r sustamablllty through JUdletous managernent otresource stocks and flows for example hydrolog1cal cydes nutnent fluxes b1omass accumulabon and agrob1od1Vers1ty Hovvever several factors senously hmder effietency of research on natural resource 1ssues There are majar dlfficult1es 1n definmg the pnnc1pal sustamab1hty and productiVIty problems espeetally 1n assess1ng the pace lnetdence processes 1nvolved causes and consequences There are dlfficulbes 1n developmg and adapbng altemat1ve solutlons whether generated by researchers or farmers and whether featunng techmcal or pohcy 1ntervent1ons There are dlfficult1es 1n understandmg the factors that shape farmer adopbon of new pract1ces or 1n us1ng such understand1ng to foster pohcy or 1nsbtut1onal changes that faethtate adoptlon of more sustamable practicas Fmally there are s1gnlficant dlfficulbes 1n the geograph1cal dehneat1on of problems wh1ch hmts the ab1hty to quantlfy the extent of constra1nts select representatiVa research s1tes and extrapolate results Wldely and effietently from a hmted number of research s1tes Th1s last shortcomng 1s part1cularly severe Wlth respect to natural resource 1ssues For example comprehens1ve rehable data on the extent degree and type of land resource degradat1on 1s lackmg LlkeWise comprehens1ve mformatlon on the continental d1stnbut1on of land use systems based for example on agroforestry or cattle graz1ng or on systerns assoaated Wlth maJar crops such as wheat ma1ze potatoes svveet potatoes and sorghum 1s not ava1lable Moreover understand1ng of the relat1on betvveen these land use systerns and the1r enwonrnents 1s based pnnetpally on expert opm1on rather than on stat1st1cal analys1s of the relabonsh1ps betvveen chmate so1ls resource degradat1on and crop d1stnbut1on NARS Pnonbes The several agroecosystem research 1n1t1at1ves are each orgamzed as consort1a Wlth a Wlde range of natlonal and reg1onal partners V\tnle the deta1ls of govemance of the vanous consortla are not 1dent1cal all essent1ally take doos1ons through penod1c formal meet1ngs of mernbers const1tuted as steenng commttees 1n wh1ch the natlonal mst1tubons compnse the maJonty of representativas The vanous agroecosystem consort1a 1n the reg1on have developed the1r research pnontles through processes that have mvolved natlonal program leadersh1p 1n setbng the research agenda Examples of such consultaban are nurnerous and ongomg but a few mstances are worth atlng 1n particular In Central Amenca gu1dehnes for 1nter mst1tut1onal cooperat1on and a research pnontles vvere set 1n a proJect planmng by obJectlves exerase 1n wh1ch DICTA (Honduras) and INTA (N1caragua) partlapated along Wlth reg1onalmst1tubons (CATIE and IICA) and 1ntemat1onal centers (CIAT CIMMYT and IFPRI) Understandmg the relat1onsh1p betvveen land use and natural resource degradat1on emerged as the h1ghest pnonty for research 5 In the forest rnargms research pnontles and s1te select1on was undertaken through a JOint corrmss1on 1n February 1992 that 1nduded three Braz~han EMBRAPA research centers (CPATU CPAF and CPAA) ISPN an NGO along wrth CATIE and IICA at the reg1onallevel and CIAT and ICRAF from the CG system W'i1le the mss1on generated agreement over a diVISion of respons1blllt1es over a Wlde range of research themes 1ndud1ng germplasm 1mprovement rnarket opportun1t1es and apphed research the s1ngle research 1ssue that was a pnonty for all part1apants was analys1s of land use changes Wlth emphaSis on understand1ng the relatJons between agnculture pollcy and natural resource conservaban Subsequently th1s work was merged 1nto the Global Slash and Bum ProJect that 1ndudes the part1apat1on of IFPRI IFDC and TSBF 1n add1tlon to those mentloned above as IJI/ell as NGOs 1ndudmg PESACRE and IPHAE In the h1gh alt1tude Andes the CONDESAN consort1um has pnont1zed four maJar research themes commod1ty systems land and water rnanagement pohcy and b1od1vers1ty The PPO part1apatory program planmng by ob]ectlves method has been used to assure stakeholder part1apatJon 1n JOint planmng and proJect monltonng A senes of planmng workshops have been conducted 1ndud1ng one on b1od1vers1ty 1n August 1992 one on land and water rnanagement m March 1993 and two plann1ng workshops m 1994 to pnont1ze actiVItles 1n the benchrnark s1tes Natural resource rnanagement research 1n the savannas 1s coord1nated through the IICA sponsored PROCITROPICOS network wh1ch 1ncludes the part1apat1on of EMBRAPA of Braz1l IBTA of Bohv1a CORPOICA of Colombia and FONIAP of Venezuela The Nat1onal Umvers1ty Bogota the Technolog1cal Umvers1ty of the Llanos and the Umvers1ty of Uberland1a have all JOined savanna consort1um act1v1t1es at the s1te leve! CIAT part1apates 1n th1s savannas research network m wh1ch 1t has been asked to contnbute part1cularly on strateg1c research on underly1ng b1ophys1cal processes that affect natural resources and also on the ut1hzat1on of geograph1cal 1nformat1on systems to charactenze enVIronments resource problems and onent the des1gn of research strategy Research pnontles across agroecosystems IJI/ere analyzed 1n detall m the process of preparat1on of the CIAT Strateg1c Plan Th1s has contlnued to be a top1c of concem to 108 and IICA, wh1ch have recently JOined Wlth IFPRI and CIAT to look more dosely at thiS ISSUe The success of these 1nlt1at1ves 1n reachmg the goals of sustam1ng the resource base and mcreas1ng agncultural productlvrty 1s cntlcally dependent on the understandmg of the1r target agroecosystems Targetlng of the development of natural resource rnanagement systems needs to be sharpened because these agroecosystems typ1cally extend over large areas and are not fully cont1guous Thus the development of problem solv1ng research processes featunng more prease georeferenced mformatlon systems at the agroecosystems leve! 1s needed to better define problems 1dentlfy 6 solut1ons accelerate adopt1on nnprove the generahzab1hty of research and facahtate the exchange of results Addressmg these 1ssues reqUires among other tools spat1al models and Informaban systems on the enwonment agnculture and the SOCio- economc context PROJECT OUTPUTS ANO AC11VITIES Based on the above analys1s of problerns and the pnont1es as they have emerged from the NARS part1capatmg 1n the agroecosystem consortla a set of proposed outputs and act1v1t1es have been denved for th1s ProJect to Enhance Research Effecbveness for Agncultural Sustamab1hty Through and Ecoreg1onal Approach 1n Tropical Amenca (See Table 1 Work Breakdown Structure) Th1s project a1ms to fully explo1t and encourage extemahbes among ex1st1ng agroecosysterns research consortla wrth an ecoreg1onal d1mens1on as vvell as prov1de v1tal h1gh payoff serv1ces for wh1ch the1r are economes 1n scale 1n the~r dehvery to the agroecosystems consort1a and other reg1onal stakeholders OveMeW The central outputs of th1s proposed ecoreg1onal proJect all focus on enhancang the capacaty to pnont1ze plan target and extrapolate research on natural resource managernent and agncultural productiVIty 1n Trop1cal Amenca Ach1ev1ng agncultural and resource sustamab1hty 1n an ecoreg1on as d1verse and cornplex as Tropical Amenca can best be atta1ned by comb1n1ng the efforts of d1verse actors at the nat1onal reg1onal and local level wh1le tak1ng advantage of 1nputs from the mtemat1onal agncultural research system. The problems and opportun1t1es of agncultural and natural resource susta1nab1hty need to be v1suahzed and dealt wtth at an 1ntemat1onal level 1n arder to ach1eve economes of scale 1n strateg1c research and transfer of results Th1s 1s done both through coordmat1on among NARS m the reg1on as vvell as through the1r effect1ve hnkage wtth the CGIAR system Recumng pattems 1n land use the b1o-phys1cal enwonment and the SOCio-economc context g1ve nse to agroecosysterns as useful un1ts of analys1s for the defin1t1on of common problerns the des1gn of technolog1cal and pohcy optlons and the 1nterchange of research results for adaptat1on at the local level Ult1mately new technolog1es and pract1ces are 1rnplernented at the local or watershed level and natural resource degradaban processes and hmts to agncultural productiVIty emerge from specafic cond1t1ons at the watershed level Thus methods are needed to des1gn solut1ons and pnont1ze problerns at these three levels 1ntemat1onal agroecosysteiTl and watershed M. the same time methods are needed to enable Informaban to be mean1ngfully exchanged across these d1fferent levels Specafic observatlons m particular s1tes 1n watersheds aggregate 1nto common find1ngs across agroecosysterns whlle agroecosysterns recur over mtemat1onal boundanes gw1ng scope for economes 1n the aSSeiTlbly of Informaban and 1n the 1rnplernentat1on of probleiTl solvmg research 7 lhe outputs of th1s project therefore are 1mprovecl methods for research plannmg and 1mplementat1on at d1fferent levels lhese methods Wlll vary WJth the scale of the analys1s but can be made 1ntercorrmumcable across scales of analys1s and across countnes by hnRing data and analys1s to georeferencecl1nformabon systems and models Strengthen1ng nat1onal and local capaaty to use such an approach 1s also a key output expectecl of th1s project Speclfic project outputs are now d1scussecl 1n tum Oulput 1 Enhanced Cross Counby Analysls lhe Objectlve of thiS output IS to 1mprove the capaaty for cross country pnont1zat1on targetlng and extrapolat1on of research results through the development of comprehens1ve env1ronmental and agncultural land use data bases and assoaatecl models for Trop1cal Amenca Analys1s at the continental scale can faahtate the capture of 1mportant cross country synerg1es and economes of scale m research that can not be ach1evecl except by worklng at the 1ntematlonal leve! SIITIIIar env1ronments resource degradatlon problems and potent1al solut1ons recur across space among countnes By 1dentrfy1ng such homolog1es the most Wldespread common problems can be pnont1zecl and effiaent paths for exchange of results mappecl Vl.tnle 1nforrnat1on systems 1nclud1ng cl1mate models so1ls data and selectecl agnculturalland uses have already been developecl at CIAT for Tropical Amenca majar gaps 1n these systems hm1t the1r usefulness Land use data for product1on systems basecl on nce beans and cassava have been developecl but mforrnat1on systems for the spat1al d1stnbut1on of productlon systems basecl on ma1ze wheat potatoes sweet potatoes sorghum agroforestry and cattle graz1ng are lackmg LlkeWJse a comprehens1ve Informaban system on land degradat1on 1s also lack1ng lh1s projed would develop the above needecl mforrnat1on systems and models lhese models and data bases would have two majar uses F1rst 1t would enable countnes m the reg1on to acoess a common and comprehens1ve mforrnat1on system that would greatly enhance the1r ab1hty to quant1tat1vely 1dent1fy themes of common concern and the1r extent across borders S1m1larly 1t would faalrtate CGIAR system Wlde plann1ng and 1mplementat1on of agncultural and natural resource management research Second these 1nforrnat1on systems and cl1mate models can be hnkecl to phys1olog1cal models of crop growth or cultivar performance data 1n arder to 1mprove s1te select1on target1ng technology des1gn and results extrapolatlon m the course of the ongomg technology development research of the mtematlonal centers and natlonal programs Models basecl on factor analys1s or pnnapal components would be usecl reveal 1nterrelatlons between env1ronmental vanables and land use L1keWJse relatlons would be explored between land use 1n dlfferent crops and resource degradatlon processes lhus CIFOR CIP CIMMYT ICRAF ICRISAT and ILRI would all gam acoess to 1nformat1on systems that would ass1st them both 1n plann1ng and 1n conductlng the1r research 8 Furthermore models and data bases at thrs level of aggregabon can be usefully hnked to pohcy analysrs Many pohaes such as trade exchange rates rnterest rates and other macro and sectoral rnstruments exert thetr rnfluence not at the agroecosystems or watershed level but across agroecosystems and even natronal boundanes lntegratron of soaoeconomc perspectrves and approaches to the analysrs of data and the development of rnodels at thrs level could therefore be hrghly frurtful and reason for a strong cornponent of sooal saences rn the plannrng and rrnplementatron of the workplan for thrs output Thus thrs output rs of potentral srgnrficance both to IFPRI and other rnstrtutrons rnterested rn examnrng the hnkages between pohcy land use and resource degradatron For 1995 full fundrng to CIAT and IFPRI for work on thrs output rs antrapated from IDB Output 2 lrTfli'OWCII\IIethods for Agroecosystems Resealdl As noted above research consortra rnvolvrng NARS and severa! IARCs are already operatrng at the agroecosystem leve! for the hrgh Andes the slash and bum forest rnargrns the Central Amencan hrllsrdes and the South Amencan savannas Thrs research wrll rncreasrngly be hnked to global CGIAR rnrtratrves such as the Systenmde Lrvestock lnrtratrve and the Sorl water Nutnent Management Program The latter for exarnple would work dosely Wlth the Savannas consortrum through the managrng aad sorls (MAS) consortrum In arder to effiaently generate rrnproved systems of natural resource management for these agroecosystems rmproved research methods are needed to address a number of key rssues To assess the rnadence of and rates of change rn resource degradatron and agncultural productrvrty at the agroecosystem leve! georeferenced temporal models that aggregate across watersheds rn an agroecosystem wrll be constructed To rmprove srte selectron and extrapobrlrty of results at the agroecosystem level statrstrcal analysrs and srmulatron models wrll be utrhzed to rdentrfy hornologous naturaVSOCio-economrc envrronments To enhance collectrve deasron maklng on land use practrces and pohcy georeferenced rnultrple stakeholder deasron models need to be developed Such methods wrll permt ex ante analysrs of the on and off srte and near and long term effects of altematrve land uses and pohaes 9 To conserve agrob1od1Vers1ty models Wlll be developed to analyze georeferenced genebc resources data 1n pnonty agroecosystems Th1s Wlll ass1st 1n 1dentlfy1ng entena for selectlng cand1date srtes for 1n srtu conservaban of ma1ze-potatoes common beans cassava andarach1s sps In 1995 fundmg from DANIDA 1s be1ng used to support th1s acbvrty Oulput 3 Research Methods fcr Wa1ershed Level Research:llt1ore effectlve methods for 1dentlfymg and pnont1z1ng research at the watershed level can enhance research be1ng currently camed out by 1ntemabonal centers and NARS m the four pnonty agroecosystem consort1a 1n Tropical Amenca Moreover these methods could be extended to other agroecosystems not only 1n Trop1cal Amenca but potent1ally world Wlde Effiaent low cost methods to 1dent1fy natural resource problerns and the1r soao- econoiTllc correlatas are st1ll needed Such methods are hkely to cons1st of a su1te of strateg1c field tnals process models and part1c1patory appra1sals Llk9Wise qu1ck, low cost methods for assess1ng the long terrn pro¡ected consequences of natural resource degradaban processes are needed Aga1n th1s 1s hkely to requ1re a vanety of approaches 1ndudmg chronosequence stud1es long term tnals and farrner monrtonng Lastly to 1mprove the generahzab1hty of results methods for adapt1ng ex1st1ng crop and process models m order to 1mprove the accuracy of aggregatlon from the farrn or field level to entlre watersheds 1s cntlcally needed Overall the development of these methods mforrnat1on and analyses would have three ma1n uses F1rst they could be ut1hzed Wlth1n a speafic watershed for defin1ng and solvmg ma¡or productiVIty and susta1nab1hty problems Second they can be transferred for apphcatlon m other watersheds reflect1ng the need for more effiaent and effectlve methods at the watershed level across watersheds Wlthm an agroecosystem Th1rd the 1nforrnabon generated at th1s level can be fed mto analys1s at h1gher levels of aggregatlon feedmg for example models and data bases used to produce Output 2 above VI/hile the deta1led workplans at the watershed level would be developed 1n the 1mplementat1on phase of the pro¡ect certam pnonty acbv1t1es can be ant1apated For example research m the h1gh Andes 1n the CONDESAN consortlum s1te of Carch1 Ecuador would be extended to 1ncorporate work at lo\11/er elevat1ons m order to create a transect that bnngs together mterconnected resource 1ssues that cross agroecosystem boundanes Input from the CIAT Hllls1des Program and CIMMYT would be sought for the 10\11/er end of the transect Ex1st1ng research consort1a hnked to SystemMde ln1t1at1ves such as the Altemat1ves to Slash and Bum prov1de a h1ghly cost effiaent mechamsm for generatlng data and testlng models essent1al for produang the outputs env1s1oned 1n th1s prOJect At the same t1me much of th1s 1nforrnat1on 1s of d1rect utlhty to the ex1st1ng agroecosystem consorba even though these act1v1t1es are not currently 1ncluded 1n the1r fund1ng 10 requests and workplans Thus hnk1ng th1s Ecoreg1onal lmtlatlve to the agroecosystem consorba both prov1des an espeaally economc means of producmg the outputs of th1s proposal whde at the same t1me prov1d1ng to pnonty agroecosystem consortla the capacrty to usefully expand the1r actMbes to generate relevant and useful outputs that otheiWise could not have been produced For 1995-96 parbal fund1ng for the dehvery of th1s output by CIP and CIAT for the h1gh Andes agroecosystem 1s ant1apated from BMZ. In add1t1on part1al fundmg for the dehvery of th1s output by ICRAF and CIAT IS antlapated through GEF Output 4 Strengthened NARS Capaaly The development of methods modets mformat1on systems and data sets IMII be frurtful largely to the degree that NARS are empowered to utdiZe them A four pronged appro'l:ich to th1s 1ssue IMII be taken \1\klrkshops IMII be held at the mtemabonal level through fora that enable NARS to develop compatible data bases and analytlcal methods for coorclmated plann1ng at the sub-reg1onal level Th1s could be cared out through IICA sponsored sub-reg1onal networks such as PROCIANDINO or PROCITROPICOS - Tram1ng on an lndMduahzed baSis or through courses IMII be conducted at the nabonal level m orcler to 1ncrease capaaty to use georeferenced data and models for research planmng Tra101ng on the use of modets and data at the watershed leveiiMII be offered through the eXIstmg agroecosystem consort1a In orcler to test the ut1hty of methods data and models developed m th1s proJect p1lot proJects IMII be 1mplernented by NARS SCienbsts to adapt general models deas1on tools and data bases to specriic actual problems and cond1t1ons faced by NARS B1bhograph1c and 1nformat1on serv1ces IMII be prov1ded on reg1onal hterature related to land use and resource degradat1on In 1995 part1al fund1ng for th1s output IMII come from IDB EXPECTED IMPACT CGIAR SyslemMde IITflbcatlons. Th1s proJect lays the foundat1on for more effectiVe 1ntegrat1on of global core systerTl\Nide research IMth the problems and opportunrt1es 1n the ecoreg1ons of Tropical Arrenca Th1s proJect complements rather than subst1tutes 11 for ongo1ng core and global research on technology development Through the development of models methods and georeferencecl data pases the outputs of SysterTliMde research can be more effect1vely targeted to and extrapolated Wlthln the ecoreg1ons of Tropical Amenca Severa! concrete outputs are env1saged that Wlll enhance the mrk of severa! centers Continental scale georeferencecl databases for ma1ze wheat sorghum potatoes sweet potatoes pasturas hvestock and agroforestry land use systems Wlll d1rectly ass1st CIFOR CIMMYT CIP ICRAF ICRISAT and ILRI 1n pnort1z1ng and target1ng the1r research for the agroecosystems of Tropical Amenca The 1ntercommun1cable data bases Wlll be compatible Wlth data already developed by CIAT on SOIIS and chmate 1n the reg1on lt Wlll be poss1ble for the Centers to hnk the1r models and Informaban on crop performance and phys1ology Wlth the georeferencecl data as a tool for the vanous Centers to a1d them 1n the des1gn and target1ng of the1r technology 1mprovement efforts - Comprehens1ve agncultural systems and land use data and models Wlll faCihtate the assessment of natural resource degradat1on processes at the continental and ecoreg1onal scales Th1s Wlll prov1de v1tal onentat1on to research on the 1mprovement or resource management and pohcy - Methods and models Wlll be developed for assess1ng the 1ne1dence and pattems of change 1n agncultural and resource problems as IJIIell as for appra1s1ng the1r consequences over time 1n pnonty agroecosystems Th1s Wlll enhance the effect1veness of ongo1ng research 1n the Global Altemabves to Slash and Bum and Mounta1n Agnculture lmt1abves 1nvolv1ng CIAT CIFOR CIP ICRAF and IFPR11n Tropical Amenca as IJIIell as ongo1ng agroecosystem consorba for the h1lls1des and savannas 1n the reg1on 1nvolv1ng CIAT CIMMYT and IFPRI along Wlth reglonal1nstltut1ons such as CATIE and IICA. GeoreferenCing of genet1c resource data for ma1ze potatoes arach1s. sps beans cassava and mult1 purpose tree spee~es 1n pnonty agroecosystems Wlll be of use to CIFOR CIMMYT CIP ICRAF ICRISAT ILRI IPGRI and CIAT Methods Wlll be developed to relate pattems of genet1c vanab1hty to so1l chmate and land use data both to target potenbal cand1dates for pnonty1n srtu conservaban and also as a potent1al tool for evolut1onary and genetlc stud1es Emerg1ng new CGIAR 1n1t1at1Ves such as the Global L1vestock lmbabve or the Manag1ng ACid Solls research theme W1th1n the Syste1'1111111de So1l Water and Nutnent Management Program muld along Wlth other such 1n1t1abves benefit from the outputs of th1s pro¡ect The research outputs of th1s pro¡ect muld prov1de tools that they could use for pnont1zat1on s1te selectton and 12 extrapolat1on Moreover the creatlon of the Ecoreg1onal Steenng Corrmttee that forms an umbrella hnkmg the four majar CGIAR aSSOCiated agroecosystems consort1a m the reg1on would prOVJde a natural channel through wtuch to these new lnltlabves could be lntroduced lnto the reglan WJthout need1ng to prohferate new 1nst1tut1onal modaht1es or duphcate contacts WJth the NARDS Expected ll'f1JEid: llf'4lltcabons for NARS - Ut1hz1ng comprehensJVe 1ntemat1onal agncultural and land use data Wlll ass1st NARS 1n 1dentlfy1ng pnontles and conmonahtles m reg1onal agncultural research plann~ng and coord1nat1ng fora such as the IICA sponsored PROCIANDINO and PROCilROPICOS networks - Methods and models developed 1n the pro¡ect Wlll ass1st NARS m the1r own work m s1te selectJon targetlng technology des1gn and extrapolaban of results by agroecosystems at a natlonal level Tra1n1ng Wlll be prov1ded 1n the use of these methods and models at the nat1onal level - Research on development of methods and models 1n pnonty agroecosystems Wlll contnbute d1rectly to the generat1on of new technology and pohcy optlons appropnate for the speafic problems encountered at these srtes Not only Wlll these components be avallable for adaptat1on and use by NARS but also tra1nmg WJII be provlded and the agroecosystem or watershed level m the use of sUJtable methods for problem d1agnos1s technology des1gn and test1ng and evaluatJon and extrapolat1on of results In combmatlon these above outputs Wlll enable NARS to 1ncrease the1r effect1veness 1n generatmg and adaptlng technology and pohcy optlons at the 1ntemat1onal nat1onal and local (regional) levels - More coord1nated 1nteract1on and less duphcatlon of effort 1n 1nterfaang Wlth mtematlonal centers Expeded ll'f1JEid: lrqJI"ICaboiiS for Fanners Enhanced effect1veness m 1ntemat1onal nat1onal and local research on agncultural and resource managernent research Wlll accelerate the avallab1hty to fanrers of technology and pohcy optlons to 1ncrease productlv1ty 1ncomes and the sustamab1llty of the1r resources However farmers are not d1rect users of most of the outputs of th1s proJect whlch does not attempt to substrtute for or add more of the same to ongo1ng natlonal and mtemat1onal agncultural and resource managernent research Rather the 1nformat1on systems methods models analys1s and data denved from th1s proJect WJII be used by the nat1onal 13 and 1ntemat1onal research systems to enable them to better meet farmers needs MANAGING ECOREGIONAL RESEARCH By 1ts natura ecoreg1onal research 1s complex lt 1s concemad not ¡ust Wlth 1nd1V1dual comrnoclltles but Wlth producbon systems 1n landscape un1ts lt a1ms to 1ncrease system susta1nab1hty as \11/ell as productiVIty lt hnks global research to speafic arcumstances lt addresses both techmcal and human d1rnens1ons of problems lt 1nvolves the efforts of a vanety of nat1onal reg1onal and mtematlonal partners 1n a collaborat1ve frarnevvor1< Due to th1s complex1ty ecoreg1onal research can not be truly implementad by a s1ngle mst1tut1on nor can 1t be effect1vely mastermndad or centrally controllad by a Single 1nst1tUt1on Success 1n ecoreg1onal research reqUires the exerase of respons1b1hty and 1n1t1atlve of a range of autonomous partners act1ng together 1n dlfferent comb¡nat1ons 1n the face of vanous problems Thus there needs to be a s1gmficant degree of decentrahzat1on 1n the operatlonahzatlon of ecoreg1onal research lmportant opportumtles ex1st though for addmg value through the prov1s1on of sorne centralizad serv1ces 1n a decentrahzad system Economes of scale can be reaped through the generat1on of key common 1nformat1on and data bases as \11/ell as methods Furthermore there Wlll be 1mportant splllovers or extemahtles to be capturad and d1stnbutad 1nclud1ng for example rnethods data or other 1nformat1on developed 1n a particular arcumstance but wh1ch have broader apphcab1hty LlkeWise m a fully decentrahzad system duphcat1on of efforts can occur or 1mportant gaps may appear S1gn1ficant value can thus be addad by an ecoreg1onal mechamsm that faahtates the ach1evement of economes of scale where poss1ble that captures and d1stnbutes sp11lovers and helps assure an effiaent diVISion of labor The Proc:ess of Ecoreglonal Researdl. Table 2 dep1cts a flow d1agram lllustrat1ng how ecoreg1onal research could vvor1< The process 1mt1ates Wlth the express1on of dernand 1n terms of a pnont1zad agenda of problems and opportumt1es 1dentlfiad by natlonal agncultural and natural resource systems broadly definad to mdude among others nat1onal agncultural and natural resource research 1nst1tutes farmer orgamzatlons extens1on and rural development agenaes nof1i1ovemrnental orgamzatlons umvers1t1es and the pnvate sector Fora are needad for th1s d1verse set of actors to reach common agenda both at the nat1onal and 1ntemat1onal levels Sorne such tora for 1denbfymg common agenda eJOst for example the IICA sponsorad reg1onal PROCI netvvor1 1 • :::::J ---- ----....- ( ) ~e ) [ SAVANNAS ...-------., - - . ~e ) DEMAND SYSTEM OUTPUTS PROTOTYPE SYSTEMS EXTRAPOLA TETO SIMILAR CONDITIONS Table 3 Budget for 1995 Workplan to support Ecoreg1onal Research 1n Latm Amenca and Canbbean Consultat1on meetmg w1th 25000 agroecosystem consortla Travel for bilateral meet1ngs 15 000 w1th semor NARS leaders Travel for attendance to meetlngs of 12 000 steenng comm1ttees of agroecosystem consortla and PROCITROPICOS and PROCIANDINO Consultants 50 000 H1gh level workshop w1th NARS Leaders 28 000 Techmcal workshop 15 000 Commumcat1ons and supphes 5 000 TOTAL 150 000 Output 1 Output 2 Output 3 Output 4 Output 5 Output 1 Output 2 Output 3 Output 4 1 output 5 1 1 1 1 1 Output 1 Output 2 1 Output 3 Output 4 Output 5 Table 4 lnd1cat1ve Budget for Project to Enhance Research on Sustamable Agnculture for Ecoreg1ons of Trop1cal Amenca ltem 1995 1996 Total Program Reqwrements Enhanced capac1ty for cross country 161 302 analys1s pnont1zat1on and target1ng lmproved methods for research on 250 276 technology and pohcy agroecosystems leve! More effect1ve methods for pnont1zmg 143 450 and extrapolatmg m selected agroecoreg1on watersheds s1tes Strengthened nat1onal capac1ty for resource 50 246 management research Coordmat1on and Informaban 150 80 Exchange Lmkages Total 754 1354 Currently ldenf¡f¡ed Fundmg Enhanced capac1ty for cross country 161 302 analys1s pnont1zat1on and targetmg lmproved methods for research on 250 o technology and pohcy agroecosystems leve! More effect1ve methods for pnont1zmg 143 198 and extrapolatmg m selected agroecoreg1on watersheds s1tes Strengthened nabonal capac1ty for resource 50 50 management research Coordmat1on and lnformat1on o o Exchange Lmkages Total 604 550 Request to TAC Enhanced capac1ty for cross country o o analys1s pnont1zat1on and targetmg lmproved methods for research on o 276 technology and pohcy agroecosystems leve! More effect1ve methods for pnont1zmg o 253 and extrapolatmg 1n selected agroecoreg1on watersheds s1tes Strengthened nat1onal capac1ty for resource o 196 management research Coordmat1on and lnformat1on 150 80 Exchange Llnkages Total 150 805 1997 1 222 276 475 223 80 1276 222 o 97 o o 1 319 o 276 378 223 80 957