Project Performance Audit Report - Brazil

85
Document of The World Bank FOR OFICIAL USE ONLY 2ep N. 9634 PROJECT PERFORMANCE AUDIT REPORT BRAZIL SECOND AGRICULTURAL RESEARCH PROJECT (LOAN 2016-BR) JUNE 10, 1991 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

Transcript of Project Performance Audit Report - Brazil

Document of

The World Bank

FOR OFICIAL USE ONLY

2ep N. 9634

PROJECT PERFORMANCE AUDIT REPORT

BRAZIL

SECOND AGRICULTURAL RESEARCH PROJECT(LOAN 2016-BR)

JUNE 10, 1991

Operations Evaluation Department

This document has a restricted distribution and may be used by recipients only in the performance oftheir official duties. Its contents may not otherwise be disclosed without World Bank authorization.

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

CURRENCY EQUIVALENTS

US$ Cruzeiro/CrusadoV

Appraisal, May 1981 1.00 82.53Yearly Average 1982 1.00 179.40

1983 1.00 576.201984 1.00 1,845.301985 1.00 6,205.001986 1.00 13.651987 1.00 39.301988 1.00 262.381989 1.00 2,834.00

FISCAL YEAR

January 1 to December 31

On March 1, 1986 Cruzado units (Cz) were introduced at one per thousandCruzeiro units (Cr).

Until February 28, 1986.

NO OMUL US 01LYTHE WO1LD BANK

Wahwnton. D.C. 20433USA.

OAc~.e fDiwS.Casual

June 10, 1991

NEDAI TO7 M11 UM1VUXV DIRECT=R AM MU PM1SDWR

SUBJECTs Project Performance Audit Report - BrazilSecond Agricultural Research Project (Loan 2016-BR)

Attached, for your information, is a copy of a reportentitled "Project Performance Audit Report on Brazil - SecondAgricultural Research Project (Loan 2016-BR)", prepared by theOperations Evaluation Department.

Attachment

This docuneat has a fercted distribution ad may be used by recpisnts only In the perfbr ncO

of their ocIa duties. Is contents may not odWorwise be disclosed without WO Oak authoution.

eM OFFCIAL M ONLY

ABDRS VIATIONS/ACRONIMS

CPAC Centro do Pesquisa Agropecuaria dos Corrados(Research Center for Agriculture in the C*rrados)

CPATSA Ceatro de PeaquIsa Agropecuatia do TropicoSemi-Ardo(Research Center for Agriculture in theSemtarid Tropics)

CPATU Centro do Posquisa Agropecuaaia do Tropico Unido(Research Center for Agriculture in theVmid Tropics)

CUPOC Centro Nacional do Pesquiss do Gado do Corte(National Research Center for Beef Cattle)

DO Departamento do Captacao de Recursos * Comercialisacion(Department for Private Sector Funding for Research andConsercialisation of Agricultural Products)

DPD Departamento do Pesquisa * Desenvolvimento (Departmentfor Research and Development, which has absorbed DTT)

DTT Departamento do Difusio a Transferencia do Teenologia(Department for Diffusion and Technology Transfer)

ENBRAPA Empress Brasileira do Pesquisa Agropecuaria(Brazilian Agricultural Research Corporation)

=NBRATER Empress Brasileira de Assistencia Tecnica eExtensio Rural(Brazilian Tecbrical Assistance and RuralExtension Corporation)

EMATERIASTER Empress etadual/Territorial do AssistenciaTicnica * Extensio Rural(State/Territorial Technical Assistance and RuralExtension Company)

ERR Economic Rate of ReturnFSP Farming System PerspectiveGDP Gross Domestic ProductIARC International Agricultural Research CenterIDB Interamerican Development BankPCR Project Completion ReportPOLONORDBSTE Programa de Desenvolviaento de Areas do Nordeste

(Program of Development of Integrated Areasof the Northeast)

PPAR Project Performance Audit ReportUZPAZ/UPAS Unidade do Zxocucio de Pesquisa do Ambito

gstatal/Territorial(Unit for Research at the State/Territorial Level)

This document has a restricted distribution and may be used by recipients only in the performanceof their oAlcial duties. Its contents may not otherwise be disclosed without World Bank authorinzation.

PmOEC PuROnanE AUDIT RIPC=

BRAZIL

, SECOND AGRICULTURAL RESEARCH PROJECT(LOAN 2016-AR)

TABLE Of CONTENTS

Page fio.

Basc Data . . . . .. . . . . . . . . . . . . . . . . .Evaluation Sumary . . . . . . . . . . . . . . . . . . v

Proiect Performance Audit

I. Background and Design . . . . . . . . . . . . . . 1

II. Implementation Experience . . . . . . . . . . . . 4- Overall Program . . . ........... 4- Staffing . . . . . . . . . . . . . . . . . . 5- Training. . . . . . . . . . . . . . . . . . 6- Consultants . . . . . . . . . . . . . . . . 7- Collaboration with Other Research . . . . . 8

Institutions . . . . . . . . . . . . . .

III. Project Outcome . . . . . . . . . . . . . . . . . 8- Production Impact .*. . . 8- Regional Allocation of Resources . . . . . . 10- Institutional Impact. . . . . . . . .o 12

o Research Programming . . . . . . . . . 12o Technology Transfer . . . . . . . . . 16o Linkages . . . . . . . . . * . . . . . 18

- Sustainability . . . . . . .tya. ... .. .. 19

IV. Findings and Issues . . .. . . . . . . . . 21- PPARs on Related Projects .. . . . . . . . 22- Research Programming and Technology Transfer 24

using FSP . . . . . * . . . . . . . . * .- Action by EMBRAPA . .. *. **. * .. 24

Tables . . . . . . . . . . . . . . . . . . . . . 29Figures . ................... so

JAW Of 2*n11M (Cont' d) Psae No.

AAA 1s FarULUg System Perspective . . . . . . . . . . 32- The Concept .. . . .. . ......... 32- Implementation of Programs at the Three . . 33

Regional Centers . . . . . . . . . . . .- Interpretation of Progress Made at . . . . 56

1achCenter .. ...........Tables . . . . . . . . . . . . . . . . . . . . 59

A 2naa . . . . . . . . . . . . . . . . . . . . . . . 62

Chart 1: Technology Generation and Adoption Model . . . 62Chart 2t Technology Transfer by Research/Extension . . 63

System to Larger Commercial Farmers .Chart 3: Technology Transfer by Research/Extension . . . 64

System to Smallholder Producers . . .

Attachment Comments from the Borrower . . . . . . . . . 67

MAPS - IBRD 15234R1IBRD 11884R2

POJECT PERFOWCZ AUDIT REPORT

BRAtlL

SCOD AORICULTURAL RESARCH PROJECTLOA 2016-R)

1. This is a Project Performance Audit Report (PPAR) on the SecondAgricultural Research Project, the Bank's second lending operation in support ofagricultural research in Brasil. The loan of US$60 aillion was approved on July16, 1981 and was closed, on schedule, on December 31, 1987. It was fullydisbursed and the last disbursement was made on June 15, 1988.

2. The PPAR is based on the Project Completion Report (PCR) 1 , of the LatinAmerica and Caribbean Regional Office, the Staff Appraisal (8AR) and President'sReports, the legal documents, the transcript of the Executive Directors' meetingat which the project was considered, the project files, the Borrower's finalproject report, and discussions with Bank staff and Government officials. An CEDmission visited Brazil in November 1990, and the kind cooperation and valuableassistance of Government officials in Brasilia and in the research centersvisited are gratefully acknowledged. The comprehensive final evaluation reportprepared by the project implementing sgency facilitated the preparation of thePCR by the Bank.

3. The PCR provides a good summary of the project's design, Implementationexperience and impact, and assesses the performance of the Bank and theBorrower. The PPAR elaborates on particular issues associated with thisrelatively successful project, including the programing and prioritisation ofresearch, the transfer to farmers of the technology developed by research, andthe sustainability of the research institution and investment supported under theproject; it also looks at measures considered necessary to further increase thebenefits of DMRAPA as th3 national agricultural research institution.

4. Following standard OED procedures, copies of the draft PPAR were sent tothe Government. Comments were received and are included as an attachment to thereport.

I Project Completion Reports Brazil, Second Agricultural ResearchProject (Ln. 2016-BR), dated November 6, 1989.

PROJECT PEPOIAMCS AUDIT REPORT

BRAZIL

SECOND AGRICULTURAL RESEARCH PROJECT(LOAN 2016-3R)

BSDIC DATA SHEE

A. Key Proieot Data

AppraisalItem Estimate Actual

Total Project Cost (US$M) 150.08 148.94Loan amount (US$M) 60.00 60.00Date components completed June 30, 1987 June 30, 1988Economic rate of retirn Not calculated for this research project.Institutional perfe-mance -- Good

B. Cumulative Disbursement

------------------------------------------------------------------------------Bank FY 1982 1983 1984 1985 1986 1987 1988

------------------------------------------------------------------------------Appraisal Estimate 6.9 19.5 34.0 46.9 56.0 60.0 -Actual 3.4 10.6 27.3 47.0 56.7 59.1 60.0Actual as 2 of Estimate 49.3 54.4 80.3 100.2 101.3 98.5 100.0

Date of Final Disbursements June 15, 1988

-------------------------------- ------------------------------------------

C. Ptel a

ITSM DATE PLANND DATE N9VIAD DATE ACTUAL

Identification (ExecutiveProject Sumary) June 1978 July 1W76 July 4, 1978

Preparatien (a) June 1976 June 1979 Nov. 1978 -July 1979 (a)

Appraisal Miselon July 19W Aug. 196W Aug.-Sept.1988Loan/Credlt Negotiatione Apr. 191 May 1911 May -12, 1981Board Approval Apr./May 19l1 May 19l1 June I, 1981Le**/CeditS e $I July 8, 19llLoan/Credit Eff1etivene* Oct. a, 161 *- Dee. $, 1681Loan/Credit Cleeing Dec. 81, 1967 - Dee. 8w, 1667Lea/Credit Completion June S, 1967 June IN, 16 June so, 1981

(a) A total of five brief miseione.

0. a

Proe Cycle Planned Revied Final

Through Appreaial - 1 19.7Appraisal through

Board Approval - - 63.5Board Approval through

Effectiveness - - 5.8Supervision - - lit.#Completion Report (PCR) - - 5.1

TOTAL - - 224.6

E. M),giegg

Date No. of Man days Specializations Performs* Type ofMission Data (mo/yr) Persons in Field Represontod!/ Ratingls Trendl/ Problems4'

Identification 7/78 2 4 A;A - -

rroaration:4/79 2 3 AlE - - -

Mission I 8/79 1 4 A;E - - -Pre-A/rsal / 6 A;A;E;ARqjRA,LS - * -Amnsisaltrrough

5a Aeoval 1/" 9 18S A;AlE;E;ARq;RS;LS;SP;RA - - -

Board Aproval

venss - - - - * - -SupervislonstSupervision 1 61/02 2 IS A;RA 2 PPSupervision 2 6-9/82 2 SO A;CE 2 PPSupervision 3 3-4/88 2 88 A;CE 2 2 PPSupervision 4 1/88 1 is A 2 2 FPSupervision 5 4-5/64 1 a6 A 2 2 PPSupervision 6 9-9/64 1 12 A 1 1 PSupervision 7 8-4/85 1 16 A 2 2 -Supervision 6 9/88 1 9 A 2 2 -Supervision 9 5/66 2 as A;CE 2 2 -Supervision 1 12/6 2 88 AiCA 1 2 -Supervision II 5-6/87 1 21 A 1 1 -

A/ A Agriculturist; RS a Research Specialist (OPS); CE a Civil Engineer; CA a ConsultantAgriculturist;

E a Economist; RA a Research Administrator; LS a Livestock Specialist; ARq a Architect; SP a SoilSpecialiet

2/ 1 = Problem freor minor problems; 2 a Moderate problems; 3 a Major problems.

SI a Improving; 2 a Stationary; S a Deteriorating.

4/ F = Financial; P w Political.

Source: Bank Files

MAUEC ZEUInM1 AUDIT REPORT

SECOND AGRICULTURAL RESEARCH PROJECT

I. Introduction The IM1lcatios for IBAPA ofacceptance of these recomendat ion

The project was the second in a are discuseed, Including Its leading**ries of three World Bank-supported role In educating the *tat* researchagricultural research projects in and extension agencies In a farmingBrasil. Despite a declining system perspective.importance relative to other sectorsin the economy over the last 30years, agriculture still is a major 2. Objectivescontributor to national exports andemployment. Government has realised The project was designed tothe importance of continued reinforce and expand the conceptsimprovement in agricultural Introduced under the previoustechnology and given priority to the research project--commodity-baseddevelopment of its federal research research programming, a regionalcorporation, EHBRAPA since its approach to research based oninception in 1973. agroecological characteristics, and

the provision of research servicesThe PPAR supports the basic for these comodity and area-specific

research strategy of VMRAPA and program. It placed increasedconcludes that the project had a emphasis on expanding EHRAPA'sfavorable impact on institutional capacity to correctly Identify anddevelopment. It recommends, however, address production problem In thethat a more complete adoption of a major farming systems, on researchfarming system perspective would programing In accordance vi.himprove research programming and defined consti-ints, and on linkagestechnology transfer, especially with with extension-respect to the smallholder farmingcommunity. Sustainability of the The components Included civilreaarch system is only likely to be works, equipment, 1irary andguaranteed with further development documentation materials, extensiveof the concept of commercial technical assistance, substantialagricultural interests having to postgraduate and short-term training,substantially contribute to the cost Incremental research and supportof applied research and technology staff, and Incremental operatingtransfer.c costs; all components except

-vi -

incremental operating costs were the Amason region, also receivedsupported with Loan funds. Insufficient attention.

The project investments were A very positive aspect ofnot restricted to particular implementation was the excellent andgeographical areas, but most of the extensive collaboration betweensupported research progras had MBRAPA anZ the internationalrelevance to the les-developed agricultural research centers andNortheast, North (Amason) and Center- many national institutes and notableWest regions of Brazil. agricultural universities.

3. Implementation Experience 4. Results

The project was implemented The research carried out bygenerally in accordance with EMBRAPA since its inception (coveringappraisal design. The Bank agreed to both he first and second researcha higher cost sharing under a Special projects) has undoubtedly had aAction Program from October 1983 substantial Impact on agriculturaluntil project completion. This productivity In Brazil.facilitated implementation in aperiod of fiscal constraint and Due to the difficulty inencouraged Government to allocate the reliable linka3o of cause and effectreduced amount of counterpart funds. In agricultural research projects, an

economic rate of return calculationGovernment restrictions on was not attempted at appraisal nor in

incremental staffing delayed and the PCR or PPAR. However, EMBRAPAlimited the number of researchers did undertake Its own assessmentrecruited under the project. EMBRAPA (using a technique previouslyutilized consultant resources to develop*4 and used In Brasil) andlargely counter this deficiency, and estimated an ERR In excess of 402;the reduced staffing did not appear this figure is consistent with mostto have any long-term negative impact other estimations of returns onon the research program. However, research projects in Brazil andit is likely that there was a elsewhere.temporary negative effect on researchoutput when staffing restrictions Production statistics alsocoincided with relatively low salary suppot the positive impact of thelevels in the period 1984 to 1987. project, and disaggregationThe training program and consultant demonstrate& that yield increasesservices were much larger than have been important in the projectprogramed at appraisal. In targeted regions (pars 3.06).staffing, postgraduate training andthe use of consultants, the audit Bank support has encouragednoted that an appropriate spectrum of research Investment in the Center-specialties was covered, but that the West; this has been the major area ofareas of socioeconomics and agricultural expansion, with largetechnology transfer appeared to be research requirements to enableunder-represented; environmental sustainablep economic production inImpact assessment, so important for the generally difficult soils of this

* vil -

sene. The Northest, with a sector In technology transfer wouldpredominance of smallholders and be beneficial.difficult socioeconomic andagroecological conditions, has alsoreceived increased allocations In S. SustainabilityOAPA so that it is now not

substuatally disadvantaged oepared WRAPA has traditionally beento the more productive and wealthier given some priority In the allocationsouthern sones. Although the North of federal budgetary resources,benefitted under Agricultural despite severe fiscal constraints.Research II, some of the gains it However$ operational support funds Inmade In relative budgetary the last two years (post-projet)allocations have been lost subsequent have been restricted. The s ofto the project (Figure 1). the recurrent budget necessary to

operate ZWIRAPA demands thatA logical, iterative research comercial farmers and agricultural

programing mechanism has been fIrms bonefitting from researchdeveloped in EMBRAPA. However, the should contribute to the costs of theinteraction phase between program technology generation and transfercoordinators, individual researchers service. EHBRAPA has accepted thisand external entities has suffered in principle and has comenced a programsome programs in recent years due to to capture private funding resources.funding constraints. The mechanismhas also not always ensured that the Although the annual nationalreal production constraints in agricultural research expenditure hassmallholder comunities have been been Impressive (pars 3.28), theidentified and addressed. relative allocat,on to research

within the federal agriculturalDespite the favorable budget has not been high (Table 19).

production impact which reflects This is due to the very large ruraleffective technology transfer, the credit and other subsidy program. Atechnology transfer units at some rationalization of macroeconomiccenters can be made much more policies whi.-h reduce the need forefficient. These units and their thee subsidies would further enhancepersonnel have often not been fully the prospects for continuedintegrated as key components in the substantial funding for research byoperations of each research center. the federal Goverment.However, as the state extensionagencies have the majorresponsibility for technology 6. Findings and Lessonstransfer to medium-sised farmers andsmallholders, the variable capability Agriculture Research Il was anof these institutions must affect the effective Instrument in supporting

final impact of the transfer process. the growth of EMBRAPA as a soundPrivate sector technical assistance research institution.dealing with commercial farmers hasalready initiated links with EMBRAPA The PPARs on the relatedcenters, but a more systematic Agricultural Extension I projectapproach to the use of the private (6/88) and rural development projects

in the Northeast (6/88 and 6/89)

- vill -

highlighted the continued problems of With larger commercial farmerstechnology transfer, including the characteristics of farminglimited linkage between research and system can be ascertained withoutextension and limited capability in suo difficulty; real problem aremany of the state extension agenciea usually readily vocalized bythese sestrainta referred mainly to Individuals or producer associations.the asslholder farming subsector. With smalholders, access to

resources and education and socialThe PPAR on the Agricultural factors often complicate the behavior

Research I project (2/85) also of farmers a&-' demand a morestressed the need for more attention comprehensive approach toto technology transfer and the understanding the dyn=mics and needsorganisatton of adaptive research to of the farming system.ensure televance of technology.These aspects of the research 1ERAPA has experimented with ageneration and transfer process were number of approaches in dealing withgiven Increased attention especially groups of farmers (para 3.15, 3.23In the latter years of the second and 3.24) to determine appliedproject, but significant Improvements research needs and to transfercan still be made. technology. The audit considers

EHIRAPA Is now at a stage where ItEMBRAPA has introduced the can expand Its role in this area and

concept of farming systems into many take the lead in introducingof its progrem. The audit considers practical, replicable and cost-that a more complete adoption of a effective techniques (using FSP) tofarming system perspective (FSP) by state research and extensionENBRAPA managemnt and at all services. This will rely on workingresearch centers would ensure that with groups of farmers with similarthe major problems of all significant Interests and concerns, and willfarming systems are considered, that employ rapid field techniques toresearch is programmed accordingly define these groups and understandand that the most cost-effective the attitudes, behavior and technicalmechanisma for technology transfer problem of their members (par& 4.07are adopted. This would be and Annex 2). The audit considerscompletely consistent with EMBRAPA's this leadership role by EMBRAPA ascomodity research strategy, and does essential, despite the majornot Imply that there should be a responsibility for extension beingmajor increase in classical farming with the states and the Increasingsystem research or that sophisticated pressure for states to manage andsystem modelling should be given finance their own services Inpriority; it means that applied agriculture.research in comodities ordisciplines should be undertaken To ensure oustainability of thewithin the context of defined research system, the role of theproblems or potentials in significant special organizational unit initiatedfarming systems -- a "demand-driven* to capture private *ator funds forresearch program -- and that proposed research will have to be emphasized.solutions mst be relevant to the However, it cannot be expected thatresource situation of farmers. resource-poor smallholders will be

able to carry a significant share of

- ix -

these service costs, and public services should be obliged tosector funding will have to be contribute substantially topreferentially directed to services their cost; it behoves a largefor this clientele. The public and services institution likeprivate funding allocations will have DEBRAPA to develop practicalto be monitored to ensure there is and acceptable schemes tonot a trasfer of research reseurces maximise the capture of privatefrom the smallholder to the larger- sector fundinggfarmer, comercial subsectorclientele. It is also important that in agricultural sectorsscientific staff are not diverted comprised of a strong,from their primary research purpose commercial, large-farmin order to promote private sector subsector as well as a morefunding. numerous but resource-poor

smallholder subsector, specialLessons which can be derived measures must be taken to

from the Proiect ares ensure relevant technologyrequirements for the latter

- an outward-looking policy in a subsector are addressed; a FSPnational research institution, in research programing and inas reflected in a willingness the organization of technologyto seek and contract external transfer can provide thisexperience, to invest heavily mechanism;in external training, and tocollaborate with the - as a FSP is not traditional ininternational research centers agricultural research, aand with universities and comprehensive adoption of itsresearch entities of principles will not occur in anexcellence, can increase the institution unless there israte of attainment of a high strong leadership in this arealevel of institutional by senior management andperformance; research programmers, and

unless priority is given to- institutional development of an appropriate staffing, training

agricultural research system is and technical assistance toa long-term undertaking, and enable its adoption.rhe Bank has adopted a correctapproach in continuing tosupport EMBRAPA whileGovernment is willing to givesome priority to its role inagricultural development;

- while Government investment intechnology development andtransfer services can yield ahigh economic return, politicaland budgetary realities dictatethat commercial interests whichdirectly benefit from these

PROJECT PERFORMANCE AUDIT REPORT

BRAZIL

SECOND AGRICULTURAL RES3ARCH PROJECT(LOM 2016-DR)

I. BACKGROUND AND DESIGN

1.01 between 1950 and 1980 the total economy of Brazil grew at anexceptionally high average annual rate of 7.12. Although the agricultural sectorperformed well in this period (at a rate of 4.4Z, which was twice the annualpopulation growth), the outputs of the industrial and service sectors increasedat much faster rates (8.5% and 7.0% respectively). Consequently, agriculturedeclined in importance relative to those sectors and its reducing contributionto the economy is indicated in the following table:

Airicultural Sector Performance 1950-1987

1950 1960 1970 1980 1985 1988

Percentage Contributionof Agriculture tos

GDP 24 18 12 10 10 9Employment 60 54 44 29 29 25Total Exports 88 87 71 49 38 43

1.02 Government policies had a major influence in the growth of thedifferent sectors. A comprehensive import-substitution, industrializationstrategy was adopted, aimed at implanting a consumer durable goods industry inthe country. This involved a variety of policy measures including an overvaluedcurrency, highly restrictive protectionist trade policies on manufactured goods,and, at times, bans and quotas on agricultural exports. These policies hadnegative implications for agriculture relative to industry. Government attemptedto overcome this to some extent through subsidized rural credit, and fiscalsubsidies (tax credits, shelters, etc.) were also introduced to encourage farmersand investors to open up new tracts of land in the country's expandingagricultural frontier. Unfortunately, although the mix of implicit taxation and

-2-

agricultural subsidies resulted in the agricultural sector expanding at areasonable rate, it also served to widen individual and regional incomedifferentials. The principal recipients of the subsidies were larger producers,and the subsidies encouraged more capital-intensive, land-extensivo undertakings.This accelerated the land concentration process, biased production methods infavor of labor-displacing mechanization, promoted the more capital-intensiveexport/industrial crops over food crops, and contributed to significant rural-to-urban migration. Wheat was an exceptional crop which had specific, high levelsof protection'/.

1.03 Despite the increasing importance of the industrial sector, at thetime of preparation and appraisal of the Second Agricultural Research Project in1979 and 1980, agriculture was still a very important sector for the Brazilianeconomy (and remains so today)." Agriculture was assigned a high priority inthe 1980-85 development plan with objectives of (a) increased production of basicstaple and export crops, with special emphasis on assistance to small- andmedium-scale producers; (b) increased efficiency through integrated production,storages, transport and marketing; and (c) enhanced agricultural research andthe adoption of improved technology. These objectives were supported by the Bankand provided a suitable framework for the project.

1.04 The Bank had made 19 loans for agriculture and rural development inBrazil by 1980, including 12 for various settlement, irrigation and ruraldevelopment projects (10 of which were located in the northeastern part of thecountry), an agricultural extension project and the first agricultural researchproject. The first national extension project had been appraised to 1977 but wasexperiencing a slow start after effectiveness in September 1978. It aimed atstrengthening the agricultural extension system in Brazil by improving theleadership and coordination ability of the Brazilian Technical Assistance andRural Extension Company (EMBRATER) and promoted research-extension liaison. Thenortheastern rural development projGcts (POLONORDESTE projects) included fundingfor locally relevant agricultural research and placed major emphasis onagricultural extension. The initial agricultural research project was notcompleted at the time of design and appraisal of the Agricultural Research II,but achievements were already substantial in the construction and equipping ofa selected set of research facilities of the Brazilian Agricultural Research

I For further review of agriculture in the economy, see Brazil - AgriculturalSector Review: Policies and Prospects. World Bank Report No. 7788-BR, July 26,1990.

2. Brazil is the world largest exporter of coffee and orange juice, thelargest producer of sugar cane, and the second largest exporter of soybeans.

-3-

Company (EWRAPA)", in incremental staffing and training, and in recruitmentof technical assistance; promising results were being obtained in the productionof higher yielding and disease resistent varieties and in biological pestcontrol.

1.05 MSAPA propered the project with assistance from an internationalagricultural research consulting firm. It was designed to reinforce and expandthe concepts Introduced under the previous research project -- comodity-basedresearch programing, a regional approach to research based on agroacologicalcharacteristics, and the provision of research support services. It wouldcomplement exisatiag research programs and increase the utilisation of researchfacilities established under the first project, but would place greater emphasison expanding ENBRAPA's capacity to correctly identify and address productionproblems in the major production systems, on research programming and on linkageswith extension.

1.06 The project was not confined to a specific geographical areas withinBrazil, but most of the supported research programs had relevance to theless-developed Northeast, North and Center-West regions of Brazil4l which wereto receive priority under the project. At appraisal, the project had thefollowing components and baseline costs.

EMBRAPA was created in April 1973 as a public corporation under theMinistry of Agriculture with a large degree of financial and administrativeautonomy. Its main functions included direction and execution of agriculturalresearch, stimulation of decentralized research to the benefit of state and localinterests, coordination of research which Involves other federal agencies,cooperation with the federal extension coordination agency (EMBRATER) inpromoting transfer of technology to producers, and the channelling of federalfunding for agricultural research activities. In 1980 EMBRAPA's research stationnetwork included three regional centers, 11 national centers for commodityresearch and 18 state/territory level research centers (UEPAE/UEPAT). Inaddition, there were 11 semi-autonomous research companies which were linked tostate secretariats of agriculture, and three completely autonomous state-ownedresearch centers.

4. Brazil is divided into five geopolitical regions -- the humid tropicalNorth (Amason), Northeast (mostly the tropical semiarid son*), Center-West(Campos Corrados), the largely subtropical Southeast, and the subtropical andtemperate South (see Map IBRD 11884R2).

-4-

(a) research development, including;

(1) civil works at 31 major locations (eightof which were not included in AgriculturalResearch I), comprising mainly ancillarybuildings such as greenhooses, storagebuildings and some staff housing in moreremote areas, and a research farm,including an Irrigated *d productionunit (US$24.5 million); and

(i) equipment, comprising mainly scientificinstruments, laboratory equipment,vehicles, farm and laboratory equipment andoffice furniture (US$17.0 million);

(iii) library and documentation servicescomprising mainly journals, periodicals andreference books, including also somemicrofilm equipment (US$8.7 million);

(b) technical assistance and training, comprising about 156man-years of consultant services, postgraduate degreetraining for 205 staff, and short-term fellowships fora further 342 staff (US$33.4 million);

(c) incremental operating costs for a five-year period,comprising salaries and associated costs of some 224incremental scientists and about 1,100 support staff, inaddition to the incremental operating costs forvehicles, equipment and facilities included in theproject (US$31.0 million).

Components supported by Loan funds included civil works, equipment,bibliographic-materials, fellowships, consultant services and contracted researchservices.

II EMPLEENTATION EXPERIENCE

Overall Proarm

2.01 The project was Implemented generally in accordance with theappraisal design. The final evaluation report by EMBRAPA and the Bank's PCRprovided a comprehensive review of Implementation of each component. Thefavorable record of project execution is demonstrated in Table 1 which comparesthe appraisal targets with the accomplishments at project completion.

2.02 The project had a slower start than anticipated, priority being givento completion of the ongoing Agricultural Research I Project (which finallyclosed in June 1983 after an 18 month delay) and the PROCENSUL 1 loan of theInteramerican Development Bank - IDB (for which final disbursements were made inFebruary, 1983). In October 1983, the Bank agreed to a higher cost sharing untilNrch 1985 under its Special Action Progrm, which ws introduced as a temporarymeasure to alleviate the negative impact of severe fiscal constraints on criticaldevelopment projects in borrowing countries. The higher disbursement percentagesfor local expenditures were subsequently extended until project completion. Thetraining and consultant services components substantially exceeded appraisalestimates and Bank approval for this was reflected in reallocation of Loanproceeds. A total of seven amendments to the Loan Agreement were made (Table 2)to reflect the Special Action Program, reallocations and the inclusion of anadditional research center (for plant protection) in the financed program. TheLoan closed with virtually full disbursement on the originally planned date ofDecember 31, 1987 (Table 3).

2.03 There were severe fiscal constraints in the years 1983 to 1985, whichwere reflected in EMBRAPA federal budget allocations (Table 18), but this did notseriously affect the implementation of the project. The higher disbursementpercentage agreed by the Bank under Special Action Program undoubtedly was anincentive for Government to continue funding its program for development of theagricultural research system. However, general restrictions were placed on therecruitment of incremental staff by Presidential Decree at the beginning of theproject and remained in effect virtually throughout the project.

Staffina

2.04 Supervision missions repeatedly reported that EMBRAPA was unable torecruit the planned incremental staff for the project and this was a regularissue in supervision letters to Government. EMBRAPA attempted to overcome thisconstraint by contracting national and international consultants to fill the roleof permanent reserchers. Bank staff maintained that this policy was leas thansatisfactory, ai the intended technology- and skill-transfer roles of theconsultants financed under the project would be limited if they were not workingwith qualified counterpart staff. Other criticisms were that a deficiency insupport staff numbers meant that senior researchers were being occupied in taskswhich could have been done by support personnel, and maximm use was not beingmade of some equipment. Incremental staff were recruited in 1984 for the plantprotection research program (which was additional to the program included atappraisal - par& 2.02) and in 1985 to support a priority given to programspertaining to agrarian reform areas. Towards the end of the project in 1987there was a staff expansion associated with the creation of a number of newcenters and programs in EMBRAPA in 1986, and many contracted personnel were givenpermanent status. At project completion about 702 of the target for incrementalresearchers and about 1202 of the target for support staff were reached.

-6 -

2.05 The PCR concluded that a staff constraint did negatively affect theresearch program, especially when this coincided with relatively low real salarylevels in the 1984-1987 period. Although these constraints are likely to havehad a temporary effect on research output, the audit was unable to identify anylong-term negative Impact on institutional development. The number of researchersand support staff (technical and administrative) have increased gradually sincethe inception of MGRAPA and the proportion of higher degrees is now veryfavorable (Table 4).

2.06 With respect to the mix of professional specialities which hasevolved through and subsequent to the project period, the audit considers thatthe fields of soioconomics and transfer of technology appear to be generallymoder-represented in the programs (Table 5), and the Implications of this arediscussed in Section III. Another expertise which needs uch greater attentionis environmental impact assessment, especially considering the importance ofenvironmental issues in the North. Some program coordinators consider thatadditional researchers in some of the highly specialized fields such as geneticengineering and animal reproductive physiology are still wanting. The ongoingAgricultural Research III should assist in this regard through its fellowshipprogram where requirements can be justified, but the audit considers that veryspecialized expertise should be concentrated in centers with specializedfacilities so that most efficient use can be made of resources.

Training

2.07 The training component exhibited the greatest increase in Loan usagerelative to the original allocation (increase of 221%). This was caused partlyby an agreed increase in reimbursement percentage for local fellowships (from 80%to 100%) in June 1983 (Table 2), but mainly due to increases of 1032 and 2192 inforeign and local fellowships respectively for Ph.D and M.Sc courses, and 4502in short-term courses (Table 1). The two World Bank research projects, two IDBprojects and an earlier USAID project which supported training in EMBRAPA havemeant a steady stream of researchers have been given higher-degree training sincethe inception of the institution (Table 6). This has led to the healthy mix ofsenior and junior research staff and supporting personnel referred to inparagraph 2.05. The audit supports the emphasis given by ENBRAPA (with full Banksupport) to higher level training to build up the professional competence in theinstitution. The extensive short-term course program (Table 7) also appears tohave been justified to enhance technology interchange between researchers andwith extension personnel.5

s. The Agriculture Research III Project also has a major fellowships componentfor programs in the Northeast and Amazon, for selected support services and fortechnology transfer.

2.08 The training was satisfactorily distributed amongst the EMRAPA units(Table 8). Researchers at the state research units (UAPAE) at Boa Vista andNacapa did not receive any post-graduate training, but these units only graduatedto UAPAB status late in the project period and still are not well developed;staff at the other UAPAEs in the North were allocated substantial fellowships,as the preferential award of additional training was one of the incentives usedby DMRAMA management to attract staff to the remote units in this region.

2.09 ENMRAPA did not have a breakdown of the major fields of trainingundertaken by the researchers under the fellowship program. However, the auditexamined the apparent specialisations of those staff with M.Sc or Ph.D equivalenttrainig at eah of the three regional centers supported under the project(CPATSA, CPAC and CPATU)O in 1990, as per Table 9. To the extent that thisreflects training under the project, it suggests that there was a wide range ofspecialisations sponsored through the fellowship program, and the mainspecializations appear to largely reflect what would be demanded by the natureof the research programs at each center. This is only to be expected, as thefield of specialization of a candidate is first approved by the management teamat the center to which the candidate pertains, and after training the researchernormally returns to the same center. However, the absence of higher-trainedstaff specialized in socioeconomics and in microbiology in CPATSA, and theoccurrence of only one staff member with higher training in technology transferat any of the three centers are noteworthy. At CPATU, higher expertise inreproductive physiology would have benefitted the buffalo program and more skillsin entomology should have been beneficial for both the plant and animal researchprograms in this humid tropical environment. Although specializations in theecological area are represented, no staff at CPATU are trained in environmentalimpact assessment, which should be a priority in this zone. Consequently,although the fellowship training program appears to have been well implementedin general, improvements could have been made in the system of programming tomore appropriately cater for medium-and long-term requirements for specialistsin each of the research programs in EMBRAPA.

Consultants

2.10 The consultant services component greatly exceeded the appraisalestimate in use of long-term consultants (Table 10). The policy of liberalrecruitment of consultants appears to have overcome the potential constraint ofreduced staff recruitment during the project period. Unit managers consideredthat, in general, both the long-tern and short-term consultants had a verypositive impact in improving the technical and scientific competence withinVNBRAPA. The consultancies covered 88 specialities in response to the

CPATSAt Research Center for Agriculture in the Semiarid TropicsCPACs Research Center for Agriculture in the CerradosCPATUs Research Center for Agriculture in the Humid Tropics

- 8 -

requirements of each program. However, it is noteworthy that little assistancein the areas of agricultural economics/farming systems and technology transferwas recruited by the regional centers and comodity programs (Table 10-A).

2.11 The contract-research component was used very little (ten contracts).Apart from contracts for the administration of the external training program andthe consultant program (national and external), specific studies were contractedin socioeconomic evaluation and technology diffusion and on the highly technicalfield of genetic engineering.

Collaboration with other Research Institutions

2.12 IMBRAPA made very good use of the resources of the internationalagricultural research centers (IARC) and of many national agricultural researchinstitutions and universities of excellence in implementing the research programsunder the project. This was not only as assistance from these institutions inthe form of short-term or long-term consulta.acies by specialists, but also incollaborative research programs. The willingness of EMBRAPA to seek externalexpertise to assist it in developing its capability, and its characteristic ofbeing relatively well-managed and adequately funded, have led to profitable andcontinued relationships with external research entities. The relationships withexternal institution were initiated and organized largely through the individualcommodity, regional and support centers, with support from EMBRAPA headquarters.

III. PROJECT OUTCOME

Production Impact

3.01 The research carried by EMBRAVA since its inception in 1974 hasundoubtedly had a substantial impact on agricultural productivity in Brazil.There is no justification in trying to separate the Impact of AgriculturalResearch I and the Agricultural Research II (nor indeed, of the on-goingAgricultural Research III) as many lines of research represent sequential orcomplementary studies which do not fit into artificial project time frames.EMBRAPA's final evaluation report summarizes the main research products for theNorth, Northeast and Center-West regions which received priority underAgricultural Research II, and for the centers providing support services. Therange and number of commodities, disciplines and production systems which werestudied is impressive. (Table 11).

3.02 In view of the difficulties in reliably quantifying the economiccosts and benefits of research investment, an estimate of an economic rate ofreturn (ERR) for the project was not attempted at appraisal. The Department ofPlanning in EMBRAPA, however, with the cooperation of the socioeconomic units andresearchers at each center (and with assistance from selected extension staff inthe states) attempted to calculate an ERR for the research investment in the

- 9 -

three priority regions. This involved estimating the impact of 249 technologiesgenerated by research as net benefit to farmers, with benefits commencing in 1981and ending in 1997 and with costs of research starting four years earlier in1977. Costs of research were actual to 1987 and then maintained constant as acost to 1997. The not benefits which could be attributed to EMBRAPA's researchwere estimated annually to 1987 in most cases [and to 1991 for a few recenttechnologies which were in a phase of rapid adoption] and then maintained to1997. This resulted in an estimated MR of 40.52 on the research investment.

3.03 To estimate benefits, the EDRAPA ZRR method involved:(a) definition of each technology developed through research which was beingadopted by farmers; (b) calculation of the not benefit to the farmer in adoptingthe technology; (c) estimation of the extent of adoption of the particulartechnology by the farming community, and so the area affected by the adoptedtechnology; and (d) estimation of the extent to which EMBRAPA was responsible foradoption of the technology (via-a-vis other research bodies, extension services,input and seed suppliers, etc.) which ranged from 20Z to 1001 of the net benefit.A number of studies and surveys undertaken during the project provided data tosupport the calculation of net benefits to a farmer in using a technology and theestimation of adoption of technology for specific crops in particular areas. Themethod, however, is heavily dependent in most cases on subjective estimation inthe degree of adoption of each technology and the extent to which EMBRAPA shouldclaim the benefit of adoption. Consequently, the calculation is subject to alarge margin of error; never-the-less, the exercise did represent an attempt toquantify benefits and costs and arrived at a return which is generally consistentwith other calculation of returns on agricultural research investment in Braziland elsewhere.

3.04 Production statistics can also provide a measure of the impact ofagricultural research. Tables 12 and 13 show rates of growth of export crops andfood crops, disaggregated for increments due to area expansion and to yieldschanges, for the period 1965-1980 and 1980-1989. Although the net productioneffect also depends on Government policies and relative prices and profitmargins, agricultural research could be expected to contribute significantly toany substantial yield increases, and to area expansion where new technology isrequired to permit economic crop production in expansion areas.

3.05 Very high growth rates in the period prior to Agricultural ResearchII occurred in soybeans, oranges and wheat; this was largely due to areaexpansion, and EMBRAPA research contributed substantially to the increase insoybean area by developing varieties suited to the different soil conditions ofthe Center-West. In the period after 1980, the spectacular growth in crop areasdisappeared, but significant production growth occurred in most of the importantcrops (cassava and coffee being exceptions). Yield increases were very importantfor cotton, wheat and rice and significantly influenced sugar, tobacco, maize andcassava production. With the exception of tobacco and sugar, EMBRAPA researchcontributed to these increases through development of varieties to meet the

- 10 -

requirements of particular production conditions, farming systems and pestchallenges, as well as through Improved cultural techniques for these crops.

3.06 Tables 14 and 15 disaggregate growth rates of export and food cropsby region for the 1980-1989 period. These show substantial growth in the yieldsof cocoa In the Northl soybeans in the Northeast, Center-West and Southeast;cotton in the Northeast; and sugarcane in the Center-West. Food crop yieldincreases are very significant for wheat in the Center-West, Southeast and South;rie in the Northeast, Center-West, Southeast and South; maise in the Northeastand Center-West; and beans in the Center-West. This disaggregation demonstratesthat yield increases have been Important in the project target regions, althoughDGNRAPA research was not significantly involved in the cocoa and sugar increases.

3.07 Livestock statistics are not precise enough to allow anyinterpretation of productivity changes.

Regional Allocations of Resources

3.08 In both the Agricultural Research I and Agricultural Research IIprojects, an important objective was to broaden the scope of the researchprograms in the less-well-developed geopolitical zones of Center-West, North andNortheast regions. Figure 1 and Table 16 indicate the application of funds tothe five Brazilian regions from 1974 to 1989. Until the late 70s, the Southreceived the highest allocation of federal funds followed by the Southeast;this, despite well-established, state-funded research institutions in Sao Paulo(Southeast), Parana and Rio Grande do Sul (South). From 1979 to 1985 the Center-West received most attention, while the allocations to the Northeast increasedto a level in the vicinity of the two southern regions. The North regionreceived improved support in the 1982 to 1986 period under the AgriculturalResearch II project although it continued as the region with the lowest fundallocations. After 1985, the Center-West lost its dominance to the Southeast,but the relative differences in fund allocation between regions decreased, albeitwith the North continuing to receive the lowest allocation. The number ofresearchers has been and is highest in the Southeast and Center-West and lowestin the North, in general coincidence with fund allocations (Table 5). Withrespect to EMBRAPA staff seconded to other research institutions, the Northeasthas been the biggest recipient, followed by the North.

3.09 It can be concluded that the continued support under the Bankprojects has encouraged research investment in the Center-West. This has beenthe major area of agricultural expansion with large research requirements toenable sustainable, economic production in the generally difficult soils of thissone. The North-East has also been receiving slightly increased allocations sothat, in absolute terms and as a region, it was not substantially disadvantagedcompared to the southern regions. However, a large number of factors must betaken into account when determining and deciding on allocation of researchsupport amongst regions.

•11 -

3.10 The folloi table cpares so0 relevant characteristics of the regSom.This show. that the North should receive high priority vith regard toenviromantal concerns, production diffticulties damanding researeb, and potenttalfor greater use of xieting farm area. It does not, however, appear to warrantspeatal attention from the viepolnt of agricultural D?, nmber of produere or

lIvsteek reseurees, although the need to iprovø fare IanGe en exst lghøldiga ls «clipoed only by the North-gant.

Bragll. lseilecd fiaeuil theaerle

Agri.ultuaral £SP (1989) I 5.9 19.5 84.7 29.5 11.4 £86

(095) N. 886,90 2,919,9W 900,9 18902,9W 16,9 W 80,S6.6 48.8 17.9 i.6 .4

Averag. FormInee~ Index (1989) 16 10 4V 828 182 -(North.a6 u 1- )

EV I rØneonte 1con~erne VN M M N N -

Pred~eln 2D0fflcultIs VN VN M li N -

Llvostek Shre (965)'Cattig N 4.2 17.5 27.9 19.4 81.0 1

Pige 9 7.1 26.1 18.2 8.7 9.9 1Poultry 1 *.A 18.1 84.2 8.7 6.2 1W

Uil lzød land as ITotal Producolve 77 89 97 97 92 wFore Land (19M)

1 S~ure: Brazil Agrieulsursi eter Revew. Werid Dank Reperl.* 778~-M,July 26, 19.

2 IeyO V = very 64gh; N w high; M derete.

8 Sour~: Draft Liveteek Se~tr R~prt. World Bank Repert * 8679-S,June 26, 199.

- 12 -

Institutional Impact

3.11 The research concepts introduced under the earlier project(commodity approach, regionallagro*cological emphasis, and service/supportresearch units) were further developed under Agricultural Research II. The auditfocussed on a number of operational mechanisms within this conceptual framework -- the efficiency of research programming, the effectiveness of the technologytransfer process, and the linkages of DBRAPA with other institutions -- all ofwhich are related to the design of research to meet real production needs and theadoption of the technology generated by the research.

3.12 Research Programina. An iterative research planning process isundertaken annually in EMBRAPAs (a) individual researchers propose subprojects,indicating how the subprojects fit into the national and local agriculturaldevelopment plans, and how they relate to past and/or ongoing research work; (b)these are reviewed at the station level initially and subsequently passed, ifconsidered feasible, to the comodity, regional or supporting service researchcoordinators; (c) the research coordinator then review the submissions and aworkshop (scheduled and coordinated by EMBRAPA headquarters) is arranged todiscuss them with representatives of the participating stations and institutionsinside and outside EMBRAPA", including the extension service and farmerrepresentatives; (4) a final program is decided when budgetary provisions areconfirmed. Funds aon-Loan) were earmarked in Agricultural Research II to ensureadequate travel ond interaction between the program coordinators and theImplementing units in the phase described under (c).

3.13 The audit found that staff were aware of and were generallycomfortable with the programing mechanism. However, in the constraints tooperating funds experienced in the last two years (par& 3.30), the interactionphase has fallen down to a large degree; there was vatiation, however, amongstthe programs depending on the importance placed on the process by the coordinatorinvolved, e.g., CPAC in the Center-West allocated more funds to the interactionphase than CPATU or CNPGC. This problem is exacerbated in CPATU in the Northbecause of the huge distances (and travel costs) involved, and where,unfortunately, there is special need for coordination and prioritization ofresearch by EMBRAPA and other research institutions because of the extensiveproblems and limited funding resources available in this region. The audit alsoobtained the impression that the research coordination mechanism did notsystematically incorporate research being done by institutions which did notreceive federal funding through EMBRAPA, so that informal personal contacts wererelied on for knowledge of lines of investigation by these entities.

The EMRAPA federal budget is also used to fund research undertaken by non-EMGRAPA units, especially the state research companies (Table 18).

- 13 -

3.14 The overall outcome of the use of the programming mechanism ast beassumed to be reasonably good, as evidenced by the favorable results beingobtained in the generation of appropriate technology. However, despite thisgenerally favorable result, there are instances where constraints in some typesof farms are not being addressed, especially those concerning smallholders.sIn the case of research geared to investigating problems and constraints oflarger, comercial farmers, the planning mechanism has been and will continueto be reinforced by the demands expressed by this type of producer directly tothe research network; this will help ensure research is responsive to theseenterprises. However, in the case of smallholders (minifundio) the problems andthe research requirements to meet farmers' needs are usually not vocalized aswell and are often not easily identified. Social, microeconomic, risk andresource factors peculiar to one type of farm can make an innovation which isappropriate in one situation inappropriate in another. It is necessary not onlyto ensure that attention is given to smallholder situations, but also that thedecision-making rationale of the smallholder with respect to adoption oftechnology and management of resources is understood by the research planner;without this knowle'ge there is a serious danger that research may not bedirectAd at appropriate solutions for this subsector.

3.15 The SAR indicated that a farming system approach would be used toensure that research is directed at the real problem of all classes of farmers(para 1.05). The regional centers fot the Corrados (CPAC), the semiaridNortheast (CPATSA) and the humid tropics (CPATU) were created in the mid-seventies to promote analysis of the agroccological environments, socioeconomicconditions and farming systems in each zone to facilitate research planning;these three centers have had extensive support under all of the Bank-supportedresearch projects. In Annex 1 to this report, the audit reviews how this concepthas developed in each of the t'.ree centers by examining the degree to which a

** Throughout this PPAR it is assumed that, despite the relatively lowoccupation of agricultural land and contribution to total agriculturalproduction, smallholders represent a major proportion of farm families (Annex 1,Tables 1 to 3) and continue to be a focus of Government and EMBRAPA assistance;this is consistent with the sectoral objectives described in par* 1.03.

* 14 -

tansiag system perspective (M8)" has been adopted in problem definition,appied research planning and taplementation, and technology transfer. Thereview coecludes thats

(s) Although all centers have a nSP to sam degree, a fullappreeation of the usefulness and ned for a PSP hasnot yet been uniformly developed among researchers InEBRAPA, and there are obvious differencs between

centers and program in the importance placed on F8PS

(b) CPATSA has allocated more attention to SP than othercenters, and this may have been associated with a moreapparent need because of the predominance ofsmallholders In this difficult semiarid environment;

(c) although CPAC has initiated work on analysis of allfarming systems in a number of districts in Golasstate, there does not appear to have been sufficientattention given to identifying the situation of thelarge proportion of suallholders who farm in thisregion;

The full use of FSP involvesa(1) an appreciation that farming represents the integrated use of

resources available to the farm managerldecision-msker, includingland, water, labor and capital, and that factors such as access toInputs, availability of markets, access to technology, assessmentof risk relative to anticipated returns, and production goals andsocial expectations are likely to influence a decision by a farmerto adopt a particular technology in his farming system;

(1t) definition and localisation of significant farming systems,Including the resources available to members of a system orsubsysteml

(IIi) production problems, and the factors influencing their managementand Investment decisions;

(iv) programing applied research to solve real production constraintswhich takes into account the accessibility to resources by thefarmers in each farming system or subsystems; and

(v) organising the transfer of technology to resolve problems andrealise potentials by making full use of the similarity in values,concerns and attitudes which are identified with members of aparticular farming system or subsystem (e.g., working with subsystemfarmer groups in a structured process).

- 15 -

. (d) CPATU has adopted a F8P in developing farming systemswhich will allow sustainable agricultural developmentin particular situations, but less attention has beengiven to clearly identifying the conditions of thelarge numbers of smallholders as old and especially newsettlers In the Amason regionj some of these situationsmay have little prospects of becoming economicallysustainable systems, but at least all should be definedand analysed as part of the research planningprocesslo and

(a) PSP Is not only essential as a research planningapproach, but is also essential for effectivetechnology transfer with amallholders; this has beendemonstrated at CPATSA and more recently at CPAC and isdiscussed later in this report (paras 3.23 and 3.24).

3.16 The introduction but less than full adoption of the FSP may berelated to the number and status of the agricultural economists/sociologists(grouped as socioeconomists) in DEBRAPA. Table 5 shows that the number ofsocioeconomists employed by EMBRAPA in the five regions represents about 4.7% ofthe total number of professional staff, with the North and Center-West having thelowest proportions. Considering the role that socioeconomists should play inresource analysis, research programming and interpretation of research results,this proportion appears low, and should probably approach 10Z of the totalresearch staff. An analysis of long-term consultancies in Agricultural ResearchII (Table 10) also indicates that this field received minor attention relativeto other aspects of research in the regions, and most of the consultants werelocated at headquarters.

3.17 The socioeconomists working in the regions usually were dispersedthroughout the EMBRAPA centers. This had the advantage that some economic inputwas present at most centers, but had the disadvantage that input was variableaccording to the capability and interest of the individual socioeconomists.There was an absence of mutual support and common direction which could haveenhanced the FSP in the agricultural research process. The text of the SARplaced major emphasis on FSP, but the working papers which detailed existing andincremental socioeconomist positions appeared to give less weight to this aspect

i' EMBRAPA, in analyzing resources and farming systems, should also take intoaccount its function as a federal institution which can provide reliable data andinformed opinion on development and settlement policies by federal and stateauthorities.

- 16 -

of research.111 This suggests that at the time of appraisal, the exact role ofsocioeconomists was not entirely clear, and, hence, the use of theseprofessionals was variable according to their capability and preferences andthose of the center management.

3.18 Tecbnolg Transfer. One objective of the project was to improvethe transfer of technology to farmers. This was to be brought about by (a) closecollaboration between EMBRAPA and 11RATER which was responsible for coordinatingextension programs in the state and territory extension affiliates(ENATERsIASTER) so that they were in accordance with national economic andsocial development objectives, and for stimulating research-extension - farmerlinkages; (b) improved liaison between extension and research personnel at localunits; and (c) special projects of collaboration with state research andextension agencies in the Northeast through rural development projects where thegreater difficulty of technology transfer to smallholders was recognised.

3.19 In general, technology transfer, as reflected in adoption byproducers, appears to be reasonably effective with larger scale, comnercialproducers. Just as this type of producer is inclined to articulate(individually, or through farmer or commodity associations) to researchers theproblems and constraints in his production system (as previously indicated in3.14), he is also naturally receptive to technology developed by research toaddress these constraints. With smallholders, the transfer process is not assimple and is inhibited by (a) research not always developing technology whichis relevant to the circumstances of smallholders, whicl, in turn, relates to notbeing fully familiar with the characteristics of the farming systems; (b) thevery poor resource base of many smallholders, which can limit the adoption ofmany technological improvements unless they are supported by some form ofgovernment assistance/sulosidyf and (c) a less-than-satisfactory capacity in manyof the EKATER state extension *Zoncies which have responsibility for extensionto smallholders. The collaborative programs with state agencies in the ruraldevelopment projects in the Northeast had mixed results (pars 4.04).

3.20 Within EMBRAPA, most research centers now have a technologydiffusion unit and a well developed system to produce technical publications.However, the importance allocated to these units and their integration into theresearch operations is very variable. In some cases, the unit is used to providetechnical advice to farmers through face-to-face contact in meetings withproducer groups, associations and other service entities, as well as through mass

u. The programs in seed production, animal husbandry, forestry, agriculturalengineering and bioenergy did not include socioeconomists in existing orincremental staff positions.

* 17 -

media W; this approach is used at CPAC, albeit mainly with larger farmers, andrelies on the staff of the technology diffusion unit having a sound knowledge ofthe research programs, of producer circumstances, and of the economics andapplicability of technology in these circumstances. At many other centers, theunit is used simply as the coordinator or facilitator of contact betweenresearchers and producers, with most resources being allocated to the preparationof technical publications.

3.21 In general, it appears that the staff of the technology diffusionunits have not yet been given the same status as researchers in the same centers.This is reflected in (a) a tendency for staff in these units to take higherdegrees in other fields and move out of the technology diffusion area; (b) therestriction of activities of the unit at many centers to production ofpublications and coordinators of contact with farmers, rather than directconveyors of technology to farmers; (c) a low proportion of staff with higherdegrees (45% at B.Sc level compared to 22Z for EMBRAPA overall - Table 5); and(d) very low use of consultants to assist in developing an effective technologytransfer system (Table 10-A). This suggests that the most effective use is notbeing made of the EMBRAPA technology diffusion units.

3.22 Apart from EMBRAPA and EMATERS, the federal and state universitiesare a public source of information for producers. However, the private sectoralso has an important role in technical services in Brazil. This occurs in theform of farm management consultants who are hired directly by commercialproducers, of technicians hired by cooperatives and farmer associations, and oftechnical staff employed by input supply firms (e.g., fertilizer, pesticide andmachinery) and agroprocessing establishments. The ratio of public to privatetechnical assistance professionals in Brazil is 1.2 overall, but ranges from 7.5in the North to 0.5 in the South region (Table 17). Private professionals dointeract with EMBRAPA centers, but EHBRAPA has not yet adopted a systematicapproach to maximize the use of the private sector in technology transfer;private services can be a very powerful tool for agricultural extension,especially for the commercial farming subsector.

3.23 EMBRAPA, in October 1990, has officially assumed most of the federalcoordination responsibilities of the former EMBRATER which was abolished in March1990. At the time of the audit, EMBRAPA was in the process of formulating itsstrategy to assume these responsibilities. This offers a timely opportunity forthe institution to develop not only the most appropriate mechanisms for transfer

is The common forms of publications issued by the centers are Research-in-Progress Bulletins (for researchers and research institutions), ResearchBulletins (for academic-scientific community), Technical C Amunications andTechnical Circulars (for extension staff and farmers, dealing with practicalrecommendations), Annual Technical Reports (for technicians and institutions),and Special Reports on specific aspects of agriculture.

- 18 -

of technology, but also to install the technology diffusion units as integralparts of the technology development and transfer process at each center. Theaudit considers that the technology diffusion units must be closely linked withthe socioeconomists at each center. This is not only because of the necessityto have all technical recommendations presented in an economic context, but alsobecase a detailed knowledge of the characteristics of existing farming systemsis essential, especially for smallholder farms, if technology is to be presentedso as to be seen as relevant to the situation of farmers; the FSP is necessarynot only for research programing, but also for technology transfer.

3.24 Through its experimentation in farming system analysis and inextension forums with farmer groups and associations, SMBRAPA has accumulatedsignificant experience on which to base an -expanded program of technologytransfer. Annex 2 presents and develops a technology development and transfermodel which could be designed from this experience, and examines its Implicatoafor EMBRAPA." This analysis implies that more effective use of SRAPAresources will ensue if (a) the FSP approach is accept4i throughout EMBRAPA asthe key to effective applied research programming and technology transfer,realizing that FSP is most necessary with smallholders; (b) the socioeconomicand technology diffusion units are closely linked at each center to enable thefull development of FSP for this purpose; (c) EBRAPA takes the lead in workingwith state research and extension bodies to establish practical and replicablemethodologies using FSP in these institutions to guide applied and adaptiveresearch and the extension of technology; (d) maximum use is made of establishedproducer groups and private sector service organizations in the technologytransfer and feedback processes; groups of farmers with similar values andcircumstances ("reference groups") would be defined and used as the contactmedium where formal associations do not exist or are inappropriate for thispurpose; and (a) EMBRAPA works directly with producer organizations representinglarger comimercial farmers and with national agriculture sector entities, has acentral mass media program on topics which should have a broad sectoral impact,and has more localized communication programs from regional centers which takeinto account the common circumstances of the farming systems to which atechnology is directed.

3.25 Linkages. The linkages within EMBRAPA are well developed, and thishas undoubtedly been facilitated by the commodity and regional orientation of theresearch program. Interaction with the international agricultural researchcenters has been excellent (para 2.12), and many universities of internationalrepute in particular areas of research have collaborative programs with variousEMRAPA centers. Linkages with the state research systems have been encouragedby the federal funding mechanisms for specific state research projects throughEMBRAPA. However, the transfer of federal funds to the states has been

W. The audit appreciates, however, that the influence of EMBRAPA on KATERservices m1st be limited by the extent to which individual state governments andinstitutions are willing to cooperate.

* 19 -

drastically reduced in the last two years along with the allocation ofoperational travel budgets in ENBRAPA (Tables 18 and 19), both of which factorswill act as a disincentive to interaction between the federal and state researchnetworks. The reduced federal transfer of research funds to the states appearsto be a conscious Government decision to encourage more state responsibility inthis area. The audit considers that it behoves ENBRAPA to strongly urge theroteation within its budget of a reasonable allocation for research to be carriedout In cooperation with the states; this would enable it to more effectivelyinfluence those state research systems which are in need of substantial1mprovement. At the same time EBRAPA should accelerate its program of technicalassistance and training to state research institutions and promote adequateresearch funding by state governments.

3.26 Linkages with EMATERS exist but their effectiveness varies. Theexisting mechanisms for linkages with CATER need to be reviewed in the contextof improved technology transfer using F as presented in para 3.24. Effectivedirect linkages between MEBRAPA and producer associations have developed, oftenalong commodity lines (e.g., beef, wheat and soybean) and these are likely to befurther developed as EMBRAPA articulates its new policy on technology transfer,and producer organizations are asked to fund a greater proportion of researchcosts (para 3.33).

Sustainability

3.27 EMBRAPA is well established as a large and capable national researchinstitution. However, with an annual cost to the federal treasury of aboutUS$200 million, the sustainability of the system must be examined. The budgetsallocated to EMBRAPA and the funds utilized annually from 1974 to 1989 arepresented in Table 18 and Figure 2. These show that the fund allocations toEMRAPA progressively rose from its inception in 1974 until 1982 and thensuddenly dropped in 1983 to 1985 before steadily rising again to 1989, withoutreaching the peak figure of 1982 in US dollar terms. The years of 1981 and 1982had very high external credit usage which coincided with the latter years of thefirst BID project and the Bank's Agricultural Research I, both of which alsoattracted heavy government counterpart funding; major expenditures on civilworks and equipment occurred in these years.

3.28 A number of expenditure ratios for the agricultural sector arecompared for 1980 to 1989 In Table 19. These show that the federal funding ofagricultural research has been between 0.522 and 0.711 of agricultural GDP,except for the 1980 to 1982 period when it was from 0.82 to 1.42. Theseallocations do not include funding by the states for which an expenditure serieswas not available; spot budgets, however, suggest that the state funding islikely to increase the total expenditure by 152 to 252. This implies an annualnational agricultural research expenditure of 0.62 to 0.92 of Agricultural GDP.This is higher than the average for developing countries as a group in the period1980-1985 which was about 0.422 of Agricultural GDP, and could be equated withthe lower income group of the developed nations (at 0.662) but would be below the

- 20 -

higher income group (at 2.43Z). The Brazilian research expenditure figures donot include existing private sector funding for commodities such as cocoa, coffeeand sugar.

3.29 When compared with total expenditure by the federal government inthe agricultural sector, it is interesting that EBRAPA was given relatively highpriority In 1983 and 1984 which were years of major fiscal constraint; relativeallocations to research in later years, however, declined. This decline occurredwhen there was a massive increase (1072) in total allocations to agriculture inUS dollar terms in 1985, and again in 1986 (722) and 1987 (42Z); the major shareof this increase was through the Ministry of Finance as rural credit and othersubsidy programs which reached nearly US$6.0 billion in 1987. m. Consequently,the allocations to research have not been high relative to other agriculturalsector allocational this suggests that a rationalization of macroeconomicpolicies could reduce the need for agricultural subsidies to correct economicimbalances (pars 1.02) while permitting continued substantial support forresearch without undue drain on total federal expenditure.

3.30 Data in Tables 18 and 19 show that in the years subsequent toAgricultural Research II the ratio of salaries to salaries plus operating costshas increased to an unhealthy 0.84 in 1989, and the audit observed that it is atleast as bad in 1990. A 60s40 ratio would be more appropriate to ensure that theconsiderable manpower and equipment resources (and cost) in research are fullyutilized. The decline in operational support occurred uniformly for all centersvisited by the audit mission. The audit has already mentioned that travel andinteraction between researchers and other sectoral participants in the researchplanning process have been curtailed in 1989 and in 1990 (para 3.13), and theconstraint has also affected the scale and supervision input on individualresearch projects.

3.31 With respect to the sustainability of the staffing establishment,following a period of severe restrictions on staff increases in the early- andmid-eighties (for all federal offices), there were significant increases inresearch staff in 1987 and 1989, and a substantial increase in support staff in1989 (Table 4). However, in 1990 all federal institutions were obliged to reducestaff. Obligatory reductions were commonly in the order of 30% (or even completedissolution in the case of EMBRATER) but EMBRAPA had to reduce by 72. This wasachieved and exceeded with a reduction of about 1000 staff including 125

" Brazil: Agricultural Sector Review, World Bank Report # 7798-BR, July 26,1990, Table A-3.30

- 21 -

researchers; the mix of higher and lower level researchers was not alteredlU.The reduction was done according to logical criteria (e.g., near retirement,poorer performance in training or operations, and voluntary resignation) and hashad no apparent negative effects, nor is it likely to have; the support staff-researcher ratio remained satisfactory at about 3.75.

3.32 Overall, DEGRAPA appears to have been given some priority in theallocation of federal budgetary resources since its inception. In the mostrecent years, however, a tight fiscal situation has demonstrated theprecariousness of having a large institution with highly qualified staff andsubstantial equipment investment being almost completely dependent on a federalbudget. With the absorption of most of the functions of EBRATMR, some increasein budgetary demand can be expected (although not through a large intake ofadditional staff) unless alternative sources of funds are found. Withoutadequate operational support funds, the salaried staff cannot produceefficiently. A rationalization of the macroeconomic policies which reduce thesectoral allocations through subsidies could enhance the availability of sectoralfunds and, hence, the willingness of the government to provide continued supportto EMBRAPA for agricultural research and technology transfer. Even with such arationalization, however, it still behoves EMBRAPA to seek the maximum non-government (private sector) support for its continued operation.

3.33 EMBRAPA management has accepted that those who benefit fromagricultural research expenditures should contribute to the cost of such researchwhen they have the capacity to pay. With this in mind, EMBRAPA, during itsreorganization in September 1990, created a department dealing with technologytransfer (DPD) and one dealing with promotion and organization of private sectorfinancing for research and marketing of the output of EMBRAPA's resources (DCO).Private sector organizations have already financed some research projects whichwould address their specific problems, and many farmer associations and firmshave indicated their willingness to fund research as long as it is seen asrelevant to their needs. The audit considers this to be a very positivedevelopment which enhances the prospects for the sustained agricultural researchprogram.

IV. FINDINGS AND ISSUES

4.01 Agricultural Research II was the second in a series of three Bank-supported research projects in Brazil, and has been an effective instrument insupporting the growth of EMBRAPA as a national research institution.

is The mix within the professional staff specialties, however, is nottompletely satisfactory in the opinion of the audit; the socioeconomist andtechnology transfer specialists need greater emphasis (par& 2.06).

- 22 -

PPA on Related Projects

4.02 The PPAR on Agricultural Research Iml Indicated favorable projectImplementation and impact but recommended improvement in the selection oftrainees and in the management of consultants, and greater attention totechnology transfer and to on-farm adaptive research. It considered that thelUnkage between the regional and commodity research centers in SIBRAPA was good,but that the linkages between these and the state research units were not welldeveloped. The second project largely overcome the deficiencies in traineeselection and consultant management, but despite improvement in the remainingareas of concern, some deficiencies persisted in technology transferladaptiveresearch and the effectiveness of linkages with national institutions varies.

4.03 With respect to technology transfer, EMBRAPA was unable to 'unctionas an independent operator and had to interlink with DNBRATER and the state andterritorial extension agencies (MATERs and ASTERs). Comments of the PPARs onthe Agricultural Extension I project and the POLONORDESTE rural developmentprojects in the Northeast (which included extensive technology development andtransfer components) are, therefore, relevant. The PPAR on AgriculturalExtension I concluded that the quantitative implementation goals wereachieved and studies on the use of farmers groups in the extension process wereuseful; the linkages between research and extension, however, were found wanting,as was the capability and commitment of many of the EMATERs, even though staff-farmer ratios were very high.

4.04 The PPARae on a cluster of POLONORDESTE projects generallyconcluded that the projects expanded the technical services available to farmers.However, they pointed out that technical advice was largely *captured* by thesubsidized credit program, so that staff became accustomed to providing adviceas packages supported by credit at negative rates of interest. This discouragedthe development of an analytical and socioeconomic capability in extensionpersonnel. Although a group extension methodology was used, the fortnightlymeeting schedule and restricted transport implied very high ratios of extensionstaff to producers, leading to high public sector recurrent costs. Attempts weremade in the projects to define the production problems confronting farmers as abasis for applied research programs, but the audit noted the danger of developing

if PPAR Brazils Agricultural Research I Project (Loan 1249-BR); ReportNo. 5474, February 19, 1985.

17 PPAR Brazils Experience in Development of the Rural Sector. World BankReport No. 7331, June 27, 1988.

U. PPAR Brazil: Supporting Rural Development in the Northeast, World BankReport No. 7910, June 28, 1989.

- 23 -

technology packages which largely depended on the use of subsidized credit. Alater study on the dynamics of rural development in the Northeast reviewed -2stand on-going experience in both the POLONORDESTE and a subsequent seria ofprojects in this area known as PAPP (Programa de Apoia so Pequeno ProductorRural). The authors were critical of the EMATER services except underextraordinary "crisis circumstances", but did comment favorably on casesinvolvi research and extension personnel working jointly with farmergroups*

4.05 Both the POLONORDESTE projects and the extension project dealtlargely with smallholders and the PPARs generally express the same concerns asthose of the current audit for this group of producerss (a) the need to ensurethat not only is applied research directed at the real production problems ofsmallholders, but that proposed solutions are applicable to their resourceconstraints; and (b) cost-effective means of technology transfer to the verylarge number of smallholders in Brasilian agriculture must be found to minimizethe demands on the recurrent budgets of federal and state agencies.

19. New Lessons from Old Projectst The Dynamics of Rural Development inNortheast Brazil by Judith Tender, et al., August 24, 1990; Report submitted toOED. The authors maintained that extension tended to perform satisfactorily onlyin circumstances where there were clearly defined and limited objectkves, aspecific time frame for execution, the results were definable, and the technologywas in response to a need realized by political and institutional bodies as wellas by the benefiting farmers, e.g., combating a disease outbreak, or changing anexisting variety to overcome a wide-spread pest challenge and provide higheryields. This was characterized as specific "demand-driven" responses as opposedto a general strategy of providing an effective technology transfer service toovercome problems and increase the productivity of the major production systemsin each extension area. This type of institutional behavior is understandable,but the audit considers that an acceptance that extension services should onlybe expected to perform in "crisis" situations would not be an effective use ofthe large recurrent expenditure incurred in maintaining an extension service.The same report commends the few attempts to introduce a system in which researchand extension bodies directly interact with farmer associations/communities andagree to provide a specific service for a defined period to meet the expresseddemands and needs of farmers in the association. This suggests to the audit thatdeveloping a capacity in the research/extension bodies to identify real needs ininteraction with farmers, and then to use their resources to respond to thoseneeds with technology relevant to the circumstances of farmers, are the keyinstitutional improvements which are necessary.

- 24 -

Research Programming and Technolo&X Transfer usina PSP

4.06 The audit considers that a more complete adoption of FSP by EMBRAPAwill help to ensure that the major problems of all significant farming systemsare identified and considered. The same FSP approach can also provide aneffective basis for technology transfer to smallholders by having the extensionservice deal with groups of producers who operate in the same farming systemsunder similar resource constraints. Rational use and frequency of meetings withgroups can reduce the need for extension staff and operating costs so that thedrawdown on annual recurrent expenditure is kept at acceptable levels.0 Forlarger, commercial farmers the advantages of F8P still apply, but thecharacteristics of the farming system are usually less complex and more easilyascertained. The use of FSP can successfully guide applied research and thetechnology transfer process in all production situations, but is essential andmore demanding with smallholder communities.

Action by EMBRAPA

4.07 Although EMBRAPA is principally a research institution, its prestigeand its newly-acquired responsitsilities for coordination of extension make it theappropriate institution to lead a program of wide-adoption of FSP by the stateagencies for agricultural research and extension. The approach is elaborated inparas 3.24 and Annex 2, and will require a number of actions and changes inemphasis by EMBRAPA if it is to be adopted successfully on the scale required:

(a) a comitment by all levels of management in EMBRAPA toimplement the concept as an essential part of its program;

(b) a realization that FSP is even more necessary in dealing withthe smallholder communities which are very dominant in theNortheast, and substantial in parts of the North and Center-West, and a willingness to allocate resources accordingly tothis activity;

"0 In many of the POLONORDESTE projects, smallholder meetings were scheduledat fortnightly intervals and had very high extension staff ratios of up to onestaff per four groups of 40 to 50 farmers. This is very costly and difficult tosustain. Intelligent programming of group meetings will permit adjustments tothe stage of interaction with a group (e.g., problem census stage, discussion ofrelevant technology for a particular problem or examination and discussion of theresults of a demonstration plot), and the specific requirements of the farmingsystem (e.g., intensive multi-crop irrigation compared to extensive low-inputsheep grazing in a semi-arid environment or tree crop agriculture).

- 25 -

(c) a close association of socioaconomists and the technologytransfer units in EMBRAPA centers to function (with theassistance of relevant subject matter researchers) as thespearhead teams which work with the state EATERs and stateresearch units to demonstrate FSP in producer groupidentification, problem definition, research requirementdefiition, adaptive researchiselection of technology, andresolution of problems in accordance with available producerresources

(d) expanded formal training in FSP in IERAPA, allocation ofsufficient EIBRAPA staff resources to the program of FSPdemonstration and practical training with the states, andrecognition by DEMRAPA's Human Resources Department of theimportance of these positions;

(e) allocation of funds within the EMBRAPA budget for developinga capability in FSP in state institutions;

(f) acceptance that detailed, frequent, physical and socioeconomicdata collection on a farmer sample over extended periods isunlikely to be a practical method of understanding farmingsystems on the scale necessary in Brasil; structured groupdiscussions following identification of farmer referencegroups and common farming systems can usually provide thisunderstanding, as can rapid informal farmer surveys by a smallsocioeconomic-technical team (see footnote 7 of Annex 2 fora brief description of this technique);

(g) programed use of the feed back concerning farm systemcharacteristics and problem definition obtained directly byparticipation in meetings with producers, and indirectly byprogramed interchange with state agencies, to guide researchprograming and on-farm verification trials;

(h) in circumstances where inadequate information is obtained fromgroup interaction or rapid informal surveys ((f) above), useof a formal sample survey among a target population to verifyfacets of the system which are important to ensure researchis properly directed; a single-visit survey method, using aquestionnaire based on information obtained from the previousinteraction with farmers, is appropriate with about a 50-

- 26 -

farmer sample -- the questionnaire is location-specific andhighly selective in contentsu I" and

(i) a review of how mass media avenues can be best utilized at thenational and local level to complement the group approach toextension.

4.08 In addition to its leadership, catalytic and training role with stateresearch and extension agencies, DEGRAPA can be expected to have considerabledirect dealings with commercial producer organisations who articulate theirproduction problems and research requirements. This will require closecoordination and liaison with state research agenctes to ensure the efficient useof resources, but will also facilitate the capture of private sector funding forresearch.

4.09 The issue of continued funding of a public sector institution of thesise of EMBRAPA is an obvious concern for EMBRAPA management and federal fundingagencies (pare 3.27). EMBRAPA in 1990 already adopted a policy of having theprivate sector share the oWts of research and technology transfer from whichthey directly benefit (pars 3.33). However, the new department (DCO) created forthis purpose will have to expand its activities into comprehensive analysis ofthe potential beneficiaries in the agricultural production, processing andservice subsectors if the maxima capture of funding resources is to be realized.The creation of DCO is very important so that scientific research staff are notobliged to devote their professional time to fund-raising. The most likelysupporters are liable to be strong local or export commodity associations,producers in specific high-value production areas (such as an irrigation schemes)and sectoral service groups (such as the plant protection industry). Experiencefrom other countries, such as Australia and the U. K., which expect agricultureto significantly support its own services, could help in this regard.

4.10 It cannot be expected that smallholders with very limited resources willbe able to make significant financial contributions to resrarch/exLension

n. See Collinson, M.P. *Farming Systems Research: Diagnosing the Problems"In Research-Extension-Farmers A Two-way Continuum for Agricultural Developmentedited by M. Cornea, J. Coulter and J. Russell. A World Bank and UNDP Symposium,1984.

m. The use of the frequent data collection from a farmer sample over more thanone season (as mentioned in para 4.07(f)) may still have a place in specificcircumstances, but is more likely to be considered as a research subprojectrather than as part of the normal process of technology development and transfer.

- 27 -

services. Consequently, private sector research funding met not be allowed tocreate an Imbalance between research directed at resolving production constraintsfor the higher-income producers (and agricultural processors and serviceindustries), and the less-organized, poorer segment of the sector represented assmallholder farmers. This Implies that some monitoring of resource allocationsat each center will be necessary by DBBRAPA management to avoid any negativetreads in this direction. A case in point is the considerable pressure which Isbeing placed on the management of EMBRAPA In Brasilia and of CPATSA at Petrolinain the Northeast to allocate substantially more resources to the problems ofcommercial Irrigated agriculture in the lo Francisco Valley. Currently, meetresources (about 802) at CPATSA are allocated to the problems of the dryland,largely smallholder faming systems in the semiarid Northeast. Further researchis justified for Irrigated farming, but this should not be done with anyreduction in the staff and resource allocations to the semiarid area. Theirrigation farmers can be expected to pay for an expansion in the researchservices to cater for their needs, while public funds should continue to supportthe substantial program of rainfed farming research for smallholders.

4.11 An effective program to increase the contribution of the comercialfarmers and supporting industries to the cost of r6bearch and extension willenable federal funds to be directed more to the benefit of resource-poor, small-scale farmers, and to environmental concerns and sustainability Issues.Rationalization of some economic policies (para 3.29) would also reduce the highcost of agricultural subsidies and encourage the allocation of federal fundingresources to those agricultural research and extension program which should besupported by the public sector.

4.12 EMBRAPA cannot be expected to develop the capability to be the majorformulator of policies that affect agriculture. Agricultural policy formulationis extremely complex in nature, and ms- take into account factors such asexchange rate retie, trade interventions, taxation, investment subsidy policy,credit controls and subsidies, domestic pricing and marketing policies and landpolicy. Nevertheless, there ls an urgent need to establish a capacity inGovernment to monitor agriculture-related policies, assess the merits ofalternative policies and programs which affect agriculture, and advise decision-making bodies accordingly. This capacity would logically be devel ed in a unitin the Ministry of Agriculture which is currently weak in the areX'. D1BRAPA,with its increasing knowledge of agroecological environments, the sustainabilityand potential of agricultural systems, and awareness of production constraintsas perceived by farmers through its use of FSP, would be in an advantageousposition to advise a Ministry of Agriculture policy unit in these crucial areas.It would, therefore, be appropriate for EMBRAPA to institutionalise this

0. For further review of the institutional framework for policy making whichaffects agriculture see Brazils Agricultural Sector Reviews Policies andProspects, World Bank Report No. 7788-BR, July 26, 1990.

- 28 -

capability to produce such advice with the approval of the Minister ofAgriculture.

- 29 -Table 1

PROJECT IMPLEMENTATION - KEY INDICATORS

-- --------- ---------- ------- f. . . .l ---------

Appralol A~lusi a6Indeeer i .Uni% tat6onoti Clowln Dolm 21

fropco (NAI. of gxffrie~t 1,319 m8Numi Tropie (CPATU, Ne. of Gåperimen6e l11 474

Camp. Crrede~ (CPAC) Ne. of exporifeMig 242 6CGMI TY SSERCH PROAtWie NO. of eaperi~*nt1s 80 96

NO. of epgrim~nte 410 1,167Vaegsbble Cree Na. of experlm~6e 710 2,067

Frei% Crepe Ha. of eperim~nte 200 00Coco~eta Ne. of experimen6e b0 13701l Pal. No. of oxporiffai~ 200 149Saba... No. of eparimmots 20 48Rbber NO. of exporImeto 440 797Foreory NO. of experlo~en 110 117sel. saede NO. of exp.eri te as 60Agri«lt,ral Entgnelfring No. of enporim~n6e 38 a1Foed Teohu~legy a. of experimente 30 822Sloonergy NO. of exporimen~ 110 1,319Anlil Nmith No. of experiffin 140 877

1WP~T SYSTESInformtion and oecu.en6tion (9) 100 100Evalmation (3) 100 100

INC~UNTAl. STAFINM nuabar 1,08 2,197Nearch Mmeff number (224) (311)Sopport and Adinibrative Staff nomber (1,079) (1,8)

CIMIJLTANTS o-years 156.8 406.40SM T-T we-yesre (99.0) 87.26

~.og-Term ~m-yeare (67.8) 800.18

SEARCH CONTRACTS number 180 7

FEL.L0 1IPS NO. of candidas 206 s1P r.77 i l Ha. of canddatow (22) (71)

Ph.D. - Abrmed Ha. of candldeaos (00 (164)M.gc. - @rall Ha. of eandideten 77 (248)

M.. - Abr~ed NO. of emndidaes (41) (61)

sMT-T CUMES MO. of porblcipot 320 1,700Era91 No. of particip~mte (97) (1,000)

Abreed No. of perliclpoff4 (228) (700)

CIVIL MIK1 CMMTRACTS, valvo (US18*000) 20,820 18,714

-- ---------------------------------.-.--.-.----.-.---.-- ----- ---

1/ The indeseersere6hose defined ~ epprolsål, and mehored duringimpleentetten.

2/ md a o inform otkn providd by MAPA In Its Pröjeet CompletionRep.r4 af Jul7 1988.

Source: PCR

30- Table 2

AMIENTS TO THE LOAN AGRE(EET

Serialso. Date Purpose of the Amndmnt

1 April 26, 1982 Reword Special Account's clause to comply withother loans approved in M .

2 October G. 1982 Include a new research program *Plant ProtectionResearch" in the project description (Schedule2).

3 June 24, 1983 To reflect that henceforth the local costfinancing of local (Brazilian) consultantservices would equal that for foreignexpenditures in this category, i.e., increasedfrom 80% to 1002 financing under the Loan.

4 December 2, 1983 Under the Special Action Program (SAP) the Bankparticipation in project financing would,effective October 1, 1983 and until March 31.1985, be increased from 35Z to 67.5? of totalproject costs (Schedule 1 revised accordingly).

5 July 11, 1984 Reallocation of Loan Proceeds among Categories(Schedule 1). Substantial changes were made toaccommodate increased expenditures for studies(fellowships) and the expanded use of consultantservices.

6 August 21. 1985 Reallocation of Loan Proceeds among Categoriesand elimination of the Special Account Categoryas for item 5 above and the Special Account(Category 6).

7 Jaanary 14, 1966 The disbursement percentages for expendituresunder of the Special Action Program - SAP (Item4) to be extended indefinitely.

Source: PCR

31 - Table 3

WITHDRAWAL OF THE PROCEEDS OF THE LOAN

AT LAST REVISION ACTUALAT LOAN SIONING (Amment of DIUS1SENTCATEGORY (7/9/81) 8/22/85) (at 12/31/88)

1. Consulting Services 12,1,0o 17, 7,gyn

2. Research Services ,gm0 0 1,189,701

3. Fellowshisps SM000 15,4m0M 3R,974,107

4. Civil Works 1 m,4 1g mgggm ,M,M

S. Equipment, Office, 26,See,at 17,71,400 17,365,875Farm, Laboratory2 Vehiclel

6. Initial Deposit IMAS - 2,614to Special Account

7. Unallocated a,g,m . .

S. Cancelled - . 46 1/

TOTAL$ gggesgsg

1/ A refund generated by a unused portion of a Latter of Credit.

Source: pCR

- 32 -

Table 4

EMBRAPA: Staff Numbers by Catexory 1974 - 1989

Research Support RatioYear B.Sc. I M.Sc. Z Ph.D. Z Total Staff Support/Research Total

1974 446 51 385 44 41 5 872 3118 3.6 39901975 542 448 47 1037 3772 3.6 4809

1976 807 61 464 35 57 4 1328 4375 3.3 57031977 795 458 58 1311 4374 3.4 56851978 543 41 702 53 91 7 1336 4698 3.5 60341979 548 777 123 1448 5497 3.8 69451980 509 33 882 57 162 10 1553 5830 3.8 7383

1981 439 941 196 1576 6105 3.9 76811982 403 25 968 59 269 16 1640 6328 3.9 79681983 355 986 269 1610 6374 4.0 79841984 320 20 1001 62 298 18 1619 6553 4.0 81721985 275 1012 363 1650 6793 4.1 8443

1986 274 16 1046 61 404 23 1724 6748 3.9 84721987 336 1075 459 1870 7008 3.7 88781988 324 17 1088 57 499 26 1911 6957 3.6 88681989 467 22 1151 53 548 25 2166 8502 3.9 10668

Source: Department of Human Resources, EMBRAPA

Sccloconoelste and Extmeson Speciallets fölativs toOther Remearehåre E£=levd la E~1P

Sel .q~latal/ tsetambr '90I Total Remrebr~Y biff=§af Te Soml41 lamIv DualIlfleatleANi Bv Funetion Operatlag In E~*PA j= Bw GUi a -IL. . nmatinala a. Ii-Tralning Ifitl. In Realom <.an 1 g L. k. (Ju.

erth 1 6 2 6 - 1 9 2M 16 9 7 -U.rtheast 4 19 2 12 8 1 16 279 11 6 6 -Snuthes - 7 9 15 1 -. 16 31 11 1 9 1S6t 2 12 5 16 2 1 19 294 19 I1 8 2Cm*tr- 4 2 12 1 - 13 fl. u * 1 -

ISTILS 9 1 26l78 7814 7 *n

1/Ari.lterel c.sseelste ed rural ecIelogiete.m.c, M.S. etc. Imply oqwaln trelning ln th, 9cle9 sclences and oxt~nlonEpcudlag had offico and mana0dd por*nnel.

Enledlag heed effico pergoomel ln Brasilla.

- 34 -Table 6

MMER O TRAKNMES ASALLY IomLiED IN E APAsPT-MADUATE PROA ACCMDINa TO REEARCN INSTIT7om Aæ

LICATION O TRAMDIDI. 19M-10I0

oA DeREURM TOTAL

9~s Beal l Abreed T I Gr«IlI Abrd TMw.I Breal I Abred ToIl BraI l Abr~ed To"I

1974 tu 47 8U7 30 18 43 1 6 1 821 811978 162 92 264 17 16 86 5 1 6 164 1o 29

19~6 283 67 3 44 22 66 12 1 18 29 9M 8791977 67 46 97 81 42 78 7 4 i 96 86 11i978 77 a* 185 74 22 96 9 2 16 19 02 1411979 76 49 119 61 48 94 36 8 124 92 2161990 69 88 167 67 24 91 & 8 6 129 6 194

1991 62 86 97 58 17 76 1 S 6 116 57 173198 a9 89 97 64 26 92 62 7 69 194 74 2U9198 6 82 gø 48 i 59 17 4 21 123 47 176194 79 28 192 69 28 62 41 14 55 179 66 289

198 76 36 100 41 16 66 82 6 46 148 53 196

1999 64 m 94 8 8 46 4 & 7 la 41 1471997 62 16 68 36 4 89 2 - 2 se 26 149196 48 21 69 26 2 28 - - - 74 28 97199 71 26 91 82 6 37 - - - 13 26 128

source: Oert~et of Nu~en ResurGe , ERAPA

35 - Table 7

M GF ADEI* M~ MMLJ. DCLM 2 R7-T14Mcms AM I Te ng"EMC 90011mU7 MO

LCATI=1 ff 1MSDSB 1M~M-18

1M4 881 18 88 18 9 18 884 18 a3im mu a m 69 a 76 M7 91 mu

iM 421 78 494 111 8 117 3 79 G111977 1,~80 76 1,~28 88 6 sa 1,83 76 1,68

1978 3,g86 188 ,18 84 1 a8 2,119 184 2,23179 1,472 151 1,3 2 26 27 1,474 176 1,6m8

18 967 a8 1,86a 4m8 28 428 1,869 111 1,4m0

1181 2,748 25 2,978 760 71 78 3,487 2«6 8,758188 1,8~7 81 1,672 376 2m8 8 1,788 624 2,267

188 88m 1m 8 977 787 - 787 1,678 18 1,084184 1,623 1n 1,712 424 67 491 1,947 26 2,2=1886 m88 1a 1,~2 47 34 81 987 19N 1,188

189 2,888 148 2,228 738 17 747 2,98 108 2,9761907 1,678 192 1,888 8G 7 843 2~,80 199 2,28

198 1,726 166 1,8~6 182 6 187 1,967 161 2,8818 782 3 8 88 6 11 61 83 01 898

Somros: Dertme~ef Nun Resource, EImAPA

Include perconeml of the techeical and cientlf le occupatfonal groupa,mdlaletrative staff, and *upport service. for research.

- 36 -

Table 8

1RMGER CP PARTICIPANTS IN STAFF TRAINING

mua2 AM~A TO7AL

asseekh 8t611. etpr% UI% wC PB C 1/ wC Pm W 1/ BC P SC 1/

CDU~ 9 1 m 2 27 9 2 - 37

CHPA1 8 a M a 2 17 1 6 167

CHPAF 17 1 5 2 9 88 19 17 8

CIPC 8 1 75 1 a 6 7 4 e1

CftC 8 8 18 6 8 11 a 8 24

5 2 16 2 8 19 7 5 34

CPF 6 9 4 8 11 4 8 11

CHPOC 8 a 2 9 4 1 6 7 16

CIP6L 19 9 8 6 19 5 16 19

CPN 4 1 t 2 9 21 6 18 46

Ci~., 9 2 lo 8 7 21 t 9 218

CPMS 6 8 2 2 19 5 22

CIPO 6 1 6 8 8 28 4 2e

CW&A 1 8 8 1 7 a1 2 16 31

CPS 9 7 5 8 8 18 7 68

CPT 6 6 2 a 9 2 6

CPAC 21 4 46 8 1 a 5 24 22 S1

CPATSA 1 18 5 9 40 1 18 64

CPATU 20 8 28 1 4 26 21 7 48

CTAA 19 2 1 8 3 58 la 5 76

UEPAE - ELM 6 8 85 8 8 2 6 9 87

utPAt - MANAUS 6 5 1 2 2 r 11

VEPA£ - PORTOV 12 1 79 1 a 13 1 79

UWAE - nIo RC 1 1 36 2 8 2 3 4 37

UEPAE uTEnTA 6 2 5 8 2 4 7

IJEA -00A VISTA 8 0 52 8 8 1 8 8 88

UPAE MACAPA 9 8 52 6 g 1 8 8 58

WU 12 8 5 9 8 82 12 9 a7

SsAPA•HQ SS 5 m8 38 42 226 64 47 276

AN (TATE 0 ) 11 1 5 8 1 16 11 2 2u

UPAL(STATE 000) 5 8 1 7 5 1 1I

TOTAL 245 71 1,= e1 164 7m0 m 225, 1,7ff

1/ *C a her$ Coure for TIhnIel ~1udes.

ures PCR

Table 9-37 -

Hiher Deare Specialiazions at Three Regional Centers of EMBRAPA (1990)

Seclalisation CPATU CPATSA CPAC

Plant production 6 7 5Plant pathology 4 3 1Plant physiology 1 2 1Plant nutrition 1 - -Plant breeding 2 1 4Seed production 3 - -

Entemology - 1 2

Microbiology 1 2

Soil science 8 4 3Soil fertility - - 8Soil physics -Soil chemistry 1 2 1Soil management 2 2 3

Agricultural engineering 1 4 3Irrigation and Drainage - 2 4Hydrology - - 1

Agrostology 5 4 2Range management - - 1

Animal production 3 3 1Animal nutrition 3 3 -Animal pathology 1 - -Animal parasitology - - 1Animal traction 1 - -

Agroclimatology 2Zoology 2 - 2Remotesensing - - 2Genetic resources/Botany 3 2 -

Fruit production - 4 1Food technology 3 - -

Forest management 4 4 1Forestry breeding 1 - 1Wood physiology 1 - -Wood technology 1 - -

Energy 1 - -Marine biology - 2 -Statistics 1 1 3

Socioeconomics/A&r. econ. 4 - 4

Production system analysis 1 1 -

Technology transfer/extension - 1 -

EIRPA± CIEITANT CoITRACTE IN AMICIL1UAL . II

ACTUAL CITRACTSINTERNATIMAL NATIMNAL TOTAL ARaL TMa. Map-yra. Ma-athe Noe. Man-yr.. Maa-menthp. No. Ma-vra. Ma-va

Long-term contracte 84 216.5 8 154.8 1w 979.9 57.3Average conkract period 2.6 2.3

hOrt-trm controcte 1gg 28.4 11.7 22g W..1 g.Average contrect period 1.4 4.7

Long-term congu itane HedOffi CAC Other StalamaRmpializad la Int. Nat. Int. Nat. Int. Nat. Iat. Nat.Agricultural «e«nemles (No.) 4 2 1 - 1 - - 2Producton syt (N.) - - 2 - - - 1 -

Techology tranefer (mo.) 1 4 - - - - 1 -

Source: EI~APA Final Evaluetie ert (There l a sligh dlscrepeecy of abee6 2 maa-yers ia the r*10 metteof losg-ter. sAd shrt-term ceeltastele la this report and the PCR, but this la m0ese q 9e*4

- 39 -Table 11

1MGM F MU0JCT8

Pg14 1era 1im 1eMi TOTAL

RieM tas a8s 178 0 176 me mb~.e. 148 107 184 t1 1 1,1 7

inla. 180 127 131 143 148 169 ma mm4 188 in mg mm4 04

101 173 17 170 0 1me gn

ser C~*o lMg m 242 m m m 1,=Döepy CGtIe 141 17 17 18 13 1in u*Vg~ ble s4 884 m 87 341 880 81 2,9?Cm~*eve 1 11 1s Ila 0l in ta m3tOIS 57 a0 67 a8 el S 107PIo al 79 el 78 es 76 48W

$orgh~m 70 77 76 e1 76 76 4MgCo~ 78 74 77 as 118 1m0 a8

Rubber 11 14 1a8 144 121 la 787ToWg. Fulto s 1s 188 11t ta 187 Mst

CItrös 08 64 e 79 e 64 491Pln«spplo 44 43 48 88 4 m8sea~en 47 41 41 44 e1 l8 27Meng. 12 18 11 17 11 11 77Vi«i1te 4re 43 47 47 a1 58 67 0011 Palm 22 10 20 34 80 19 149

F9ed T*hley 51 54 61 49 49 w0 m88Gene6Ie Resemric 120 147 181 18 28 2n 1,642

SeIl Urveye 47 40 49 46 8 49 m90Cp Cor~ed. st.dIs 1e 18 16 23 8 28 in

Se-Arid Triepic at a 41 30 0 8u 249Hmid Tropies StdI 1a 16 19 26 2n 88 13

Crop Adpta~ea Ca~. Corrmd 81 37 48 a5 e0 118 414Cfet A~~66. eaI-Arld Tr~ple 8 114 118 la et 7 mCrop Ad9s 1tf. Trepl 8 38 8 48 41 87 m

Coppodo sy*~ Cap. C~rred. 28 n7 43 8 a8 8 ga88Cepplag tystei Aeel-ArId Trple 18 17 18 19 18 18 aCrepog Sysee Huaid Tro~pe 11 19 17 e1 a5 17 11Fnrstry ng 87 11* 1g7 1g1 tu gga

&erg cudi* 244 86 m88 m83 107 07 1,818Positry 81 37 81 47 44 6m 3.0

C~ 9 8 88 8n 48 17PlafPrf6lIen ~ ~ 8 18 28 1 1t?

8.1l Sile 9 6 84 8 88 al 1gSebe.. Pela 6 g g 19 1 12 48AnlaIMtIh g 6 02 178 148 377W.1i Cone~va6. 6 g 8 0 e1 44 78Crop Ov~lfloat 24 8 1 8U7 81 870 1,=8

TOTAL 8,8 8,#00 8, 4,0 4,188 4,188 88,10

Source: PCR

- 40 -

Table 12

DISAGGREGATION OF RATES OF GROWTH OF PRODUCTION BY

AREA EXPANSION AND YIELDS

VISLD$s 196-1990

DISAOEATD RATES OF iROWTHRATES OF IN AREA EXPANSION AI YIELDS

GROWTH OFPRODUCTION Area Yield .

EMPORT CROPSobeens 22.95 20.71 ( 90.24 ) 2.24 ( 9.76 )Oranges 10.15 6.75 ( 36.21 ) 1.40 ( 18.79 )Sugar 4.25 2.69 ( 68.29 ) 1.56 ( 86.71 )T*be*e 8.09 0.91 ( 29.45 ) 2.13 ( 70.55 )Coo". 4.87 0.00 ( 0.00) 4.87 ( 100.00)Coffee -8.87 -2.54 ( 75.37 ) -0.38 ( 24.68 )Cotton -1.08 -0.49 ( 48.28 ) -0.57 ( 58.77 )

FOOD CRPSgh"t 9.95 9.17 ( 92.16 ) 0.78 ( 7.64 )Melae 8.29 1.8 ( 51.22 ) 1.60 ( 48.78 )RICe 1.61 1.90 ( 100.00 ) -0.29 ( 0.00 )Caesave -0.85 1.70 ( 0.00) -2.05 ( 100.00)Sees -0.96 2.18 ( 0.00 ) -8.08 ( 100.00)

No6ets rowth rates are derived using the least square, method. Numbers inpesietheIs are the relative contribution to rate- of growth from aresand yield ineres*.

Source: Brazil - Agricultural Sector Review, World Bank Report # 7798-BRJuly 26, 1990

- 41 -

Table 13

DISAGGREGATION OF RATES OF GROWTH OF PRODUCTION BY

AREA EXPANSION AD Y1ELDS BY MAJOR CROP: 1980-1989

018M AIE ATES GF UTHRATISfOF IN MEA EAIIM AS M YIELDS

PRBUC7IIM Ar«. Ylleid -

seybee 4.17 8.37 (90.9m!) 0.90 (19.101)Dramge 5.06 4.71 (98.27) 0.84 (6.730)Suger 6.77 8.47 (80.0) 1.80 (19.20)Tb~ 1.03 -1.11 (0.00 ) 2.14 (100.00!)Coco. 1.84 3.90 (100.00) -2.16 (0.00%)Coffee -2.5 2.71 (0.00 ) -5.25 (100.00m)Ce~ 2.26 -. 93 (0.00^) 11.19 (100.001)

Fa. cmat~ha 0 6.19 1.14 (1.2)7.0s (6.

me izo 2.9M 1.70 (67.~% 1.26 (42.9Rico 2.98 -0.9 (0.00) 8.94 (100.001)Ca~v -0.13 -1.84 (100.00) 1.71 (0.00 )Semn. 1.18 1.42 (100.00) -0.29 (0.00 )

* Growth rat for wh~t wer* caiculated for the period 1~90-19B.

Note: OreG h r~t ar* derived sing the 1staquers ahd. Persen~agefigurw refer to the relativ. contribution to res of growU6 fre= aremsnd ylield incresa.

Source: Brazil - Agricultural Sector RevieM, World Bank Report # 7798-BR

July 26, 1990

- 42 -Table 14

DISAGGREGATION OF RATES OF GROWTH OF PRODUCTION BY AREAEXPANSION AND TIELDS FOR INDIVIDUAL EXPORT CROPS BY REGION 1980-1989

IIsaOUnSAIS RAlUs UP m11s1MhW. SAKoAll~M mma ep VUI#~ss mms a aa~us a n~sa

in PnMCI Are (he) (1) Yleld. (kgjbe) (C)

Ceees 8.u0 14.84 (60.17) 14.74 (40.0)cffee O.n3 10.40 (100.0) -4.w1 (0.30)

Cä~ee. 1.18 8.n1 (100.00) -2.18 (0.00)eLne t.7 IN. (98.77) 0.17 (0.1)

suger Ca.. 8.14 2.10 (880) 1.14 (40.70)Tebeeca -7.74 -. 7 (4010) 1.8s (0.0>)

soyb.ee 1o.71 5.12 (47.45) 5.07 <13.5s)C... -0.11 2.17 (0.00> -2.10 (100.00)Cerf.. 1.83 0.05 (100.00) -0.78 (0.00)

ceas.. -2.s8 -10.12 (100.00) 18.10 (0.00)• emag.. 8.80 8.17 (100.00) -0.27 (0.00)Suet Cea 4.1n 8.48 (14.07) 0.08 (15.88)Tel.... -7.70 -7.3s (14.48) -0.48 (s.53)

soben 0.3s -1.as (0.00) 7.1e (100.00)41?e-09 1.2e (100.00> -4.81 (0.00)

Cete. -.. 0 8.80 (100.00) -7.8e (0.00>c.10.. 3.8s 2.80 (0.00> -0.10 (100.00)

Simus.. 5.48 1.01 (08.2?) 0.43 (7.78)Sue cea 0.14 5.72 <53.42) 1.81 (17.50)

ebse-10.1e -4.78 <10.01) -1.8 (10.50)

setee-1.75 -1.88 (70.00) -4.43 (24.00)cete. -2.3* -1.22 (5s.n) -o.17 (44.31>

ceas.. 4.81 8.87 (70.10) 0.14 (n.11)sreg..e -0.10 9.14 (0.00) -1.14 <100.00)

suger c..an 0.54 7.8o (71.00> 1.14 (30.84)Tel.... 1.10 8.04 (100.00) -1.18 (1.00)

soybes.e 10.17 1s.10 (54.01> 8.01 <1(1.10)ese48.78 28.07 (40.se> 30.71 (53.10)

<ettee 0.41 1.0e (100.00) -1.18 (1.00)<e1soa 8.1s 4.00 (100.00) -o.24 (0.00>Oregse. 3.78 4.51 (100.00) -1.71 (1.00>

suger cee 35.10 21.25 (54.80) 8.14 <11.04)Tese , -17.10 -11.0? (30.51) -2.10 (11.48)

flse.s Oceuth este ee deelved esIas the eet eqseee ehed. Pereuaeeligeren 'elt 8e the '.1e1ive ctelbtIe. te este..e4 gru48h free see.and pleld les.e.

Souf ce: IgreAn - Aew.wieuwal Sector Reviev; Wor1,4 lanik Report 1 7798-EP.July 26, 1990

- 43 -Table 15

DISAGGREGATION OF RATES OF GROWTH OF PRODUCTION BY AREA EXPANSION.AND YIELDS FOR INDIVIDUAL EXPORT CROPS BY REGION: 1980-1989

*ssn8 ia0m M es S V sa

auseN cmv IF 3sFsasomk PNM1IT.e 1 n.> (.> S.é (s.s<)

Sees. 1241 11.1M (00> 0.8 <.0Casse., 4.38 8.84 (77.01> 0.3e (11.88)

Malm. 12.8 10.40 (1.00> 1.04 (14.00>Rlei g.03 8.84 (01.n1 0.40 (1..9s)

Se... 0.04 s.84 (70.07> 1.80 (81.00>ceva-1.78 -2.70 (100.00) 0.0? (0.00)

mea* 18.19 4.00 (40.00) 7.28 (se.81)RIs 4."s 0.3s (10.24) 8.01 (01.70)

Seerne -1.44 -. 88 (100.00) 1.01 (0.00>Carneeve -4.1M -8.2 (7.48) -1.so (80.s?)Mela. 2.8s 0.08 (1.0s) 8.87 (07.00>

Rl. 2.48 -1.18 (0.00) 8.01 (00.^1.04 0.01 (0.01) 0.08 (00.08>

0.... -2.44 -1.87 (80.11) -1.07 (48.01>Ceseeve *.wo 0.04 (88.07) 1.1 (0s.48)

khl. -o.42 -0."8 (100.00) 0.90 (0.00>Si~ 4.01 0.31 (10.94) s.l (01.>

111~ 0.01 0.41 (0.10) 7.0 (W.03

MEM-wTS0ees 4.M8 -0. (0.0") 4.47 (100.0>

Cee~e 4.10 8.18 (74.m0) 1.31 (^.18>kbl.. n.1? 0.0e (M0.41) 4."s (41.0>

i~ -0.1M 4.1 (00.00 8.1 (0.1.1111~ o.18 18.40 (79.88M s.87 (.17

Noe eroebh re ar* delved slg the "0 6 sq~e tm hed. poreesgefIgr rfer to he roel"6ivo e ui6a . røtes *f greffh fre eemmd yleld ine91 ss.

Source: Brasil - Agricultural Sector Reviw, World Bank Report # 7798-BRJuly 26, 1990

APMLICATION OF YNDS BY !MBRAPA BY REGION 1974 - 1979 (US$)

Hea4.d offi~ g f film ,s

:1974 2,35,134 2,913,944 4,343,932 3,3#4,551 1Ø,449,387 23,316,9Mi 01 ti1975 4,464,146 6,956,8 11,4#4,878 8,365,244 19,869,268 5#,159,341 14,6m,171 5,21,f9711976 6,32,99 10,737,968 14,#39,89 12,842,725 31,491,356 75,432,561 21,955,892 9,535,4

:1977 7,231,744 12,?68,794 16,511,676 14,437,164 39,237,334 9#,366,7121 27,263,4451 11,973, 91978 8,956,58 18,168,845 23,526,475 17,583,379 43,177,194 111,411,9511 26.356,6961 16,2,4911979 1#,239,431 18,976,?64 20,#28,971 24,753,392 76,915,79 146,913,1271 49,742,4351 27,172,6449

119"8 12,367,942 21,321,217 21,987,62 27,#91,755 64,268,22 146,937,324. 29,919,391 34,349,7931981 15,232,654 24,421,373 25,221,331 31,399,228 78,45,75 173,91,336 35,324,3551 42,721,39511982 21,979,223 30,366,604 32,762,f19 33,179,719 1*8,917,691 227,1#4,2541 58,658,21s 50,259,41$*1993 15,479,8 9 19,123,656 19,263,967 16,624,643 52,397,415 122,899,M941 27,854,535: 24,542,448

l le11984 13,624,316 16,489,794 17,727,141 14,683,627 48,153,86M 11,278,7» 22,979,819 25,174,04111995 14,2^2,51# 15,796,862 16,224,282 15,249,922 57,368,871 118,83,4471 33,494,785 23,984,86l1996 17,91,373 21,823,524 25,267,116 21,549,256 51,419,168 137,154,4331 21,440,616 29,978,5521987 18,418,948 24,792,326 27,485,419 23,127,236 63,84,559 156,9^8,4411 29,318,7361 33f765,82211988 16,216,Ø56 23,530,874 30,314,387 24,429,9#2 58,162,866 152,645,M1 19,566,68 39,596,79811999 21,366,867 29,559,756 36,343,398 29,8,398 72,861,95 19,479,3161 29,484,181 44,71,724:

TOTAL 205,337,129 296,247,595 342,348,348 313,610,116 875,82,124 2,33,363,3121 446,613,$54: 418,757,683

MAZIL PRIVATE AAM PI~C SECTO AM!CULTARAL TECMIICAL ASSISTMEE B_MI

Bagien Number øf Farm R*lativ* Tachn e nia AnOn&tane Prep& m RI21Ee*abihmn t Farm Income Number Establ/Prof. PrIlVMe Publie Public/

lu omte Sbar Privamæ

153 1,281 39 168 1,11 7.5

2,1,w 18 7,2M8 88 2,1M ,168 2.6

SOmh~s 9, 489 7,81 132 s,13m 4,446 1.4

South 1,262,0 323 9,432 127 6,116 8,822 6.6

CO #ter-set 318,w 632 a,88m 93 1,989 1,441 6.7

1i,48m 15,497 1.2

1 Frem Brazil Agricultural Sctr Revi.« Report No. 7798-8R Table A.2.18 using th. Northeut *~6bIl*m I Iness 00 l.3 Collated by EWAPA from prof~oulonal assocations.

M ^P OCKM Am AKULUCTM W em 1974 • 1=0

3 3 1974 1975 1976 1977 19981 821I sms ss ses eet -red. Gavt; i ~ M51 i0tM 4,, ,,NA ,i i:1- .VEe Pregram 3 1381 1,*8,19 3*.115 5*,.1 MMM 4

3 ,

:4- Other contraectet 6M,B 5,71,4I3 N,4I,S *,43,0 1., 135 - Leon Support t i Sh,u,% 2,65,36 19,43,96 8,118,*8 D*,MIV 3i i

6 - prior Tear 341.8 a,2,, 414e,* 3,M 5,wt7# T A L i 36414 låen,te f,i.W1 1,.I,. S,48, 'I

Application

1 - Civil works 1 0 ,8 ., 4,8,12 - Eeuiment e3,491M6

248,9 4,5,40 9S3- Salaries 1 8 3,51,8 - 1,40M ,M,IU 1,4,56 33Line 3/3+4() 47 52 64 7114 - Operating Co.ta g a,1, 2,6,43 ,363N 4~MM 5,4095 - Trasfers to 1 1,1711 6,1.M 9,73,918 14,9, 34,8844States

1 VIT0 L 1 23,316,93? 38,iW,63 7.4. 98,88.N 111,41,.95B EKchange fat.?! 1 6.856 8.2 10.701 • 14.132 18.206 1

teg

9 I 1979 1980 1981 1962 1983 goI 3rc I

lo IIt -Fed. Govt. I 91,W,W0 16,29,141 12,7M,6tt 13,W,9R ,.1 II I

I12 - Spe. Programs I 13,697,173 7,613,721 3,12,1 46, 3 6WI.412I I I13 - Income I ,1, 8,3W,7143: 21,923,8 ,6,A70 a.,4.

1- II14- Other Contracts I 6.446,94 17,219,143 6,2,672 4.54t1 5,M,712 II I

II5 - than Support 1 25,96,366 30, 0, f,164,990 41,A4.0 1,711,711 1I - Internal I 18,412.34 14,1,05 2,W,M 2,466,W W.W I- External I 1,31 15,9,9N 19,441, XoM,41 U.A.I I

I16 - Prior Year al.* I S,M6,463 2,WAR 3,18,19 3,3,2. . 2,33,,1. II T0 TAL 1 13Z,4,91 1,32,7 IM,5N,t 23,417,4 1,564,3 . .I

If-- -- - -I Application 1

f1 - Civil Works I 7,01,6" t9,192,1 15,49,131 18,4,11 1,29,4 *11 I I12 - Equipment 5,413,319 5,423W2 9,,95 a,90.3 5691,445I I I13 - Salaries I 11,tst,314 12,W2.976 13,454,679 IU,W.67 73,191,317 I1 1 I8 Line 1/3+4(M) 61 71 64 65 71 1I I I14 - Operating Costs I 44,W1,4 29,93,54 7,6,3 St,91,MS 3,214,94 II I IIS - Transfers yo I 21,MU,339 19,969,122 18,W,939 a,W, M,W,21States-

TO TAL 1 146,913,1K 146,9V,324 13,911,33 W,t14,t 0919,090I Exchage Rate_ 2 27.131 54.471 94.23 183.686 606.572 I

000

1 I 1984 1985 1986 1U7 IM8 ugg II Source II I I55 I55 331 . IS ISS 331

st - Fed. Govt. I p,W,W 9,3, tie, 13k,ox HIM.= 8.X.k4 nI I:2 - Spe. Prorfq iAU",

loom ,4,32II. mw .,13- Income I 13,F4,47 13,Wom 19,W,A 3,W,3 n,6, a.0IS II I

I:4- Other Connects 36lade 2,154,43 1,MW,96 1,1 3,55,9I5 - Loan Supportl N,a,W 1t.W,954 R,3, g,n 1.,91 I,,P low. 8II

I 1I -Internal I 2,0 M I aI 6 0I -External I 42AP6

I6-Prior.Year 2O. 3," MIN 3,N,0 5,, ,MA

II TM VL 1 11M"1, INAA,2 "3,1 WARM4 3,,1 .UApplication

II'11- Cri Vorvs 1 4,,1V A,t.W 0,W,W 3.,R 13,W 3.3.01 I <SI

I12 - Equipment I 5M,4i 5,59.3A 6%W,AG4 1,M NA,R A,191,W 1I I13- Salaries I 5,A,a M,41,W M2,m,m 91,411,96 U,, 14e 1

Line 3/3+4(%)h 56 66 65 71 71 84 I14- Operating Coies ,A a,, 3,,1 5,3,13 a.. 3MM.W." 1I II5- Transfers to I 1a,Ml 6,1a,, 9,7,116 9,M9,491 Mud,t 1.W,tN I'kRates!! II

M I

r Exchange Rate/ 1941.917 6476.917 13.77 41.262 290.165 3.269 1

This refers to support for research carried out by the state research companies.The average exchange rate was calculated using the official exchange rate at the end of each month of expenditure. 'I 'I

0 CoI,%

MAZIL ARICULTURE EXPENDITMR RATTOS £197E-129M

1ELJILLIM

1975 19~6 1961 1962 1983 1964 1986 196 1967 19 19

1. EMAPA F~mdinu1/ 6 147 174 227 123 116 119 137 167 158 1M2. Agricultral øP2/ 8766 18179 1836 1624 18946 2116 23642 2W612 23911 2416 264~3. Line 1/Lin 2 (S) 6.6 6.8 6.95 1.4 f.6 6.52 6.62 6.55 6.66 *.68 #.714. Gewrmnt expend ituro

lo agricultural sector NIA 3759 2776 2828 1261 1771 3678 6318 89W M/A N/A5. Lie 1/ Lin. 4 (5) N/A 3.9 6.3 8.6 9.8 6.2 8.2 2.1 1.7 M/A N/A6. SeIarie/SelOpr.Ceete(N)

-EEM A i 47 71 64 65 71 66 Co 65 71 71 94- St. Rem..archå/(av*rag 10 istitutln) - - - - - - - - - 78 se

1/ Seere. Tablø 182/ Seurce ERAPA/ Brazil Agricultural Sctor Revl«e M Rport 7798-BR July 26, 199d/ SAR Ibird Agriculturel Resarch Projoct 7991-GR Sopteuber 26, 1989.

00

D 0 8 8

1974

1975

1976

1977

197B

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

£ nTI

T SZ"UTos

ENAPA xECEIPT AND APPLICATION OP PUNDS 1974 - 1989 (US$ 000)

361200 -

REC

323700 -AP'APL

286100 -

248600

211000

173500

136000

98400 -

60900 -

23300

~~~~~~~ Ln ø r-. m~ <bo ~ w f c ~ æ

The large difference between budgeted and utilized funda in 1989 Vas due to a very latødrop of the bulk of funds in 1989: unutilized funds are carried over into the next year.Source: ENRAPA

- 52 -

ANNEX 1

fAMING SYSTE PERSPECTIVE

The Concept

Umlemetation of the Programs at the Three Regional CentersCPATSACPACCPATU

Interpretation of Progress Made at Each Center

The Concept

1. In addition to national commodity and research support programs,Agricultural Research II supported EMBRAPA's three regional centers, CPATU,CPATSA and CPAC - which were to concentrate on "farming system research specificto the three broad agroecological regions of the North (humid tropics), Northeast(arid and semiarid tropics), and Center-West (Campos Corrados or Savannah)",respectively. The three centers evolved a common strategy for this researcht

(a) a first phase as an inventory and partial evaluation of theavailable natural and socioeconomic resources of the region,including a study of the farming systems currently in use;

(b) a second phase of more detailed evaluation of how the availableresources could be utilised, in which the production potentials ofa range of innovations in the existing systems are examined, makingmaximum use of the findings of the commodity centers as well asregion-specific, conventional research; and

(c) a third phase of bringing together the results of the second phaseinto farming systems which permit greater productivity andprofitability from available resources with acceptable levels ofincome stability, risk, environmental stabiliy, and any other non-monetary factor important to the farming community.

There was to be a high degree of interdependence between the regional and thecommodity research so that th4 contributions of each would reinforce or redirectthe other. This interdependence was the key element in the strategy, as itsefficient implementation would ensure that all applied research, whether regional

W51/DOC/U216ANX

- 53 -

or comodity-oriented, would be directed at solving real problems in the majorfarming systems, i.e., a farming system perspective in all applied research.

2. Adoption of the farming system perspect.11vre has been initiated but is notyet completely adopted in EMBRAPA, and the implementatLon of the strategy by thethree regional centers has been unev6n.

Implementation of the Programs at the Three Regional Centers

CPATSA

3. At CPATSA, the first Phase of inventory evaluation has been carried outwith support from French bilateral assistance (ORSTON/CIRAD). This has been avery comprehensive undertaking which characterizes natural resources (topographyand soil series, natural vegetation, rainfall quantity and period of incidence,surface and subterranean water quantity and quality), socio-economic conditions(production systems, principal products, population densities, and land ownershipstructure) and factors which indicate the potential and limitations of each unit.A total of 170 units are described, and are grouped into 20 land units or landsystems. These are being presented in map form (1:2.0 million) with accompanyingdescriptive matrices; publication awaits funding for the maps.' The secondphase dealing with resource utilization has concentrated on the drierenvironments of the Northeast. This has resulted in technologies designed toincrease productivity within the constraints of the environment (introduction ofdrought tolerant grass, legume and wood-producing tree species, crop and animalhusbandry improvements, and soil moisture conservation techniques), andtechnologies designed to substantially modify the environment (water harvestingand irrigation development, with associated crop selection and appropriatemanagement). There is a 150 ha "model farm" at CPATSA which is used to verifyprimary research data, and to generate input-output data which should be closeto actual farming conditions. This facility is useful for the third phase of thestrategy. CPATSA, however, is adopting an even more appropriate system of on-farm validation or verification ("testes de ajuste") as a component of thefarming system approach to applied research/technology transfer that it has beendeveloping since 1982. Some of the major schemes implemented by CPATSA in thethird phase are elaborated in paras 4 to 7 below.

4. The Ouricuri area in the state of Pernambuco was chosen as a test area forstudy of natural resources and socio-economic conditions in the semiaridNortheast, with the objective of developing a system to ensure research is

The varzea areas (seasonally flooded, lower-lying areasin the vicinity of water courses) are mapped at1:500,000 because of their higher development potential.

W51/DOC/L1216ANX

- 54 -

directed at solving local problems. In the years 1982 to 1984, 15 agroecologicalunits were defined and sapped with descriptions of physical resource andsocioeconomic conditions in each, which could be used as a basis for researchplanning. Subsequently, it was realised that greater interaction with producerswas necessary to verify the importance of problems and constraints and to ensurethat a particular technology was relevant in the farming situations. This ledto programaed visits of about three days duration in each conamunity by CPATSAstaff and extension officers with groups of producers to more accurately defineproduction systems and producer attitudes and priorities in each. This wasfollowed by an analysis of technology expected to meet producers' concerns, andfollow-up meetings by CPATSA and extension staff with the farmer groups in whichaction plans were agreed and Implemented. This project was completed in 1989'.

5. Detailed farming system analysis has also been ini'ated with one group of14 producers who were land reform recipients and witi. 109 smallholders inirrization communities, both in the state of Pernambuco. These involve detailedmonthly recording of input-output data to gain in-depth knowledge of the farmingsystems, in addition to working with the communities to define problems, andagree on implementation plans. However, the state extension service, EHATER, hasnot been involved in the land reform community, so that this training opportunityhas so far been missed.

6. In the state of Bahia in the Massaroca district, CAPTSA has been workingwith nine community groups since 1985. The state research institution has notbeen involved as it has no facilities close to the area, but ATER extension hasbeen involved from the outset; the continuity of interest and input by EMATERstaff has suffered, however, due to discontent with their employment conditionsand periodic work stoppages.

7. The Bahia exercise is a practical example of what can be done on a largescale with the cooperation of research and extension personnel. Based on a

knowledge of the agroecological situation and informed opinion on existingfarming systems and community identification, the interaction was initiated with

2 Five farms were monitored in detail with monthly recordsof input/output data as a tool to gain a more thoroughunderstanding of the dynamics of the farming systems.This was useful as a learning process, but is notusually essential to understand the systems; sufficientinformation on producer circumstances to be aware ofreal constraints and producer attitudes can usually beobtained by structured interaction between theresearch/extension staff and the communities in groupmeetings.

WPS1/DOC/L121ANX

- 55 -

the commnities. This allowed problem definition, discussion of how problemscould be addressed with available technology, which technolomies needed to beverified in the farming systems, and which problems needed further investigation;plans for verification and/or demonstration and/or Implementation of thetechnologies were then drawn up and followed. This is not what is commonlyidentified as farming system research, but uses the farming system perspectiveto identify the serious technical constraints faced by farmers and to determineresearch needs, and then to introduce technology to key components of the farmingsystem to increase productivity and assist farmers to meet their objectives.

2C-

8. At CPAC, the first phase of Inventory evaluation has been undertaken to theextent that 59 agroecological zones have been defined in the region. Theoreticalcalculations of potential biological yields of crops in each zone based on solarenergy and moisture determinants are being made, and ground water availabilityis also being plotted to add this factor to the resources available for increasedproductivity. There has been major progress in defining how the major progressin defining how the major corrados soil groups must be managed to supply plantnutrients and avoid structural problems, and in selecting varieties anddeveloping appropriate management practices in annual crops, in perennial fruit,industrial and forest crops, and in pastures for animal production; theseinvestigations have been under the second phase of the strategy. The thirdphase has not taken a distinct form and appears to be represented inrecommendations made for utilizing resources in the major commercial farmingsystems. However, a special study concerning farming systems was undertaken byCPAC and is described in the following paragraphs.

9. With French bilateral assistance (ORSTON/CIRAD), in the last year of theproject (1987) CPAC commenced an intensive farming system analysis project(Silvania) in a 3,500 km. area of Goias state with 2,100 farms in whichcontinuous soybean production was the major activity. The primary objective wasto develop and demonstrate a methodology for effective transfer of technology toimprove the productivity of existing production systems. It was not specificallydesigned to determine research needs, nor to evolve an "ideal* farming system,although the former would be a natural consequence of the approach. A 15% sampleof farmers was selected in which factors such as main activities, ownershipstatus and level of technology used were defined, and subsequently a stratifiedsubsample of 35 farmers was selected for continuous monitoring on a monthly basisover two years (1988/89).' Inputs and outputs in the systems were monitored,including labor, and annual inventories were made.

The sample was eventually reduced to 16, largely becauseof change of ownership or administration on the samplefarms.

Ws1/Doc/L21ANX

- 56 -

10. This accumulated data on the farming systems have been progressivelyanalysed by the Silvania team (though not published as yet) and led to the teamworking after September 1989 with 5 major farmer groups, four of which werealready established as farmer associations in particular localities. Within eachmajor group, subgroups were formed according to farming circumstances and theteam worked with these subgroups with common interests. The farmer groupapproach was similar to that adopted by CPATSA in Bahia state (para 6). The teamconsisted of two French experts, four technical staff from EMBRAPA, arepresentative from the Goias state research company, and EMATER staff. In 1990,the Silvania team expanded its farmer groups activities to five moremanicipalities in Goias outside the original subproject area as atraining/demonstration exercise for state research and MATER staff; this did notinvolve continuous monitoring of input-output data of a farmer subsample, butrelied on direct interaction with the groups to ascertain the real circumstancesof the farming systems.

CPATU

11. At CPATU, the inventory evaluation of the first phase has been done to theextent that data have allowed, but this will be progressively improved as morestudies yield further data. Climate, vegetation, soils and demographiccharacteristics are described, but there is only the broadest definition offarming system for this vast area. The state of Para has had more attention inresource analysis than other states in the North. Substantial progress has beenmade in the second phase in collections of genetic resources of the unique Amazonenvironment; in enhancing the productivity of the native Brazil nut; indeveloping system of tree crops in association (e.g., rubber and pepper) and asmonocrops (e.g., oil palm); in selection of annual crops suited to the commonhigh rainfall, high temperature, low fertility conditions; in forest management,reforestation and agro-forestry; in development of new products from nativefruits; in pasture development and pasture rehabilitation; and in buffalonutrition, husbandry and external parasite control. The third phase of thestrategy does include efforts to develop production systems for a number ofspecific situations through agroforestry, crops in association, and buffalomanagement on inundated and dry land pastures. However, there has not been acomprehensive program at CPATU t* identify all of the major farming systems andthe problems faced in these systems. There also appears to have been inadequateattention given to the situation ftced by smallholders in general, and by manyof the settlers who have moved to the Amazon in the late 70s and early 80s andwho are faced with minimal productivity on degraded land.

Interpretation of Progress Made at Each Center

12. CPATSA appears to have given the most attention to development of the farmsystem perspective to enhance its capabilities in research programming and tocreate an appropriate mechanism for transfer of the technology generatedthrough research. It has also taken positive steps to make maximum use of its

%P51/DOC/L1218ANX

- 57 -

collated resource information on environment and farming systems with the stategovernments in the Northeast. This includes state planning departments, thebanking system, the agricultural secretariats and the agrarian reform agencies,all of which can greatly benefit from an improved data base.

13. CPATSA initially attempted to involve the state research and extensionservices in farm system analysis using intensive continuous monitoring of samplesof farms in identified production systems. This was considered to be toointensive and unsuitable for replication on a wide scale, and was not acceptedby the states. A start, however, has recently been made with four states toimplement methods of rapid field assessment to define communities and farmingsystems followed by working with groups along the lines used with the Bahiacommunities (para 7). This is a very important development and warrants the fullsupport of EMBRAPA management and the Bank. To be useful as a researchitechnologytransfer tool, the use of farming system principles must take account of time andmanpower requirements; its use must be cost-effective. Interaction betweentrained extension personnel (with strategic input from researchers), and groupsof producers in the same farming system and techniques of rapid field surveyoffer the most practical means to achieve this (PPAR para 4.07(f).

14. The Northeast has the largest proportion of smallholders (minifundio) ofany of the regions (Table 1), the harshest production environment, the highestproportion of subsistence farming and the most difficult socioeconomicconditions. These are circumstances in which the real production constraints andfactors involved in an apparent lack of adoption of available technology are notalways obvious. This means that the use of a farming system perspective indealing with the farming communities to define research requirements and toassist in introduction of appropriate technology is most essential. It is,therefore, appropriate that CPATSA has been the most innovative in this area.

15. Despite the use of a farming system perspective, technology generation andtransfer in the Northeast faces an important constraint. This relates to thenear-subsistence status of a large proportion of farmers who are unable toaccumulate sufficient capital to undertake many of the appropriate technologiesfor this semiarid area. Most farmers are loathe to borrow at fully indexed,positive interest rates in a highly inflationary environment with a significantrisk of lower real prices for the products of the investment. Traditionally,many of the technological innovations in the Northeast have been introduced assubsidy or even grant components in area-development programs or with heavilysubsidised credit. This has resulted in many technologies being defined asappropriate to address defined problems in the farming systems, provided someform of government financial assistance is provided; An periods of fiscalconstraint this assistance must be limited. It appears that greater emphasiscould be given to define methods to introduce income increasing technologieswhich are more within the common resource base of farmers, and which do not haveto depend on government financial support.

W51/DOC/L121GANX

- 58 -

16. CPAC has concentrated its efforts on the obvious problems in the dominantcommercial farming system in the corrados and produced very significant results.This relative success in dealing with obvious production constraints appears tohave led to less emphasis to ensuring that all the important farming systems havebeen identified and their characteristics understood. Excluding the FederalDistrict of Brasilia, which would include many part-time farmers, the proportionof farms of less than 10 ha. in the Center-West is very substantial at 272, 332and 162 for Nato Grosso do Sul, Mato Grosso and Goias respectively; farms ofless than 100 ha. are 612, 712 and 592 respectively (Table 2). Considering theirnumerical significance, the analysis of the situation on smaller holdings in theCenter-West does not appear to have received adequate attention to date.

17. The Silvania exercise (para 9) which covers small, medium and large farmsin more than one farming system was a one-off operation. The expansion of thisprogram, however, into additional municipalities in the state of Goias doessuggest that more attention will be given to use of the farm system approach toassist in research programing and technology transfer in the future,particularly for medium scale farmers and smallholders. The less-intensiveapproach to identifying and working with producer groups in the expansion phaseof the Silvania project appears practical and replicable.

18. At CPATU, the numerical importance of smallholders dependent on a livingin the humid tropics does not appear to have been appreciated. In Maranhao, 842of holdings are less than 10 he in area, while in Rondonia, Amasons, Para andAmapa the farms in this small sise group make up from 282 to 462 of the totalfarms (Table 3). Apart from working on the development of sustainable farmingsystems for specific agroacological situations, CPATU needs to promote andcoordinate attention to the smallholder and recent-settler situation in theNorth.

W51/DOC/L1AnWX

”蔥妒低勺馴冷戶”肉爾d 臘購戶 .妒嗎跑馴開江一“他網細 肉 馴口電.馴開.,�‘“嗎�一 ld輪 寫劇“劇唱

.�� 馴磚.辟奮�� ��騰 馴磚�個總.霄 .�綢 ,話t個唱t.騰 .�館 騙日豐馴開.霄 .�, 電啊.劇嗡二口魄 個開.劇磚.論 ”馴鵠” 個論,. 磚, 馴圈 ,” 狗電.戲 t 二 馴購.觔電 ��韶 dJ �以0 .�以 口巳,�劇細�,口網 劇開�‘開�, ’妒口訌化 ..... . ⋯ ⋯ →雙杷 _ ‘二, 鰓電�馴口., �.寫 載,�綢0,&, .驢 ,魷方個瀾一,, ��啊 遞開仕馴口�. ,&. 遞鈴豐網諦,,口網 個鰓■口暱,.翰 ”閑蓄臢 ”’ 馴才, 〞 。 ,” �, ,&.&.,, .�” 榭 �鵬 ,�。 襯 �馴悻 �網 號�.觔遞 ’祠喝 開哺�皂寧 口

館· 口婉�,, ,�” 召村�點召 二計 蔥儲巒,開 一t, 個歸豐劇陰 二戲 綢個‘卹, 口網 以口‘,開., �叫U ” 戲 e 電開 , ,.& , 視蓄 �二 以召二 �,朧 �屹.�t �一 劇螂�胎 口戲 騖藝.戲 t ”妒口闢 籐口“唱

,· 個口.訂 t· 遞! ,開 .U仲 �·” 鱸 t .個個�.tg· 儲 勰 ,方闢� 一 tt 儲�‘寫胡 �戲 開 細.劇總.寡 ·叫- 二 細�” t 啊 , U 寫 牌 .甚 t 騙 .� t �,, 產U ,霄 ,�神 電編 .&,, 口統 劇魷�鈴 , ’妒 劉網州細 I甲

騙 . 胡 0 .牌 審.粤 ,幼 .馴購 �.t合 電闢 .口開 .召 ,. !釁 t鴉 .戲 t .召 L.戲 劇勵.�颼 口戲 盧鰓 .tt 乙.� .閑�t” 婦 � 開�” . �,’ 颼 , 二 , 乙挺 ‘,t .�偶 會辟 .以 t .魷 騙 ��露U 口口t 開口.口開 ”妒 劉綢馴戶開神 d

蓄·, ,糒 .臼 ⋯ 馴仲��鄴 �.” 以 乙��鄴 �召 二紹 開召豐鈴 ��t ,�� 症� ‘劇開 ��, �纏,.�語 ., ·叫神 , 以 細開 ” ,,, ,婦 寡”, 以 9�� ,�斗 鰓 蓄.鈴 二 以 轉 ,�個中電 口戲 開�.耗唱 ”妒 劉. 叫 l叫 個d飾

留奮 ----&. ‘。-一 _ _ -一 _--二 _-.- -一 。 一。..一 ‘- 一。一 ‘...-. -⋯ 。 。.-. -一 一 -一 ‘._. ‘ 一_一’ L�寡 觀纏�以,.,, .�認 辭召,繪,” ��寡, 以口吃個卹.t .�韶 乙,’仕召坤�, �”計 ‘婦.劇寫 口戲 耗仲.馴仲二 ’叫l 神 , 計 論 ” � ., 開自 �二 �開 .� t �歸 劇M�神 二 黝 開 認.魷 自戲 胡 自�朧 , ’妒 寫

劇開魚開悶 啊戶婦勺 �皂.�二 �,.奮粈 ,二 “.. ”口�. t�” .t, ,�”: 寡�“ &O豐。t.,t 二 �論、�� �戲 ”�.&t.,,.&一 ” 轉 , 騙 , 戲 ., .細 ,,, ,� 口神 ,儲 �.斗 魷 �,編 二 ” 飾 ��個口 口戲 劍磚�馴開 .妒 劉

馴 徊0念一 , 劉總�,颼., ,�啊 劇..�”二 �·“ ,�,.乙” ., 一‘, 乙.,豐胡口.⋯ , �寫.馴開 口戲 “口�開0., , �馴& ” 胡唱 t.&,&. , 網開,t ⋯ 」 �儲 , .&t 寡 奮勰 �」 ,&t必 .,��轉 , 口戲 露編 ,奮歸 ”開 劉

!闢 他J二 � 口口 .馴總�, ,·朧 乙卹 .‘一 �, ��驢 開 露.,神 ., 一,, 斗 召豐開 日.召 ,二 乙弱 .配 � 口婉 轉 言.ttt.,t �馴騙 ” 馴總,,I �” 騖 ,” 綢 挪�召 ,,, 開 口�粈 ‘’�, �磚 �」 ��購 ,t.�L汁 口戲 劉磚.闢 個 ’開 劉

叩甲劇領.�����.�.�.■�申叫�,■.� .�..鬨�.�闖■�,�闢���.■.口‘■k�. �■�■圈���咽■���■“���’間 �個�唱��..■目�,�..�‘ �馴�闢目悶�..�.�.���.�“■目��.�細口”��.�.�,�.�‘..�寫 J馴戶馴 寫 '馴戶馴 , J網戶馴 寫 J啊戶嗡 寫 J馴戶馴 寫 ‘�甲�啊 , 馴悶�馴 ’��‘-〕闖間戶甲開鬥 --鬥開n 用鬥一 ��,牌個不綢開r - 一 開馴開開r -一 �- 口頂觀掬吋--間 --鬥啊r- -界甲r 憫神一 妒

一 , ’ 柄 職 , ,r 一 一 一 一 一 一 一 一口閱中個闔口咽劇�囑團口唱馴口圈

馴口「頂悶�膩下�門�調戶呵不同網阿祠硎物〕�啊首爾叩肉r 下間戶悶門閑奮寫p 頸

C9M WST REGION* AMMUJIUR&L ESTABLISHMEMTS Eff 1129

Stöt* Tota I 51= of Fotablighemeta (ha)ila 19 lm ju - 10= l~ - la-CMIN 3,16-mö -medaklaced-

N~r N~r 9 ~ör--- - ---- -----

Meta 4 en. Do sel

E94. sålds l619.69 14976 27.18 legio 34.16 14716 28.89 6768 19.46 484 9.64 379 6.49ma 31231131 IN.M 84529 9.21 873898 2.18 6~ 99 17.32 15419336 49.34 987~1 99. M - -

mu Gr~

- Ett. 78376 169.0 25993 33.64 29610 37.85 17331 22.11 4922 8.26 du 6.93 81 6.69Ha 87962436 l199.60 114625 9.310 1192589 2.99 59~ 2 13.29 14112963 37.17 17~ 26 48.83 - -

Gala*

- Est. 179378 IN.O 291119 18.27 78138 42.46 64539 35.98 9179 5.12 led 0.11 161 19.619Ha 47"~ l00.60 18~ 9.35 3456159 7.25 19734118 41.47 29664906 43.69 879~ 7.94 - -

Födera l DIgtrIct orgel Ila

Ewt. 3434 l610.601 1114 92M 1766 49. N 554 18.13 47 1.37 - - 19 19.56mm in.m 49919 1.83 89179 19.87 148117 47.76 9~ N.94 - - -

------------- ------ 0%

TOta l

- Est. slom l66.610 71163 22.59 128187 30.99 97184 30.71 lom 8.29 107 6.41 tig 6.19- ma 117668328 109.619 30798 61.319 62MO22 4.52 9~ 78 26.91 U1219M 42.61 999~ 26.47------------- ---------- ---

AgrIemItäral Conemo 19M by Broxillan ImtItutc of Goograpby and Statlotlet.

AmIflluAL ES7ABLIinMTE BY LU2f Aim .

7____________ta 1 Sig& of Estabhamalt tha)o<16 .l--

- lom 1,532 11.19 22,88 27.92 42,771 62.44 1,81 19.18 441 6.64 29 .M3 al 9.97- . 8,N,647 If.f9 10,168 1.64 2,621,994 33.28 2,168,496 ff.81 961,567 16.02 o0,687 18.98Aero

- E9. 3f,32 . 1.1o 2,9~8 8.27 18,168 61.26 13,86 39.5 m .86 28 d.8 6 6.01- . ^,26,444 1M.g0 8,896 8.17 973,116 10.42 2,26,718 3.46 M7,178 1.24 1,«8,479 U1.58

- Eat. 117,114 lo.M 64,66 46.69 52,88 46.16 8,798 7.61 22 8.4 86 .0 m .26- Ne. 6,987,552 1M.ff 282,826 3.39 1,485,371 24.89 1,817,861 3.46 967,64 16.22 1,44,~8 26.04

- Ewt. 6,424 1a.m9 64 8.78 2,849 86.67 2,8 45.78 m57 8.67 17 6.26 1 6.62- N%. 2,167,128 If.f9 1,119 N.m6 145,116 6.73 48,529 22.69 1,238,48 67.3m 4,414 13.18

- Ee6. 254,=18 1M.1 83,I9f 32.76 131.861 67.73 36,185 14.34 2,268 .00 1M l.18 58 N .22-f.. 2N,i3,850 119.1 301,~88 1.28 4,5689,423 19.42 6,828,784 28.84 6,426,447 28.W 6,887,28 29.81

- E"6. 4,832 If.O 1,8~5 28.87 1,632 38.79 1,88 84.83 18 2.18 19 8.39 -- Na. 1.219,~81 18N.M 6,482 8.45 83,768 6.27 288,324 23.92 279,5= 22.69 79,624 48.87

- E"6. a88,8 la.m 447,313 83.78 67,829 1.8 22,89 4.29 2,28 8.42 71 8.81 4,368 8.32- Nm. 16,811,124 1N.O 879,267 4.44 2,3@,222 16.36 5,845,295 88.M. 4,987,887 82.44 1,^0,=56 8.94

- Es1. 78,878 1NG.M 26,8~ 88M. 29,618 37.16 17,381 22.11 4,22 8.28 ma8 9.88 el D.M8- Nm. 87,82,4»5 lo.M 114,626 8.38 1,1N2,639 2.9W 6,848,42 13.29 14,112,^53 37.18 17,~8,325 46.33

- Ej*. 1,112,41 1f.o 88,719 67.44 885,938 3.21 119,8~6 18.78 11,31= 1.82 1,882 N.m8 f,3~6 8.48- m. 8,167,911 1N.O 1,414,041 1.44 12,711,409 12.96 24,561,418 25.81 28,642,28 29.W8 ^,88,772 31.62

Sere: AgrIcult4@rgl C*~ess 198m by oraxillen I~6tute of Ge~graphy and Statistiie.

- 62 -

A 2

De r~ gar~ rollT~ amd pro~M Eg

ta å t~totgA---* - ef Ltmies and descriptiiepssr

La srl tlgral eroamd related seoctora

-- of predetlom (eredit, etc.)prblms and -

Aah41pS Di feto» ; Batiomal and regoal- aa.s -a.d witb problemo objectLve la

and po Ual agricultural soctor

coa mderatios of availabl grod national andt~ ~logis. to addreas regio~al r»esarch

probl and po~ta,l atrategy

gesearc ~esarcb to addroe1

i adaptverifyVerfflatiom Veraftoag:0 of research technology

seletis of geloct relevantT oloy tcnolog

Treafer of Transfer o technology goproducer»

Eeedhaek peed irc of produetta problem. andeatra~ la lateractiom with producera

- 63 -

AmE 2

chare 2

TMM n~ TRAzu~iz~= REERC~ETn1 Sist TO A e ma~ ydma

Ages~emm~eelö ta" "gelgPusblem I stfy and dweith #Mur a e~"*As..emen an

S v**spegte, n ochar rese~ u~*. todefin M and maalys blE-

andseetg with private t«b~heUml esiass

<U>

Prpee amd dicus oolmelca ith grop-ud M 87 81TfB

~reksc andDefime reffach geestremac By SEIT-fltE

~rpe, rwim amd eamete reseankprejecce (at Center amd ubeantorG) U7 -f

'iOa-faem verificatios rials with co- 87 8-38 WIt

Te*iicUat opesattm s of a@netao~, gup and afteaef

Selsets 1@Fö~* toch w for use in a 3arn withSSP~ie fanrmig y tta e

i ~ ~ --- fWPCO-22CEM

emsg with prodcer tiene andga~p and PT and giasi radie and

Taaefer ad televslm, al~ays direcedfso~se to equirl** to speuific fa~ 7S

by andm 37 Tr

edhsek -- Im al. face-to-face interam_a0t

Snr - tn"amsanaömd wchenin== :ranfer ~ita at MBoAm A Centers.*B - pese msesrehere

* - .lem C«bmm~ rae*aeboraiT - Techeg trmöfer wui~m la APA

- 64 -

ANNIEE 2

Chart 3

TECEnG TRAUFE ty USEARCU/ETENSOU STSTEN TO SMALLHODE PRODUCES

Agreselegleel Zening

Id=nicy farmer group. s clienele - foimalProblem sammtt4es. asseuia~ions, or i.nexnal safereneAsa-sememu grps of fSBIt who aa meet fea interaesV a

vith te8h~aL staf - thrugh an PersofaaUfMM * amd ava~able statelreso dy ET-~0.E'

U~drand far~ ee&mad identify problem.+ a Geck. throush Interaction ich gr~ in

Analyste ~CeagN, complemeted as nacissa by aWafoal famer gurvev techiquas gy 3ETT-fl.IT

~ Aalyta problem. and arrive at solutiona relevantto resource sicuation n easch raming systemwhere possible 97 ET 48-ET

| | andDefina research rquiremuts y SET-S-1

Rserch Prepare, review andeecute research 57 SE-S

Verification Of-fari verifcatlon ar WES Tickasaieta=* of sT

Fontatehology

fng use in *aEkT-aT

Sel4ctio Dememtrate wtgroup couceaus

Transfer 71=..TAC EM~uar meIng with o and regional

pr~educr reterece groupa radio and television,SU_*=T -almyn directed to

Sy Tf-ETa~pc f&~i *yactise

Feedback Fr. all faca-to-face Itceractione

SET - ocIoeconoIc and echnology transfer unke at E APA Ccnte.R - s6t se.eerchers

RE - EMAPA technical resesrchereET - EATER *tffTT - Techelgy transfer un1te a ERAPA

- 65 -

Footnote* for Charts 2 and 3 of Annex 2

1. Infosmed Person Interviews This involves an organised program of interviewswith persons selected for their knowledge of agriculture in a specific area(large or small)i persons could be farmers active in producer asociations,entemeem agemes with leag experience, beak managers, experienced staff inservice fime (Iputs, marketing) and agroprocesesiag, or socioeconomicresearcher with eperience in a particular area.

2. Reference Groupi This is a group of producers who have generally similarconerns and attitudes with respect to their production system, and who havesilar soees to resourcess this embeases the likelihood of member reactingIn a siailar way to overcome constraint& and uss opportunities. In the caseof miaituadio, a reference group usually ha. stronger norme than that whichexists with larpr. comercial producerel here, a strong slilarity in valuesand attitude encourages them to use each others opinions in arriving at adecision to reject, adopt or partially adopt agricultural technology.

3. In deftiing problem, established techniques of structured problem census inmeeting with groups of farmers permit a reliable asseesent of what mostfarmers consider as their problems (rather than the possible expression of theviews of a vocal minority).

4. STT-RS: It is likely that EKBRAPA will have to take the lead in introducingthe farming system perspective (PSP) approach to state research agencies, andwill continue to have a major direct role in technology transfer to largercamercial farmers through SITT.

5. Private technical assistance entities can be a very useful source ofinformation, especially for the commercial farming subsector.

6. SETT-RS-ETs In the case of smallholders and medium size farmers, the EMATERshave the major responsibility in technology transfer. EMBRAPA will have aleading role in educating MATER staff (and state research staff) in practicaltechniques using FSP approach to understand farmers circumstances and to adopteffective meas of technology transfer with farmer groups and in mass mediacomnication. EWATERs and state research agencies should progressivelyassme full responsibility, but mast be obliged to interact with ENBRAPA forresearch programiag and formulation of recomendations for farming systems.EMBRAPA centers would continue to undertake formal farm system surveys inspecific situations.

7. Rapid Informal Farmer Survey: Structured discussion with producer groups canbe used to reliably obtain an understanding of farmer circumstances, andEMBRAPA (at CPAC and CPATSA) has some experience in this technique. However,

another rapid appraisal device is the informal survey by a smallsocioeconomic-technical team of a number of individual farmers in a farmingsystem. These discussions occupy about six to ten days for a farming system.Unstructured interviews are based on a set of guidelines that are divided intosections and designed to elicit a description and understanding of the systemand to identify local problems and development potential. The team interviewsa farmer on one or more sections of the guidelines. At the end of each day,after each member has talked to perhaps three farmers about the same ordifferent sectors of the guidelines, the team mets and jointly evaluates theInformation obtained. Additional interviews are carried out until the teamis convinced that they understand the system. A report is prepared by theteam describing farmers' activities, the faming system and problemidentification.fAnt Collinson, M.P. (1982) "Farming Systems Research in Eastern Africa: The

Experience of CIMI and Some national Agricultural Research Services1976-81". International Development Paper No. 3. East Lansing:Michigan State University; and

Collinson, M.P. "Faming Systems Resarchs Diagnosing the Problems" inRasearch-Extension Farmer: A TvorWay Continuum for AgriculturalDevelopment edited by M. Cernea, J. Coulter and J. Russell. A WorldBank and UNDP Symposium 1984.

- 66 -

Dhmonstrations. Dstratglas diffet fre on-fam verification trialø (testesde ajuet) i that they are designed to show producers how an ifproved*Ueelegy oa redily fit into their produatim syøtem aad bo rølevant to theesgeaes at their disposal. la the case of mif~udis it is very i rportsat

that the de stratios is i cicumtamc*s which are co~s to the referencegroup targetd in the destratioal this is esured by allowig group øberto selet the demastratiom cooperator and atte, withi aay ~ecessary

sidelimes p ~ided by the exteasiom agent.

67 - ATTACDENKTCgENTs nCHE BoRnwER

EMPRESA BRASLEIRA DE PESOUISA AGROPECUARIA - MRAM

C.PR.N2 0/1 rsll,300 a

GRAHAM DONALDSONAgriculture, Infrastructure andHuman Resources Division, Chiefoperations Evaluation DepartmentWorld Bank

Dear Sir,

This is to inform you that we have received adraft of the performance report of the Second AgriculturalResearch Project (LM 2016-8R) prepared by your OperationsEvaluation Department and submitted for EMBRAPA's pertinentcomments.

On the whole, the Report describes in aconsistent way the conditions under which the project wasdeveloped and allows for Important suggestions as to howEM8RAPA can perfect future actions developed together withthis Bank in other contracts.

Comments concerning aspects which limited totalsuccess of the project proceed and do concern EMBRAPA whichwill take the necessary steps to remove these limitations.

Thus, the Third Agricultural Research Projectsigned January 1990 emphasizes those aspects inherent forthe conservation of the environment and the diffusion of

the results. Besides, EMBRAPA has transformed six research

centers in the North of country into five Agroflorest

Research Centers, whose principal aim is the development of

knowledgs and techonologies which will allow for the

rational and sustainable use of resources in these areas.

VINCULADA AO MINISTRIO DA AGRICULTURA

MA 66-MA

EMBRAPA has also received the incumbency to

coordinate and supervise farming extension activities and

has begun work to define estrategies which will integrate

research and extension in such a way as to diffuse betterthe results of research generated.

All these are necessary conditions for a

sustainable farming program which will take into

consideration environmental impact and undoubtedly will

lead to further development over the next few years.

Sincerely,

MURI(0 X IER FLORES

President

DCV/CRMS/msb

MAP SECTION

(v: EZU ELA, lFRZ ,,SURINAME\GIAN'

COLOM/B I oavitmAZIt> RORAI A 7AAT

LL. NTER ITORY EMP\..~.. x ITRY ,-TERRITOR 11 PROWECj ,-A MacapåLo~ohof % Rsech 5~on

AMA O P A R A ARANHAORn Å CE RA GR ONORTE

SMira *mø P1 A4d orte u. -in NA B

Af C leh

-RONDONI OC TINS

B A H R A SERGIPERi e Cruz das AlmisG R 0 S 5 O)

cuiaba GOIÁSSOL i; BRA5IIA& nrK@tina7. -t.odonciol

B 0 L. l V I A . . .°"

MATO GROsso*RA -{Grande Leopf P ýTO SANTOAWpo Belo Hrzne~

DO SUL

~~~A T PA A N0 Få Fbchca- ....... %. SÃO PAULOasI.RG~"^"ý*@ os ROPARA AYao0 c F A NCHILE

Cur io E Nn

as""VConco"d"oALIs cjSOUTH ss Foran' . ASA A~ mAU m uBRAZIL A R G E N T I' A F^udo O A TIOWAcoOOOfty cImun10 GRANDE DO SULZ 1Juson saTroWs

O TAT! AIN TERRITORY CATA"AMERCA P8rtoAlegre o Tows

i s rh. heBts a.AND. - ST A RITO 0o AMn1%. kh .. wh. » i. W.M Bo..d h. 3p4p !

r &".d k " i- URUGUAYj.*..-~ 6T

60^ m 1i-

v E N E Z U E L A5. (?ENEZULA <4 ~ FRENCHm

.. s,. . ~ SUR INAME

«ýRORA IAA 4 k1RRA7 __ANÄn

t OAMÅPÅAAMAP

P E R

AMA R A CEARÁ ooR

Am Z 0 N S

_ O

A C R E 'Ove Eý«RNA

- - M.o4P7 TOCANNRONDÖNIA

MA T 0 6 AP E R u - G R 0 S S 00~

C E N T E R rW E STG. A 5 F. . N ABRAZIL

G E \ A l SAGRICULTURAL RESEARCH fI PROJECT .-MAT0 GROSSý'Agro-Ecological Regions

.. n ESPIR ITOom~/SANTO

Aoto-EcotoGICAI ÄggAs D 0 5 U L S O U TMm

SÃO PAULO ° -'Mg - ~T - og -. - m RO DE J4 ~o

~-AMDp A R A N Á

-e AMEAG NYN CATARINA

'Me- Me illlIll ge.a,uo.s

S O

RIO GRANDEDO) SUL o"MAEE I ¥/ w -c

ENO SAIRESå OREVIDE D

Ym a lir