An evaluation of the CARDI/USAID


Material Information

An evaluation of the CARDI/USAID small farm multiple cropping systems research project no. 538-0015
Physical Description:
64 leaves : ; 28 cm.
Michigan State University -- Dept. of Agricultural Economics
United States -- Agency for International Development
Caribbean Agricultural Research and Development Institute
MIchigan State University and the office of Multi-sectoral development, Bureau for Science and Technology, United States Agency for International Development, Washington, D.C.
Place of Publication:
Arbor, Michigan
Publication Date:


Subjects / Keywords:
Agriculture -- Caribbean Area   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Includes bibliographical references (leaves 60-64).
General Note:
"Prepared by USAID Evaluation Team which visited St. Lucia, Antigua, St. Kitts, Montserrat, Dominica, St. Vincent and Trinidad, March 17 - April 8, 1982."
General Note:
Cover title.
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 668366361
System ID:

This item is only available as the following downloads:

Full Text



NOA. 538-0015*

Prepared by a USAID Evaluation Team
which visited

St. Lucia, Antigus, St. Kitts, Montserrat,
Dominica, St. Vincent and Trinidad

March 17- April 8, 1982

Team= Members

Everett Everson, Agronomist/Plant Breeder, Michigan State University and
Joseph Beausoleil, Economist, AID/ST, Wastington, DC., Co-Leaders,
Robert Deans, Livestock Specialist, Mic-ugin State University,
Russell Freed, Agronomist/Plant Breeder, Michigan State University, and
Daniel Gait, Production Economist, University of California, Davis.

*This evaluation was carried out urder the "Alternative Rural Development Strate-
gies" Cooperative Agreement AID/ta-CA-3 between the Department of Agricultural Eco-
nomics, Michigan State University and zte Office of Multi-Sectoral Development, BEreau
for Science and Technology, United S:ates Agency for International Development,
Washington, DC.





Description 1
Status 2
Logical Framework 2
Inputs 2
Outputs 2
Purpose 3
End of Project Status 3
Goal 3


Data Collection and Interpretation 4
Identifying Farm-Level Production Problems 5
Identification of Crop Production Constraints 6
Current Design of Farm Trials 8
Linkage With Extension 10
Summary of Implementation Problems 10
Training 11
Summary 13


Agricultural Research Policy in the Caribbean 15

Organization and Management Issues 16
CARDI Administrative Structure 16
Policy Direction 16
Technical Direction 16
Administration and Financial Management 18
Project Coordination 19
CARDI Staff Resources and Financial Issues 19
CARDI Staff Performance and Evaluation 20
Linkages with the University of West Indies 20
Financial Issues 21
Project Organization 21
Project Leader 21
The Regional Technical Coordinator 21
The Country Team Leader 22
The FSR Field/Territorial Network 23
Overall Assessment 23

Revision of FSR Methodology 23
Identifying Target Farming Systems:
Import Substitution and Export Commodities 25
Carrying Out Surveys of Homogeneous Farm Groups 25
On-Farm Trial Design and implementing Farm Trials 27
Summary 32


Strengthening Technical Research: Crop Production 33
Crop Production in the Eastern Caribbean 33
Production Constraints 34
Agronomic Environmental 34
Socioeconomic 35
Areas of Research 36
Intercropping 36
Agronomy Nutrition 36
Water Management 36
Regional Research Stations 37
Staffing 37

Strengthening Technical Research: Animal Production 38
Situation 38
Types of Animal Production Systems 38
Constraints on Animal Production 39
Nutrition 40
Integration of Livestock into the FSR Program:
Research Agenda 40

Training 43

Technical Assistance to CARDI 44






























Agency for International Development

Analysis of Variance

Bureau of Census

Caribbean Agricultural Extension Program

Caribbean Rural Development And Training Services

Caribbean Agricultural Research and Development Institute

Caribbean Economic community

Center for Tropical Agricultural Research and Training

International Maize and Wheat Improvement Center

Country Team

Country Team Leader

Farm Profile

Farming Systems Research

Field Trials

International Agricultural Development Service

International Agricultural Research Institute

International Center for Research In Semi-Arid Tropics

International Institute for Tropical Agriculture

International Rice Research Institute

Small Farm Multiple Cropping Project

United States AID Mission in Barbados

University of the West Indies


The Small Farm Multiple Cropping Systems Research Project No. 538-0015 provides
grant funds of $2,210,700 in financial and technical assistance to the Caribbean Agricultural
Research and Development Institute (CARDI) to establish a research program which
addresses the problem of low productivity of small farmers on the island territories of the
Eastern Caribbean. CARDI was not quite four years old when the Project was approved in
1978. It had just begun to decentralize its activities and to emphasize food crop rather than
plantation export crop research. The Project reinforced both these new thrusts by
facilitating the organization of cooperating country/CARDI field research teams to conduct
on-farm research with smallholder food producers.
Prior to this initiative, the AID Mission in Barbados (USAID/B) made available in
FY 76 and FY 77 at a total of $285,000 in grant assistance to CARDI to establish three
regional agricultural research stations, one each in Belize, St. Lucia, and St. Kitts. The
establishment of these stations provided CARDI with the facilities to conduct traditional
agricultural research work in agro-ecolgical zones very distinct from that of CARDI's home
base in Trinidad.
The present initiative goes furthe- in that it brings the research work to the farmers'
field. The Project incorporates two distinct stages in its design. First, a diagnosis is made
of the natural (physical and biological) and social (economic and cultural) conditions under
which the small farmer operates. Then, after studying the crop and livestock activities
within this total farm context and their relationship to this external environment,
experiments are designed to apply or adapt technology which will improve the farming
system. The approach, commonly referred to as farming systems research (FSR), is not a
substitute for traditional station research but rather requires the support of station
commodity and disciplinary researchers. The CARDI core staff is then an essential element
of the Project.
The goal of the Project is "to increase the value of agricultural production in the less
developed countries of the Eastern Caribbean through the improvement of small farm
profitability, nutritional productivity, and employment generation" (Project Paper). In
designing the Project, it was recognized that the accomplishment of this goal would depend
on factors beyond the scope of the Project. The research results had to be disseminated,
inputs had to be made available, and market channels improved. The Project was expected,
however, to influence this supporting institutional environment by providing an information-
al flow to agricultural policy makers and planners.

The Project is essentially an adaptive research effort whose purpose is to "develop
recommendations for improved farming systems through adaptive, farm-based research
which farmers can and will use, extension agents can explain, and credit institutions will
finance" (Project Paper). The approach used was first to understand the characteristics of
the farming systems and then to adapt improved technology that fits the characteristics of
the farming system to the system.

The Project was evaluated in order to assess progress to date and to determine
whether further financial and/or technical assistance would be necessary beyond the life-of-
project date of November 1982. The terms of reference prepared by the AID Mission in
Barbados (USAID/B) specified that the evaluation team should examine the "effectiveness"
of the project, the "appropriateness" of the model, and provide "specific recommendations
concerning further assistance" (see Appendix A, Terms of Reference).

Logical Framework
The logical framework is attached as Appendix C.

The project has been extended through November 1982, at which time all AID funds
will have been spent. The line items were adjusted in August 1981, to reflect programmatic
changes. The CARDI contribution was not reviewed, nor was that of the LDC countries.

(A) The most significant output of this project was the establishment of eight
country teams (CTs) in seven territories (St. Kitts and Nevis each have one team).
Approximately 25 farmer cooperators were interviewed by these teams for the diagnostic
survey in each country. On-farm trials are being conducted with about 20 percent of
farmers included in the diagnostic survey. In most territories, a Ministry counterpart
worked as a team member. The evaluation team was impressed with the enthusiasm of the
country teams. With some additional training most of the personnel on the country teams
will be able to carry out their Farming Systems Research (FSR) responsibilities.
(B) Baseline surveys are available for six of the territories. The baseline surveys and
the weekly interviews with farmers over a year-long period have provided CARDI cae and
country staff with a better understanding of farming in the region and the special problems
of small farmers. A number of specific studies have been completed which provide imnights
into the resources, constraints, and objectives of the farmers.
(C) Farm profiles are being developed for 5 to 10 farmers in each territory but these
will not provide information which can be used to develop extension recommendations for
larger groups of farmers.

(D) To date, the project has not developed improved smallholder farming systems.
The massive data collection exercise and the preparation of farm profiles have delayed
moving into replicated on-farm testing. Most of the on-farm testing has involved varying
only one component of the system. The on-farm trials have been mainly observation trials.

(A) The project has not met its purpose, nor will it do so by November 1982. If any
recommendations for small farmers are forthcoming from the ad hoc on-farm trials, they
will need to be validated through at least one more cropping season.

End of Project Status
(A) Country teams have been formed in seven territories and are starting to
function. CARDI now has a visible presence in the seven territories, and its country teams
have developed good rapport with farmers and Ministry personnel.
(B) There have been no formal research recommendations that have emerged for the
extension service as a result of the research activities of the project to date.
(C) There is an indication that the territories will have difficulty in covering the
cost of all field personnel included in the project.
(D) Cooperating farmers in several territories have planted virus-free yams, and
black belly sheep were introduced in St. Lucia. However, the project has had very little
impact on non-cooperating farmers.
(E) CARDI has not formulated recommended practices for the credit institutions
because they have not yet carried out the full sequence of FSR activities.

The purpose has not been met and, therefore, the goal has not been affected by this
project as yet. A word of caution is in order. Some of the shortcomings of the project to
date are directly linked to the false expectations which were built into the design of the
original project. The evaluation team is of the conviction that the original project was
unrealistic in assuming that a functioning on-farm research program could be established in
the territories in a three- to four-year period. This process will require a 10-to 15-year
effort and this time frame should be taken into account by the Phase If design team.
Since there was not enough time to accomplish the Project purpose and certainly
insufficient time to fulfill its goal, the Project should be evaluated at this time with respect
to its outputs, especially with respect to its institutional development aspects. The
following sections will focus on this aspect of the Project, i.e., the establishment of an FSR
program in the cooperating countries. Specifically, the evaluation will consider what has
been done by CARDI and what needs to be done to further establish an FSR capacity in the
research systems of the cooperating countries.


Data Collection and Interpretation
The project was implemented by first undertaking baseline surveys in six of the
territories between March and August of 1979. A total of 120 farmers per territory were
interviewed under a contractual arrangement with the University of the West Indies (UWI).
The next step was the initiation of the diagnostic phase which included weekly data
collection on a sub-sample of farms (+ 25 per territory) included in the baseline survey
between June 1979 (St. Lucia) and January 1981 (Dominica).1 For this diagnostic phase, the
following criteria were used to select the sub-sample of farmers from the original sample of
(a) Farm size between 1-5 acres (or 1-15 acres in St. Lucia);
(b) The farm should be representative of those near it;
(c) The farmer must be a willing participant; and
(d) The farm should be near a major road (logistical consideration).
No attempt was made between the baseline survey and the diagnostic survey phases to
identify homogeneous sub-samples of farmers based on usual selection criteria (e.g., major
cropping systems or predominant cash crop, part-time versus full-time farmers, rainfall
and/or soil type classifications, marketing problems, etc.).
During the diagnostic phases, each country team (CT) was to collect agro-socioeco-
nomic data for one year. Analysis of this data by the economic resource group in Trinidad
was to provide the quantitative basis for selecting interventions2 for farm research.
Analysis of this data was to reinforce the observations of each country team and assist them
in identifying potential farm production improvements. Problems encountered in the
diagnostic data collection phases (Manteiga, 1980, 1981; Rosen, 1981; Cuevas and Weber,
1981; Goodhue and Ferraiuolo, 1981; CARDI, March 1981) led to a modification of both
timing of farmer interviews (to fortnightly) and format (BUCEN forms, 1981).
In August 1981, a decision was made to revise the diagnostic methodology by having
the country teams concentrate on the preparation of farm profiles (FP) (Manteiga, 1981).
These FPs, which are not specified in the project paper, are case studies of individual
farmers taken from a sub-sample of the 25 farmers in the diagnostic sample (3essee, 1981,
p. 16). The CTs and other CARDI staff have focused on preparing five FPs per territory as

IHurricane David was responsible for setting back initiation of the program in

2An intervention, as defined by CARDI, is a simple non-replicated farm-level trial on
a limited number of farms, designed to confirm to researchers that a potential improvement
should be considered in a set of formal trials in the following year.

the major visible project output before the end of the Project in November 1982.
Unfortunately, the non-systematic selection of the FPs means that the synthesis of farm
profiles for a territory will not necessarily "represent" the farmers of that territory, nor will
a summary of FPs across countries necessarily represent the farmers of the Eastern

Identifying Farm-Level Production Problems
A key to farming systems research is the rapid identification of major farm-level
production constraints in order to design farm trials. In the CARDI project, two general
types of data were collected. The first type of data included the two baseline surveys which
were completed relatively early in the project life. The final baseline reports (Henderson
and Gomes, 1979; Rankine et al., 1980) are quite long on tabular presentations and quite
short on analysis. In addition, neither set of data was available until the project had
progressed beyond the need for the general information they provided. Both surveys were
undertaken by UWI researchers, but data processing and report preparation were slow.
Moreover, because CARDI CTs did not participate in the baseline surveys, they have very
little personal feeling for the farmers surveyed in their territory.
The second type of data generated was the diagnostic, or in-depth farmer interviews.
If done at all, such interviews usually take place only after the researchers have stratified
the farm universe into homogeneous groups. Such interviews usually coincide with, or follow
(not precede) the implementation of farm trials. Diagnostic data are available in the
memories, notebooks, and worksheets of each CT enumerator, and in UWI's computer and
subsequent printouts. Members of the evaluation team discovered that the CT members who
collected the data either consider their own personal recollections more "valid" than
CARDI's centralized (computerized) data, or they believe the subsequent analysis based on
computer printouts will corroborate their impressions of the farming realities of their sub-
sample of farmers. Neither view is totally correct. The facts are that (1) the complexity of
data collected during the diagnostic stage is such that personal recollections cannot be
reconciled with computer printouts (Chang, personal communication, 1982), (2) the reason
for collecting the diagnostic data was never explained clearly to each CT, and (3) the
incredible quantity of data collected emphasized the complexity of the components of each
farming system. This emphasis on components tends to obscure the project objective of
identifying major farming systems and ranking major farm-level constraints within each
The mass of data collected during the diagnostic stage has been processed after a
considerable delay. This delay has prevented the CTs from designing Farm Trials (FTs) to
confirm or reject potential improvements in current farming systems. Moreover, CARDI

staff and CTs have not used the data collected to characterize either the Caribbean farmer
in general (sample size too small and non-random), or to identify homogeneous farmers. The
only types of on-farm trials identified to date have been called intervention trials. Such
interventions are derived from informal observation and/or prior findings of researchers.
CARDI staff have added an intermediate step between diagnostic and intervention trials.
This step is known as "exploratory intervention," and refers to a simple change in the
farmer's crop system practice, which is unreplicated and ad hoc in origin and intent (i.e., it
is exploratory). Exploratory interventions are now being carried out on a small number of
farms in each territory. The number of farms where exploratory interventions are being
carried out varies from seven to fifteen per territory.
The evaluation team saw very little data on the results of the exploratory interven-
tions. Appendix D lists the titles of the interventions for the eight territories and the status
of the trials and number of farms. The interventions, started in St. Lucia in 1980, involved
planting legumes on farmers' fields. These legume trials, including blackeye peas, peanuts,
kidney beans, and bodie beans, were conducted on ten farms. Some of the farms had just
one legume while others had two or three. Data collected included yield, consumption on
the farm, and planting and harvest dates. There was no replication nor was plant population
recorded. Many of the legumes were planted into mixed cropping areas. The major
objective was to increase protein consumption of the farm family. The amount of protein
consumed per family was calculated and ranged from 2.6 to 28.8 pounds. The objective to
increase protein consumption is a good one, but the data collected to date will not enable
the CT to make recommendations for other farmers. Is it better to grow tannia and sell it
to buy imported legumes, or should the farmer grow cowpeas? If comparisons were made of
the amount of protein produced by growing cowpeas as compared to tannia, the results
would be more dramatic. With a little refinement in technique, some useful information
could be obtained from these on-farm trials.

Identification of Crop Production Constraints
In May 1981, several core CARDI staff members met with all members of the Project
staff to identify crop production constraints in each territory. The publication for St. Lucia
(George et al., 1981) lists general as well as specific constraints. The workshop participants
agreed (CARDI, 1981, Appendix D) that three of the major constraints facing Eastern
Caribbean farmers were (1) labor, (2) water, and (3) marketing-related issues. As a result of
this meeting, each country team was to identify intervention trials which could overcome
these constraints.
Table I shows the 40 intervention trials (of eight major agronomic types) planned by
CARDI staff and the CTs for the 1981-82 growing season. Broadly speaking, trials based on

Table 1.

Types of On-Farm Agronomic Trials
CARDI, May 1981

No. of No. of No. of
Type Trials Farms Territories

Improved Variety 6 34-36 4

Fertilizer Levels 5 24-282 4

Pest Control 4 18+ 3

New Cropping Systems 2 20 2

Livestock Intervention 2 5+ 2

Plant Density 1 6 1

Pattern 9 48+2 4

Tech-Pack 11 8-12+2 5

TOTALS 401 163-173+

IThis total includes counting six combination trials twice. Actually 34 trials were to
have been planted in 1981.

2These trials exclude all combination trials. Thus, the total of this column represents
the total number of farms (locations) with trials in 1981.

SOURCE: CARDI. 1981. "Summary of Intervention Workshop Held on May 18-23, 1981."
Small Farmer Multiple Cropping Systems Research Project CARDI/USAID 538-
0015, Trinidad, W. I.

the tech-pack concept (n = 11, or 27.5 percent of the total) and intercropping and/or
planting pattern (n = 9, or 22.5 percent of the total) represent 50 percent of the proposed
interventions (n = 20). Other traditional trials, such as introducing new varieties (n = 6, or
15 percent), fertilizer trials (n = 5, or 12.5 percent), and pest control (n = 4, or 10 percent)
represent about 31.5 percent of the remaining trials. The remaining trials (12.5 percent)
represent new crop system introductions (n = 2), livestock interventions (n = 2), and density
(plant population) trials (n = 1). With the exception of the weed control trials, the livestock
trials, the density trials, and the intercropping/planting pattern trials, the remaining trials
could have been designed before the baseline and diagnostic surveys had been carried out.
More details of CARDI's proposed farm trials are presented in Table 2. CARDI
proposes to examine 19 different crop and/or livestock systems during the 1981-82 season.
Since this is the first year of farm trials, the evaluation team feels this is an overambitious
undertaking because the CTs will be spread too thin. The CTs' proposing interventions in
five or six different crops and/or livestock systems on St. Lucia, St. Vincent, Antigua,
Montserrat, and St. Kitts are particularly overambitious.

Current Design of Farm Trials
Intervention trial design is based on the introduction of simple technology or variety
changes affecting a very small area of a farmer's holding. With the exception of the CT in
Montserrat, trials are placed in farmers' fields without replication. In Montserrat, three
replicates each of corn + peanuts and peanut spacing were observed on two farms.
Intervention trials are in fact viewed as preliminary observation trials for new technology by
CARDI staff.
Intervention trials to date have not all been accepted by the recipient farmers. For
example, one of the more promising interventions, in Antigua, involves mulching with
Guinea grass. This intervention conserves moisture on an island which received relatively
little rainfall, helps keep weeds in check, and may reduce vegetable losses to diseases.
However, some of the farmers do not even spread the riulch supplied by the CT, arguing
that it is too much work. In fact, it would not be too much work if each farmer had ready
access to a supply of Guinea grass at the beginning of the cropping cycle before weed
growth began. Thus, these farmers perceive the risk of spending their time to look for a
source of mulch as being too great when compared to the known risk of weeding every day.
Thus, they choose the known drudgery of weeding.
In summary, farm trials in different territories are a mixture of exploratory
intervention and intervention trials. Many interventions consist of "shelf" technology or
tech-pack components which could have been proposed without any of the length data
collection phases--the baseline and diagnostic surveys- -which preceded thle farm trial

Specific Crop and/or Livestock Systems for On-Farm
Intervention Trials by CARDI, May 1981

Specific Crop and/or No. of No. of Specific Territories
Livestock System Trials Farms to Host Trials

Sweet Potato

Grain Legumes

Hot Pepper
Forage Legume




St. Lucia, Antigua
St. Lucia
St. Lucia, Antigua
Antigua, Montserrat,
St. Kitts
St. Lucia, St. Kitts
St. Lucia
St. Vincent, Nevis,
St. Kitts
St. Vincent
St. Vincent
St. Vincent
St. Vincent
Grenada, Montserrat
Antigua, St. Kitts
Montserrat, Dominica
Dominica, St. Kitts

SOURCE: CARDI. 1981. "Summary of Intervention Workshop Held on May 18-23, 1981."
Small Farmer Multiple Cropping Systems Research Project CARDI/USAID 538-
0015, Trinidad, W. I.

Table 2.

phase. A more serious problem is the lack of thought going into the design of the trials
themselves. Moreover, more attention should have been directed to marketing issues as an
integral part of the process of defining research priorities. The universe of farmers was not
stratified into homogeneous groups before the trial design phase. The minimum number of
farms to maximize the probability of statistically significant results across homogeneous
groups was not derived. Analysis of intervention trials was not considered at the time of
trial planning. Given the ad hoc approach to intervention trial design, analysis across farms
is not possible in the 1981-82 growing season. As far as the evaluation team can ascertain,
the individual CTs need guidance from CARDI on how to set research priorities and then to
move beyond the identification of constraint and ad hoc interventions, to the rest of the
steps in FSR, including analysis of the results of on-farm trials, redesign of farm trials for
the next season, testing of improved technology under farmer conditions (verification stage),
and the diffusion of improved practices through extension agents.

Linkage With Extension
The evaluation team sensed that a new dimension has been added to the agricultural
research-extension relationship. The team views the FSR effort of CARDI in the Eastern
Caribbean as an applied research infrastructure to the bridge between research and
extension. The bridge is built by getting the involvement of the extension service in all
phases of this project from planning to execution and evaluation. By involving extension
workers in the on-farm testing program from the onset, technology dissemination is
accomplished with a minimum of delay. Once the farm trials have been evaluated, the
extension team member is familiar with the treatments and can communicate the
appropriate technology to a larger group of farmers.

Summary of Implementation Problems
Lengthy sections of several documents are devoted to many of the implementation
problems of SFMCP (Jessee, 1982; Manteiga, 1980, 1981; Rosen, 1981; Cuevas and Weber,
1981; Goodhue and Ferraiuolo, 1981; CARDI, 1981; Chang, 1981). In view of the evaluation
team, the most serious of the problems encountered in implementing farm research were:
(a) Early and systematic evaluation of the SFMCP by USAID, while specified in the
Project Paper, never occurred. The team found no evidence of systematic
reports, which are the usual method of tracing a project's progress.
(b) The data collection process has been allowed to dictate project objectives and
manpower development, not vice-versa.
(c) Partly exacerbated by the decision to begin initial work in all eight territories,
the field teams were spread too thin. The teams tried to document too much
detail about a sub-sample of farmers which represents neither a homogeneous

group within a territory, nor a random sample of the farmers of all the
(d) Ad hoc exploratory interventions have not emerged from the constraints identi-
fied in the baseline or diagnostic data collection phases, nor have they been
systematically replicated enough to represent either within-farm or intra-zonal
(e) Too much up-front emphasis on data collection and detailed analysis, coupled
with a lack of implementation flexibility, has delayed field trials.
(f) Failure to attain a true interdisciplinary interaction of CARDI core personnel
has led to minimal benefits from the potential interaction of the several
disciplines involved in the research project.
(g) Centralized decision making, as distinct from research team consensus, has led
to fragmentation (or compartmentalization) of expertise and to the formation of
the attitude of "we-them," particularly between (1) central leadership and the
socioeconomics resource unit, (2) the CTs and the rest of the project, and (3) a
given CT and other CTs.

As required in the project grant agreement, a number of research-oriented workshops
and seminars were conducted for the SFMCP staff to acquaint them with the project.
Subjects covered included the farmer survey and assessment phase of the project, the
methodology of farming systems research, crops and livestock constraints, on-farm re-
search, and other topics. There is a critical need for FSR staff to have additional training.
The following list illustrates the range of topics covered in the training sessions.
15/1/79 Training of Enumerators for Baseline Survey of Small
17/4/79-6/5/79 USAID/CARDI Small Farm Multiple Cropping Systems
Research Project Workshop held in Trinidad.
22-23/10/79 Seminar on Activities to date in Small Farm Multiple
Cropping Systems Research Project held at Hotel La Toc.
21/1/80 Seminar on Small Farm Systems in India conducted by B.A.
Krantz, Consultant to CARDI/USAID Small Farm Multiple
Cropping Systems Research Project.
9-15/3/80 Data Collection Workshop for the Multiple Cropping Project
in St. Vincent.
16-18/4/80 Animal Production Workshoo in St. Augustine, Trinidad.

21-25/4/80 Weed Control Workshop for the Small Farm Multiple Crop-
ping Systems Research Project held in Grenada.
1-7/6/80 Intervention Workshop for the Small Farm Multiple Cropping
Systems Research Project held in Trinidad.
25/6/80 Plant Diagnosis Course as part of Small Farm Multiple
Cropping Systems Research Project conducted by B.A.
Krantz, Consultant, Laxman Singh, Systems Agronomist, and
Winston Small, Plant Pathologist.
17-24/5/81 Small Farm Multiple Cropping Project Intervention Work-
0, shop, Trinidad.
22-26/2/82 Enumerator Training Workshop for New Questionnaire con-
ducted by Carol Weber of BUCEN.
The review committee commends the CARDI-SFMCP staff on the number and quality
of their workshops and seminars. However, the evaluation team found little evidence of
application or follow-up after these meetings. Furthermore, these activities would have
been more effective with the addition of monitoring tours as a part of each workshop and
the preparation of a trip report following the inter-territorial travel of each scientist.
The following personnel were provided training abroad in multiple cropping research
3. Hammerton June 1978 CATIE, Turrialba, Costa Rica
S.Q. Haque June 1978 CATIE, Turrialba, Costa Rica
St. C. Forde June 1978 CATIE, Turrialba, Costa Rica
S. Parasram June 1978 CATIE, Turrialba, Costa Rica
3. Hammerton Julv 11. 1979- ICRISAT. IITA. IRRI

C. George
A. All
N. Kirton
R. Carew
R.H. Phelps
K. Donawa
3. Hammerton
V. Sargeant
3. Cropper
S.Q. Haque
M.M. Alam
C. Madramootoo

August 9, 1979
November 26-30, 1979
November 26-30, 1979
November 26-30, 1979
November 26-30, 1979
November 26-30, 1979
February 11-18, 1979
January 2-5, 1979
January 2-5, 1979
January 2-5, 1979
August 1979
August 1979

CATIE, Turrialba, Costa Rica
CATIE, Turrialba, Costa Rica
CATIE, Turrialba, Costa Rica
CATIE, Turrialba, Costa Rica
CATIE, Turrialba, Costa Rica

CIAT (Colombia)
CATIE, Turrialba, Costa Rica
CATIE, Turrialba, Costa Rica
CATIE, Turrialba, Costa Rica
International Congress, WA
International Congress, WA

3. Lowery 1979 CATIE
S. Parasram November 1979- IRRI, AVRDC, ICRISAT,
January 1980 IARI, IITA
H. Harricharan July 1980 Dairy Development Center,
Guyarat Anan, India
3.A. Bergasse September 17-19, 1980 CATIE, Turrialba, Costa Rica
M. Alam November 17, 1980- India
December 13, 1980
R. Carew 1981 WINROCK International,
Arkansas, USA
A. Ali 1981 WINROCK International
Arkansas, USA
G. Mohammed 1981 WINROCK International
Arkansas, USA
R. Carew 1982 CATIE
L. Singh 1982 CATIE
3. Hammerton 1982 CATIE

In view of the evaluation team, the SFMCP, as designed, was far too ambitious. The
design team failed to take into consideration the complexity of farming in the Eastern
Caribbean and to realize that the state-of-the-art of Farming Systems Research is still in
its infancy. Moreover, data collection was allowed to dictate project objectives and
manpower deployment, not vice-versa. Personnel who designed the baseline and diagnostic
questionnaires did not consider how the data were to be analyzed before they were
collected. Moreover, the type of computer and the availability of computer software should
have been considered when the questionnaires were designed. At that time in the life of the
project, the data collection process should have been in the hands of a capable FSR data
management specialist. Too much up-front emphasis on the year-long diagnostic survey
delayed the introduction of farm trials. The ad hoc field trials to date have not been
systematically replicated enough to represent either within-farm or intra-zonal variability.
Furthermore, the project (partly exacerbated by the early decision to begin working in eight
territories) has spread itself too thin and has tried to capture far too much detail about a
subsample of farmers who are not representative of a homogeneous group of farmers within
a country.
The evaluation team had difficulty in obtaining a good overview of the training phase
of the project. From all appearances, training was assigned a low priority and most training
was ad hoc. A master plan for training needs to be developed.

Since this was CARDI's first major project, it is understandable that there were initial
problems of coordination and reporting. A number of miscellaneous reports have been issued
by the SFMCP but the project does not appear to be well documented. Authority for
approval of inter-island travel is not defined and the evaluation team found few trip reports
which were of use to further project objectives.
In the area of the SFMCP technical guidance and coordination, the evaluation team is
of the opinion that advisory committees should be established at several levels to improve
communication and the overall management of scarce technical personnel and resources.
The Small Farm Multiple Cropping Systems Research Project (SFMCP) was established
to develop recommendations for improved farming systems through adaptive, farm-based
research. Although the ambitious objectives of the project paper were not fulfilled and
many of the expected results were not obtained, there is good rapport between the field
teams and farmers. One of the unexpected outcomes to date has been the strong support of
farmers for the program. Moreover, in every territory visited, it was obvious in
conversations with the Ministers of Agriculture, the Permanent First Secretary of Agricul-
ture, and/or the Chief Agricultural Officers that they considered the CARDI Research
Program and the FSR Program as their program. In several cases, this is the first tangiL.e
Ministry research effort in their country and they plan to support it.


The evaluation team has identified the following major issues that need to be
addressed in considering further USAID assistance to CARDI:
Agricultural Research Policy in the Caribbean
Organization and Management Issues
Revision of FSR Methodology
Strengthening Technical Research: Crop Production
Strengthening Technical Research: Animal Production
Technical Assistance

Agricultural Research Policy in the Caribbean

The starting point for Phase II discussions is to address the question of what is a
feasible research mission for CARDI in the Caribbean? This question involves technical,
financial, political, and logistical dimensions which can be framed in the context of
preparing an agricultural research strategy and policy statement for the Eastern Carib-
bean.I While there is no standard model of developing regional and national research
systems, there is substantial empirical evidence which shows that an effective agricultural
research system needs to be (a) decentralized and (b) "articulated." In order for regional
and national research systems to address regional and local problems, it will be necessary to
build centers of excellence (mainly research on major commodities and interregional and
international trade) in the region before decentralized research teams can be effective in
carrying out applied research on local problems.
The evaluation team is of the opinion that CARDI should (a) develop its capacity for
research on problems common to the region and (b) help departments and ministries of
agriculture strengthen their own applied research capability in each of the eight territories.
CARDI's field teams (CTs) represent the first step in helping develop a modest local applied
research capacity.
Building local and regional research capacity should be viewed in a 10-20 year time
period. The Phase 11 design team should urge CARDI to convene a seminar to discuss
agricultural research strategy and policy questions for the 1980s and 1990s in order to

IThese issues are addressed in Ruttan's recent book, Agricultural Research Policy

2Articularion refers to an effective set of linkages and feedback between internation-
al, regional, national, and research systems.

clarify which problems are regional in scope and require CARDI's core staff and which
problems should be addressed by local research teams. In the final analysis, growth in
territorial support for CARDI will likely be sustained only to the degree that CARDI can
respond efficiently to territorial needs. Support also depends on the extent to which the
economic benefits of CARDI's programs can be effectively measured and demonstrated.I

Organization and Management Issues

Organization and management were primary concerns of the evaluation team. The
absence of the key CARDI staff people, when the team was at the headquarters in Trinidad,
made it difficult to thoroughly review these two aspects. A number of issues, however,
were raised by the teams. Some relate to CARDI as a regional research institution and
others to the Project as an outreach effort on the part of that institution.
CARDI was a new organization when the Project began. The first executive director
had just been appointed. The FSR effort was CARDI's first major Project. In three years
(1979 to 1982), the professional staff doubled in number. At the time of the evaluation, half
of the staff was assigned to the SFMCP Project.
This rapid growth and emphasis on a decentralized program put a tremendous demand
on core staff both technically and administratively. The organizational structure (see
Figure 1) that has emerged, though, seems adequate. There is a standing committee of
representatives of the member states, a board of directors responsible for policy determina-
tion, and an executive director for day-to-day operations. Three directors (research and
development, finance, and administration) report directly to the executive director.

CARDI Administrative Structure

Policy Direction
The Executive Director's role in policy making and liaison with regional governments is
clearly defined. He has established communication linkages with numerous donors to
broaden the resource base for CARDI.

Technical Direction
The Director of Research and Development is responsible for the research programs in
general and the specific research projects. Chiefs of research sections, leaders of country
units, and project directors report directly to him. He functions as supervisor and

IFor information on helping improve national research systems with small pools of
trained scientists, see CIMMYT's experience in working with national research systems in
Eastern Africa over the 1976-81 period (Collinson, 1982).

Figure 1

CARDI Administrative Structure

Policy Level -

Guidance and


I Comnodity
Support Unit Projicts
Service 'ltams EDR, IDRc

- Director Unit 'Team Project
Leaders Directoraders


Research &

coordinator of an array of research specialists scattered over a large geographic area
accessible reliably only through air travel. Supervision and coordination of the research
program is the number one organizational and management issue for CARDI. The Director
of Research and Development needs assistance in carrying out his functions. He needs to be
informed at all times so that he can allocate scarce resources of CARDI to best accomplish
its objective.

A number of suggestions have been made on how this may be accomplished. One idea
that CARDI might consider is the establishment of a Technical Management Group to assist
the Director of Research and Development in the overall management of technical
personnel and resources of CARDI. This group would consist of each project leader and the
head of each research section. These people would function (1) as a group for interdiscipli-
nary communication among themselves and with the Director of Research and Development,
and (2) as individual leaders to evaluate and monitor the various staff in their own
disciplines, including recommendations for salary increments and reporting achievements of
progress of each staff member to the Director.
In Trinidad, there is an activity that complements this idea, although it is country
specific. It is a policy and review committee composed of representatives of CARDI, the
University of West Indies (UWI), and the Ministry of Agriculture. The committee has the
authority to coordinate, review, and approve all research activities in the country. The
evaluation team recommends that CARDI consider establishing similar committees in each
country where it has major research activities. The CARDI representative on each of these
committees could keep the Director informed of the proceedings of their meetings and this
would alleviate to some extent the need for personal presence.
There are many areas which need to be improved. Position descriptions need to be
written, employee regulations developed, incentive plan determined, performance evaluation
system introduced, administrative procedures documented, accounting procedures improved,
and internal audit established.

Administration and Financial Management
It is planned that the Director of Administration and the Director of Finance will be
responsible for general administration and financial control. The specific responsibilities of
these positions should be clearly defined and communicated to all CARDI headquarters and
field staff. There is also a need to refine the process of fund authorizations and
accountability. The team recognizes the difficult communication problem between Trinidad
and the various territories. The establishment of the SFMCP leader position in St. Lucia
with an administrative assistant should strengthen the system. Ar. important task foi this

new project leader will be to develop a sound system for travel authorization, accountability
of results of travel, and documentation of expenditures.

Project Coordination
As a Regional organization, CARDI staff members interact with Ministries, Universi-
ties, Private Sector Organizations, Cooperatives, and individual farmers. This interaction
occurs both at the institutional/administrative level and at the individual staff level. There
are presently two coordinating groups which guide and coordinate CARDI activities.
A Policy and Review Committee has been established at the Trinidad Unit and similar
committees are being considered for Guyana and Jamaica. This committee is composed of
policy level representatives of CARDI, the University, and the Ministry of Agriculture; the
committee has the authority to coordinate, review, and approve all project activities in the
particular country.
A Territorial Advisory and Review Committee functions at the country level and
involves the Ministry of Agriculture, CARDI, agribusiness groups, and farmers. This group
agrees on project areas and reviews ongoing project activities. The team did not review
CARDI charter nor by-laws. The team did see a young and growing organization determined
to improve the quality of research in the Caribbean.

CARDI Staff Resources and Financial Issu s
CARDI staff members are located in units in Trinidad, Guyana, Belize, Jamaica,
Barbados, and Antigua. Country teams (CTs) are located in Antigua, St. Kitts-Nevis,
Montserrat, Dominica, St. Lucia, St. Vincent, and Grenada. Currently, there are 65 CARDI
field staff, including the technical officers. CARDI's professional staff consists of 3 animal
scientists, 5 economists, 5 entomologists, I pesticide specialist, and 20 plant science spe-
cialists, including forage agronomy, plant breeding, plant virology, and nematology. Some of
these professionals are in central administrative positions and are not actively working in
their specialties (i.e., 2 economists, I entomologist, and I soil scientist). In addition, the
support services for CARDI include 3 biometricians and 2 analytical laboratory specialists.

CARDI needs to consider strengthening its core staff in plant protection, agricultural
engineering (hydrology and small implement development), post-harvest loss technology,
plant breeding, and weed control. As the animal production component becomes integrated
into the CARDI SFMCP project, additional animal scientists may be needed.
CARDI needs to reevaluate the mission of its agricultural economics unit. It appears
to need more analytical skills. At present, the agricultural economics staff members are
not involved on a regular basis with other resource staff or CTs in project planning,

execution, and evaluation. The team likewise recommends that CARDI consider allocating
two economists to regional positions--one to the Windward and one to the Leeward
Islands--and to include the economists into policy analysis and the setting of research
priorities. As CARDI accepts more outside funded projects, there will be growing pressure
on its scientific staff to administer these projects. This is a potential problem which could
result in reducing CARDI's technical capacity to support field activities.

CARDI Staff Performance and Evaluation
The evaluation team believes that there is a need for CARDI to search for ways to
improve communication and foster a more challenging attitude to excellence and diligence
in research.

One approach to improving communication at the local headquarters of CARDI is to
hold regularly scheduled monthly meetings for all staff members. All resource staff should
also have an understanding of their responsibilities and how they will be evaluated annually.
The team recommends that CARDI develop precise job descriptions, which include lines of
accountability and methods of evaluation. Furthermore, it is recommended that the annual
review include resource people and administrators in the management group who are
affected by the activities of the resource staff.

Linkages with the University of West Indies
There are a number of active linkages between UWI and CARDI. In most instances,
these linkages stem from the needs of a given CARDI project to secure the support and
counsel of the expertise available in UWI. These linkages often involve individual consulting
arrangement for UWI staff. This arrangement is preferred by UWI staff in lieu of a
collaborative inter-institutional agreement because UWI faculty members do not receive any
portion of the overhead on UWI sub-contracts with CARDI.
An important linkage exists between CARDI and UWI's graduate student program.
UWI graduate students involved in CARDI research can improve relationships between
CARDI and UWI. The university statutes require that graduate student research is under the
direct guidance of UWI major professors. Graduate students may work on CARDI projects if
the project meets the research standards of the University and if the research activity
furthers the goals of a CARDI project.

The team recommends that long-term funding be included in Phase II to provide
support for both graduate student research and UWI staff travel for research and extension
activities consistent with the objectives of the project. The PP team should examine how
more effective institutional arrangements should be developed between UWI and CAilDL.

Financial Issues
A difficult question involves the financial capability of the constituent members to
sustain the growth of CARDI. At the present time, CARDI member territories are severely
limited in their ability to increase their financial contributions. In fact, CARDI is already
suffering cash flow problems as a result of the failure of some member territories to
forward their cash commitments on a timely basis.

The team recommends that a system of overhead fees be developed and budgeted for
each externally funded project. Such funds would be earmarked for direct support of basic
research relevant to the new project and as a contingency fund to finance some staff
salaries between contracts. The PP team should examine the overall question of long-term
financial support for CARDI, including projections of feasible contributions from the
individual territories.

Project Organization
This section will examine the SFMCP project as it is presently organized.

Project Leader
The Project Leader is responsible for achieving the goals of the project; he is
accountable to the funding agency, USAID. The Executive Director and the Director of
Research and Development are informed on all reporting.
An Administrative Assistant is needed to help the Project Director achieve effective
budgetary control and report/data coordination. The Administrative Assistant would report
directly to the Project Leader and would be office with the Project Leader.

The Regional Technical Coordinator
The Coordinator has the general responsibility to:
1. Ensure technical and methodological soundness in the project.
2. Provide linkages and exchange ideas among country teams.
3. Facilitate regional communications with the Project Director.
The specific responsibilities are to:
1. Certify the research soundness and accuracy of field trials.
2 Plan the regional research station programs and certify their technical
3. Secure needed technical skills for various country projects.
4. Coordinate all reporting and data submission to the Project Leader.
The organizational structure of the Project was recently modified by adding two
technical coordinators, one for the Leeward and one for the Windward Islands. This change

was made in order to have technical support more accessible to the CTs. It is too soon to
assess the value of the change. It would appear, though, that if there was an established
methodology for the FSR program and if the CTs had acquired the skills to conduct FSR
using this methodology, there would be little need for this kind of general technical
assistance. What the CTs need is specialized technical assistance in response to particular
problems. The technical coordinators can serve as a link with such specialists at CARDI.
As the Project develops, the organization structure should change in order to be more
responsive to the specialized needs of the CTs. The evaluation team makes no recommenda-
tions with respect to the organizational structure of the Project. The organizational
structure should be kept flexible and changes be made only after consulting with the CTs.

The Country Team Leader
The CTL is responsible for:
1. Directing the country team in implementing all Farming Systems Research
(FSR) activities within the country.
2. Communicating with the Regional Technical Coordinator on the design of
on-farm experiments, FSR methodology, data collection and analysis and
3. Managing all FSR funds for in-country use.
4. Communicating with the Project Director on all budget items and authoriza-
tion for regional travel, etc.
5. Serving on the FSR Coordinating Group described below.
Most of the management decisions were made at the CT level. The CT leaders all
commanded the respect of their team members. All seemed to function well providing
direction, supervision, and supporting assistance as needed. The Project Director allowed
the CT leaders to manage their program and yet kept them accountable. Problems that had
arisen in the past because of too much interference or too little attention were not apparent
at the time of the evaluation.

In order to develop effective communications between country FSR projects and
extension workers, the team recommends that a FSR Coordinating Group be formed. At a
minimum, it should consist of the Country Team Leader, the two Regional Technical
Coordinators, an Extension Advisor from UWI, and the Project Leader. The group would be
advisory to the Project Director. Since the field teams are developing innovations for use
by extension, it is recommended that a member of the UWI extension staff be included in
the FSR Coordinating Group. We recommend that this Program Operations Group meet
once a month for two days to consider project programs, methodology, and progress. We
feel the group should meet in a different project territory each time, prioritizing the visits
on the basis of the needs of the territory programs.

The FSR Field/Territorial Network
An important characteristic of CARDI is its "field presence" through a network of
experienced field staff. The network for the country or territorial units consists of (1) a
local committee (the territorial Advisory Review Committee) composed of key people in the
host territory/country and (2) linkages between the territorial units and the Regional
Coordinator, the Project Director, and the Director of Research and Development. The
team recommends that discussions presently underway on radio tele-communications be

Overall Assessment
CARDI has taken a number of steps to improve its administration and management.
The above recommendations are suggested as ways to assist CARDI in promoting improved
communication and coordination. In the final analysis, excellence in research management
can only be achieved through a long process of gaining experience in the design,
implementation, and management of on-going research projects and programs.

Revision of FSR Methodology

A major issue confronting the project is the need to revise the FSR methodology in
order to deemphasize data collection and the preparation of farm profiles and to shift to
standard FSR methodology. This redirection of the FSR methodology is needed in order to
develop more efficient and systematic farm trials which will lead to the final goal of FSR:
research recommendations for small farmers. This section lays out a plan of work for
achieving this final goal.
The need to reorient the project FSR methodology can be shown by comparing the
methodology in use with what is commonly accepted as the standard FSR methodology.
Unless CARDI is willing to bring about a major reorientation of their FSR methodology,
there is little justification for USAID to consider funding additional activities.

Comparison of CARDI's SFMCP Farming Systems Research
With Standard FSR Methodology

SFMCP Farming Systems Research
Methodology: 1979-82

1. Baseline survey of 120 farmers
in seven territories

2. Diagnostic survey of sub-sample
(25 farmers) in seven territories
through weekly farm interviews
over 12-month period

3. On-farm intervention trials on
small sample of farms in each

4. Preparation of farm profiles for
a small number of farms in each

Standard FSR MethodologyI

1. Informal survey (Sondeo) to
understand problems of small farmers

2. Design and monitor on-farm trials
(experiments in farmers' fields
under farmers' conditions)

3. Analyze results of on-farm trials
and redesign on-farm trials for
the next season

4. Verification trials to test
improved technology under farmers'

5. Diffusion of improved practices
through extension agents

IFor standard references on farming systems research, see the state-of-the-art paper
by Gilbert, Norman, and Winch (1980).

2Implicit to this step is the review of background papers and reports on market
potential of agricultural commodities, including interregional and international trade.

Table 3.

This section discusses specific steps to revise the FSR methodology in order to carry
out the full sequence of FSR activities.

Identifying Target Farming Systems:
Import Substitution and Export Commodities
The first step is to identify major import and export crops and animal products by
territory (Mohammed, undated, pp. 4-5; Mohammed, 1981, p. 5, item 5; Mohammed, 1982;
Ali, 1982). CARDI has direct access to this information for years during the 1970s (Barker,
1981), but a 5-year trend should be employed for decision making. Phase II should support
research to identify commodities which can be successfully produced in the Eastern
Caribbean, but which are now in short supply. This will include comprehensive studies of
imports by commodity across the region. Attention should be directed to removing the
barriers for increased regional trade of unprocessed and processed fruit or vegetable
products. CARDI researchers should cooperate with scientists in other regional organiza-
tions like the CARICOM Secretariat and CATCO (Caribbean Agricultural Trading Company)
who are also carrying out market opportunity analyses for the Eastern Caribbean. Project
economists should examine the question of how to develop an agro-industry which would
generate more employment through- intensive farming systems and better utilization of
fruits and vegetables. CARDI should use this information to identify the major crops which
can be grown locally with a fairly high probability through successful on-farm research. In
addition, current export crops with an assured market overseas (or inter-island) could be
added to this list of research priority crops.
Once a more systematic approach to identify priority crops is in place, the CARDI
multidisciplinary staff can assist each CT in identifying the farmers in the sub-sample of
each territory who produce these crops or animal products. Finally, groupings of
homogeneous farmers can be formed around these priority crops for the implementation of
farm trials. Whenever necessary, the sub-sample of farmers can be augmented from the
general farm universe to form a reasonable approximation of a statistically representative
sample (Carew, 1982, pp. 2, 6).

Carrying Out Surveys of Homogeneous Farm Groups
When the crops or animals have been identified for research priorities, the CTs can
select homogeneous groups of farmers. The homogeneous groups can be selected from the
sub-sample of approximately 25 farmers and augmented by selecting additional farmers
from either the baseline survey group, or the territory universe of farmers, or both
(Figure 2). The number of farms to include in each homogeneous group varies between 8-25
according to (1) the underlying homogeneity of the new sub-sample, (2) the manpower of a
pa, ticular CT, and (3) the expected magnitude of the improvement being introduced into the
system. From a statistical point of view, items (1) and (3) should weigh most heavily in

CARDI Methodology

Universe of Territory Farmers

Farmer Sub-
Sample, n=+25',

Intervention Group 1-3, na"I-12, may be
heteraeniipus or homogeneous


- Farm Trial Group 1-3, n-6-25, Is
homogeneous by selection


Baseline Survey Farmer
Group, n=120, hetero-


determining the number of farm trials per improvement. From a practical point of view,
most CTs should begin by assessing their own internal capabilities, determine the total
number of farm trials they believe they can manage, and then decide how many homogene-
ous groups of farmers they can work with initially.1
CARDI personnel need to address the issue of statistical representability: are six
farms a large enough sample to allow the necessary statistical inferences to be drawn from
on-farm trial results? While CARDI administration may believe that the proposed
improvements will increase yield (or decrease labor inputs) by a factor of two or three
times, experience in other FSR programs indicates that such magnitudes of benefits are
rarely, if ever, attained at the farm level. A more feasible increase for an improvement
would be 10-30 percent. In this situation, in contrast to the former, many more farm sites
are required for the statistical advantage of the improvement to be affirmed than in the

On-Farm Trial Design and Implementing Farm Trials
CARDI staff must also consider the issue of trial replicates per farm site. Two or
three replicates of each trial add to the time required for design, layout, planting,
observing, harvesting, and analyzing each trial, but no other mechanism will enable the
research team to evaluate the importance of intra-farm variability, or to provide a
compensating mechanism if a section of the trial is inadvertently destroyed. Some good
general outlines for addressing particular trial design and replication are contained in the
CARDI publication, "Design and Layout of Field Experiments in Conditions Experienced in
the CARICOM Area" (Lauckner, 1980).2
Figure 3 provides a simplified summary of the CARDI methodology flow from project
inception through March 1982. Given the fact that the farm characterization stage should

'For example, if a CT consists of four members, it can probably effectively manage a
maximum of 20 farms with replicated trials. Thus, the choice would be between working
with one homogeneous group of farmers (n = 20) on a single farm trial type with a very high
chance of statistical confirmation of results, or of working in two teams of two CT
members, each with two homogeneous groups of farmers (n = 10), still with a fairly good
chance of observing statistically different results across farms. However, working in three
homogeneous groups with three teams on 6-7 farms, each will (1) dilute the effort of the CT
too much by increasing travel, and (2) make statistical conclusions from any of these sets of
farm trials less likely.

The general methodology for on-farm testing was also outlined by Dr. K. A. Gomez in
1977 ("On-Farm Testing of Cropping Systems"). Dr. Gomez discusses the technology-
development research as well as technology-adoption research. The first seeks to develop
technology while the second tests the acceptability of the technology to the farmners.
Among the major testing issues addressed by this paper are: (1) the need to test on several
farms, (2) selection of test sites, (3) precision of on-farm trials, (4) measurements of
environmental factors, (5) technology to be tested, (6) choice of factors tv be tested, (7) test
criteria, (8) farmer participation, and (9) data to be collected.

Figure 3

CARDI SFMCP Methodology Flow,
1978 March 1982

Step In Methodology Implementation of SFMCP

Year (& Month)

Col lcct ion

(Or Analysis)

(Farm Trials)


1979 (Aug.)

1980 (Mar.)

1981 (Mar.)

1981 (May)

1981 (Aug.)

1981 (Sept.)

1982 (Mar.)


Forma 1
(Weekly or Fort-

One Year Of Data
Collected In St.
Lucia & St. Vin-

Farm Character-
ization in --
-Farm Profiles
-Computer Print-
out Summaries



I initiate
Formal Data
s \(n-+25)



Complete Farm
_ {Profiles; Initi-
ate Exploratory

/ Analyze

,Survey Data

Ini tiate
Intervent ions

Second Season
! Phase Of
L Jne.cyeptjons -

be less a formal exercise (via farm profile development) and more of an interdisciplinary
workshop exercise, the evaluation team recommends that this step be completed soon. The
evaluation team also recommends that each CT summarize their observations about their
exploratory intervention and intervention trials as soon as possible (Figure 3).
Table 4 demonstrates how CARDI might begin Phase II, given where the project
currently stands. This table lists each step, the type of decision format (workshop, etc.), the
time required to complete the step, and the CARDI personnel to be involved in the step.
Table 4 assumes a much heavier emphasis on field actions--farm trials and interaction with
collaborating farmers about such trials--in the future. The project cannot function
efficiently unless greater emphasis is placed on the efficient use of manpower and time and
greater reliance is placed on multidisciplinary team consensus.
Following identification of research priority crop and livestock systems and the
definition of constraints by territory, each CT will be able to accomplish the next three
steps: (1) selection of homogeneous farming groups, (2) definition of major constraints in
these groups, and (3) design of farm trials (Table 4). Following this timeframe, each CT will
be ready to implement and monitor replicated farm trials which are statistically representa-
tive of the underlying homogeneous sub-sample of farmers.
Each CT needs to carry out the following activities while farm trials are in the fields:
(1) Monitoring trials to observe and measure predetermined crop parameters (inci-
dence and severity of diseases; infestation and damage by insects; weed types
and densities; rainfall; animal damage; sales or consumption of thinned plants,
(2) Collecting necessary socioeconomic data (date of sale, quantity sold, and price
received for each crop sold, for example); and
(3) Deciding whether any special surveys (on land preparation, tenancy history,
cropping history, marketing constraints, etc.) are needed to assist a CT to
understand the farmer group better, or to assist in the redesign of the FTs for
the following season.
The next three steps (two analyses and one planning session) are crucial to pass from
the testing stage to the verification stage. A quick, well-executed analysis of the
agronomic and economic aspects of each trial has to be accompanied by a rapid analysis of
the crop or livestock sales data collected during the year. These two types of information
must then be brought together by the CT to allow design of verification trials. A synthesis
of agronomic and economic data must be made at this time because an improvement causing
a doubling of yield is only beneficial if the farmer can sell his product at a price which is
sufficiently higher than his increased costs. If not, the agronomic improvement may cause


Table 4. Proposed Implc:ncntation, of On-Farm Trials. Phase il

Step Decision Formiat Time Frame Personnel Involved

1. Identify priority Workshop (2 wks) Country Team (CT)
crops/livestock systems

2. Discuss major farmer Workshop (1 wk) Country Team (CT)

3. Select homogeneous Reconnaissance
group(s) of farmers Survey (1-2 wks) Country Team (CT)

4. Identify specific
constraints of Sondeo followed by (1-2 wks) Country Team (CT)
homogeneous farmer Workshop

5. Develop list of Workshop (1 wk) Technical Coordinators,
parameters to Project Coordinator,
monitor in trials CTLs

6. Design Farm Trials Workshop (1 wk) Country Team (CT)

7. Review Farm Trial Workshop (2 days) CT with Biometrician
designs and oojectives and Technical Coordinator

S. Finalize Farm Trials Workshop (2 days) Country Team (CT)

9. Acquire needed inputs,
supplies, and equipment Purchasing (1-2 wks) CT, Administrative Assistant
for field teams

10. Set detailed planting
schedule and reconfirm On-farm (1 wk) Country Team (CT)
dates and times

11. Plant (and replant) On-farm (2-4 wks) CT, Technical Coordinator,
Farm Trials Agricultural Economist

12. Monitor Farm Trials On-Farm (weekly until CT and disciplinary
using agreed-upon harvest) experts as the neeo arises
parameters, and observ- in the field
ing any unanticipated

13. Collect necessary socio- On-farm (weekly until CT and Agricultural
economic data (date of harvest) Economist
sale, amount sold, and
price received)

14. Schedule harvests precisely On-farm (1 wk) Country Team (CT)
with farmers and reconfirm

15. Perform pre-harvest On-farm (I wk) Country Team (CT)

16. Harvest Farm Trials On-farm (3-6 wks) Country Team (CT)

17. Analyze Farm Trial results Office (2-4 wks) CT plus Biometrician
statistically: ANOVA if needed

IS. Analyze Farm Trial results Office (1-3 wks) CT plus Agricultural
economically (CIMMYT method) Economist if needed

19. Draw conclusions from tne Workshop (I wk) Country Team (CT)
Farm Trials

20. Write up results of Farm Office (3 wks) Country Team (CT)
Trials and send to
Technical Coordinator

21. Present results of On-farm (2 wks) Country Team (CT)
(1) his particular farm
and (2) the homogeneous
area to each farmer

22. Determine which treatments Workshop (1 wk) CT with Technical
to advance to verification Coordinator,
phase, which to drop, and Bioinetrlcian,
which to modify (and how) Agricultural Economist

23. Redesign Farm Trials to Workshop (1 wk) Country Team (CT)
become verification trials

(From here on, implementation details fNllow above, .gimngin with Step 7,
except levied trials kbcore uemo,.:,;ration trials)

the farmer to be economically worse off than before, especially if his input, harvest labor,
and transport costs are higher.
The rest of Table 4 shows the progression of on-farm research to the point where
tested research recommendations are available for extension agents to diffuse to farmers.
In conclusion, it is important to stress two major issues which have not been given
enough emphasis by project leadership during Phase I.
(1) Implementation of all stages of any FSR project should be flexible. This flexibility
includes the ability to implement farm trial research in line with the available field staff in
each territory. Such flexibility is also needed in the identification of specific farm
parameters as constraints in the different homogeneous zones or farm groups of each
territory. Evidence of flexibility in implementation policy should be the willingness of
project leadership to go ahead with the field work, even if the background "homework" may
be incomplete. The key to FSR is hands-on experience (Krantz, April 1981), which can only
come about if each CT receives explicit directions from central project leadership to
proceed with the farm trial stage.
(2) FSR is an iterative process allowing closer and closer approximations to the
farming reality across a predefined homogeneous group of farmers and the identified
constraints which are common to farming systems. The field team must first admit it does
not know what the farmers actually do, then devise methods (questionnaires, farm trials,
specific surveys, verification trials, demonstration trials, etc.) to collect information, step
by step, which will lead to a better scientific understanding of the practices and decision-
making criteria of farm households.
A farmer (and his/her spouse) does not know all there is to know about the total reality
of their farm operation because they have access to limited information about (a) technical
agronomic/livestock relationships and interactions, and (b) macroeconomic relations (e.g.,
territorial demand for each crop and demand for substitute commodities produced by other
farmers and imported from external sources). In addition, the amount of knowledge which is
actually transferable from the farm family to others (the CT) is less than the total family
farming knowledge because of time limits, forgetfulness, the personal nature of certain
types of information, etc.
Thus, regardless of the amount of time spent questioning farmers, just so much
knowledge representing the reality of farming can ever be extracted by CT teams. This is
true whether the CT uses an abbreviated sondeo instrument or a more formal and lengthy
questionnaire. Many FSR researchers believe that the shorter survey (sondeo) generally
provides nearly as much information as the questionnaire, while avoiding the potential
problem of collecting too much data for quick and proper analysis (see Collinson, 1982).

The next stage of FSR involves focusing on solutions to overcome the constraints from
the homogeneous farm group. This stage is called the farm trial (FT) and subsequent
monitoring. Again, being based on either a sondeo or questionnaire with no prior experience
in farming in the area, the FT in the first year addresses some of the constraints which may
affect farm productivity, but must inevitably contain certain misinformation (e.g., exact
depth of seeding, depth of seed coverage, exact amount of pressure used to cover seeds,
etc.). The CT should view this step as a critical learning experience to improve farm trials
during year two.
The results of the FT (and the optional special survey) are used by the CT, along with
their informal observations and farmer suggestions, to design the verification trial (VT).
The VT contains fewer experimental treatments and larger treatment plots than the FT, and
should be an even closer approximation to the solution of a particular constraint to farm
productivity. This is demonstrated by showing that the VT covers a greater amount of the
farm operation reality than does the FT, thus containing slightly less misinformation about
the farm. Therefore, it is absolutely crucial that the CARDI CTs be urged, and assisted as
needed, to begin replicated, systematic FTs as soon as possible.

(1) All host territories, CARDI CTs, and collaborating farmers are anxious to get on
with farm trials but the trials should be laid out within a system of clearly
defined crop and livestock research priorities.
(2) Adequate data already exist for each CT to select homogeneous groups of
farmers and identify perceived major production constraints.
(3) Each CT should decide, in consultation with the respective technical coordinator,
how many farms will be needed to represent the homogeneous area so that
meaningful statistical differences may be expected.
(4) CARDI core staff must resolve the statistical issue of whether the faran trials
are viewed only as replicates across a homogeneous area, or are also to be
designed to gather information on intra-farm variability (replicates within
(5) Since less emphasis is to be placed on immediate analysis of the sociseconomic
data in Trinidad, the evaluation team recommends that members of the
economic resource unit join either the Leeward or Windward research groups.
Such a step will allow more interaction between economists and the CIIs (and
farmers), as well as providing CTs, when needed, additional manpower inr field
trial implementation, monitoring, and analysis.

(6) Little consideration has yet been given to the key issue of the farm logistics of
implementing full-scale farm trials. This is a vital, complicated step which
requires a good deal of coordination and advanced planning. To address this
potential bottleneck, the evaluation team recommends providing each CT with
the following: (a) guaranteed transportation to each homogeneous group of
collaborating farmers; (b) a set of basic equipment and supplies (twine, cutlasses,
tapes, seed bags, plastic bags, tags, markers, field scale, etc.) for each sub-group
handling a set of homogeneous farm trials; and (c) a programmable, hand-held
calculator capable of performing at least these functions: (a) traditional ANOVA
analyses of replicated agronomic trials, and (b) CIMMYT-type economic analyses
of the net benefits of each trial treatment. To introduce this step, a workshop
should be given by the CARDI biometrician and resource economists to familiar-
ize each CT member with the use of these calculators.

Strengthening Technical Research:
Crop Production

A vigorous and productive FSR .program requires strong supporting research programs
in. crop and animal production. The evaluation team feels that research is the most
important component of the project. The generation, adaptation, verification, and ultimate
dissemination of the improved technology which will benefit the small farmer is the general
purpose of the project. If appropriate technology is not generated, the project will fail to
achieve its goal. This section outlines the present situation of crop research and identifies
major issues which need to be addressed.

Crop Production in the Eastern Caribbean
Agricultural production in the Caribbean region varies from the sugar estates of St.
Kitts to the complex mixed-cropping patterns of the small Montserratan farmer. The crops
grown are mostly tropical and include: Banana, Papaya (Pawpaw), Vegetables, Coconut,
Breadfruit, Avocado, Citrus, Macambou, Tannia, Sweet Potato, Mango, Dasheen, Cassava,
Lettuce, Yam, Plantain, Cucumber, Guava, Tomato, Golden Apple, Spice, Cocoa, Cabbage,
Carrot, Sugar Cane, Cashew, Pineapple, Corn, Pumpkin, Beans, Sweet Pepper, Pigeon Peas,
Blackeye Peas, Celery, Eggplant, Onion, Peanut, Christophene, Okra, Coffee, Ginger,
Elephant Grass, Turnips, and Chive. The farmer is usually a part-time individual who is
working with one to several small plots of land. The agro-ecological zones also vary
considerably within the region. This mix of elements makes agriculture a very complex
phenomenon in the Eastern Caribbean.

Surveys have revealed that most farmers on the islands are over 40 years old. This
will have a serious impact on the region if something is not done to bring young people into
farming. The two problems which need to be addressed are (1) the drudgery of farming and
(2) the economic viability of farm households. If these two problems are not addressed
within 20 years, most of the region's food will need to be imported because of the lack of
farmers. The complexity of the farmers' needs demands a holistic and realistic problem-
solving approach.
The constraints to crop production are as follows (CARDI, Undated, "Major Con-
straints Identified from Farm Characterization Phase").

Production Constraints

Agronomic Environmental
(1) Total dependence on hand labor for all farm operations, especially very hard
time-consuming activities such as land preparation and weed control. Very often
land preparation is inadequate for the crop to be grown. There is a complete
lack of labor-saving devices for crop harvesting, handling, and post-harvest
(2) A wide range of crop protection problems including:
(a) The abundance of a wide diversity of weed species.
(b) Rapid regrowth and survival of many weed species, especially during the
rainy season.
(c) High incidence of pests and disease, especially during the rainy season.
(3) Inefficient utilization of crop protection measures such as:
(a) Variable rates and frequencies of application of chemicals.
(b) Improper timing of application in relation to incidence and/or level of
infection or infestation.
(c) Use of wrong chemicals.
(d) Lack of reliable information on the choice and use of chemicals.
(4) The frequent unavailability of input supplies for efficient crop production and
management. This is particularly the case with fertilizers, vegetable seed, and
(5) The very variable unselected and poor quality of planting material used in cereal,
legume, root crop, and fruit crop production.
(6) Poor sowing, planting, and nursery techniques resulting in the:
(a) Waste of funds for purchase of larger quantities of seed than actually

(b) Use of higher or lower plant populations than the optimum for a particular
(7) The use of marginal land for agricultural activities leading to:
(a) Drainage problems in low-lying areas of heavy soils.
(b) Poor soil fertility in certain areas.
(c) Severe erosion in steep areas under high rainfall conditions.
(8) The predominance of rainfed agriculture with little control of water resources.
There can be an oversupply of water during the rainy season, leading to drainage
and disease problems, and an undersupply during the dry season, leading to
drought conditions.
(9) Unplanned and/or unorganized systems of production often result in a mosaic or
amalgam of crops which makes proper management difficult.
(10) Poor handling at harvest, poor post-harvest handling, packaging, and transport-
ing, leading to damage and loss of marketable produce.
(11) The complete lack of proper on-farm storage facilities resulting in high storage

(1) The dominance of "family land" as the major tenurial system has led to severe
fragmentation of holdings often in areas too small for economic utilization.
(2) The scattered nature of land holdings or parcels belonging to a single farmer
results in a substantial amount of time traveling from parcel to parcel.
(3) The relatively high average age of farmers reduces labor productivity for on-
farm activities usually arduous.
(4) Dependence on off-farm employment for supplementing income reduces time
devoted to on-farm activities.
(5) Overall unfavorable attitude toward farming and provision of incentives/tourism,
industrial and manufacturing subsectors leads to the young generation moving off
the farms and thus creating a shortage of labor in some rural areas.
(6) Low educational level of farmers making it difficult for farmers to understand
the scientific principles involved in new, improved practices. This leads to low
adoption rates of improved technology.
(7) Poorly organized marketing systems to meet the demands of domestic agricul-
tural operations--both food crops and livestock.
(8) Lack of agricultural information channels for effective communication of input
and output prices resulting in inefficiencies of operations at the farm level.

(9) Absence of on-farm roads and feeder roads linking parcels to each other or to
main roads.
(10) Low utilization of available credit facilities due to strict collateral conditions
required before disbursement of funds, as well as trouble and time involved in
obtaining loans from agricultural banking institutions.
(11) Insufficient numbers of trained agricultural extension personnel and inadequate
back-up facilities required to have an effective delivery system to meet the
varied and multiple needs of widely dispersed farms located in difficult terrain.

Areas of Research

The evaluation team observed slides of an on-farm experiment on intercropping in
cotton in Nevis and unpublished yield data from that experiment for the major cotton crop
and the short-term crops--peanuts, corn, and beans. The cotton yields were not reduced,
whereas considerable income was generated by the interplanted crop. In the case of cotton,
the intercropping is initiated when the cotton is planted. In the case of sugarcane, the
intercropping is initiated following cane planting or follo-'ing cutting when the crop is being
ratooned. Further research needs to be done, however, to test intercropping under farmer
Considerable effort should be exerted on on-farm research in St. Kitts-Nevis to extend
intercropping. This could provide a major breakthrough in crop production in a territory
with a considerable need for food legumes and corn. Possible land rental arrangements for
intercropping should be explored for laborers on the NACO controlled land. This example is
cited as a Farming Systems Research output which is ready for verification on a larger
group of farms.

Agronomy Nutrition
The area of production (including multiple and intercropping of food and fodder crops)
should be extended wherever possible. Great care should be taken to examine the
nutritional balance (or imbalance) on each island so that crops which will enhance the diet of
the population (particularly those crops high in protein) can receive higher research
priorities. The FSR project should play a major role in this area of nutritional enhancement.

Water Management -
The reports of Bert Krantz (Krantz, B.A., 1981, Consultant Report No. 6) consistently
stress the need for identifying areas where small, low-cost catchment reservoirs can be
constructed. The general problem in the region is not lack of sufficient rainfall, but poor
distribution. Thus, the major thrust of a water management project would be to even out
water availability during the crop year by making use of supplemental irrigation.

Regional Research Stations
The project evaluation team recommends the further development of the CARDI
Regional Research Stations. The evaluation team visited the Small Farm Research Center
near Betty's Hope, Antigua. Dr. B.A. Krantz (Report of Advisor to CARDI-USAID Project
#538-0015, April 1981) has given a complete description of this site, the objectives of the
research to be conducted on the station, and the sub-programs on farming systems research
which could be conducted on that station. Reinforcing the conclusions of Dr. Krantz, the
evaluation team encourages USAID to assist in the funding of these developmental research
programs at the Antigua Research Station: (1) Soil and Water Management Systems,
(2) Implements and Power Source Systems, (3) Cropping Systems and Crop Management,
(4) Forage Crops and Livestock Systems, and (5) Socio-Economic Systems Research. The
evaluation team recommends funding of crop research at St. Kitts-Nevis and Antigua on
drought resistant legume species, such as Dolichos lablab and the improved legume species
identified by Dr. John Keoghan for leaf meal production. Interdisciplinary studies on solar
drying and the production of high protein leaf meal for livestock feed should be conducted
with the objective of reducing the importation of feed supplements. Low cost methods of
drying legumes, reduction of legumes to leaf meal, and storage should be developed to
permit the construction of local farmer-owned facilities financed by low level investments.

Professional research staff must be added in the areas of plant protection, agricultural
engineering (hydrology and small implement development), postharvest physiology and crop
In the area of plant protection there should be an individual trained in crop protection
for each of the islands, or at least one in the Windward and one in the Leeward Islands. This
specialist would monitor pest populations and be able to diagnose insect, disease and weed
problems and suggest control measures. Agricultural engineering research is needed in the
layout of watersheds and the design and construction of check dams and water impounding
structures. Research on small implement development is also urgently needed. Postharvest
losses are a serious farm-to-market problem requiring research on time of harvest, produce
handling and preservation. Disease and insect resistance, improved plant type and greater
adaptation are characteristics which can be readily added to most crops through plant
breeding. Initially, special emphasis should be spent on the improvement of food legumes
and vegetable crops.
To enhance the potential of trained and experienced research scientists and thus their
research productivity and value to the project, CARDI should have a research technician
(B.S. degree) working with each senior scientist. Several scientists currently in

administrative roles or directing projects could continue productive research if provided
with a research technician.

Strengthening Technical Research:
Animal Production

The inclusion of livestock in the farming systems research project is important
because more than one-half of livestock products consumed in the region are imported. Also
livestock production provides local employment and makes an important contribution to soil
There are marked territorial differences in animal production in the region. Antigua is
markedly higher in milk production, Antigua and St. Lucia excel in beef production; pork and
poultry production are important in the Windward Islands. Pork accounts for one-half of all
the red meat production in the Eastern Caribbean. The potential for growth into these
markets is obviously substantial given the fact that less than 50 percent of the market is
being supplied from local production. However, not all of this market can be penetrated by
local animal products considering the specialty demands of hotels/restaurants and the
customary local systems of animal production which yield a different type of product.

Types of Animal Production Systems1
There are a number of different systems in which the animal is used in the Eastern
Caribbean. The most predominant is that in which the animal is a simple complement to the
cropping activity and converts wastes and residues into a high value animal product. The
animal has a very important function in these systems, being a savings or cash building
mechanism readily converted into cash or barter for other products. Most of the baseline
studies give data on animals in this type of system. Farmers who produce animals in this
type of system are more likely to commit inputs to the cropping system and allow the
animal component of the farm to seek a balance with.this. They are not looking for
substantial increases in animal numbers but are interested in increased output per animal if
the requirements to do so are not high. Increases in animal populations will be considered if
farmers can maintain them on their cropping output. The manure produced by animals is
very important to these farmers. It is obvious that interventions with these types of
farming systems must involve simple low capital requiring concepts.
A second type of animal production system is the herd or group type in which capital
and labor investments are more intensive and from which output levels are higher.
Examples of these are the small dairy schemes in the Leeward Islands and the swine and
poultry production units in the Windward Islands of St. Lucia, Dominica, and Grenada.

IFor a standard reference on livestock in the region, see Archibald.

Archibald, et al. (1981) have suggested the following animal use pattern for the
Eastern Caribbean:
Cattle are used primarily for generating income. The general absence of
refrigeration and meat preservation techniques precludes slaughtering of cattle
entirely for home use.
Sheep and goats, because of their smaller size, are used to a greater extent than
cattle for home consumption.
-- Pigs raised in semi-confinement or confinement by small farmers are used
primarily for income generation purposes.
Poultry kept in the loose, backyard system are used primarily for home consump-
tion with occasional sales of birds or eggs. Small farmers utilizing more intensive
poultry production systems, particularly if imported feed is used, do so for revenue
purposes rather than home use.
Rabbits are kept mainly as pets. End-use is principally for home consumption.
Milk production at the small farmer level is geared for home consumption. Excess
production is either left for the calf or sold to immediate neighbors.
Hides, manure and draft power are of little or no commercial importance to the
small farmer that keeps livestock.
When considering the overall goal of an FSR program to increase food production, the
development of animal production schemes should be an integral part.

Constraints on Animal Production
Osuji (1982) has outlined the following constraints on animal production:
1. Poor nutrition due to the seasonality and unavailability of local feed, poor quality
pastures, and high feed costs.
2. Parasitism in animals. Both ecto and endo parasites are a problem. Lack of
information among farmers, lack of an adequate supply, and high cost of
antiparasitic drugs have limited the productivity of most animal species.
3. Marketing. The market structure, lack of adequate transportation, big differ-
entials between farm gate prices and retail prices, and the absence of market
information to farmers are the areas needing attention. Animals are sold by sight
and pricing policy is consumer-oriented.
4. Distance to processing facilities. The availability of processing facilities may
increase the farmer's earnings through the value added to processed products.
5. Unavailability of suitable land for pasture expansion.
6. Unavailability of suitable breeding stock, especially in ruminants.
7. Predial larceny and predation by dogs, especially with the small ruminant
8. Poor management expertise' on the part of the farmers. Housing, nutrition, and
health seem to be the main areas of concern.

Research at the various stations and projects in the region indicate that a very marked
response in productivity can be obtained with the indigenous animal by inputs in feeding
levels both in quantity and quality. Seasonality of grazing and the resultant variation in
grass nutritive content seriously limit the output levels of growth and milk production of
Monogastrics (swine and poultry) require more nutritionally complete and complex
diets in order to function efficiently. The feed supply sector in the Caribbean poorly
services this need for monogastric diets in its failure to utilize local feeding materials and
in not supplying supplemental critical nutrients which allow the farmer to effectively use his
basic farm-produced feeds.
A number of private sector feed mills import food grains and protein concentrates to
prepare their complete mixed feeds and thereby become subject to world market prices,
high transportation costs, foreign exchange availability, and product quality fatigue factors
in attempting to supply feedstuffs to local farmers in a consistent manner. Such a system is
subject to external manipulation. The result often translates into higher costs of blended
livestock feed (imported feed costs are two and one-half times the U.S.-based cost) and
inconsistent animal performance, thereby cutting into the profitability and interest on the
farmer's part in continuing his/her production program.
In most instances, the present commercial feed supply system requires the farmer to
use substantial levels of capital for a livestock production system because it requires the
farmer to purchase both bulk energy and concentrate requirements.

Integration of Livestock into the FSR Program: Research Agenda
Baseline data and profile information on livestock have been included in the surveys
conducted in some of the territories. While an organized approach to this was not included
in the Project's FSR plan of work, there seems little justification at this point in embarking
upon such an extensive exercise for livestock as there have been a number of livestock
surveys conducted (Archibald, Singh, and Osuji, 1981; Ahmed, 1980; Osuji, 1980). Further-
more, the country teams and the livestock coordinator for CARDI have an excellent
perception of the livestock industry in the region.
Livestock production systems in the Eastern Caribbean are two major types and each
requires very different strategies for farming systems research. The following provides the
specific research requirements for each type of system.

A. Analysis of Farming Systems Where Animals
Subsist on Cropping Residues
1. Develop several simple dynamic models of livestock cropping systems which show
how the system functions over time and describe the interactive linkages. The
models should be selected so as to provide representative farm types and
especially multi-faceted activities in cropping/animal production (multiple crops,
several species of animals vs. mono-crop single species). This should include
seasonality effects. These would be included in the overall sampling picture of
the project.
2. Estimate the present constraints in terms of:
a. the performance of individual animals.
b. factors affecting species, or herd or group type animal systems as if they
were presumed to exist.
3. Determine at the station level, the growth and feed utilization response patterns
of animals to interventions which are developed from the study of constraints.
This would involve measuring the actual productivity of indigenous type animals fed the various cropping residues or mixtures of these produced on the small
model farms.
It would also include varying nutritional levels and measuring the responses
obtained from appropriate supplemental inputs. New interventions which have a
mandatory animal component such as forage legumes for land conservation should
be involved in getting measurable responses at the station level.
4. If the intervention developed in "3" indicates the need for a supply requirement
(minerals) or a simple processing input, then these must first be developed and be
accessible for the introduction of these innovations to the farm system.
5. The "tech-pack" and intervention can be diffused into the sector through "model"
scheme or by simply letting the farmer introduce them to his animals. It will be
difficult to obtain comparative measures of intervention responses in such single
animal systems because we cannot set replicates.
B. Analysis of Farming Systems Where
Animals Are Raised in Groups
1. The station testing could be done with groups similar in size, sex, age, etc. and
environment (housing). These interventions can be quantitatively identified
because of the availability of large enough numbers for more reliable estimation.
A number of different treatments or variables can be measured in this research.
2. Field application can involve dividing animal groups on the farm for comparative
testing. "Parallel" trials in which similar animals and other inputs being used by

the farmer are actually sampled and monitored on the station level serve to
provide excellent guides for measuring farmer management and to provide a
measure of defending the validity of a new intervention. Farmers can observe at
the station what levels of performance occurred with the same source of animals
and feeds that they were using.
C. Special Studies
There are other various experiments that could and need to be done. Again,
information from the FSR program should continually help to set priorities. The lists that
follow are meant to be illustrative.
1. Animal utilization of cropping residues and marketing surpluses.
2. Production of lower cost animal feeds through utilization of locally available
feedstuffs and industrial by-products.
3. Developing a supplemental-type feed supply system for small farms.
4. The utilization of soil conservation-type plants by animals.
5. Emphasis on increasing the reproductive animal function in all species through
nutrition and management inputs.
6. Developing mini-management "tech-packs."
7. Developing solutions to the problems of marketing and credit.
Windward Research Station
Station Trials
1. To develop a practical feeding program emphasizing the use of copra meal,
bananas, fish meal, legumes. Note: This research will identify supplemental-
type interventions which can be provided in small packet form for farmer
supplementation of "succulents" and other farm feeds.
2. To determine the feeding values of crop residues from trials.
3. Develop feeding programs to utilize larger quantities of surplus crops.
4. Evaluate forage legume feeding values and develop feeding systems.
Note: Leaf meal production should be included.
5. Measure response in the reproductive function of both ruminants and mono-
gastries with varying feeding levels of crop residues and supplements.
On-Farm Trials
1. Develop legume planting and other erosion control plant systems and use
ruminants to monitor yield, seasonality of supply, mineral supplementation,
and parasite control response.
2. Test feeding of crop residues to reproducing animals and monitoring of
reproductive functions.

3. Initiate a mini-herd production scheme with farmers in poultry and swine.
This would involve feed supply, tech-pack, and marketing support.
4. Develop mini animal-processing units.
Leeward Research Station
Station Trials
1. Determining animal feeding values of crop residues.
2. Develop supplemental packs for animals utilizing these residues.
3. Measure reproductive animal response to feeding/management inputs based
upon cropping residue and industrial by-product feeding systems.
4. Develop mini-group systems from poultry based on by-product policy system.
1. Develop mini equipment for watering grazing animals.
2. Introduce schemes to utilize cane residues with molasses urea supplementa-
3. Distribute daily supplemental feed packs for animals.
4. Develop mini animal-processing units.


Training issues that need to be addressed are: long-term degree training; FSR
training; and one-year certificate training. These decisions should be made only after policy
and strategy determination has been made.
The survey work has developed certain skills for collecting data which now needs to be
complemented with skills in analyzing the data collected. The same can be said for the on-
farm testing. The members of the CTs need to be trained in the design, layout, and
monitoring of field experiments. Certain decisions, however, have to be made with respect
to methodology before such a skill training program can be prepared.
A number of farm systems research workshops, seminars or conferences will be held
throughout the world during the next several years. CARDI needs funding for its scientists
to participate in international travel to these conferences and visits to FSR projects in other
countries. FSR scientists need this exposure.
The project evaluation team also recommends that funds be allocated for study at one
of the regional agriculture diploma schools by secondary graduates of the LDCs of the
eastern Caribbean. Candidates for this training would be required to spend at least six
months as a field assistant in one of the national FSR programs to be eligible for this
training. The returning trainees would be required to work as a field assistant for at least
one year- before being considered for further training. Funds should be provided for these
field assistants ns this is considered part cf his/her training.

Technical Assistance to CARDI

The team believes many of the problems of the project to date could have been
avoided if the project had received some technical assistance from a Title XII group from an
American university or consortium of universities. Technical assistance from an Agricultur-
al Research Administration Specialist and a Farming Systems Economist or Agronomist with
hands on experience in conducting farm trials is needed. The Agricultural Experiment
Station Administrative Specialist should have broad experience in multidisciplinary research,
program and budget development. He should have experience in the management of multi-
location research operations with a wide array of activities, including extension. He should
have the ability to listen, work with people, and should have management and administrative
skills and experience in fiscal and personnel matters.
The Agricultural Experiment Station Administrative Specialist would assist CARDI
Administration in the establishment of a FSR Coordination Group to assist the Project
Leader, the establishment of a system of overhead fees on externally funded projects to
support core staff between projects, the completion of job descriptions and annual
evaluations of each CARDI staff member, the support of graduate student research and UWI
staff travel associated with the graduate's thesis work, and the development of more
effective inter-territorial communications.
The Farming Systems Economist or Agronomist would be a person with broad
experience in applied or adaptive, hands-on field research who would assist the CARDI staff
(a) determining priority constraints for homogeneous groups of farmers;
(b) the identification of possible adaptive research needs to overcome production
and marketing constraints;
(c) the development of the methodology of coordinating a multidisciplinary approach
to solving the above constraints.
(d) designing on-farm research;
(e) training field support assistance in statistical methods;
(f) interpreting data, validation, and demonstration of recommended practices; and
(g) developing research-extension linkages to extend cropping recommendations.


The purpose of the Project as stated in the Project Paper was "to improve
smallholders' farming systems" by developing improved management and production recom-
mendations through adaptive farm-based research. Specifically, "at least 12 significantly
improved smallholder farming systems were to be designed based on the integration of crop
and livestock-specific proven technology." The approach used in developing the improved
recommendations was to deploy a three- to four-person CT headed by an agronomist in six
(later seven) of the territories. The methodology was partially derived from the experience
of the Center for Tropical Agricultural Research and Training (CATIE) in Central America.
The CT teams were supported by CARDI's core staff, both administratively and technically.
This evaluation covers the first three and one-half years--January 1979 to mid-
1982--of the implementation of the Project. The implementation of the Project consisted
of four major activities: baseline surveys, year-long diagnostic surveys of a sub-sample of
farmers, a series of on-farm trials of promising technology, and the preparation of a set of
case studies or profiles of individual farmers. The four steps completed to date represent an
incomplete approach to farming systems research.
The baseline surveys were carried out on approximately 120 farms on each territory by
the Department of Extension at UWI under contract with CARDI. The baseline survey work
began in January of 1979 on St. Vincent, Dominica, and St. Lucia and was completed in
December while the surveys for Antigua, Montserrat, and Grenada were completed in
October of 1980. As soon as a CT was in place on an island, diagnostic surveys were carried
out on a sub-sample of 25 to 30 farmers over a period of one year. The diagnostic surveys
consisted of weekly interviews which helped to identify some of the production constraints
faced by small farmers and helped the team gain the confidence of farmers. The third stage
was the establishment of on-farm trials on 5 to 10 farms in each territory. The final phase,
beginning in September 1981, was the preparation of case studies/profiles of some of the
farmers who were interviewed during the diagnostic surveys.
The major problem in implementing the Project to date has been the overemphasis on
farm surveys and data processing. The evaluation team observed that preoccupation with
processing the massive amount of longitudinal data from the diagnostic survey has delayed
the introduction of on-farm trials of promising technology, the analysis of results of on-farm
trials, the redesign of on-farm trials for the next season, the testing of improved technology
(verification phase) under farmers' conditions, and finally the diffusion of recommended
technology and changes in farming systems by extension agents. In summary, the Project
team became "swamped" in data and were unable to carry out the full sequence of farming
systems research which culminates in a set of research recommendations which can be

diffused to farmers by extension agents. The ad hoc on-farm trials that are now underway
need to be designed and conducted using standard FSR methodology. These on-farm trials
are not being carried out on farms which are representative of a homogeneous group of
farms. The evaluation team has provided guidance on how to select homogeneous groups of
farmers, criteria for selection of a sample of representative farmers for on-farm trials, and
guidance on how the CTs can move more quickly from the diagnostic phase to on-farm
A major output of the Project was to be the establishment of "small farm research
programs in six countries" by incorporating farming systems research into the national
research program in each territory (PP, p. 20). It was assumed that each host government
would take over the multiple cropping research program and make it part of its overall
agricultural research and extension development program. The Project has not yet
succeeded in accomplishing the first part of this sequence, i.e., in establishing small farmer
research programs in any of the countries. In fact, the field teams have functioned as
CARDI "outreach" teams. Nevertheless, some progress has been made. For example, CTs
have been deployed in seven territories, the CTs have gained experience with several phases
of farming systems research, and there is strong support among farmers and agricultural
personnel in the territories for on-farm research. In spite of this progress, however, if
Project funds are terminated in November of this year, the field teams will most likely
disappear. This will occur in spite of the general appreciation of the Project by the host
governments. The evaluation team is of the conviction that the team designing the original
Project was unrealistic in assuming that a functioning on-farm research program could be
established in the territories in a three- to four-year period. This process will require a 10-
to 15-year effort and this time frame should be taken into account in planning for the
Although there have been administrative and managerial problems in launching the
Project, it should be kept in mind that CARDI was only four years old when the small farmer
research project was initiated. Moreover, the Project required managing a number of
decentralized teams on eight islands and introducing an approach to agricultural research
that had not been tried in the area. CARDI has gradually resolved many of the problems
associated with implementing the Project and management has improved over the last three
years. The evaluation team, however, has identified a number of management issues that
need to be addressed. These issues include: (a) communication between the field teams and
headquarters, (b) administration and control of Project funds, (c) core staff technical
support of country teams, (d) training and supervision of the country teams, and (e) feedback
from the country teams to CARDI in developing its research agenda.

The evaluation team identified several major issues which need to be addressed if an
effective small farmer research program is to be established in the region. These issues
include the following: (a) agricultural research policy and the mission of CARDI in
furthering agricultural research in the region, (b) the division of labor between CARDI's
agricultural research programs and the research and extension programs in the territories
served by CARDI, (c) the level of research capacity that is to be developed by CARDI and
on each of the islands, (d) the revision of the FSR methodology in order to move more
quickly to research recommendations which can be diffused by extension agents, (e) training
of the CTs, (f) technical assistance, and (g) the question of recurrent costs after the Project
is completed. The Project requires assistance in research management and in refining the
FSR methodology used in the diagnostic, design, and on-farm testing phases and in applying
micro-computer technology for processing data at the field level. The introduction of
micro-computer technology, particularly for survey data and field trial results, would
greatly facilitate the preparation of timely results.




I. To evaluate the effectiveness of the Small Farm Multiple Cropping Systems
Research Project in improving the income and well-being of small farmers by
development of appropriate management and production technologies by

i. acceptability of proposed interventions by experimental groups and the
potential of these interventions for wider application;

ii. methodology and the results of small farm surveys and analyses; and

iii. net benefits to small farmers of project interventions.

II. To evaluate the appropriateness of the Project, as a basic model for applied
research in small farm agriculture in the Eastern Caribbean, including the
institutional framework at both the regional and national levels.

III. Provide specific recommendations concerning further assistance in the area of
applied agricultural research, particularly as it relates to improving the
income and livelihood of the small farmer in the Eastern Caribbean.


I. To achieve objective I, the evaluation team will:

i. assess the effectiveness of CARDI's efforts to date, to collect data,
interpret this data, and determine appropriate interventions for project
target groups;

ii. examine interventions underway and recommend improvements, if
needed, or changes in agronomic approach; and

iii. analyze the ability, to date, of CARDI to transmit information on
improved technologies to extension personnel, farm groups, and other

II. To achieve objective II, the evaluation team will:

i. examine the ability of CARDI to coordinate and adapt its institutional
structure to perform appropriate small, farm adaptive research, particu-
larly as it relates to the CARDI multi-disciplinary approach;

ii. examine the institutional and absorptive capability of public and private
agricultural organizations in the LDCs of the Eastern Caribbean to
utilize existing applied research;


iii. examine the priority needs of various islands relating to applied
agricultural research; and

iv. discuss the effectiveness of the project in addressing these needs.

III. To achieve objective III, the evaluation team will:

i. make recommendations for appropriate areas of applied research, both
regional and country-specific, for AID involvement in the future; and

ii. recommend appropriate institutional arrangements and procedures for
such applied agricultural research activities and programs.





3/16-17 Washington, DC

3/17-20 St. Lucia

3/21-25 St. Kitts

3/21-24 Antigua



3/25-28 Dominica

St. Vincent

3/29-4/6 St. Augustine


Beausoleil, Gait
Everson, Freed

Entire Team

Freed, Deans,

Beausoleil, Gait

Beausoleil, Gait

Beausoleil, Gait

Everson, Deans

Entire Team


Filipe Manteiga, Bill Baucom, Jeff Rosen,
Carol Weber, Miguel Cuevas, Mike Weber

Calixte George, John Hammerton,
Jim Hughes, Vasantha Narendran,
Julio Chang, Arthur James,
Ron Pilgrim, Burnette Sealy,
Greg Avril

Laxman Singh, Charles Williams,
Jennifer Lowery, Austin Farrier,
Howard Batson, Roger Francis,
Ken Martin, George Bradley

Lenny Daisley, Leo Nicholas, Vincent
Belle, Daryl Roberts, Francis Henry,
John Keoghan, Roberta Anthony,
R. Edwards, Robin Yearwood, Ted
Burleigh, C. Young

John Pittman, Marcus Pitter, Peter
Lake, Jammi Kumar, Jasmeed Adam,
Claude Gerald

Colin Bully, Herman Adams, Ismanie
Roger, Gregory Robin, Earl George,
Munir Alam, Roger Harris

Noel Kirton, Clairmont Cordice, Glen-
roy Browne, Carlton Williams, Lenford
Sampson, C. Antrobus, D. DeFreitas,

Joe Bergasse, Sam Parasram, Richard
Carew, Ghiasudeen Mohammed, Ashraf Ali,
Ralph Phelps, Brian Cooper, St. Clair
Forde, Syed Haque, Pascal Osuji,
Haymchal Harricharan, T.U. Ferguson,
Lawry Wilson, T.H. Henderson, M. Patton,
P.I. Gomes





To increVse value of agri-
cultiral production in the
LDCa of. the Eastern Carib-
b Ann through the improve-
ment of smnll farm profi-
tability, nutritional
pr-ductivity and employ-
Wmen generation.

a) The quantity and value of agri-
cultural production in the I.DCn
will increase significantly by
the end of project, and the
proportion of regional food
requirements supplied from
extra-regional sources will
decrease significantly.

b) A significant proportion of small
farm families will earn an income
from agriculture of at least
US$1500 per yenr in the LDCs.

c) The rate of rural unemployment
and underemployment will not
increase from present levels
in the LDCo.

National and regional
agricultural production

External trade

National employment

Extended periods of
drought and other inter-
uptiona of natural phe-
nomena do not occur.

Close coordination with
planned extension pro-
ject is maintained.

Supply distribution
agencies, credit insti-
tutionn, and marketing
agencies take appro-
priate actions to accom-
modate implementation of
research findings.







To develop recommendations
for improvred forming syatema
though adaptive, fari based
raiearch which farmers can
and wlll use. extension
agetn can explain and
credit in 'l.'irono will
f .-ancte.


1. The cooperative CARDI/Country
farm-baped, agricultural systems
(multicropping) research method
will be established as a continuing
and productive programme on six
Caribbean Islands.

2. Member country contributions to
CARDI'a budget will have increased
to absorb all programmed core and
Country Field Tenm personnel re-
quired to sustain the multiple
cropping research prograume.

3. At least half of the cooperating farmers
will be uning CARDI recommended practices
and systems on at least half of their
sultnble land area with evident economic
benefit to the farm family.

A. Objective evidence can be found that
some CARDI recommended practice or
system is being used on at least twenty
percent of the non-cooperating farms in
the target ecological zone where the
adapted research ia conducted.

5. Credit will be offered on a preferential
basis by the local credit Institutions
to farmers following CARDI'recommended
practices and system.

1) CARDI Budget and CARDI/
Country agreement.

2) Review of farm records,
interviews with farmers
and observations.

3) Interviewn with farmers
and observations.

4) Interviews with bankers.

Governments of the
Region continue to
place high priority
on increasing food
production to meet
regional demand.






The expected outputs of this a)
activity include:
a) The establishment of co-
operative Country/CARLDI
small fanner systems re-
search programmes in six b)

b) The better understanding
of smallholder farming
systrmn.resources and ob-
jectives through the crea-
tion of a soclo-economic
Information base,obtained
through surveys and on-
farm research. c).

e) The'denign of at least 12
aignific.ntly improved
fmvllholder farming systems
based on the Integration of
crop and livestock-specific
proven technology with em-
pirically based economic
analysis that take into
account profitability,canh
flow,nutritional contribution
and labour utilization charac-
d) The transmittal of smallholder
characteristics and improved
farming systems recommendations
to extension officers,credit
officers.planners,and other
agricultural officials through
and field-day activities.

Six cooperative CARDI/Country
research programmes, each in-
volving approximately 25 small

Records, analyses and reports
about the small holding farming
system which reveal technical.
economic and social characteris-
tics of the farm family and
farming syotems and provide in-
sights into the farmers objec-
tives, limitations, resources
and values.

Description, testing and economic
assessment of technological
packages of at least ten crops in
at least two systems applications
in each inland.
Description, testing and economic
assessment of at least two multi-
cropping systems on each island
which represent significant Im-
provement In Lerms of profitability
and/or labour use and/or nutrition
and/or risk reduction and/or cash
flow pattern when compared with
traditional patterns.
Publication of a minimum of
twelve extension Bulletins per
island on technological packages
and production systems in a form
suitable for ready extension to

Organization of a minimum of
eight field days per itland
(two per year) for extension
credit and other personnel to
demonstrate the programme and
its benefits.

CARDI research records,
analyses and publications.

Interviews with CARDI core
staff and country project
field teams and national
and regional extension and
credit personnel.

AID project management
projects and evaluations.

CARDI staff, Country
Agricultural personnel
and cooperating farmers
are able to develop and
sustain a harmonious
working relationship."
T11is will be a matter of
continuous concern to
CARDI management, the
CARD[/AID Programme co-
ordinator and the country
Project Field Teams. All
CARDI core and field per-
sonnel will receive in-
struction in the formal
relationships involved
in the collaborative mode
and the limitations qnd
sensitivities of farm
baaed research.






Consulting Services


Capitol Coat"

Other Coats





























TITLE # of # of Status
farms Reps. Idcnt. Ongoing Conpl t

Eval. of NPK fert. on Eggplant production 2 3 x
Eval. of NPK fort. on Yamn production 2 3 x
Eval. of increasing pl prop. on yield of
S. Potato 2 2 x
Eval. of NPK fert. on Banana Production 2 2 x
Eval. of increasing pl prop. in Bush Beans 2 2 x
Eval.of NPK fert. on Okra prod. 1 1 x
Eval. of NPX fort. on S. Potato Production x
Control of Pests in Cotton x
Intercropping of Corn with Legumes- 3 4 x
Intercropping Sweet Potato with Legumes x
Intercropping Yam with Legumes 1 1 x
Control of Cylas in Sweet Potato x
Eval. of mulching in Vug. Crop production:
Thyme 1 1 x
Onions 2 2 x
Sweet Peppers 3 3 x
Cabbages 4 4 x
Squash 1 1 x
Tomato 2 2 x


TITLE # of # of Status
fu rms Reps.
SRps. Ident. Ongoing Complete

1. Introduction of Protein/Energy feed
Banks for Cattle Production 3 J x
2. Yield exploitation in Peanut Produc-
tion by increasing plant prop. x
3. Intercropping Sea Island Cotton with
Legumes and Corn 2 2 x
4. Evaluation of fertilizer application
on Sweet Potato Production 1 1 x
5. Evaluation of ridge and bed systems
for Thyme Production x
6. Intercropping Sweet Potato with short
duration legumes. x
7. Banana Intercropping System 1 1 x
8. Reduction of infestation of Sweet Potato
weevil by a field sanitation/cruiI
rotation approach x
9. Intercropping Peanuts wi th Beans x
10. Reduction (if Weod Puopulation in lHt-L
Pepper by intercropping 1 1
11. Yield response of peanuts cultivated
on beds vis-a-vis ridges 1 5


1. Evaluation of Year Round Vegetable Prod. 7 7 x
2. Evaluation of Yam/Legume Crop mix 1 1 x
3. Evaluation of Multiple crop mix Dasheu.n
and Cucumber 1 1 x
4. Evaluation of Vegetable Crop mix
5. Evaluation of staking in Tomato Prod. 4 1


Ident. Ongoing Complete

1. Intercropping Rootcrops with Legumes
2. Evaluation of Mulching on Sweet Pepper
3. Evaluation of Staking on Tomato Prod.
4.Plant Population Control in Carrot Prod.

5. Evaluation of Yams Round Homeyard
Vegetable Production Systems
6. Introduction of Improved NC-2 Peanut
7. Introduction and evaluation of short-
duration Pigeon Peas


1. Establishment and evaluation of
Protein/Energy Bank to Cattle feed
in dry season
2. Evaluation of fertilizer application on
Cotton Production
3. Evaluation of an Improved Peanut Produc-
tion Package
4. Control of Sweet Potato Borers
5. Intercropping of Sweet Potato with short
duration Pigeon Peas
6. Evaluation of an Improvcd Sea Island
Cotton Production Package

10 10



# of

# of

12 17


TITLE # of H of Status
farms Reps. Ident. Ongoing Complete

1. Introduction and evaluation of
Recommended Virus Tested Yam with
White Lisbon 10 10 x
2. Introduction of Black Dolly Rams to
upgrade local sheep 8 8 x
3. Control of Fusarium Wilt in Tomato
by introduction of wet-resistant
cultivars 5 5 x
4. Validating Control of Black bug of
Cabbage by hot water treatment 5 5 x
5. Introduction and evaluation of
Improved Cassava Cultivar 1 1 x
6. Introduction and evaluation of short-
duration legumes with existing Multiple
Cropping Systcms 10 10 x
7. Poultry Management Improvement x
8. Evaluation of Multiple Crop Systems
based on Yams x
9. Intercropping of Sweet Potato with
Logumes and Corn x

10. Intercropping of Cassava with short-
duration legumes x
11. Evaluation of alternate crop protec-
tive methods for control of Swuct
Potato weevils x
12. Evaluation of Small Machines and
implements and tools in land pre-
paration activities for crop prodn. x
13. Evaluation of alternate weed control
methods on Multiple Cropping Systems x
14. Evaluation of mini-dams for irrigating
vegetable crops x
1i. Evaluation of Kuccmuiended High Yieldingj
Sweet Potato Cultivar A 26/7 x


TITLE # of # of Status
farms Reps. Ident. Ongoing Ccmplet(

1. Introduction of new Banana Variety 1 1 x
2. Intercropping Bananas with Legumes 3 3 x
3. Introduction and evaluation of now
Peanut variety 8 8 x
4. Evaluation of Increasing peanut
Population on Peanut Production 4 4 x
5. Minimum Tillage of Aroids 3 3 x
6. Validating use of insecticide on
Control of Diamond Back Moth in
Cabbage 5 5 x
7. Cowpea Population Increase in Sweet
Potato/Corn Intercropping Systems 4 4 x


1. Improved Eygplant Production Package 10 10 x
2. Poultry Management Improvement 6 6 x
3. Variety/Fertilizer introduction into
Sweet Potato 4 4 x
4. Evaluation of fertilizer application
in Cocoa 4 4 x
5. Evaluation of fertilizer application
in Tannia 4 4 x


Ahmed, B. A Report on the Livestock Component of CARDI's Small Farmer Multiple
Cropping Project in St. Lucia. CARDI Memo, June 1980.

Ali, A. "Suggestions for Phase II Small Farm Project", Memo to Calixte George, 15 March

Archibald, K., R. Singh, and P. 0. Osuji. Animal Production Systems in the Eastern
Caribbean. CARDI Consultant Report #7, 1981.

Barker, G. H. An Agricultural Profile. CARDI/USAID Research Project 538-0015:
#1 -- Montserrat, January 1981;
#2 -- St. Vincent, January 1981;
#3 -- St. Kitts-Nevis, January 1981;
#4 -- Dominica, January 1981;
#5 -- St. Lucia, February 1981;
#6 -- Grenada, February 1981;
#7 -- Antigua and Barbuda, February 1981.

Bradfield, S. "Appropriate Methodology for Appropriate Technology", Paper presented at
the American Society of Agronomy Meeting, Chicago, Illinois, December 1978.

CARDI. Literature on Project Available in St. Lucia. CARDI/USAID Small Farm Multiple
Cropping Systems Research Project 538-0015, Undated.

CARDI. Major Constraints Identified from Farm Characterization Phase. CARDI/USAID
Small Farm Multiple Cropping Systems Research Project, Undated.

CARDI. Small Farm Multiple Cropping Systems Research Project -- St. Lucia: Agronomic
Practices. Undated.

CARDI. Small Farm Multiple Cropping Systems Research Project -- Schedule I. St.
Augustine, Trinidad, W. I., Undated.

CARDI. USAID/CARDI Small Farm Multiple Cropping Systems Programme -- Farm Inven-
tory. Undated, CARDI, Report of the Chairman 1976-1981.

CARDI. St. Lucia: Chronology of Events. Caribbean Agricultural Research and Develop-
ment Institute, Undated.

CARDI. Staffing-January 1982. CARDI/USAID Small Farm Multiple Cropping Systems
Research Project, Undated.

CARDI. On-Farm Tests/Interventions. CARDI/USAID Small Farm Multiple Cropping
Systems Research Project, Undated.

CARDI. Farming Systems in St. Vincent: Preliminary Analysis. Undated.

CARDI. Activities Pre-On-Farm Tests/Back-up Research, 1979-82. CARDI/USAID Small
Farm Multiple Cropping Systems Research Project 538-0015, Undated.

CARDI. Small Farm Multiple Cropping Systems Programm0e--Work Sheet. Undated.

CARDI. "Notes on the Meeting of Staff of CARDI/USAID Small Farm Multiple Cropping
Systems Research Project," La Toc Hotel, St. Lucia, October 22-23, 1979.

CARDI. CARDI/Agro-Socio-Economic Survey, 1980 Enumeration Schedule, 1980.

CARDI. A Summary Report on the Meeting Held to Discuss the Small Farming Systems
Research Project, Barbados, 7-10 January 1980. Caribbean Regional Food Plan
Project RLA/78/013.

CARDI. A Technical Report on Work Executed by CARDI Up to 31st March 1980.
Caribbean Regional Food Plan Project RLA/78/013, March 1980.

CARDI. Report on "In-House" Workshop on CARDPs Technical Assistance Programme in
Support of the Caribbean Regional Food Plan. Project RLA/78/013. March 18-21,

CARDI. "Notes of a Meeting Held on June 2, 1980 between Dr. S. Parasram and the
Grenada Country Team Comprising Messrs. K. U. Buckmire, R. L. Benjamin, G.
Julien, and D. 0. Rennie", CARDI Secretariat, 6 June 1980.

CARDI. Work Programme 1980 Supplement. University Campus, St. Augustine, Trinidad,
W. I., August 1980.

CARDI. CARDI: Caribbean Agricultural Research and Development Institute. St.
Michael, Barbados, W. I., September 1980.

CARDI. "Minutes of Meeting Held on November 4, 1980, at CARDI Office, National
Provident Fund Building, Castries", Territorial Advisory Committee, St. Lucia.

CARDI. "Summary of Salient Work Done During 1981 by CARDI, St. Kitts in Collaboration
with the Ministry of Agriculture, St. Kitts-Nevis," Undated.

CARDI. Report of an "In-House" of the CARDI/USAID Project 538-0015. Fort Thomas
Hotel, St. Kitts, March 12-14, 1981.

CARDI. "Minutes of Meeting Held on April 2, 1981 at CARDI Outreach Station, La
Resource, Dennery", Territorial Advisory Committee St. Lucia.

CARDI. "Summary of Intervention Workshop held on May 18-23, 1981," Small Farmers
Multiple Cropping Systems Research Project CARDI/USAID 538-0015, May 1981.

CARDI. General Principles Followed in Compiling the Country Work Programs.
CARDI/USAID Small Farm Multiple Cropping Systems Research Project, 1981-82 Work

CARDI. CARDI COURIER. Vol. 1, Nos. 1, 2, and 3 (January, April, and July), 1981.

CARDI. Caribbean Agricultural Research and Development Institute Professional Staff
List. CARDI Internal No. 21, December 1, 1981.

CARDI. Small Farm Multiple Cropping Systems Research Project Semi-monthly Question-
naire 1982.

CARDI. Small Farm Multiple Cropping Systems Research Project Beginning Inventory
Questionnaire 1982.

CARDI. Work Programme 1980-85, CARDI Supplement to Work Program 1980-85, January

CARDI. Proposed Interventions for Implementation on the Small Farms of the Small Farm
Multiple Cropping Research Project in Montserrat. Undated (1982).

CARDI. CARDI Small Farm Multiple Cropping Systems Research Project Antigua Small
Farm Profiles (Nos. 422, 468, 416 and 431). Undated (1982).

CARDI. Small Farm Multiple Cropping Systems Research Project Semi-monthly/4-Month
Supplement 1982.

Carew, R. "General Comments on CARDI/USAID Small Farming Systems Research Project
Phase II", Memo to Calixte George, Project Leader, 23 March 1982.

Chang, 3. "Report on (1) Quality and Quantity of Data for St. Vincent and St. Lucia, (2)
Computer Outputs and Available Information for St. Vincent and St. Lucia, (3) Some
Suggestions for Analysis", October 17, 1981.

CIMMYT Economics Staff. Assessing Farmers' Needs in Designing Agricultural Technology.
International Agricultural Development Service Occasional Paper, New York, NY,

Collinson, M. P. "Farming Systems Research in Eastern Africa: The Experiences of
CIMMYT and Some National Agricultural Research Services, 1976-81," MSU Interna-
tional Development Paper No. 3, East Lansing, Michigan, Department of Agricultural
Economics, Michigan State University, 1982.

Cuevas, M. and C. Weber. Trip Report: Trinidad, St. Lucia, St. Kitts, Dominica and
Barbados. U.S. Bureau of the Census, November 6-December 11, 1981.

Daisley, L.E.A. CARDI Small Farm Multiple Cropping Project. CARDI, St. John's,
Antigua, W. I., March 23, 1982.

Fiester, D. R., et al. Agricultural Development in the Eastern Caribbean. A Survey,
USAID, October 16-November 23, 1977.

George, C., R. Pilgrim, B. Sealy, and A. James. Guidelines for the Determination of
Interventions on Small Farms in St. Lucia. CARDI/USAID Small Farm Multiple
Cropping Systems Research Project Inventention Workshop. Trinidad, W. I., June 1-7,

George, C., A. James, B. Sealy, and R. Pilgrim. Identification of Crop Production
Constraints in St. Lucia. CARDI/USAID Small Farm Multiple Cropping Systems
Research, May 1981.

George, C. Guidelines for Problem Identification and Interventions Year I. CARDI -
Windward Islands Unit, Small Farm Multiple Cropping Systems Research, Undated.

Gilbert, E. A., D. W. Norman, and F. E. Winch. "Farming Systems Research: A Critical
Appraisal," MSU Rural Development Paper No. 6. East Lansing, Michigan, Department
of Agricultural Economics, Michigan State University, 1980.

Gomez, K. A. "On-Farm Testing of Cropping Systems", presented at Symposium on
Cropping Systems Research and Development for the Asian Rice Farmer, IRRI, 1977.

Goodhue, 3. and N. Ferraiuolo. An Assessment of the Small Farm Multiple Cropping
Research Project. Prepared by the U.S. Bureau of the Census for CARDI, February

Henderson, T. H. and P. I. Gomes. A Profile of Small Farming in St. Vincent, Dominica and
St. Lucia. Department of Agricultural Extension, UWI, St. Augustine, Trinidad,
December 1979.

Hoekstra, D. A. and B. B. Goel. CARDATS Evaluation Report, Small Farmers Project,
Dominica. UNDP/FAO/CARICOM Project CAR/77/007, Grenada, September 1981.

International Bank for Reconstruction and Development. Country Profiles for Antigua,
Dominica, Grenada, Montserrat, St. Kitts-Nevis, St. Lucia, and St. Vincent. Undated.

Jessee, D. L. "Farming Systems Research at CARDI: Observation and Suggestions
Presented for Discussion", St. Augustine, Trinidad, March 26, 1982.

Koeghan, 3. M. Forage Grasses for Caribbean Livestock Systems, CARDI.

Krantz, B. A. Report of Advisor to CARDI/USAID Project 538-0015, CARDI, August 1980,
Consultant Report No. 1.

Krantz, B. A. Report of Advisor to CARDI-USAID Project 538-0015. CARDI, April 1981,
Consultant Report No. 6.

Krantz, B. A. "Report of the Second Visit to the CARDI/USAID Small Farm Multiple
Cropping Systems Project", March 15-27 (undated).

Krantz, B. A. "Report of Visit to the CARDI/USAID Small Farm Multiple Cropping Systems
Research Project", January 11-29 (undated).

Lauckner, F. B. Design and Layout of Field Experiments and Conditions Experienced in the
CARICOM Area. CARDI Technical Bulletin No. 7, August 1980.

Leacock, St. C. A. Preliminary Report on Socio/Cultural Factors Affecting Small Farmers
Production and Marketing of Arrowroot and Sweet Potatoes in St. Vincent. Undated.

Lowenthal, A. F. "The Caribbean", The Wilson Quarterly, Spring 1982.

Lowery, 3. E. "Report for Evaluation Team", CARDI/USAID Small Farm Multiple Cropping
Systems Research Project 538-0015, March 1982.

Manteiga, F. P. Small Farm Multiple Cropping Systems Research Project Implemented by
CARDI. Pre-assessment, St. Augustine, Trinidad, W. I., November 12-14, 1980.

Manteiga, F. P. USAID/CARDI Internal Review of the Small Farm Multiple Cropping
Research Project. August 19-September 2, 1981.

Mohammed, G. "Marketing Economist-Work Programme, 1981." 1981.

Mohammed, G. "Market Characterization Methodology of the Market Sub-Sector of Target
Countries Under the Small Farm Multiple Cropping Project of CARDI." Undated.

Mohammed, G. "Suggestions for a Phase II of the Small Farm Project", Memo to Calixte
George, Project Leader, SFMCSRP, 12 March 1982.

Narendran, V. Farming Systems in St. Lucia: An Anthropological Perspective. CARDI
Project 538-0015, June 1981.

Narendran, V. Dominica: Some Observations. CARDI Project 538-0015, Undated.

Osuji, P. 0. Animal Production Activities of Small Farmers Involved in the Cropping
Systems Research Project A Report on a Visit to St. Vincent. CARDI Memo, 1980.

Osuji, P. 0. and S. Parasram. Crop-Animal Systems Research The Experience of CARDI
Card I, Memo, April 1982.

Parasram, S. "Report on a Visit to Antigua and Montserrat July 1-8, 1980," 1980.

Parasram, S. "Report on a Symposium on Farming Systems Research," Washington, DC,
December 8-9, 1980. CARDI, January, 1981.

Parasram, S. "CARDI's Current and Projected Programme in Cropping Systems with Special
Reference to Bananas", paper presented at Workshop on Banana Intercropping in the
Windward Islands, Castries, St. Lucia, June 1-4, 1981.

Parasram, S. "A Survey of Small Farms and Farming Systems in Dominica and St. Vincent",
paper presented at Workshop on Banana Intercropping in the Windward Islands,
Castries, St. Lucia, June 1-4, 1981.

Rankine, L. B., P. I. Gomes, T. V. Ferguson, and K.A.E. Archibald. A Profile of Small
Farming in Antigua, Montserrat and Grenada. CARDI Farming Systems Base Data
Series No. 3, Trinidad, W. I., October, 1980.

Rosen, J. Recommendations Pertaining to the Functional Structure and Staffing Require-
ments for the Small Farm Multiple Cropping Systems Research Project (St.
Kitts/Nevis and Dominica Phase). Prepared by the U.S. Bureau of the Census for
CARDI. August 1981.

Ruttan, Vernon. Agricultural Research Policy. Minneapolis, University of Minnesota Press,

Singh, R. H., K. Archibald, and P. 0. Osuji. "Integrating Goat Production in the Small Farm
System in the Caribbean", Mini-Paper, Third Conference on Goat Production and
Disease, Tucson, Arizona, 1982.

USAID. Project Grant Agreement Between the Caribbean Agricultural Research and
Development Institute and the United State of America for Small Farm Multiple
Cropping Systems Research Project. Project No. 538-0015, August 31, 1978.

USAID. Project Paper, Caribbean Small Farm Multiple Cropping Systems Research, Project
No. 538-0015, August 5, 1978.

Zuvekas, C., Jr. A Profile of Small Farmers in the Caribbean Region. General Working
Document #2, Prepared for USAID, Bureau for Latin America and the Caribbean, by
USDA/OICD, September 1978.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd