A Classification of Farming Systems

in the Eastern District of Dominica

 M. Genthon
 French Technical Cooperation Office (F.T.C.)
 c/o La Plaine Post Office
 Dominica, W.I.


 A preliminary classification is presented as a working docu-
ment for development agents on the agricultural extension
staff. The research work for this classification is based on inter-
views conducted by F.T.C. staff with the assistance of other
agents from 1981 to 1984. An attempt is made to place
farmers in homogenous groups to allow more relevant inter-
ventions to be proposed for each group. This is hypothesized
to increase the chances for adoption of new ideas and prac-


tices. Other applications include (1) identifying target groups
for training and for work elements; and (2) identifying rele-
vant interest groups of farmers. The classification, subject to
refinement, includes five main elements: (1) "Petit Planteur";
(2) the progressive farmer; (3) diversified farmers (3
categories); (4) part-time farmers; and (5) limited agricultural
income farmers (four sub-categories).


The Training-Research and Development
Project in Dominica
 This project started in 1981, in the southeastern district of
Dominica. Its prinicples are those of a "Research-Training-
Development (TRD)" process:
 1. Research, training, extension and development must be
 closely linked. They are generally separated from both
 structural and geographical points of view.
 2. Research must operate close to farmers. Farmers' participa-
 tion is required. The researcher must know the farmer, so
 research can really focus on farmers' needs. The technical
 information given to extension must not only include
 isolated topics, but the coherence and the insertion
 capability in the present system must be checked.
 3. The Research-extension-development process much be
 continuous. The usually vertical structure

 Research

 Extension

 Farmer

 is replaced by the following one, which includes feedback
 at each level.

 Research


 Extension Farmer

 At an institutional level, a small team of three agonomists, two
French and one Dominican, are working in close relation with the
Extension staff of the district (two extension officers, one
agricultural officer). It provides technical support to Extension,
training activities and realises a part of the research work needed
(on the farming systems of the district especially). This district
covers six main villages and about 1,200 farmers.
 Other similar projects, supported by the French Regional
Cooperation Mission in St. Lucia, have been established in other

VOL. XX-PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY


Windward Islands (St. Lucia, St. Vincent, Grenada). The com-
parisons of their results and evolutions helps to generalize the
results of these projects, and gives ground to regional cooperation
on these methodologies.
 A new university degree in Caribbean agricultural develop-
ment, in collaboration with UWI, has recently been established
in the University of French West Indies and Guyana. This is ex-
pected to deliver training on research/development processes and
farming system analyses to students of French or English-
speaking countries of the West Indies.

Methodology
 The basic assumption of the TRD process is that the farmer's
decisions are rational, given the objectives and the set of con-
straints he faces, whether at the field level (for example, fertility
problems), at the family level (allocation of work, land available),
or at the regional or national level (marketing of agricultural pro-
ducts, transport facilities, support from extension services). In
other words, the farmer gets roughly the best results possible,
given his environment and objectives. His objective might not be
to get a maximum level of income and production, but to secure
what he considers as "his basic needs" and to engage in other ac-
tivities during his spare time.
 So the first step of this approach is to identify the general
agrarian system of the area. The notion of agrarian systems in-
cludes not only farming systems, but also the general environ-
ment, socio-political as well as natural. Then, different agro-
ecological zones have to be identified within the area. An agro-
ecological zone is an area where the natural environment and the
combination of different farming systems are homogenous,
leading to the same type of geographical characteristics, as land-
scape, production, etc.
 In the case of the southeast district, the first data suggested
that the district could be considered as one homogenous area,
climatic and pedological variations being limited within the dif-
ferent villages of the area.
 The next step is to identify the different types of farming
systems encountered within each homogenous area. Even if
farmers belong to the same agrarian society (which in this case
may be considered as peasantry), each individual has a different


111