Front Cover
 Table of Contents


A farm system development strategy
Full Citation
Permanent Link: http://ufdc.ufl.edu/AA00008152/00001
 Material Information
Title: A farm system development strategy one component of a natural resource management project for the Dominican Republic
Physical Description: ii, 29 leaves : ill. ; 28 cm.
Language: English
Creator: Hart, Robert, d. 1999
Publisher: CATIE
Place of Publication: Costa Rica
Publication Date: 1981
Subjects / Keywords: Agricultural systems -- Dominican Republic   ( lcsh )
Genre: non-fiction   ( marcgt )
Spatial Coverage: Dominican Republic
Summary: This paper discusses farming systems in the Ocoa watershed and is concern with development and conservation.
Statement of Responsibility: Robert Hart, Tito Jiménez Ch.
General Note: Cover title.
General Note: "February, 1981"
General Note: Typescript.
 Record Information
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 669500971
Classification: lcc - S476.D65 H37 1981
System ID: AA00008152:00001

Table of Contents
    Front Cover
        Front Cover
    Table of Contents
        Page i
        Page ii
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
Full Text




Robert D. Hart
Tito Jim~nez Ch.

Turrialba, Costa Rica
February, 1981



I. The Problem .......................................... 1

II. The Project .....,..................................... 1

III. The Role of Research ................................... 1

IV. A Farm System Research and Development Strategy ........ 2

A. Preliminary Characteristics of the Regional
System and Important Farm Systems and
Agroecosystems ...................................... 7

B. Regional (Watershed) Studies ....................... 7

C. Farm Register Studies of Representative Farm
Systems ..........~................................ 7

D. Agroecosystem Experiments .......................... 7

E. Design of Alternatives ..,.......................... ~8

F. Evaluation of Alternatives ......................... 8

G. Mass Media Technology Transfer ..................... 8

V. A Preliminary View of the Agricultural Systems
in the Ocoa Watershed ....~............................. 10

A. The Watershed ................,.................. 10

B. Farms .. . . . .. . . . .. . . .1

C. Agroecosystems ..............................~.... 1,3

D. Soils .. . . . . . . . . . . .14

VI. The Farm System Research Strategy Applied to the
Preliminary View of Ocoa ............................... 16

VII. Some Ideas About Costs .................................. 27




The Relationship Between Training and Research ... .. .

Future Problems .......................***************..

A Positive Note ..................--...------**********

APPENDIX: Informe General sobre los Sistemas Agricolas de Ocoa.





Rapid-deforestation and a shift in land use to pasture and cropland,

along with road building and urban development, has produced massive soil

erosion. An evident result is a decline in productivity of the land and

severe rural poverty. In the long term the faster-than-estimated sedimentation

rate behind dams reduces irrigation potential and the total energy that will be

obtained from hydroelectric sources.


The project goal is to "increase and/or sustain the productivity of the

natural resource base" by improving the natural resource management capabilities

of the Dominicans. Pilot watersheds will be selected, institutional machinery

set up, and erosion will be reduced by encouraging small farmers to use con-

servation-consistent management practices with existing crop and animal

production systems, and to shift from annual to perennial crops.


While the overall success of the project will ultimately be best judged

by evaluating the natural resource management capabilities of the Dominicans,

the medium-term priority is finding out how to do it. Research isi one sub-

process in the larger finding-out-how-to-do-it effort.

To find out how to reduce the erosion and increase the productivity of

small hillside farmers, and obvious start is to look at what has worked else-

where. No documented case of a successful development project within a water-

shed with similar ecological and socio-economic condition is available to

copy; what is available is some relatively successful pieces. Some institu-

tional structures have been more successful than others; some management

practices produce less soil-loss and maintain productivity better than


The research activities can be distinguished from other activities,

such as demonstration or training, by the inclusion of specific testable

hypothesis and objective monitoring, but it should be emphasized that training

technicians, demonstrating a potential technique to farmers, and quantitative

measurement in order to understand how something works (research) are in no

way mutually exclusive. In fact, in this project they will often be combined,

making it difficult to design a research strategy without making some

assumptions about the training and demonstration strategy.

In summary, the role of research within an overall strategy to find out

how to reduce erosion and increase income, get this information to farmers,

and help them adoptit, is to speed up the identification of ecologically and

socio-economically viable alternatives by making this process as objective as

possible. Research is often assumed to slow down the process of getting tech-

nology-to farmers because it requires objective monitoring, but in the long

run there is simply no doubt that the inductive-deductive process is more

efficient than subjective technology selection and demonstration. At the same

time, the research component of the overall effort should be expected to

identify general non-site-specific principles that can be applied in other

areas of the country.


"Farming system" is a label being applied to many different approaches.

There is general agreement between most farming system groups as far as the

general philosophical necessity of conceptualizing the farm as a system formed

by a set of interacting components. Most groups do not go beyond using the

conceptualization as a framework to inhance communication between specialist

and to insure the relevance of the research. Fewer groups actually analyze

whole farm systems and test and evaluate potentini alternatives.

Farming system groups that do not plan to quantitatively analyze farm

systems can get away with crude conceptual models, such as models that include

both abstract and real components (for example,an arrow is drawns between

"religious beliefs" and a "cropping system"). If the farm system is truely

going to be the focal point of a research-effort and system analysis techniques

are going to be used, then a conceptual framework that orders reality in such

a way as to be a useful tool (as opposed to a general guideline) is obviously

needed. The general strategy described below is consistent with an ecological

system conceptual framework. The framework has been found to be a useful

conceptual guideline as well as an analytical tool.

An ecosystem is a biological community and the physical environment

with which it interacts; communities are composed of plant and animal popula-

tions; population in turn are formed by organisms, composed of organs, composed

of tissues, etc., forming a hierarchy of ecological systems. If this framework

is applied for agricultural phenomenon, a set of hierarchically-related

agricultural systems can be identified (see Figure 1).

In Figure 1 a region (watershed) was selected as the largest system.

Of the many components (subsystems) forming a region, a farm system was

selected for closer inspection. Farms are conceptualized as a set of agricul-

tural ecosystems (agroecosystems) that interact with a socioeconomic subsystem.

/ 56 IL5
'1 (4\cRL- atGA-rges


F3Vuee 1, 7i- E bRreatulcot RadamiI~Sb;P (J5sytebSI?3
64Or~M /G "cupm, rmes, pclw w. t~\,ts ArJ


In Figure 1, a crop agroecosystem was selected for further dissertion and the

soil subsystem is emphasized, because of the obvious importance of these

systems in the design of a research project for a watershed with soil erosion


In general, agricultural researchers have studied the phenomenon

included in the hierarchy summarized in Figure 1, as separate units. There

seemS to be two important weaknesses to the traditional approach: (a) the

units are often studied in isolation; any interaction between systems is often

ignored, and (b) the lower-level systems theirr agroecosystem components) have

received more attention than have agroecosystems, farms or regional systems.

In order to analyze any system, at least three hierarchical levels

must be studied, the system itself, the components that form the system, and

the system in which the system of interest functions as a subsystem (component).

Applying this three-level-minimum principle to a project for improving farm

system, means that information on regional systems, farm systems and agroco-

systems must be studied and integrated. If emphasis is going to be given to

improving farm systems by improving agroecosystems (as opposed to improving

housing, for example), research with agroecosystem subsystems will have to be

included. If soil erosion is an obvious problem, specific emphasis will

obviously be placed on the soil subsystem.

Figure 2 is a summary of a general farm system research strategy with

activities carried out on the regional (watershed), farm, agroecosystem, and

agroecosystem component levels. The strategy can be divided into the following

basic steps:

ArGno- sat~
GCo5v 37/^3 Sv ox w illm

F=qutEa 72. CITM cakyotopu adjeQr-Apli .p mAnd~ SErvro ?CPeS
of' R Et 64/200 Artiyj~ lE&1~ tIN A IRQt l?~~Cr)M 8q
A pocut7

A. Preliminary Characteristics of the Regional System and Important
Farm Systems and Agroecosystems

The objective of this phase is to produce general qualitative

models of these systems. A methodology has been developed and tested in Costa

Rica and Honduras (see documents: "Region, Farm, and Agroecosystem Characteriza-
tion", and "Crcerzc" inicial de la Regi~n de la Esperanza"). The

methodology makes extensive use of systems diagrams; they serve as an aid to

forming mental models to be used during information collection and as a way to

summarize information.

B. Regional (Watershed) Studies

Using the static qualitative model produced during Phase 1, climate,

soils, marketing and credit studies necessary to understand the important farm

systems are identified and carried out. It should be noted that these studies

aremnot selected arbitrarily or at an arbitrary level of detail; the selection

and level of detail is determined by what is necessary to understand the farm


C. Farm Register Studies of Representative Farm Systems

Using the static qualitative farm system descriptions produced

during the cha3recterization phase, representative farms should be selected for

dynamic studies. Farm register methodologies are available (see documents:

"Una Finc'a de Honduras Como un Sistema" and "An$1isis Dinimico de Dos Sistemas

de Finca Predominante en el Cant~n de Turrialba, Costa Rica", Ms. Thesis: 0 .


D. Agroecosystem Experiments

Ideally, this phase should begin with exploratory experiments

(many factors, few levels) in order objectively identify the modifications in

the systems with the most potential and to identify the components of the

systems that most strongly interest and therefore have to be studied as an

unit. The exploratory phase should be followed by an analytical phase

(experiments with few factors and many levels). However, political realistic

usually require that subjectively designed alternatives be tested during the

first. year.

E. Design of Alternatives

The regional and farm system studies and the exploratory and

analytical experiment should produce enough information to move from a qualitative

to a quantitative description and understanding of the farm systems being

studied. The quantitative models can be used to design potential agroecosystems

and farm system alternatives.

F. Evaluation of Alternatives

Potential alternative designed in step E can now be tested on a

large scale within the region.

G. Mass Media Technology Transfer

Potential alternatives adopted by farmers in step F can now be

transferred to farmers on a massive scale. If the proceeding steps have been

at all successful., the potential alternatives will have already begun to be

adopted by farmers through a diffusion process that occurs naturally when one

farmer convinces his neighbors. It must be emphasized that if changes in the

credit or marketing system on the regional level are necessary for an

alternative to be successful, these changes must be made before a massive effort

to transmit the new alternative is made.

There are two general criticisms that are often made of this strategy:

(1) there is too much emphasis on understanding the system and not enough on

getting the' new technology already known to exist out to the farmers, and

(2) the strategy requires a multidisciplinary team with capable people that

don't exist in many countries.

The first criticism has more merit if the goals of a project are very

site-specific, since many times a trial-and-error approach will produce

something that works and nobody can really explain why it works. If a project

is directed at a problem in general and not just in one site, it can easily be

demonstrated that, in the long run, designing alternatives on the basis of

understanding how a system works is more efficient than trial-and-error, since

the project produces not only site-specific-technology, but also design

principles that can be applied elsewhere.

The second criticism has obvious merit. The paucity of people trained

in farming systems research is undeniable, however most countries do have

specialists in the different components that together form the systems under

study. Institutional changes necessary to form interdisciplinary teams is

probably more difficult to remedy the lack of trained personnel, but it can be

done. IRRI uses a three man local team with agronomy, economics, and plant

protection specialities represented. The team works with farmers within a

specific "land type" (usually defined by topographic and soil criteria). The

teams within a community are supervised a regional team that is composed of the

same specialities represented on the local team but with more experience and

training. The IRRI scientist interact with and train the regional team.

The research strategy described above and the general comments on

institutional problems are expanded in the section~of this document in which

a hypothetical strategy for the Ocoa watershed is described.



Three man/days were spent in the Ocoa watershed interviewing farmers,

talking to technicians in "Tierras y Aguas", and talking to the local Catholic

priest. While some facts inefuded in this summary also come from reading

reports from other consultants and talking to AID staff, most information comes

from first-hand interviews. In addition to this summary, more information is

presented in Appendix 1, a report on a trip to Ocoa made by Tito Jim~nez.

To be consistent with the hierarchical systems conceptual framework,

the information collected on visits to Ocoa is presented for the watershed,

farm systems, agroecosystems, and soil system.

A. The Watershed

The Ocoa watershed can be viewed as system with components that

interact (people, soils, plants, animals, etc.), inputs (rainfall, radiation,

winds, materials produced outside the area, information, etc.) and outputs

(people, agricultural products, water, soil, etc.).

The human component of the watershed has had a massive impact on the

other components. In 1846 there were 40 families in the area (religious book

own by Father Quinn); today there are 18,000 families. A large percentage of

families are small farmers. In addition to these farms other components of

the primary sector include unimproved grazing land that, in addition to

producing feed for cattle, produces the majority of the wood for charcoal.

Some land is still in forest.

'The components of this secondary sector that were noted during

the visits include bakeries, cement block factory, tailors, shoemakers, rice

hullers (rice is not produced in the areas, but is hulled there), charcoal

producers, and others.

The service sector includes banks, schools, clinics, gasoline

pumps, cooperatives, stores, markets, churches, transportation, communication,

electricity, local government, and agricultural extension and soil and water

representatives of the national government.

The three sectors in the watershed interact to produce, in addition

to the outputs of soil that is last from the watershed, exports of pigeon pea,

beans, peanuts, potatoes, cabbage, onions, eggplant, goats, cattle and a small

amounts of coffee. The region has to import rice, sugar, and flour, besides

fuels, medicines, tools, fertilizer, insecticides, fungicides, and seeds

(principally potato seed).

B. Farms

The natural resource components and the different crops and animals

listed under this primary sector of the watershed system are not found on all

the farms; different combinations of these components are grouped together

to form different types of farm systems. In the limited time spent in the

watershed, it is and possible to devise a farm system classification system for

Ocoa, but there seems to be at least 4 natural groupings of farm system types;

the larger high-income cattle-base farms, and small farmers with very limited

resources, limited resources, and medium resources. The relationship between

types of small farmers and the components included on the farms is shown in

Table 1L.

--Table-2 : The-ReTETtlETOTTHERE getw-6f Tye r IT-Yarms and the Components
included in the Farm System
Farm Type
Very Limited Limited Medium
Components Resources Resources Resources

1. Area 10, 10-80 > 80

2. Annual Crop Intercropping Pigeon pea mono- Pigeon pea mono-
Systems with pigeon pea, culture inter- culture
corn, and beans cropping with
pigeon pea, corn
and beans

3. Perennial Banana, mango -Banana, mango, Banana, mango,
Crop papaya, avocado, papaya, avocado,
Systems coffee coffee

4. Vegetables Squash, chayote Potatoe, onion, Potatoe, cabbage

5. Animals Chickens Chickens, goats, Cattle, goats,
donkeys horses

C. Agroecosystems

Components that form agroecosystems include crops, animals, weeds,

pests, diseases, and soils. The farmer in managing this unit is most

interested in the performance of ,crops and animals, but he has to devise~a

plan to till the soil, kill weeds, avoid losses to insects and diseases, and

order the crop or animal populations in space and time.

The animal based agroecosystems are important components of the

Ocoa farm systems. The goats, cattle, oxen, donkeys, mules, and horses are

fed, in most cases, in the natural vegetation areas, and, in a few cases, in

enclosures with managed pastures.

The goats eat garbage from the house, crop residues and almost

anything. As one farmer said "when they are hungry, they suck on rocks".

Another farmer said he manages his goats like you would a dog; they run loose

all day and come back at night. Goat kids (1-3) are born twice a year, and

milk (1 It/day) is produced for 3 months after the kids are born. Cheese is

produced, but its importance was not determined. Animals are sold or eaten

after one year. The principal health harzard for goats is cars on the


Donkeys, mules, and horses are used to haul cargo to and from the

fields. Smaller farmers seem to have more donkeys and larger farmers mules

and horses.

Oxen are seldom owned by the farmers with very limited resources.

They are rented by farmers with limited resources, but farmers with very

limited resources do all tillage with a hoe. No cases of tractor-powered

tillage was noted in the watershed.

Chickens are found on almost all farms, but in a few cases with

-irrigated-vegetables, farmers did not have chickens. Farmers with very limited

resources tend to have 10-15 chickens; farmers with medium resources might

have 30-40.

Cattle aie found on large farms and on some farms with medium

resources. In terms of area and erosion problems, this agroecosystem is one

of the most important systemsin the watershed.

A striking feature of the crop-based agroecosystems in the Ocoa

watershed is the importance of pigeon pea. It is planted alone and inter-

cropped with maize, beans, or bananas., It is usually left for two years

unless disease problems occur (witches broom?). Maize is almost always

intercropped (no case of maize monoculture was found), usually with beans and

pigeon pea. Beans are often planted twice a year; often intercropped with

maize and pigeon pea at the beginning of the rainy season, and almost always

in monoculture during the second planting. Vegetables, such as potato,

cabbage, and onions are planted as part of rotation systems; for example,

potato-cabbage or potato-beans. Perennial crop systems are usually complex

mixtures including papaya, mango, avocados, citrus, banana and coffee.

Garden systems for home consumption are planted near the house and sometimes

mixed-in with the annual and perennial crop systems. Crops included are

cassava, squash, chayote, fruit trees, sweet potato, figs achiote, oregano,

and ornamentals (and others).

D. Soils

The soil subsystem of the agroecosystems that are part of the farm

systems of the Ocoa watershed deserve specific attention. The soil is '

managed in different ways in different agroecosystems and on different farms.


Farmers with very limited resources practice a farm of minimum

tillage--(but not zero tillage). The soil is prepared for planting by hoeing

up the first 5 em of soils. In pigeon pea, the weeds are controlled in the

same way, sometimes up to 5 times in one year. Farmers recognize that large

quantities of soil are lost when the rains begin, the largest amounts in May

("mayor arrastre") and in October.

Farmers are very aware of the relationship between fertility and

rotation; both among crops, between crops and pasture, and between crops and

fallow. The choice of crops to plant a given point in time depends on fertility

("si estd cansado"). When production gets very low (usually 2-3 years), the

plot is abandoned for at least 2 or 3 years. The rotation is primarily within

the farm; the farmer does not prove his entire farm, as in the case typical

flash-and-burn agriculture practiced in areas with low population density.

Farmers with enough resources to plow with oxen manage their soil

differently than farmers with very limited resources, but most also use hoes

to do the weeding, and on plots with very steep slopes, the initial land

preparation is also done with hoes. Fertilizer is applied to vegetable

(potatoes, cabbage, etc.), sometimes ,to beans and maize, but never to pigeon


The soil is managed differently within the perennial crop agroeco-

systems. Hoes are sometimes used to control weeds, but in many cases a mulch

is found on the soil surface and soil erosion is obviously less.

The soil in the proved pasture grazing areas is "managed"

indirectly by allowing or not alowing grazing. No direct weeding to remove

less desirable species is done. Overgrazing is a major problem and in many

areas the soil loss from these areas may exceed soil loss from the pigeon pea

intercrop systems.

The soil-loanproblem in the Ocoa watershed is not an isolated

phenomenon; it is part of a complex socioeconomic problem. The introduction

of "systems jargon", such as components, inputs, outputs, limits and

hierarchies has its disadvantages; but the problems some how must be tangled-

with as a unit. In the next section, an attempt is made to guess at what a

research effort in Ocoa might look like.


In order to better explain .the general research strategy proposed

in this section, the general steps are applied to a preliminary view of the

Ocoa watershed. The three man/days we spent in Ocoa, even with the information

provided by other consultants, is obviously not enough information to actually

recommend a specific strategy. In start, the following hypothetical, semi-

functional strategy is presented for illustrative purposes only.

The 7 general activities included in the proposed methodology are:

(see Figure 2).

-Preliminary characterization

-Watershed-level studies

-Farm register studies

-Agroecosystem experiments

-Design of alternatives

-Evaluation of alternatives

-Technology transfer

The methodology suggested for consideration within the preliminary

characterization phase has three basic steps: (a) analysis of existing

information; (b) intensive field studies; and (c) synthesis of information

gathered in steps (a) and (b). The output of the methodology is a report

that includes qualitative diagrams of the watershed, farms and agroecosystemE;

and quantitative data presented in written and tabular form. The document

"Caracterizaci6n de La Esperanza, Intibrea, Honduras", is an sample of the

type of report produced by this methodology.

The characterization methodology could be applied to the Ocoa

watershed in 2 weeks by 10 people (or weeks by 5 people). The methodology

requires motivation, and could best be used as part of a short-course to

teach agricultural systems concepts. If ten people participate, 3 days

could be spent learning concepts and how to use diagrams as analytical tools,

2 days summarizing existing information, 3 days in the field, and 5 days

writing the report. If 20 people instead of 10 participate, the time required

would probably not be less, but more people would be trained.

The first report produced would have chapters on the watershed, farms

and agroecosystems. Each chapter would include written description, available

data and diagrams. In order to give an idea of what the diagrams might look

like and their role in orienting the research, we have used the preliminary

information gathered in 3 man-days to draw a semi-real-semi-hypothetical

diagram of the watershed, and designs of the different types of farm systems

identified, and the agroecosystems that seemed predominant.

Figure 3 is a diagram of the Ocoa watershed as a system. The components

are grouped into a primary sector (processes that use the output of the

A petuv4/)RYj I/FNu G t j jp
/4 A S Q,.dy;*1,~ CicLF AnF raq~r i t ,t/ u n RE
-Fg!,e u- P.~ugu~ me-j-UI, 710 /tJ AQF f #tt,; Ci
MA.rAIALs PHE/7GY o t? tt??1

F; pe 3 .


primary sector), and a tertiary sector (private and government services).

The Ocos watershed is an integral regional system. This is not the case for

all watersheds; in some cases a watershed will include only primary sector

components (this is an important methodological consideration for future

watershed studies).

The characterization methodology requires a farm system classification.

In order to illustrate the methodology, use classified the farms into 4 basic


-Very limited resources, basic grain; and chickens.

-Limited resources, basic grains and perennial crops; chickens,

goats, and donkeys.

-Medium resources; basic grains, perennial crops, vegetables;

chickens, donkeys.

High resources; vegetables, pasture; cattle.

Figures 4 and 5 are diagrams of the first two types of farm systems.

The diagrams are of real farms and were drawn after conducting interviews of

approximately 1-2 hours/farmer. Representative of these cases is, of course,

completely unknown.

Figure 6 is a summary of the different agroecosystems identified in the

watershed. Intercrop systems based on pigeon pea are extremely important.

Following the strategy proposed, the next step chronologically would

be to simultaneously begin regional studies, from register studies, and

agroecosystem experiments; but these phases can not be begun before forming

an interdisciplinary team to carry out the research. The success and problems

SOCI -("~ u lds~ 14 0

1?, ~ iI! ;

)I 1~'


LI .4

~I_ __~__



OE e ;

C .

c- Hn


A sRai;cocsT Fnos

r dt Mt`7 P 9misnd

Pl Goerd PCLA

Lt A RmA

CAMr~ + FRvt tRF ,

P84 (T ven r; +- oATS

~n;~-is~r, Tt ac-t aj
C 'rJkUU


idon\\ \ MP

c ~--nC

I __

-~ ~ A/ ;jt, o
TS~v7 -p~ G TA


j 1_

AM IT r s O r b TrFMf
Ma144ks -4

C~ty two AreadL /iibqflagenfews OcDA

PIG90ta PtA
F\1P7 ) / MnlL 2
OR -

plGict pA

/L^Pii i) e6i~ 8 A

147r lg~l 1 yd ei)

I A Wr/Ap

FSur t y1 R EE 2 ( f-14;v C rimJ P',w4 A<, A)th .


GGA7-y fppowit~ wE FbT

C> Ccrr
ft lot n11 .

loo -

Ecl 017 FAR#JJ

L(r7) Limr~ey

L on dfld

L 1 All 1

Lt 7 ~IV 1


'3, it op77i $)aE (1,90up~

c/144 R F, [14, CPJe reh t
oj` (Ira: t- NEW;P Ca r &#: r' h illr (Chffrc ;ff ~p_ c

vTi\~~ .


of Plan Sierra, the IRRI cropping system network, and the CATIE experience

in Central -America and Panama should be considered in setting up the teams

for the Ocoa watershed.

The basic unit of Ocoa project could be Farm System Development

(FSD) teams. Taking ;he farm system-agroecosystem-soil system hierarchy as

a reference point, the FSD team could be composed of an extension agent with

an interest in economics (from level studies), a general agronomist

(agroecosystem studies), and a soil and water specialist (soil system level

studies). The number of FSD teams would depend on the farm system classifica-

tion agreed upon and resources available. Assuming, for sake of discussion

that the project directs its interest toward 6 farm system types (e.g., very

poor, medium poor, and medium farmers on very steep slopes and less steep

slopes), 6 FSD teams would be needed.

These 6 FSD teams could be supervised by a watershed studies committee

composed of the same three specialities, but with MS training or BS with

extensive experience. This committee in addition to coordinating and advising

the FSD groups would be expected to conduct (not delegate) the necessary

studies of climate, soil, credit and,marketing on the regional level. This

committee would be administratively and technically responsible to an agri-

cultural systems' coordinator.

The farm systems' coordinator should be a member of the watershed

development council composed of local leaders. Ideally, he should not be

a direct employee of either "Tierras y Aguas", "Investigaci~n" or "Extensi6n",

but named directly by the Council. The members of the watershed studies

committee and the FSD teams should ideally be technicians from "Tierras y

Aguas", "~Investigaci~n" and "Extensi~n", on leave of absence for two years

with jobs waiting for the~aki hi epcieisiuin hnte

quit or the project ends. A less ideal choice is have them remain in their

institutions and be coordinzted by the ag. systems' coordinator. In this

case everyone would have two bosses and be continuously traveling to Santo

Domingo doing paper work. Even less ideally, would be to have the technical

staff break all ties with their old institutions. In the short-run things

would be easier institutionally, but in the long-run experience gained would

be less likely to be institutionalized.

Now that the team is set up, how do they find ways to reduce erosion

and increase income? The watershed, farm, agroecosystem and soil directed

activities can begin almost simultaneously, however the selection of farmers

is probably the first step. The on-farm activities will include:

-Detailed dynamic case studies; 2/FS type (let say 12).

-Detailed static interviews; 10/FS type (60).

-Specific question surveys; depends on the information needed by

watershed studies group.

-Exploratory and analytical experiments; the number of farmers

depends on the number of agroecosystems selected for study. If

3 are selected (for example pigeon pea and maize,maize and beans

followed by beans, and potato followed by cabbage), 15-25j farms/

FSD team might be a logical number.

-Model farms: one/FS type (or less) during the first year, more

each year as the number of exploratory and analytical experiments


At the end of each year, when the studies are analyzed each FSD group

would be expected to present the FS register studies, the agroecosystem

experiments and demonstrations, and the soil system experiments and demonstra-

tion, as an interpreted report and identify the best model farm plan avail-

able at that point-in-time and the subjects requiring more studies. The

watershed studies group would combine the results of their watershed-level

studies with the FS studies and develop a plan for next years' activities.

As an example of what these studies and experiments might look like,

for illustrative purposes, lets look at the very limited services, basic

grain, chickens FS type. The extension-economist would take the initial

characterization of this FS type and develop a questionnaire that would alow

him to quantify the qualitative diagram of the farm. He would use this

questionnaire to interview a sample of 10 (or more) farmer with this type of

system. He would, in this process, also look for farmers that would be

willing to participate in a weekly interview and fill out registers of daily

labor, amounts bought and sold, etc., (see methodology described 0.

Rockenback thesis). He would begin weekly visit to 2 farmers.

The agronomist on the FSD team might pick the pigeon pea and corn

agroecosystem as an important system for modification. Along with the soil

specialist he would design simple experiments to test possible changes in

tillage, weed control, varieties, etc. In partnership with the soil specialist

conservation technique would be exhausted. Along with minor changes in the

agroecosystem, the team would also consider the steps necessary to begin a

evolution from pigeon pea and maize to a perennial crop system. For example,


permanent shade trees could be planted in the pigeon pea along with banana

as a temporary shade for coffee. During the first year more emphasis should

be placed on minor changes, while in second year the farm register and the

watershed studies would be available with information on labor availability,

marketing, credit, etc., that would make the design of perennial crop system

much easier.

One important aspect with regard to a move towards perennial crops

should not be overlooked; sometimes they are more risky than annual crops.

When a hurricane or storm hits, a farmer can loss years of investment in a

few hours. Avocados may have a good market price, but if -the first winds

blow down the trees a farmer gets nothing; mangos can withstand wind much

better, but may have no market. Coffee prices are notoriously unstable; a

good year in Brasil could wipe out an Ocoa farmer. While if a farmer plants

beans and a storm comes he looses only two months of investment and in 2 1/2

months he can be back to where he started.

As alternative agroecosystems and better farm system become available

more and more resources would bE devoted to demonstration and technology

transfer. At this point the quantitative measurement of yield, farm income,

and soil loss become very useful. If an extension agent can say go talk to

Don Fulano L~pez, he made changes A and B and in two years was stable to go

from 2 to 6 goats and eat meat once a day, the impact is much greater than

going around giving talks on contour plowing. The role of mass media

obviously also become more and more important, but word of mouth success

stories are probably the ideal way to transfer technology to all the farmers

in the watershed.


-Salary of 22 technicians (6 FSD teams of 3;
3 man watershed coordinating group, 1 head) $

-30 vehicles over a 5 year period $

aintenance of vehicles $

-Surveys and farmer registers ($5,000/year,
5 years) $ 25,000

-Experiments and demonstration ($150/year,
5 years $200 each) $150,000

-Cost of physical conservation practices $

-Offices, equipment $

-Credit (?)

-Reducing risk for participating farmers
(everybody buys rice in Ocoa, how about
giving participating farmers a 3ack of rice?) $


-Many-training activities can be combined with research objectives. As

mentio-ned earlier, the prelimi~nary characterization phase car (ane probably

should) be done as part of a short-course. One way to remedy the lack of

farming systems experience within national institutions is have the economists,

agronomist, and soil positions on the local FSD reams filled by young Ing.

Agronomos in their first year with the Secretariat; after two years with the

project, those who have demonstrated the most promise can be sent to get a

MiS and return to do their thesis on a subject directly related to the project.

An experiment currently being tried by the Ministry of Agriculture in

Honduras is to use a pilot area project as an area for students to do Ing.

Agranlome theses. In the university system, stude-rs in their last year are

expected to conduct an experiment and write it up before they get degrees.

The Secretariat provided support so that these experiments could be done on

farmers land with the farmers participation. Some sort of similar effort should

be considered as part of this project.


There will be many problems that will have to be face if a farming

systems strategy is used in this project. Some obvious ones are: (a) m~any

will say that you are making a simple erosion problem more complicated than it

really is, why not just go out and make terraces and plant coffee. The answer

is, of course, if things were that simple people would already have done it;

they are not and probably won't. The problem is not soil erosion; it is a

socio-economic system that keeps flat land in sugarcane in the hands of the

government and leaves small farmers in the bills,

b) There are few trained technicians and the institution structure Is

not appropriate. This project has the primary objective of training people by

giving thepte first-hand experience, Thel only way to change in~stitution is to

take actions that will force institutional analysis and change, nor trr to

change the institutional structurre -atakin, action,

c) Forming multidisciplinary teams that really work is verp difficult.

An, important aspect of the methodologies suggested to make this strate~gy

work, is the use of system diagrams. These diagrams can be continuously used

as a communication tool. Teach technicianL should be able to identify the

phenomten he (or she) is work~ing on and explain how it relates to other studies..

The hierarchical relationship between systems encourage. integration. The soil

specialist will have to use aerg-roeclsyste level information in order to jesigr

and evaluate alternatives; the agronomist will have to use farm level informa-

tion in order to design and evaluate agroecosystem studies. The FSD teams will

need information from all levels i-n order to 3eiser model farms.


The good-will and interest in finding answers among farmers, local

religious leaders, and the field-level technical staff of "Tierras y Aguas"' is

undoubtably cause for optimism. Seeing a technician and a farmer greet each

other with an "abrazo", or listening to a farmer talk about how things used

to be, how bad things are now, and how ;ood they are going to be when changes

are made, makes one think that in sp-ite of the gargantuan problems, something

still can be done.