• TABLE OF CONTENTS
HIDE
 Front Cover
 Title Page
 Table of Contents
 Famers' approaches to soil-fertility...
 Increasing the adoption rates of...
 Characteristics of improved technologies...
 Interaction between nitrogen level...
 Farmer-first qualitative methods:...
 Sustainable agriculture and farming...
 Crop-enterprise selection of farming...
 Agroforestry adoption: The role...
 Measurements of economic viability...
 Reorientation, not reversal: African...
 Instructions to authors






Group Title: Journal for farming systems research-extension.
Title: Journal of farming systems research-extension
ALL VOLUMES CITATION PAGE IMAGE ZOOMABLE PAGE TEXT
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00071921/00003
 Material Information
Title: Journal of farming systems research-extension
Alternate Title: Journal for farming systems research-extension
Abbreviated Title: J. farming syst. res.-ext.
Physical Description: v. : ill. ; 23 cm.
Language: English
Creator: Association of Farming Systems Research-Extension
Publisher: Association of Farming Systems Research-Extension
Place of Publication: Tucson Ariz. USA
Publication Date: 1990-
 Subjects
Subject: Agricultural systems -- Periodicals -- Developing countries   ( lcsh )
Agricultural extension work -- Research -- Periodicals   ( lcsh )
Sustainable agriculture -- Periodicals -- Developing countries   ( lcsh )
Genre: periodical   ( marcgt )
 Notes
Dates or Sequential Designation: Vol. 1, no. 1-
General Note: Title varies slightly.
General Note: Title from cover.
General Note: Latest issue consulted: Vol. 1, no. 2, published in 1990.
Funding: Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
 Record Information
Bibliographic ID: UF00071921
Volume ID: VID00003
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 22044949
lccn - sn 90001812
issn - 1051-6786

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Title Page 1
        Title Page 2
    Table of Contents
        Table of Contents
    Famers' approaches to soil-fertility maintenance under reduced fallows in the south west province of Cameroon, by S. W. Almy, C. Ateh, T. Woldetatios, M. Mboussi, C. Poubom, and M. Besong
        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
    Increasing the adoption rates of new technologies with a new technology-transfer model, by Ramiro Ortiz and Adlai Meneses
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
    Characteristics of improved technologies that affect their adoption in the semiarid tropics of eastern Kenya, by A. P. Ockwell, L. Muhammad, S. Nguluu, K. A. Parton, R. K. Jones, and R. L. McCown
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
    Interaction between nitrogen level and Hessian fly protection of wheat in Morocco, by John Ryan, J. P. Shroyer, and M. Abdel Monem
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
    Farmer-first qualitative methods: Farmers' diagrams for improving methods of experimental design in integrated farming systems, by Clive Lightfoot and D. R. Minnick
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
    Sustainable agriculture and farming systems research teams in semiarid west Africa: A fatal attraction?, by J. L. Posner and Elon Gilbert
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
    Crop-enterprise selection of farming systems in eastern Gambia, by M. B. Kabay and Lydia Zepeda
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
    Agroforestry adoption: The role of farmer associations in Senegal, by F. A. Caveness and W. B. Kurtz
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
    Measurements of economic viability in Cape Verde, by Mark Langworthy
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
    Reorientation, not reversal: African farmer-based experimentation, by Doyle Baker
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
    Instructions to authors
        Page 148
Full Text
Volume 2, Number 1
1991


journal
for Farming Systems
Research-Extension







Journal
for Farming Systems
Research -Extension


Volume 2, Number 1, 1991


Published by
the Association for Farming Systems Research-Extension








Journal for Farming Systems Research-Extension

Editorial Board


Timothy R. Frankenberger, Editor
Office of Arid Lands Studies
The University of Arizona, Tucson

Peter E. Hildebrand
Food and Resource Economics
Department
University of Florida, Gainesville
Harold J. McArthur
Office of International Programs
University of Hawaii, Honolulu



The Ford Foundation
Inter American Foundation
The Kellog Foundation
Michigan State University Foundation
The Rockefeller Foundation


Timothy J. Finan
Bureau of Applied Research in
Anthropology
The University of Arizona, Tucson
Donald E. Voth
Agricultural Experiment Station
University of Arkansas, Fayetteville

C. David McNeal, Jr.
Extension Service, USDA

Sponsors:

United States Agency for International
Development
United States Department of Agriculture
The University of Arizona
University of Florida


The Journalfor Farming Systems Research-Extension is published by the Association for
Farming Systems Research-Extension (AFSRE), an international society organized to
promote the development and dissemination of methods and results of participatory on-
farm systems research and extension. The objectives ofsuch research are the development
and adoption through participation by farm household members of improved and
appropriate technologies and management strategies to meet the socioeconomic and
nutritional needs of farm families; to foster the efficient and sustainable use of natural
resources; and to contribute toward meeting global requirements for food, feed, and
fiber.
The purpose of the Journal is to present multidisciplinary reports of on-farm research-
extension work completed in the field, and discussions on methodology and other issues
of interest to farming systems practitioners, administrators, and trainers. The Journal
serves as a proceedings for the annual international Farming Systems Symposium from
which selected and refereed papers are included. It also welcomes contributed articles
from members of the AFSRE who were unable to attend the symposium. Contributed
articles will be judged by the same review process as invited articles.
Technical Editors: Nancy Schmidt, Daniel Goldstein, and EmilyE. Whitehead, Office ofArid Lands
Studies, The University of Arizona
Design: Paul M. Mirocha
Production and Layout: Nancy Schmidt, Delphine Keim, Diedre Muns, and Robert S. Breckenridge,
Arid Lands Design, Office ofArid Lands Studies, The University of Arizona
ISSN: 1051-6786









Journal for Farming Systems Research-Extension
Volume 2, Number 1, 1991


CONTENTS


1 Farmers' Approaches to Soil-Fertility Maintenance Under Reduced Fallows
in the South West Province of Cameroon
S.W. Almy, C. Atch, T. Woldetatios, M. Mboussi, C. Poubom, and M. Besong

19 Increasing the Adoption Rates of New Technologies With a New
Technology-Transfer Model
Ramiro Ortiz and Adlai Meneses

29 Characteristics of Improved Technologies That Affect Their Adoption
in the Semiarid Tropics of Eastern Kenya
A.P. Ockwell, L. Muhammad, S. Nguluu, K.A. Parton, R.K. Jones,
and RL. McCown

47 Interaction Between Nitrogen Level and Hessian Fly Protection
of Wheat in Morocco
John Ryan, J.P. Shroyer, and M. Abdel Monem

57 Farmer-First Qualitative Methods: Farmers' Diagrams for Improving Methods
of Experimental Design in Integrated Farming Systems
Clive Lightfoot and D.R. Minnick

71 Sustainable Agriculture and Farming Systems Research Teams in Semiarid West
Africa: A Fatal Attraction?
J.L. Posner and Elon Gilbert

87 Crop-Enterprise Selection of Farming Systems in Eastern Gambia
M.B. Kabay and Lydia Zepeda

97 Agroforestry Adoption: The Role of Farmer Associations in Senegal
F.A. Caveness and W.B. Kurts

109 Measurements of Economic Viability in Cape Verde
Mark Langworthy

125 Reorientation, Not Reversal: African Farmer-Based Experimentation
Doyle Baker








Farmers' Approaches to
Soil-Fertility Maintenance
Under Reduced Fallows in the
South West Province of Cameroon

S.W. Almy, C. Ateh, T. Woldetatios, M. Mboussi, C. Poubom,
and M. Besong'


INTRODUCTION
The principal problem of agricultural sustainability in Africa is the lack of
adequate technologies for maintaining soil fertility when the practice of
leaving fields fallow for long periods is abandoned. Increases in population
densities have forced farmers to make more intensive use of land that was
cultivated only in 6- or 25-year cycles (Nye and Greenland, 1960; Ruthen-
berg, 1976; Prinz and Rauch, 1987). Reduced fallows have led to an increased
nutrient-depletion rate, erosion, changes in soil texture, and increases in insect
pest populations and crop diseases. This paper describes farmers' soil-
management practices in the African humid forest zone and evaluates the
potential for alley cropping and green manuring techniques for sustaining
fertility.
A variety of biological solutions have been proposed, including legume
intercropping, improved fallow crops, mulch crops, and alley cropping
(Hartmans, 1981; Padwick, 1983; Watson, 1983; Balasubramanian and Egli,
1986; Prinz, 1987; IITA, 1990). However, experience shows a slow response
by farmers to agronomic recommendations. The most popular have been
applications of chemicals (fertilizers, insecticides, fungicides, and herbicides),
which give quick results but have troubling long-term consequences if used
alone. In this era of economic crisis, they are also increasingly difficult to
obtain.
A large part of the difficulty of convincing farmers to use new, more-
sustainable practices lies in their complexity, which is a key constraint in

Testing and Liaison Unit-Ekona, Institute of Agronomic Research, PMB 25, Buea
South West Province, Cameroon.





ALMY ET AL.


adoption of innovations (Rogers, 1962; Rogers and Shoemaker, 1971).
Sociologists have debated the relative importance of negative profitability and
technical information in explaining farmers' reluctance to use conservation
technologies. Nowak (1987) used evidence from the United States to argue
that a better technical fit of recommended practices to localized environments
would increase their adoption. A technique that is adapted to an existing
situation becomes effectively less complex because fewer of its elements are
new to the situation. This type of technique is usually less costly and more
profitable, which also influences its adoption.
To reduce complexity and cost, farmers must be brought into the process
of developing the solutions, and, at the beginning, their perceptions of both
problems and solutions must be understood. Only then can new or altered
practices be suggested that will better the farm environment without increasing
the farmers' constraints.
This paper presents the results of studies undertaken by the Testing and
Liaison Unit (TLU) in Cameroon in order to understand farmers' approaches
to soil-quality maintenance, focusing on soil fertility and compaction. It looks
at the methods of, and the reasons for, the preparation of planting surfaces and
disposal of weed and crop residues in four different ecologies. The potential
adaptability of new, sustainable techniques of alley cropping and mulching is
compared.
TLU is a Farming Systems Research and Extension (FSRE) unit at Ekona
Agricultural Research Centre within the national Institute of Agricultural
Reserach (IRA) of Cameroon. It is funded by the U.S. Agency for Interna-
tional Development (USAID) under the National Cereals Research and
Extension Project and is technically supported by the International Institute
of Tropical Agriculture (IITA), Ibadan, Nigeria. TLU is responsible for the
interface between food-crop research, farmers, and extension and carries out
surveys, on-farm testing, and extension training. Its mandate is the high-
rainfall coastal lowlands of Cameroon, particularly the South West Province
(SWP). During the first two years of general farming-systems surveys, on-farm
variety testing, and development of research-extension links, TLU identified
the problem of sustainability under reduced fallow periods as a major issue.


METHODOLOGY

After completion of statistically representative surveys of the entire SWP
(Almyand Besong, 1987, 1989a, 1989b;Almy, 1988), four composite zones


Journal for Farming Systems Research-Extension





SOIL-FERTILITY MAINTENANCE IN CAMEROON


were defined that represent much of the ecological and agronomic diversity
of the province's lowlands and which became the focus of further work: the
Lower Volcanic (or Volcanic), Kumba Corridor (or Kumba), Sands, and
Mamfe. These four "key zones" contain half the farming population and two-
thirds of the food production of the province.
In 1989, TLU started a study of food-crop farmers' land-preparation and
weeding practices in the four key zones. The methods used, described below,
include issue-focused rapid appraisal tours, monitoring of weed growth and
weeding, confirmatory discussion in village meetings, and quantification of
selected farmers' opinions and practices in a 1990 questionnaire survey. The
principal method contributing to this paper is rapid appraisal (Hildebrand,
1981; Chambers, 1985).
Rapid appraisal tours were conducted during land preparation (about 2 to
4 weeks before planting) and at about 8 to 10 weeks after planting. Seven
villages, representing relevant variations in soils and practices, were visited by
a group of two to four agronomists (including a botanist and a weed scientist),
an anthropologist, and the local extension agent, who acted both as a key
informant and as an interviewer. The agent picked four to six farmers with
varying land-preparation methods and/or soils for farm visits.
During the initial tours, the team characterized neighboring fallows and
the methods used in land preparation. Discussions with the farmer and agent
covered the relationship of fallow period to fertility and weediness, reasons for
choosing methods of disposition for different cleared residues, and reasons for
using different types of burning and seedbed. Special attention was given to
farmers who had changed from one practice to another, because of migration
or influence from migrants. Chemical analysis was done on soil samples from
raised seedbeds, alleys, and neighboring bush fallows.
At 8 to 10 weeks after planting, the team revisited most of the farms. Weeds
were identified, weed growth and crop vigor were evaluated and compared
with land-preparation method by the team and farmer, and the farmers'
opinions on the worst weeds and their growth and management were
solicited. The conclusions and hypotheses from both tours were combined in
a final written report.
Several methods were used to confirm and develop the observations from
the rapid appraisal tours. TLU monitored weed growth and weeding time and
methods in 80 fields in four villages in the Sands and Volcanic zones.
Discussions evaluating farmers' experiences with "modern" technology,
including soil-management techniques, were held in eight villages (two per


Vol. 2, No. 1, 1991





ALMY ET AL.


zone) later in the year. Finally, a questionnaire about the adoption of IRA
technologies, which included confirmatory questions on farmers' opinions
and practices, was applied to a random sample in 16 villages in the four zones.


THE ENVIRONMENT
The South West Province (SWP) contains a complex mix of soils, rainfall
levels, altitude ranges, and population densities. Mean fallow periods vary
from 1.4 to 5.1 years in the different zones. Many of these zones contain large
numbers of immigrant farmers with different cultural practices than those of
the original inhabitants. The diversity makes an ideal laboratory for the study
of farmers' approaches to maintaining soil quality.
The province contains four major soils types: new and old volcanic,
sedimentary, and granitic. Mean rainfall varies from 1,800 to 9,800 mm per
year, although 2,500 to 3,500 mm is the most common, allowing two
growing seasons. Altitude of farms ranges from 0 to 2,500 m above sea level,
but the majority of the province is lowland. Farm population densities range
from 0.5 to 9.3 nuclear households per km2.
There are 18 indigenous ethnic groups and at least as many immigrant ones,
each with its own language or dialect. Most people live within walking
distance of other language speakers, and Pidgen English has become the
common tongue. Willingness to considerand adoptother peoples' technologies
is a noteworthy characteristic of local farming.
The SWP does not depend on a single staple; instead, most farmers give
nearly equal attention to plantains, cocoyams, maize, and cassava. Almost all
lowland farmers grow cocoa and coffee. Most fields are planted in the same
season both with three-month crops (maize, melon, and groundnuts) and
later-maturing crops that will remain for 9 to 12 months or even up to 2.5
years. All cultivation is manual, and fertilizer and other chemical inputs are
scarce and rarely sought after. Animals (goats and chickens) are few and free
ranging.
The four zones vary sharply (Table 1). The Volcanic zone has highly fertile
soil of very recent volcanic origin. It is farmed almost entirely by immigrants
from the highlands, where intensive intercropping on beds is common. In the
Volcanic, however, they have learned to monocrop different sections of the
field and to plant mostly on flat ground.
Most Kumba-zone soils are of old volcanic origin, heavier than Volcanic-
zone soils, of medium fertility, and usually deficient in phosphorus. Kumba


Journalfor Farming Systems Research-Extension






SOIL-FERTILITY MAINTENANCE IN CAMEROON


Table 1. General Characteristics of Key Zones in South West Province, Cameroon

Kumba Lower
Corridor Volcanic Mamfe Sands

Ecology:
pop./km 48 67 18 29
altitude (m): 100-400 0-600 0-300 0-300
rainfall (mm): 2.2-2.500 3-4.500 (3.200) 1.8-3.000
soils: base: old volcanic new volcanic granitic sedimentary
mean org. C: 2.8% 4.8% 1.9% 1.6%
mean N: 0.26% 0.36% 0.11% 0.10%
mean P: 10 29 8 7
mean K+ meq: 0.26 0.80 0.17 0.12
mean pH (H20): 5.3 5.7 4.7 5.6

Major crops: maize plantain cassava cassava
plantain maize maize plantain
cassava cocoyam plantain yam
cocoyam taro taro egusi

Land preparation:
fallow: 0-1 yr. 0-2 yrs. 3-5 yrs. 1-2 yrs.
clearing: burn/mulch bum/mulch bum/mulch burn
form: beds flat mounds flat/mounds

Cropping system:
densities: high medium high medium
associations: maize+taro maize+ cocoyam+taro cassava+
cocoyam+ cassava+ maize+cassava maize+(egusi/
groundnut+ cocoyamm/ groundnut+ groundnut);
(cassava/ taro); egusi+yam; egusi+yam+
yam); maize sole cocoa sole; maize;
plantain+ coffee+ cocoa+ cassava,
cocoa plantain plantain cocoa sole


is farmed by both highland immigrants and local farmers who have put most
of the surface area in cocoa, creating land shortages. Food crops are usually
intensively intercropped on beds, with the local farmers (who traditionally
planted on mounds) gradually shifting to the highland system.
The Sands zone comprises two subzones, both with low-fertility, sandy
clay-loam sediments. The short-fallow-period subzone supports a heavy
population of immigrants from elsewhere in the coastal lowlands; they


Vol. 2, No. 1, 1991





ALMY ET AL.


intercrop at medium densities, usually on mounds. The long-fallow-period
subzone, which also experiences higher rainfall rates, has a population of local
and ethnically related Nigerian farmers, who practice mixed cropping on the
flat ground.
The soil of the Mamfe zone is granitic, low-fertility, and acidic (pH<5.0),
and is in some places very sandy and in others very clayey. The purely
indigenous population lives at relatively low densities. They practice long
fallows and intensive intercropping on large mounds. Unlike the other zones,
marketing outlets are not available to spur increased production.


RESULTS


Farmers' Concepts of Fertility and Plant Growth
Throughout the province, farmers say the fertility of their land has declined
since they began their farming careers. Except in the Volcanic zone, all farmers
state that there is a definite minimum fallow time to regain fertility, a time
expressed in years but judged for each field by the type and maturity of the
vegetation growing there. Even in the Volcanic zone, many farmers regard a
6-to-12-month fallow period as necessary. Fully developed fallows, as defined
by the farmers, consist of Chromolaena odorata ("acha cassara") and young
hardwoods in Mamfe, Sorghum arundinaceum ("mandola") and young
hardwoods in the Sands, and Sorghum and Pennisetumpurpureum ("elephant
stalk") in the Volcanic and Kumba zones. Leguminous cover crops, Pueraria
and Mimosa invisa, are also present in most fallows.
In practice many farmers do not respect their own fallow times. Fields in
the more-populated zones fallow for 0 to 2 years on average, 30 percent
shorter than the farmers say they fallow in general. In Mamfe and the less-
populated part of the Sands zone, actual fallows are 3 to 5 years, or 5 percent
shorter than described. Many immigrants are restricted to the land they buy;
if they cannot buy more, or pay a high enough rent, they must fallow less.
Women who are sick or pregnant farm close to home, where land pressure is
high. Those who cannot get enough help in clearing and mounding reuse land
that has been followed for less time, because the vegetation is easier to cut and
the ground is less compacted.
SWP farmers differentiate between five types of vegetative matter on their
fields: "sticks," "strong grass," "soft grass," "chop," and "cover crop."


Journal for Farming Systems Research-Extension





SOIL-FERTILITY MAINTENANCE IN CAMEROON


"Sticks" are trees, living or felled. "Strong grass" includes woody-stemmed
shrubs, small branches, and thick leaves. "Soft grass" is any nonwoody weed.
"Chop" is a food plant. "Cover crop" is any food crop or weed that spreads
along the ground, covering the surface. This classification guides farmers'
treatment of the plants. Further differentiation in Pidgen English is made only
for crops and fallow indicators ("mandola, elephant stalk," and "acha cassara").
The factors commonly thought by farmers to contribute to plant growth
are "manure," "sun" (light), "air," and "water" (rain). "Manure" is plant
food, any natural material from which the crop's roots can feed to strengthen
the crop. The term usually is applied to weed mulches, but can be extended
to woodash, woody branches, and ordinary soil, especially when it is molded
around the crop stem. Animal manure is not referred to as "manure" and is
used only rarely as an olfactory deterrent to goats and pigs that might uproot
special crops. Perhaps due to its rarity and low level of use, few farmers have
observed the effects of animal manure on crops.
Soil regains nutrients lost to crops through the passage of time, which is
marked by the growth of certain grasses or saplings. However, these species
are not of themselves considered to create fertility. Only a few farmers believe
that groundnuts enrich the soil, and only a few identified a specific weed as soil
enriching. Coffee husks are the most widely known enhancer, but few farmers
are able to get them from the factories.
Inputs are rarely brought from outside the farm to improve fertility.
Chemical fertilizer is known to be useful as manure. Farmers, however, are not
willing to spend cash on it, and some disapprove of its effects (rotting and
taste) on root crops. Only 10 percent have ever used chemical fertilizer on
their food farms; most people who used fertilizer did so only because a small
quantity was given to them. Animal manure is occasionally used by 9 percent
of the farmers. None bring extra plant residues in from outside the field.
Competition among plants for light and breathing room is visualized as
taking place primarily above the ground. Crops are usually planted at very
high densities, and the major concerns are to avoid shading those that need
the most sunlight (egusi melon, groundnuts, and, at early stages, maize and
yams) and to clear weeds away to allow established crops still in production to
obtain "air": specifically, an empty space around the stem where the wind can
pass.
Although the ecology is perhumid, some farmers express concern about
adequate water, especially in the first few weeks after planting when rains may


Vol. 2, No. 1, 1991





ALMY ET AL.


be sporadic. Better water absorption is felt to result from tilled soil. Standing
water (waterlogging) is also recognized as a problem for root crops and
groundnuts in some areas.
Finally, farmers whose soil is overly compacted (primarily in Mamfe) refer
to "strong ground" as an important factor in their soil-management decisions.
Reduction of compaction is also a concern for small-seeded specialty crops
(okra and leafy vegetables).

Farmers' Practices
Practices affecting soil quality include leaving living stumps in fields,
gathering of surface soil in raised seedbeds, burning or incorporating weeds,
and using leguminous crops. These are considered in turn.
Living stumps in food-cropfields: One factor common to all the zones is
the presence of small, truncated trees throughout most food fields, at densities
of 300 to 1,200 plants per ha. Farmers reduce competition with crops by
cutting them off at the base and pruning regrowth during weeding. In most
cases the stump and root system remain alive and regenerate young, 2-to-
3-m tall trees during a 3-to-4-year fallow, or shrubs during a 3-to-12-month
fallow. In areas with yam culture and short fallow, saplings are preserved and
pruned yearly to serve as yam stakes; when they grow older and broader, they
are destroyed.
The existing trees, except for a Cassiasp. seldom found inside fields because
it attracts ants, are not thought to be nitrogen contributors. Although the
stumps contribute to soil quality by blocking run-offand probably by drawing
nutrients from the subsurface to the pruned leaves, no farmer interviewed
voiced any benefit from the practice. Some farmers in all zones try to destroy
established saplings by intensive burning and uprooting, with only partial
success. The majority who leave stumps alive say that they are too difficult to
uproot and that as long as they are cut back they do not deprive the crop of
nutrients.
Leaving living tree stumps inside the fields performs the same functions as
alley cropping, without imposing a linear design. The practice is largely
accidental, the result of the difficulty of removing trees and of the lack of
strong reasons for doing so. This difficulty affects the feasibility ofleguminous
alley cropping, as the existing stumps must be removed to make room for the
new.
Seedbed preparation: Mounding or bedding is practiced by at least some
farmers in all zones, and by 80 to 85 percent of all farmers for root crops. Beds


Journal for Farming Systems Research-Extension





SOIL-FERTILITY MAINTENANCE IN CAMEROON


("ngong") are flat surfaces about 10 to 20 cm high, 60 to 80 cm wide, and
100 to 150 cm long. Mounds ("heaps") are convex surfaces, originally built
as cones but rapidly compacted into rounded heaps 30 to 40 cm high and 60
to 200 cm across.
Most commonly, part of the soil is pulled from three adjoining mounds, or
from two adjoining beds, to make the new one in the former alley. The pattern
is created around existing tree stumps, the mound placed at the side ofa stump
so that after it is moved to the alley the stump will remain at the edge. Any
surface litter is pulled in, although some farmers remove large pieces of "hard
grass" (twigs, roots, and leaves). The old mounds are largely intact up to three
years after fallowing begins, even on sandy soils, and could still be seen in a
clayey Mamfe field after a 15-year fallow.
Making mounds or beds is an extremely labor-intensive task, occupying
more time than any other operation in the farming calendar. However, a
raised seedbed significantly increases fertility and reduces weed growth and
weeding labor costs. Comparison of nutrient levels between mounds, alleys
(flat areas between mounds), and adjacent fallows of nine fields in the lower
three zones (Table 2) shows consistently and significantly higher levels of
organic carbon, nitrogen, phosphorus, and potassium in mounds than in alleys
or fallows. Nutrient levels of different Mamfe fields sampled on flat or raised
surfaces (Table 3) also support this trend. Acidity was not reduced on
mounds, except possibly in the highly acidic Mamfe soils. Weeds were
decreased: in the 80-field sample from the Sands and Volcanic zones, planting
on mounds or beds was significantly related to lower weed levels.
The decision whether to plant on a raised surface is related to seed
establishment and soil compaction. It seems not to be related to fertility
considerations. Farmers using them say raised surfaces are necessary for the
good establishment ofsmall-seeded crops groundnutss, maize, and okra) and
for tuber expansion in root crops. Farmers in the zone with the richest, least-
weathered soil (Volcanic) plant tuber crops directly into their rocky soil,
undertaking the laborious task of bedding only for small-seeded crops, which
cannot get a grip on the coarse-grained new volcanic particles. But a third of
those in the medium-fertility, more-compacted soil of Kumba zone seed
maize directly on the flat ground and plant all their tuber crops in beds or
mounds. The farmers in the low-fertility Sands zone use mounds where the
soil is prone to water-logging due to underlying hardpan, but plant on the flat
elsewhere. (Sands farmers explained that crops planted on mounds die if the
early rains are interrupted.) The practice of cultivating on raised surfaces may


Vol. 2, No. 1, 1991






ALMY ET AL.


Table 2. Relative Increases in Acidity, Organic Matter, and Nutrients in Mounds
(Raised Surfaces), Alleys, and Fallows in Nine Farms of South West Province,
Cameroon, March 1989

Mound/Alley Mound/Fallow Fallow/Alley

mean s.d. mean s.d. mean s.d.


pH (H0O) 1.05 .08 1.02 .05 1.01 .09

Organic C 1.54a .43 1.18 .29 1.27a .19

Total N 1.37a .22 1.18b .20 1.11b .07

Avail. P 1.89b .69 2.25b .74 0.94 .40

K+ meq 2.00b 1.09 2.28b .88 0.99 .40


a Significant at .01, t-test.
b Significant at .05, t-test.


Table 3. Chemical Analysis of Soil Samples from Farms in Manyu Division (Mamfe
Central), on Flat or on Ridges (May 1988 and March 1989)

Flat at 2-4WAPa Flat at OWAP Ridge at OWAP

mean range mean range mean range

pH (HO0) 4.8 4.2-5.3 4.9 4.7-5.3 5.3 5.0-5.8

Organic C 1.9 1.0-3.2 2.5 2.5-3.7 3.5 2.5-4.6

Total N .11 .05-.17 .21 .12-.31 .30 .22-.39

Avail. P 9 4-20 7 3-9 15 7-30

K+ meq .16 .10-.21 .16 .13-.19 .35 .19-.59

a WAP-weeks after planting


Journal for Farming Systems Research-Extension





SOIL-FERTILITY MAINTENANCE IN CAMEROON


also help avoid competition between tree stumps and crops (Jonsson et al.,
1988).
The choice between mounds and beds depends on tradition, field history,
and plant architecture. Beds are a highlander contribution to the province,
and many indigenous farmers in Kumba zone have adopted them as better
planting surfaces for groundnuts. Some Sands- and Volcanic-zone farmers
also use them, especially when planting groundnuts or leafy vegetables.
Mamfe farmers object, saying that they are not accustomed to the body
motions necessary to build and weed them. Egusi melon (Citullus lanatus),
a common Sands and Mamfe crop, adapts better to mounds than to beds,
creeping down to expand through the alleys and leaving the higher surfaces
to other crops. Finally, a field planted originally in mounds cannot easily be
converted to beds, or vice versa, because the original surface must be
destroyed completely and a new pattern erected. The difficulty of field-
pattern revision impedes the adoption of alley cropping, which presupposes
linear arrangements.
Residue management: Weed/crop-residue management at land prepara-
tion is of four types: clearing from the field, burning, surface mulching, and
incorporating in beds or mounds. Weeded residues are usually dumped in the
alleys. At harvest of early-maturing crops, the weeds and crop residues are
incorporated into the soil, which is molded over the remaining crops. At later
weedings, residues are left in the alleys for incorporation or burning at next
clearing.
Although farmers in the Volcanic and Sands zones admit the possibility of
residue incorporation during land preparation, only Kumba and Mamfe
farmers practice it (and only about half of these). Volcanic- and Sands-zone
farmers say that the weeds in their areas (Sorghum, Pennisetum, and Pueraria)
compete too strongly with the crops, resprouting from inside the mounds and,
in the case of Pennisetum, blocking tuber development. In Kumba, farmers
with the same weed species choose to incorporate them, with the intention of
improving fertility.
In Mamfe, the dominant weed, Chromolaena, must be well minced to rot
enough to allow good tuber formation. In the forested areas in Mamfe the
soil is more water-retentive, leading to quicker decomposition, and farmers
often choose to incorporate weed residues. Western Mamfe farmers burn
instead. Most farmers incorporate "soft grass" if Chromolaena has not taken
over the fallow and claim that before the advent of this weed, burning was rare.


Vol. 2, No. 1, 1991





ALMY ET AL.


Throughout the SWP, surface mulching is usually limited to mid-season
moulding. However, where a good soft-grass fallow has been established, it
is cut during the height of the rains and second-season crops are planted
through it, suppressing weeds. Vines are cleared out of the field. Clearing the
weeds from the field is a strategy for wet-season planting only, practiced by
those who fear resprouting. Often the cleared weeds are reserved for burning
after the rains stop.
Burning is practiced by almost all farmers before the first-season planting.
Even those who incorporate some weeds bur sticks, hard stems, noxious
weeds, and old crop residues. Sands- and Mamfe-zone farmers cut down all
the fallow growth in the midst of the dry season, let it dry in situ, and set fire
to the field. If the fallow weeds are not numerous they bum poorly, and the
farmer gathers the residue and resprouted grass into heaps and completes the
burning. Volcanic and Kumba farmers seldom have enough material to bum
and gather the weeds into heaps for the first bur. Volcanic, Sands, and
Kumba fields are often only partially covered with ash because of irregular
burning, and no attempt is made to spread the ash.
Station trials on burning versus incorporation are inconclusive with respect
to effects on fertility and crop production. In a five-year trial in the Cameroon
highlands (NCRE, 1989), burying residue gave significantly higher maize
yields in all years than burning it on the surface of the ridges. A four-year trial
in Sands and Kumba sites (Programme Pedologie, 1985, 1986) compared
maize, cassava, and groundnut yields under slash and bur plus "minimum"
Nitrogen-Phosphorous-Potassium (N-P-K) fertilization with total clearing
plus "medium" N-P-K. Significant differences between the treatments were
obtained only for cassava in years 3 and 4. In a trial by Faulkner and Doyne
in Ibadan, Nigeria (cited in Padwick, 1983), annual burning or incorporation
of a previous leguminous green-manure crop made little difference to maize
yields over a 5-year period; the incorporated plots contained more total N and
available P after 5 years, but had lower pH, CEC, and less soluble salts than
the burned plots.
The effects of burning and incorporation in TLU's nine-field sample were
also ambiguous. Four fields had been burned completely and the rest either
cleared or burned in isolated heaps. Soil analysis showed higher mound-to-
alley P levels in the burned fields than in the unburned ones, but insignificantly
higher mound-to-fallow levels. Levels of K were significantly higher in alleys
relative to fallows but insignificantly higher in mounds relative to fallows.
The farmers regard burning as important to create ash for manure for


Journal for Farming Systems Research-Extension





SOIL-FERTILITY MAINTENANCE IN CAMEROON


fruiting crops, melon, and okra especially, but defend it primarily on labor-
saving grounds. Not only does it dispose of the cut branches and grasses, but
itis the only efficient nonchemical wayknown to combat Chromolaena. Another
difficulty with alley-cropping introductions stems from cross-field burning,
which is incompatible with permanent intercrops.
Leguminous crops: Nutrient deficiencies are found in all zones except the
Volcanic, with N being the most critical. There are few locally occurring
leguminous plants that could be used in soil amelioration. The only one in
present use is groundnut (Arachis hypogaea), but random observations of
nodule color across the province showed that groundnuts at 4-to-6-leaf stage
were fixing little nitrogen. An indigenous leguminous vine, Pueraria, has been
used for weed suppression since early in the century in the rubber plantations
and invades fallows and fields across the zones, primarily in the less acidic soils
(pH>5.0). Although it produces larger nodules than groundnut, it also fixes
little nitrogen, as evidenced by the dull grey color of the nodules in young 4-
to-6-leaf, as well as older, vines. Farmers try to eradicate it because it entwines
and smothers the crops.
During the appraisal, TLU identified Mimosa invisa growing throughout
the province and producing large numbers of medium-sized, extremely pink
nodules in all soils types. Mimosa is regarded as a particularly noxious weed
because of its thorns, but a thornless variety (M. invisa var. inormis) is being
tested.
Intercropping with high densities of groundnuts is practiced by most
farmers in the Kumba and Sands zones, but few farmers surveyed regarded
groundnuts as a source of "manure." Groundnuts instead are used for weed
control and as an economic crop. Kumba and the central Sands section are the
only areas with rainfall under 2,500 mm, and groundnuts are grown in both
seasons. Farmers in Western Mamfe and high-rainfall Sands zones substitute
the nonleguminous egusi melon, which is also a weed suppressant and better
adapted to the rainfall pattern, in first season. From observation of groundnut
beds in low-fertility areas, the associated maize does not profit from the
nitrogen fixed in that season. At the end of the season, the residue is used in
moulding the tuber crops or in the following season's beds.
In summary, groundnut residues are used in ways that profit the develop-
ment of the slow-maturing tuber crops but not the more demanding early-
maturing crops. Nitrogen-fixation seems to be slight, but may reduce nutrient
depletion by tuber crops and thus strengthen fallow regrowth. No other local
leguminous food crop does well under the province's heavy rainfall, and all the


Vol. 2, No. 1, 1991





ALMY ET AL.


local leguminous mulch plants have a tendency to twine, forcing the farmers
to try to eradicate them. However, the concept of soil improvement through
weed growth and incorporation/mulching is widely accepted, and the as-
sociated practices add little to labor costs where mounding/bedding is
practiced.


DISCUSSION
The principal techniques for organic fertility enhancement now under de-
velopmentin West African research stations involve nutrient recycling through
alley cropping, improved fallows, or low-growing intercrops such as mucuna
(Hulugalle and Lal, 1986; IITA, 1987, 1990; Prinz, 1987; Hulugalle, 1988;
Rocheleau et al., 1988; Sumberg and Atta-Krah, 1988; NCRE, 1990). The
remainder of this paper considers how or if these technologies can be adapted
to farmers' systems in the South West.
Alley cropping involves the growing of food crops in wide alleys left
between permanent rows of leguminous trees, which are pruned for mulch
material. In the SWP, nutrient or space competition of short trees is not an
issue for the farmer. Rather, problems arise with reference to the labor
involved in destumping to make room for the new system, the shading of
growing crops, the disruption of the existing field pattern of mounds or beds
required by linear alleys, and the conflict with annual burning. Alley cropping
is a replacement system rather than a complete innovation. Thus trials should
be designed to test the efficiency of patterns using mounds or beds and to
compare the effects of those tree species already in farmers' fields with those
recommended for introduction. The most acceptable form of agroforestry
might abandon intercropping altogether in favor of small woodlots or borders
ofnitrogen-fixing species that could be harvested for mulch. This, however,
demands more labor than the alternative, nonpermanent green-manure crops.
Green manuring, or the cultivation and incorporation of improved fallow
crops, is criticized as an excessively labor-intensive technique for low-input
farmers, who cannot use herbicides and rototillers to kill and incorporate the
mulch (DAI/USAID/Yaounde, 1989). In the SWP, farmers who make
mounds or beds already put as much effort into land preparation as they would
under a green-manure system. Farmers know that dead vegetative matter
provides "manure" to their crops. Farmers also welcome cover crops that
suppress weeds during cultivation. After the short-season crops are harvested,
the empty spaces in most tuber-crop fields can be planted again with cover


Journal for Farming Systems Research-Extension





SOIL-FERTILITY MAINTENANCE IN CAMEROON


crops. IfIRA can provide a mulch crop that does not increase current clearing
and weeding costs and does produce better yields at least for the late-maturing
crops, many farmers will try it.
The most immediate advantage (with, therefore, the potential for the most
rapid and widespread adoption) would be fora mulch crop that could compete
against the most common fallow plants (Sorghum, Pennisetum and Chromo-
laena), yet would be easier to cut down and incorporate during land
preparation, and would not twine around food crops on resprouting. Can-
didates such as Mimosa and Crotelaria are now being screened at the IRA in
Ekona.
Incorporation methods used in trials will have to be based on the hoe,
rather than on chemical or mechanical aids, and trials will have to be
conducted in each zone, with an expectation that results will vary greatly. A
companion study must develop labor-cost estimates for incorporation and
burning. Further trials might be carried out on-station in low-fertility areas
to determine ifsurface mulching of mounds is sufficient to replace the annual
or seasonal requirement of rebuilding the mounds, a technique that would cut
land-preparation costs so drastically that it might revolutionize agriculture in
half the province.


ACKNOWLEDGMENTS

The authors would like to thank V. Balasubramanian, K. Dvorak, A-M. Izak,
and I.O. Adobundu of the International Institute of Tropical Agriculture for
their comments.


REFERENCES
Almy, S.W. 1988. Farming systems surveys of Meme Division, South West Province,
Republic of Cameroon. Technical report. Institute of Agricultural Research (IRA),
Ekona, Cameroon.
Almy, S.W., and M.T. Besong. 1987. Farming systems surveys of Fako Division, South
West Province, Republic of Cameroon. Technical report. IRA, Ekona, Cameroon.
Almy, S.W., and M.T. Besong. 1989a. Farming systems surveys of Manyu Division,
South West Province, Republic of Cameroon. Technical report. IRA, Ekona, Cam-
eroon.
Almy, S.W., and M.T. Besong. 1989b. Farming systems surveys ofNdian Division, South
West Province, Republic of Cameroon. Technical report. IRA, Ekona, Cameroon.
Balasubramanian, V., and A. Egli. 1986. The role of agroforestry in the farming systems
in Rwanda with particular reference to the BGM region. Agroforesty Systems 4:271-
290.


Vol. 2, No. 1, 1991






ALMY ET AL.


Chambers, R 1985. Shortcut methods of gathering social information for rural develop-
ment projects. Pages 399-415 in M.M. Cernea, ed., Putting people first: Sociological
variables in rural development. New York, N.Y.: Oxford University Press.
DAI/USAID/Yaounde. 1989. Mid-term evaluation of the National Cereals Research
and Extension Project Phase II (631-0052). Washington D.C.: Development Al-
ternatives, Inc.
Hartmans, E.H. 1981. Land development and management in tropical Africa. Rural
Africana 10:41-53.
Hildebrand, P.E. 1981. Combining disciplines in rapid appraisal: The sondeo approach.
Agricultural Administration 8:423-432.
Hulugalle, N.R. 1988. Effect ofcover crop on soil physical and chemical properties of an
Alfisol in the Sudan Savannah of Burkina Faso. Arid Soil Research and Rehabilitation
2:251-267.
Hulugalle, N.R., and R Lal. 1986. Root growth ofmaize in a compacted gravelly tropical
Alfisol as affected by rotation with a woody perennial. Field Crops Research 13:33-44.
International Institute of Tropical Agriculture (IITA). 1987. IITA annual report and
research highlights 1986. Ibadan, Nigeria.
IITA. 1990. IITA annual report 1988/89: Toward sustainable agriculture for Africa.
Ibadan, Nigeria.
Jonsson, K., L. Fidjeland, J.A. Maghembe, and P. Hogberg. 1988. Vertical distribution
of fine roots of five tree species and maize in Morogoro, Tanzania. Agroforestry Systems
6:63-69.
National Cereals Research and Extension Project (NCRE). 1989. Annual report 1988.
IRA, Yaounde, Cameroon.
NCRE. 1990. Annual report 1989. IRA, Yaounde, Cameroon.
Nowak, P.J. 1987. The adoption of agricultural conservation technologies: Economic
and diffusion explanations. Rural Sociology 52:208-220.
Nye, P.H., and D.J. Greenland. 1960. The soil under shifting cultivation. Technical
Communication No. 51. Commonwealth Bureau of Soils, Harpenden, England.
Padwick, G.W. 1983. Maintenance ofsoilfertilityin tropical Africa: A review. Experimental
Agriculture 19:293-310.
Prinz, D. 1987. Improved fallow. Ileia 3(1):4-7.
Prinz, D., and F. Rauch. 1987. The Bamenda model: Development of a sustainable land-
use system in the highlands of West Cameroon. Agroforestry Systems 4.
Programme Pedologie. 1985. Annual technical report 1984. Centre National des Sols,
IRA, Ekona, Cameroon. Pp. 16-18.
Programme Pedologie. 1986. Annual technical report 1985. Centre National des Sols,
IRA, Ekona, Cameroon. Pp. 9-11.
Rocheleau, D.,F. Weber, and A. Field-Juma. 1988. Agroforestry in dryland Africa. Pages
114-122 in International Council for Research on Agroforestry (ICRAF), ed., Science
and practice of agroforestry 3. ICRAF, Nairobi, Kenya.
Rogers, E.M. 1962. Diffusion ofinnovations. New York, N.Y.: Free Press.
Rogers, E.M., and F. Shoemaker. 1971. Communication ofinnovations: A cross cultural
approach. New York N.Y.: Free Press.
Ruthenberg, H. 1976. Farming systems in the tropics. Oxford, UK Oxford University
Press. 2nd edition.


Journal for Farming Systems Research-Extension






SOIL-FERTILITY MAINTENANCE IN CAMEROON


Sumberg, J.E., and A.N. Atta-Krah. 1988. Potential of alley farming in humid West
Africa-a re-evaluation. Agroforestry Systems 6:163-168.
Watson, G.A. 1983. Development of mixed tree and food crop systems in the humid
tropics: A response to population pressure and deforestation. Experimental Agri-
culture 19:311-332.


Vol. 2, No. 1, 1991








Increasing the Adoption Rates
of New Technologies With a
New Technology-Transfer Model'

Ramiro Ortiz and Adlai Meneses




INTRODUCTION
In the last two decades there has been an increasing effort within the
agricultural sectors of Third World countries to design technological models
and to identify research methodologies that would guide technological-
innovation institutions and programs. This is so these countries can generate
and validate production technologies appropriate to the needs and expecta-
tions of limited-resource and resource-poor farmers. This attitude has
produced substantial changes in the approaches of national and international
institutions, which have rapidly come closer to research models based on on-
farm activities and oriented toward the development ofappropriate technology.
International financing institutions and international centers have also been
involved in this effort, playing an important role in promoting this new
attitude by financing projects with the new approach and by training large
numbers of this new type of scientist.
Nevertheless, agricultural-extension institutions have been left out of this
effort, and the technology-transfer programs have become second in impor-
tance, one of the reasons being the belief that appropriate technologies
disseminate and are adopted very easily. This may be the case, but only up to
a point. The numbers reached in adoption still remain relatively low when
widespread, massive dissemination is expected; this is true, more often than
not, in areas where limited-resource farms are predominant and far too
numerous. To increase adoption rates it is necessary to identify other


1 Paper presented at the Ninth Annual Association for Farming Systems Research-
Extension Symposium, University of Arkansas, Fayetteville, October 9-11, 1989.
2Technical Consultant, Directorate General of Agricultural Services (DIGESA), and
Coordinator, Project for Crop-Livestock Technology Generation and Transfer, and
Seeds Production (PROGETTAPS), DIGESA, respectively.





ORTIZ AND MENESES


mechanisms, incorporate other institutions with wider coverage, and design
the models through which this technological infrastructure could reach a
much larger number offarmers, that is, in contrast to the ones reached by on-
farm research activities.
In the last decade, Guatemala began a very intensive effort to get better
adoption results by incorporating existing technical institutions into the work
conducted by on-farm researchers at the Institute for Agricultural Science and
Technology (ICTA). The best alternative was the agricultural-extension
institution, the Directorate General of Agricultural Services (DIGESA),
which has wide coverage and conducts its activity through extension agents
at the community level. This strategy worked very well, but the impact was
still not what was expected. Isolated successful cases in different parts of the
country (because this integration was not implemented in every place) showed
the potential this joint venture had in terms of reaching larger numbers of
farmers. This provided the higher authorities within the agricultural public
sector with the necessary evidence to establish a new project to formally
integrate these institutions into the generation and transfer of agricultural
technologies.
Since 1986, a new technology-generation-and-transfer project has been
implemented within the research and extension (RE) system in five regions of
the country, providing RE institutions with a formal opportunity to establish
integrated work on a large scale to achieve massive dissemination. This paper
details the main technological ingredients of that project and what has been
achieved in the first three years ofthis experience, emphasizing the contribution
of the agricultural-extension institution, DIGESA.


LINKING RESEARCH AND EXTENSION
In the second half of 1986, the Project for Crop-Livestock Technology
Generation and Transfer, and Seeds Production (PROGETTAPS) was estab-
lished and implemented to strengthen the research and extension system of
Guatemala (Fumagalli et al., 1985). This project began activities in five
regions (seven departamentos) of the country, where small, limited-resource
farmers comprise the largest part of the population. Overall goals were to
increase production and productivity of the existing farming systems that
generate mostly food crops (basic grains and vegetables).
Within this project, technology for crop production is generated and
validated by ICTA and transferred by extension agents of DIGESA. A new


Journal for Farming Systems Research-Extension





INCREASING ADOPTION RATES OF NEW TECHNOLOGIES 21

model for technology generation and transfer was also involved, and this is
what has made PROGETTAPS a project with outstanding results. The design
of this technological model and its implementation were based on the farming
systems research-extension (FSRE) methodology, since it is strongly based in
a true-existing linkage between on-farm researchers and extension agents.
These researchers and agents plan, conduct, and evaluate on-farm activities
jointly (farm trials, farmer-managed tests, and transfer plots), and promote
active participation of farmers in technology testing, adaptation, integration,
and transfer activities conducted on their own farms (Ortiz, 1987).
Before this project, ICTA developed a technological model with strong
emphasis on on-farm research that was successful in developing appropriate
technologies relevant to farmers' needs and easily integrated into their existing
systems (Fumagalli et al., 1985). ICTA's technologies were well-accepted and
adopted, but the adoption rates still remained below expectations. By
incorporating DIGESA with its wide coverage, the adoption rates could
increase dramatically. These were some of the ideas that gave rise to great
expectations, but this improvement was only possible if DIGESA could adopt
the new technological model. To achieve this, all extension agents went
through a "crack-down" course in FSRE methodology ("Technology Within
the Farming Systems Approach") that enabled them to understand the new
model, the technological system that operated for research, the contribution
that they would make in the joint effort, and the methodologies that they
would use to transfer new technologies. This course was offered in 1987, after
the first year of the project (1986), which used the traditional extension
methodology (demonstrative plot). The numbers, shown in Table 1, for
transfer plots and promotion activities (transfer days, farmer tours) in 1987
and 1988 show the tremendous increase achieved when DIGESA understood
its role and began making a substantial contribution to the effort.


THE NEW MODEL FOR TECHNOLOGY TRANSFER
In the design and execution of PROGETTAPS, the RE authorities and field
teams have adopted a new attitude by orienting their joint effort toward the
active participation of farmers in the different phases of the model, rather than
educatingthem. In other words, for technologyproduction in food crops, the
extension effort to "educate" farmers in the use of new technologies has been
replaced, for technology-transfer purposes, by joint participation of farmers,
extension agents, and on-farm researchers in all the phases of technology


Vol. 1, No. 2, 1990





ORTIZ AND MENESES


innovation (Ortiz, 1987). A brief discussion ofthe outstanding features in the
new transfer model follows (Ortiz, 1987).
1. Development of a strong and broad interface between research and
extension. This allows joint participation of both groups in the planning,
execution, and evaluation of RE activities. It also creates the opportunity for
extension agents to participate in on-farm research activities and to get to
know firsthand the characteristics and management of the new technologies.
On the other hand, on-farm researchers also have the opportunity to get
involved in extension activities (transfer plots and transfer days) in order to get
direct feedback from farmers on the performance of new technology and to
bring a technical/scientific support to this effort (Figure 1).
2. The direct involvement of farmers in all phases of the technology-
innovation process (TIP). This requires the active participation, rather than


Table 1. Activities Conducted by the Research and Extension
Integrated Effort in Guatemala, PROGETTAPS, 1986-88a

Number of each type of activity by year


Type of activity 1986 1987 1988 1986-88 1989

1. Farm trials 193 99 242 534 422

2. Farmer-managed tests 274 724 248 1,246 368

3. Transfer plots 506 2,876 2,547 5,929 4,630

4. Seed plots 11 719 730 1,722

5. Communal gardens 445 445 898
(vegetables)

6. Transfer days 23 122 678 823 1,241

7. Farmers' tours 11 76 275 362 415

8. Agricultural encounters 13 63 267 343 342

9. Farm records 367 4,580 2,735 7,315 12,500

a From 1986 to 1988, only seven departamentoswere covered by PROGETTAPS. There
are 21 departamentoswhere DIGESA has activities, out of a total of 22 in the country.
In 1989, PROGETTAPS activities were extended to a total of 15 departamentos.


Journal for Farming Systems Research-Extension






INCREASING ADOPTION RATES OF NEW TECHNOLOGIES 23

education, of farmers. This is probably the feature that needs to be stressed
the most in order to objectively guide the design and execution of the RE work
plans. Besides having farmers participate in field trials and transfer plots and
be the main speakers on the transfer days, they have been incorporated into
decision-making on what the annual RE work plans will be. This is done
through a methodology known as "consultative groups" (grupos de consult;
Ortiz, 1988), which is a systematic procedure to obtain information from the
rural population, thus promoting their participation in making decisions more
relevant to their needs and problems in terms of RE activities.
3. Participation of rural leaders generating a "multiplying effect" in
technology transfer. These rural leaders have been hired by the Ministry of
Agriculture to work part-time as representantes agricolas (RAs); that is, they
are representatives of the agricultural public sector (APS) and serve as linkages
between their communities and the institutions in the APS. These people have
been selected by their communities based on their leadership qualities, their
ability as farmers, and their spirit of service to the community; they are well-





World Stock of Agricultural
Science and Technology





Technology Tecolo Te g Tc Technology Diffusicui
RESEARCH Generation Testing Adaptation Integration DisseminatI &


I.SCIENCE P--- Technology Development ---
Research Organization -1 ---------....----

----------..........Extension Organization I-
---------4---r- Research & Extension Interface --- .--------4
Adapted frcn: Reseach and Ext~nion(Emph=izing arming Systems Research and Extenaio) Working
Drat No.3. Prming Sysm Support Project Univawity of Florid Ginsville. February 1985.


Figure 1. The Technology Innovation Process Within the
Agricultural Research-Extension System of Guatemala


Vol. 1, No. 2, 1990





ORTIZ AND MENESES


known and respected by their peers. The RAs constitute a key element in the
transfer process: they conduct some transfer plots and transfer activities with
groups offarmers they have formed, and they are the main speakers, along with
cooperating farmers, on the transfer days.
This feature has had great success in terms of achieving wider dissemination
and increased adoption rates because the rural leaders have (1) the "know-
how" to get the message across in a clear and appropriate language, (2) a well-
established credibility in their communities and with their groups, and (3) a
great sense of responsibility in executing their work.
4. A stronger emphasis on transfer and promotion of new technology
rather than providing technical assistance to a reduced clientele. This is
possible within this project, because in terms of food crops, the technologies
selected in the transfer phase are simple (easy to understand), profitable (with
low investment), and have high acceptability indexes.
The decision to more strongly emphasize transfer and promotion activities
is based on the fact that the technology-transfer approach allows an easier
working strategy with the groups organized around the RAs, thus generating
a large "multiplying effect," with wide coverage and dissemination. This far
surpasses the expectations that produced the technical assistance approach.
5. Looking for a sharper increase in the adoption rates, DIGESA has
entered the field of facilitation of new technologies by developing within and
among small, limited-resource farmers a nonconventional seed-production-
and-distribution system for the improved varieties that reach the transfer
phase (Ortiz, 1989). This decision originated from the fact that new and
improved materials of food crops were adapting well to farmers' circumstances,
the seed was demanded by large groups of limited-resource farmers in the
areas where PROGETTAPS was conducting RE activities, and no seed
industry existed to cover those demands. Here, the extension agents work
closely with farmers who have been chosen as seed producers, taking into
account their ability as farmers and their slightly better economic status, which
permits them to invest a little more than what they put into their subsistence
crop. Through barter and monetary transaction, this seed is reaching large
numbers of farmers (Ortiz, 1989).


ORGANIZATION AND OPERATIVE METHODOLOGY
Within this new project, ICTA and DIGESA are jointly responsible for the
organization and operative methodology of a modular system for technology


Journal for Farming Systems Research-Extension





INCREASING ADOPTION RATES OF NEW TECHNOLOGIES 25

transfer (MSTT). There is one of these in each of the five regions covered by
the project, and its activities are coordinated by one representative of ICTA
and one from DIGESA; in the case ofICTA, it is the leader of the technology
testing and transfer team (on-farm researchers), and in the case of DIGESA
it is a university graduate with experience in extension. The MSTT has a total
of four on-farm researchers (university graduates in agronomy) and to each
one of these is assigned a minimum of three to a maximum of seven extension
agents (agronomists at high-school level), depending on the number of
extension agencies within the area covered by the researcher. Each one of the
extension agents is assigned an average of 10 to 15 RAs, and each one of these
has at least one organized group of 20 farmers. The expected coverage-the
number of farmers involved in the transfer activities-is far larger than
originally planned, because each MSTT reaches more than the expected 2,400
farmers each year.
The flowchart of PROGETTAPS' technological model (Figure 2) shows
the different RE stages where ICTA and DIGESA are sharing responsibilities.
The first two stages, technology generation and technology testing (valida-
tion), are ICTA's responsibility, but the extension agents get involved in
them, beginning with the farm trials (generation) and more in the farmers'
tests (validation). Here the extension agents come in contact with new
technology as it develops, participate in the evaluation of its performance, and


Public and
Private Agro-Socioeconomic and Cultural Information
Sectors





T_ Technology Technoloy Pmomotion and
Generation Yes Testingand Yes Dissemination Yes
Bank Validation --
of Farm Trials Transference Rural
Technology Fanner Plots P t Population
Appropriate Acceptability Mullying
Technology Evaluation ff

No No

Feedback


Figure 2. Flowchart of PROGETTAPS' Technological Model


Vol. 1, No. 2, 1990





ORTIZ AND MENESES


enthusiastically make it their own, given the credibility and the knowledge of
how it performs and is managed. Then, in the next stage (technology
transfer), DIGESA's agents take charge, with ICTA's on-farm researchers also
participating, to get direct feedback from the farmers and extension agents,
and to technically and scientifically support those activities (transfer plots and
transfer days).


TECHNOLOGY-TRANSFER ACTIVITIES

As shown in Table 1, from 1986 to 1988 DIGESA worked with farmers in
operating a total of 5,929 transfer plots, 719 seed plots, and 445 communal
vegetable gardens (huertos comunales). This served to train groups of small,
limited-resource farmers in the diversification of their farming systems for
profitability and nutrition purposes. DIGESA's extension agents counted on
the technical/scientific support of ICTA's technology-testing-and-transfer
teams (on-farm researchers), who also conducted on-farm activities of their
own: 534 on-farm trials and 1,246 farmer-managed tests. All of these on-farm
transfer activities required the effort of 72 extension agencies (92 extension
agents) and 18 on-farm researchers, as well as the backup of approximately
800 RAs who, along with cooperating farmers, participated as main speakers
or communicators on 823 transfer days, 362 farmers' tours, and 343 agricultural
encounters. DIGESA also developed 7,315 farm records with cooperating
farmers to establish the costs and profitability of the traditional and the new
technologies, which along with sondeos in each of the five regions helped
characterize the target populations and their farming systems.
A monitoring and evaluation group within the project has estimated, based
on their sampling of population in all the five regions, that there exists a ratio
of 8 to 10 adopters of new technology for each transfer plot operated. This
results in a total of over 50,000 farmers who are using new technology
transferred within PROGETTAPS. There are also records with the names of
21,045 farmers who obtained new seed varieties within the nonconventional
seed-production-and-distribution system established by DIGESA. With
these two types of adoption, the total number of farmers with new technol-
ogies reaches close to 80,000; this is, in only three years, twice what the project
goals are for four years (40,000 farmers). The adoption of those new
technologies was projected to produce in 1989 an estimated increase of


Journal for Farming Systems Research-Extension





INCREASING ADOPTION RATES OF NEW TECHNOLOGIES 27

16,200 metric tons in food crops in the areas covered by the project, which
is enough food to cover the annual needs of slightly over 26,000 additional,
typical rural families.
The success achieved by this integrated effort of farmers, RAs, researchers,
and extension agents has encouraged the RE institutions to expand this
transfer model in 1989 to eight more departamentos, making a total of 15 out
of the 21 nationwide where DIGESA conducts activities. The numbers of RE
on-farm activities being conducted by 149 extension-agency teams and 40 on-
farm researchers, also are shown in Table 1. In this effort, a total of 1,884 RAs
are also participating with their organized groups, and it is estimated that close
to 100,00 farmers will be adopting and using new technologies in 1990.


SUMMARY

The involvement of the national agricultural-extension institution, DIGESA,
in support of on-farm research conducted by ICTA has resulted in massive
dissemination of new technologies and has provided the agricultural public
sector of Guatemala with the necessary evidence to establish an integrated RE
effort in two-thirds of the country. This is in order to achieve higher
production levels of food crops, as a direct result of higher adoption rates of
new technology. A new technology-transfer model, where research and
extension interface, has been established, and it has become effective by
incorporating rural leaders (RAs) and the groups of farmers that they have
organized. It was also necessary to change the attitude of the extension agents
by training them in farming systems research-extension methodology and to
emphasize the promotion of new technologies to produce a wider and massive
dissemination.



REFERENCES


Fumagalli, A., R. Ortiz, and M. Castillo. 1985. A new model for technology transfer
within FSRE. Paper presented at the Farming Systems Research-Extension Manage-
ment and Methodology Symposium, Kansas State University, Manhattan, October
13-16, 1985.
Ortiz, R. 1987. Transferencia de tecnologia en Guatemala. Paper presented at the
Seminar-Forum Mecanismos de Investigaci6n y Transferencia y sus Implicaciones en


Vol. 1, No. 2, 1990






ORTIZ AND MENESES


la Adopci6n de Tecnologfas, Proceedings of the XXXIII Annual Reunion of the
Cooperative Central American Program for Crops and livestock Improvement
(PCCMCA), Guatemala, April 1-4, 1987.
Ortiz, R. 1988. Los grupos de consult: Una estrategia para incorporar agricultores al
process de innovac6n tecnol6gica. Proyecto de Generaci6n y Transferencia de
Tecnologfa Agropecuaria y Producci6n de Semillas (PROGETTAPS). Guatemala:
Direcci6n General de Servicios Agricolas (DIGESA). Mimeograph.
Ortiz, R 1989. Developing anon-conventionalseedproductionanddistributionsystem
for limited-resource farmers in Guatemala. Paper presented at the Association for
Farming Systems Research-Extension Symposium, University ofArkansas,Payetteville,
October 9-11, 1989.


Journal for Farming Systems Research-Extension








Characteristics of Improved Technologies
That Affect Their Adoption in the Semiarid
Tropics of Eastern Kenya

A.P. Ockwell, L. Muhammad, S. Nguluu, K.A. Parton, R.K. Jones,
and R.L. McCown'



INTRODUCTION
Over recent years, increasing recognition has been given to the role that on-
farm research can play in the development of technologies for smallholder
agriculture (Byerlee et al., 1982; Harrington and Tripp, 1984; Ashby, 1986).
The linkage between on-station research and on-farm research has received
greater prominence in the development process through the widening accep-
tance of farming systems research as an appropriate methodology for identi-
fying problems for technical research (Collinson, 1981). One important
component of the overall approach has been the emphasis on farmer partic-
ipation in on-farm trials of emerging technologies (Farrington, 1988).
Another useful component of on-farm research has been researcher-managed
experiments on farmers' fields, which provide a broader base ofenvironmental
conditions (e.g. soil fertility, rainfall, etc.) than those offered at the research
station (Collinson, 1982). These on-farm approaches should provide oppor-
tunities for researchers to better understand the reasons for the divergence
between farm and station yields (Ghodake and Walker, 1982).
On-farm trials of technologies from the research programs of the National
Dryland Farming Research Centre (NDFRC), Katumani, in Eastern Kenya
commenced during the short-rain season of October 1980 and continued
until March 1982. The improved techniques of production in these trials

1Formerly of KARI/Australian Centre for International Agricultural Research (ACIAR)
Dryland Project, Kenya, and now at the Bureau of Transport and Communications
Economics, Canberra, Australia; Department of Agricultural Economics and Business
Management, University of New England, Armidale, Australia; KARI, National Dry-
land Farming Research Centre, Machakos, Kenya; Department of Agricultural Eco-
nomics and Business Management, University of New England, Armidale, Australia;
CSIRO, Cunningham Laboratory, St. Lucia, Australia; and CSIRO, Davies Laboratory,
Townsville, Australia.





OCKWELL ET AL.


included both those that had been in use by some farmers for many years (e.g.
the use of farmyard manure, dry-season ploughing, and intercropping) and
others recently developed by the NDFRC.
As part of the postproject evaluation of these on-farm trials, the extent of
adoption of the new technologies was evaluated at the farms where the trials
were held. A case-study approach (Doorman, 1990) was used to define clearly
the technologies in use and to provide an insight into the characteristics and
extent of the problem of nonadoption. This paper discusses the features
endemic to the technologies that affected their rate of adoption. In contrast,
a related paper (Ockwell et al., 1990) examines those features endemic to farm
households that influence adoption. Both papers use data from the same set
of 16 farms involved in the on-farm trials. In addition, the current paper
provides a contrast by including two more farms not influenced by the on-farm
trials.


BACKGROUND

As shown in Figure 1, the pattern of farm activities follows the bimodal
distribution of rainfall. Although seasons are extremely variable and crop
failures occur frequently, an average rainfall year permits two cropping
seasons. Early maturing food crops, such as maize, beans, cowpeas, green
gram, and millet, are grown during both seasons, whereas late-maturing
crops, such as pigeon pea, are generally sown at the beginning of the short
rains (October-December) and harvested at the end of the long rains (March-
July). The major nonfood crops in the region are cotton, tobacco, and
sunflower.
Eighteen farmers participated in on-farm Pre-Extension Trials (PETs).
They included seven farmers in the Machakos, five in the Kitui, three in the
Embu, and three in the Meru Districts. The selection of farmers was based on
advice from local extension officers according to criteria specified by the
Farming Systems Section at the NDFRC (Bakhtri et al., undated). The
farmers who were included in the trials were considered typical or represen-
tative of those in the region.
Technologies were disseminated to participating farmers over several
seasons, commencing with the short rains of October 1980 and concluding
with the long rains of March 1982, so that different technologies were tested
in different seasons. Crop technologies included early land preparation, early
planting of crops (i.e., dry planting), depth of planting, correct spacing of


Journal for Farming Systems Research-Extension






TECHNOLOGY ADOPTION IN EASTERN KENYA


50






25





0











S 1.0






0.5






0.0


Aug.


Nov. Feb.


May Aug. Nov. Feb.
Figure 1.
(A) Rainfall Distribution (From Jaetzold and Schmidt, 1983)
(B) Distribution of Labor Requirements
Makueni
Abbreviations: LR long rains; leg legumes; cer cereals; pip = pigeon pea;
FMY farmyard manure; and SR short rains.


Vol. 2, No. 1, 1991





OCKWELL ET AL.


crops (i.e. ranging from 65 cm to 80 cm for pulses and 75 cm to 90 cm for
cereals), optimal plant populations (e.g. 40,000 plants per ha for maize),
intercropping, use of manure, use of fertilizer, use of improved crop varieties,
chemical pest control, and weeding by the use of oxen-drawn implements.
Improved crop varieties that were tested on farmers' fields were Katumani
Composite B maize (KCB), Katumani 80 (K80) and Machakos 66 (M66)
varieties of cowpea, line 26 green gram, early maturing pigeon pea (lines 422,
423 and 109), Sorghum IS76 and ISB8595 and Serena sorghum, and bulrush
millet Katumani PM-I. Farmers were provided with free packets of seed of
each variety.
The technologies directly relevant to livestock that were introduced to the
participating farmers included improved grass/legume pastures and improving
the quality of livestock by replacement with cross-bred animals. In addition
to the goal of upgrading fodder resources, different grass species were tested
for their effectiveness in stabilizing terrace banks against water erosion.
Rekindled interest in terraced banks in the 1970s followed a period of
dormancy relative to the emphasis given to terracing during the colonial
period of the 1950s (O'Leary, 1984).


METHOD
Technologies that had been fully and partly adopted by the case-study farmers
were recorded and their characteristics were examined. Next, the technologies
used by PET and non-PET farms were compared. Finally, the effectiveness
of application of the new techniques was assessed by closely monitoring farm
practices and yields during the 1986 short rains.
Contact with the 18 farmers participating in the PETs concluded in
December 1982 at the end of the short rains. In order to complete the first
step of investigating the long-term use by farmers of the technologies
disseminated under that program, particularly without further research and
extension input, contact with 16 (of the 18) farmers who had remained on the
same farms was resumed during the long rains of 1986 and continued through
visits at intervals of six weeks over a 15-month period. A formal questionnaire
was used to record the observed farming practices. In addition, informal
discussions were held with farmers to establish their reasons for use or non-
use of technologies. Emphasis was given to the question of nonadoption,
providing important feedback information to the scientific research staff who
were responsible for developing the technologies.


Journal for Farming Systems Research-Extension





TECHNOLOGY ADOPTION IN EASTERN KENYA


To complete step 2, data were collected from two more farms. By
contrasting them with the PET farms, they were used as "controls" to show
some of the influences ofon-farm trials. These farmers were chosen at random
according to the criteria used in the original selection of the pre-extension
farms and hence were considered representative of farms in the study region.
Neither farmer had previous contact with research staff, nor had they
participated in any on-farm trials.
Next, the yields on farmers' fields in a particular season were documented.
A series of plots was laid out on the case-study farms in January 1987 (i.e., not
long before the commencement of harvesting operations). [The approach of
marking out the plots at this late stage prevented the farmer from varying the
management of these areas, so the results would represent actual yields from
farmers' fields under normal levels of management.]
A 5-m by 5-m plot was used for all sole-cropping and intercropping
arrangements, whereas a 10-m by 10-m plot was used for mixed cropping
arrangements where there was no definite row structure. A total of 134 plots
was established on the 18 case farms, with the number of plots per farm varying
according to the range of crops grown, the types of crop arrangement, the
use/nonuse of manure and/or fertilizer, and the planting date. There were
two or three replications of each treatment. Plant populations were determined
when the yield plots were marked out. Harvesting was done by the farmers.
Fresh-weight yields for each crop were taken in the field by the researchers on
a plot basis. Subsamples of each crop on each farm were then collected for dry-
weight analysis.


RESULTS
Four topics are discussed in this section: (1) use of improved techniques of
production on the 16 PET farms, (2) adoption on the two control farms, (3)
yields during the 1986 short-rain season, and (4) additional considerations.
Summary statistics on the principal farm characteristics of the 18 case-study
farms are presented in Table 1, in which farms 17 and 18 are the non-PET
farms. There was considerable variation in the size, crop area, grazing area,
and number oflivestock on each farm, as well as stocking rate per ha of grazing
land. Although common grazing land is not included in the table and
comparison between the last two columns exaggerates the degree to which
carrying capacity is exceeded, it was observed nevertheless that pressure on
grazing land resulted in an overall erosion problem. More detailed informa-


Vol. 2, No. 1, 1991






OCKWELL ET AL.


Table 1. Principal Farm Characteristics of Case-Study Farms

Farm Loca Farm Crop Grazing Livestock Technical
No. size area area Capacity
(ha) (ha) (ha) Oxen Cattle Goats Sheep LSUb LSU/ (LSU/ha)c
graz ha

1 MWA 7.2 4.0 3.2 2 15 8 0 13.4 4.2 0.25-0.40
2 MWA 8.0 3.5 4.5 3 6 3 2 7.2 1.6 0.25-0.40
3 WAM 16.0 3.5 12.5 2 6 10 0 7.3 0.6 0.20-0.35
4 MAK 6.4 3.2 3.2 2 10 6 2 9.8 3.1 0.50-1.00
5 MAK 8.4 2.4 6.0 4 6 6 0 8.1 1.4 0.50-1.00
6 MAK 6.0 4.0 2.0 2 9 10 0 9.4 4.7 0.20-0.35
7 KIT 32.0 6.5 25.5 4 8 20 5 12.5 0.5 0.50-1.00
8 MUT 8.9 4.9 4.0 4 1 10 2 5.5 1.4 0.25-0.50
9 MUT 25.0 10.0 15.0 3 8 25 0 11.8 0.8 0.25-0.50
10 MUT 3.0 3.0 0.0 0 0 0 0 0.0 0.18-0.30
11 EMB 3.0 1.0 2.0 0 0 6 0 1.0 0.5 0.67-1.25
12 EMB 1.2 1.2 0.0 0 2 0 0 1.4 0.67-1.25
13 EMB 18.6 3.3 15.3 0 15 72 32 27.2 1.7 < 0.33
14 MER 11.7 2.7 9.0 2 2 30 13 9.7 1.1 < 0.33
15 MER 5.0 3.0 2.0 1 7 46 2 13.3 6.7 0.33-0.83
16 MER 15.6 5.8 9.8 3 7 4 0 7.8 1.2 0.33-0.38
17 WAM 7.0 3.0 4.0 2 2 14 5 5.9 1.5 0.25-0.40
18 MUT 7.0 3.0 4.0 2 13 10 0 12.3 3.1 0.30-0.55

a Locations (Loc) are Mwala (MWA), Wamunyu (WAM), Makueni (MAK), Kitui (KIT),
Mutomo (MUT), Embu (EMB), and Meru (MER).
Estimates of livestock units (LSU) were derived from conversion factors used by
Tessema et al. (1985).
SEstimates of technical carrying capacity are derived form Jactzold and Schmidt (1983).


tion on the farming systems of these farms is contained in Ockwell et al.
(1990).

Improved Techniques on the PET Farms
The improved techniques were categorized into 16 types, and the farms
were ranked according to the number of techniques adopted (Table 2; see
Ockwell et al., 1990, for a description of the ranking process). The use or
adoption of a technology ("A" on Table 2) was taken to indicate full
commitment and regular application.
The various techniques are classified as (a) noncash using, (b) low-risk cash
using, and (c) high-risk cash using. The first category includes early land
preparation, dry planting, terracing, and farmyard manure. Although the last


Journal for Farming Systems Research-Extension






TECHNOLOGY ADOPTION IN EASTERN KENYA 35

Table 2. Levels of Technology Adoption

Type of technology
Farm No.a Noncash using Low-risk cash using
FYMb TER EPT ELP FOD OXW SCM

7 Ac A A A A A A
16 A A A A A A A
1 A A A A A
8 A A A A A A
12 A A A A A A
4 A A A A A
9 A A A A A A
14 A A
13 A A A
17 A A A A
3 A A A A
5 A A A A
6 A A A
15 A A A
2 A A A
10 A
11 A
18
High-risk cash using

KCB MMB CWP SOR PIP FCM FTZ MLT GRG

7 A A A A A A A
16 A A A A A
1 A A A A A
8 A A A A
12 A A A
4 A
9 A
14 A A A A
13 A A
17 A
3 A
5 A
6 A
15 A
2
10
11
18

a Farms listed in descending order of adoption index
b FYM-farmyard manure, TER-terraces, EPT-early planting, ELP-early land
preparation, FOD-improved fodder grass, OXW-oxplough weeding, SCM-crop
storage chemical, KCB-Katumani composite B maize, MMB-Mwezi Moja bean,
CWP-cowpea, SOR-sorghum, PIP-pigeon pea, FCM-crop protection (field)
chemical, FTZ-fertilizer, MLT-millet, GRG-green gram.
c A-adoption.
Vol. 2, No. 1, 1991





OCKWELL ET AL.


two techniques in this category are labor-intensive, they have the potential to
result in a more efficient use oflabor. The second category includes livestock
fodder, crop storage chemicals, and oxplough weeding. The high-risk cash
using group is composed of fertilizer, improved crop varieties, and field-
protection chemicals.
As shown in Table 2, the techniques with the highest level of use by farmers
were terracing and farmyard manure, followed by dry planting of crops.
Terracing and farmyard manure are complementary techniques for improved
soil and water management, which place high demands on available labor but
do not necessarily require cash for their implementation.
The next most-used techniques were Katumani Composite B maize,
livestock fodder, oxplough weeding, improved bean varieties, and early land
preparation. Where farmers owned oxen and tillage equipment, oxplough
weeding, early land preparation, and dry planting placed no demands on the
cash resources of the farm. The benefit that flowed from their use was the
potential to smooth out the peak demand for labor at planting and weeding
times and to improve the timeliness of cropping activities.
Those techniques that had a low level of use by farmers included fertilizers
and field-protection chemicals. Although a relatively limited extension effort
may partly explain this, both of these techniques required cash for imple-
mentation and were characterized by a high degree of risk. Fertilizer,
particularly di-ammonium phosphate, was regarded by farmers as a seasonal
item of cash expenditure with a low probability of return in a high-risk
production environment.
Overall, farmer behavior is risk averse and constrained by cash requirements.
A ranking of the techniques from highest to lowest use reveals that, except for
maize and beans, the noncash using techniques have been adopted by many
farmers, whereas the low-risk cash using techniques have an intermediate level
of adoption and the high-risk cash using techniques are hardly used. The
significant departures from this trend were the wide adoption of Katumani
Composite B maize and, to a lesser extent, improved bean varieties. Two
reasons are advanced for this. First, there was a complementary relationship
between dry planting of maize and the use of "fresh" Katumani maize seed
because of the chemical treatment of maize that allowed it to remain in the soil
for longer periods of time without risking deterioration before germination.
Second, there was a dietary preference for maize that may have stimulated early
interest in a higher-yielding variety. Improved bean varieties also realized a
high rate of use among farmers, perhaps for a similar reason.


Journal for Farming Systems Research-Extension






TECHNOLOGY ADOPTION IN EASTERN KENYA


These adoption levels were observed for maize and beans despite farmers'
complaints about the high cost of seed (a 10-kg packet of the Katumani
Composite B maize costs KShs70 ($US 4.25) from the local cooperatives,
whereas Mwezi moja bean costs KShsl50 ($US 9.10) for 10 kg). Adoption
also had proceeded despite some of the risks perceived by farmers who
advanced arguments related to maize such as "it does not do as well as the local
variety under high rainfall conditions; it does not store as well as the local
variety;" and "it is more susceptible to insect damage." With improved
varieties of beans, the concern was with their level of drought tolerance and
susceptibility to flower shedding under conditions of moisture stress.
Figure 2 shows the results of a cluster analysis across the improved
techniques. It confirms the complimentary relationships observed above
between maize and early land preparation, with storage using chemicals being
an associated technology. [This dendrogram was obtained with a Pearson
distance metric and a furthest-neighbor linkage (Anderberg, 1973; Everitt,
1980), using the SYSTAT program (Wilkinson, 1987). Similar dendrograms
resulted from the use of a number of alternative distance and linkage
methods.]



Early planting
Pigeon pea
Fertilizer
Green gram
Crop-protection (field)
chemical
Cowpeas
Katumani composite B
maize
Early land preparation
Crop-storage chemical
Mwezi Moja bean
Farmyard manure
Millet
Sorghum
Improved fodder grasses
Terraces
Oxplough weeding

0.0 1.0 2.0



Figure 2. Dendrogram Showing Associations Between Improved Farming Practices,
Using Pearson Distance Metric and Furthest-Neighbor Linkage

Vol. 2, No. 1, 1991





OCKWELL ET AL.


Examination of producers' opinions about other improved crop varieties
reveals further components of perceived production riskiness. For example,
those farmers who had grown improved cowpeas stated that they were more
susceptible to pest damage, particularly from apion weevil. The NDFRC had
recommended that crop-protection chemicals should be applied to reduce
this problem, and, as shown in Figure 2, there is an association between
adoption of improved cowpeas and use of chemicals. However, discussion
with farmers revealed that they did not adhere closely to the recommendation
because of the high cost of chemicals. This may be a reason why other farmers
did not adopt improved varieties of cowpeas. Insect pest problems were also
reported for improved varieties of pigeon pea.
It was clear from the interviews with farmers that production risk was an
important consideration in the adoption process of improved varieties and an
issue to be given serious consideration by scientists and extension personnel.
Despite the fact that a majority of the crop types grown by the 16 farmers
included some component of improved varieties, crop yields were generally
low to very low (see the section below in which yields are discussed). It could
reasonably be argued that none of the farmers had fully adopted the package
of inputs necessary for producing high-yielding crops. An important reason
for this was a basic lack of cash. In addition, there were many partial-users who
grew a combination of improved and local varieties because they preferred to
reduce the risk associated with growing only one variety. Some of these
farmers argued that they recognized the differential response of crop varieties
to seasonal conditions.
Disregarding improved crop varieties and their associated technologies,
technologies were in common use across most farms. This does not mean,
however, that these technologies need no further development. For instance,
most of the farmers had extended their area ofimproved grasses. Nevertheless,
grazing lands on almost every farm visited showed signs of sustained overgrazing,
which resulted in severe depletion of ground cover and in gully formation.
Although past mismanagement had obviously contributed to the current
status of grazing lands, there appeared to be few alternative technologies for
grazing lands to offer farmers.

Non-PET Farms
In an attempt to observe a group that had no contact with the PETs, two
additional farms (numbers 17 and 18) were included in the survey. These two
control farms were different from each other in their management and in their


Journal for Farming Systems Research-Extension





TECHNOLOGY ADOPTION IN EASTERN KENYA


responses to improved techniques of production.
The owner of farm 17 recognized the existence of problems such as low soil
fertility, erosion, overgrazing of natural pasture land, and the effect of pests
on crop yields and acknowledged the role of improved techniques ofproduc-
tion in easing the impacts of those problems on farm performance. However,
the farmer's ability to adopt improved techniques of production, such as
fertilizer, were severely constrained because the farm supported a large
number of dependants and had no sources of off-farm income (see Table 1).
Despite this, the farming system observed on farm 17 was quite innovative in
terms of its response to environmental constraints. Most of the cropland was
terraced, although corrective action was needed in places to preserve cropland
exposed to gully erosion where terrace banks had been damaged. The farmer
distributed farmyard manure to two of the ten terraces during the long dry
season and then proceeded to dry-plough those terraces in preparation for
timely planting at the beginning of the short-rain season. The cropping
arrangement consisted of maize and cowpeas interplanted within the crop
row, with a reasonable 90-cm spacing between rows to allow for oxplough
weeding. Limited use was made of improved crop varieties because funds were
not available to purchase the seeds. (Most farmers save their own seed from
season to season.) However, the farmer also stated that the local variety of
cowpea was preferred for its ability to provide edible vegetable material over
a longer time period. In conducting on-farm experiments on this farm at a
later stage in the project, it was clear that the farmer was not averse to trying
alternative techniques ofproduction. Indeed, this farm resembled some of the
medium-level adopters in the PET category. It ranks at position 10 in terms
of number of practices adopted (see Table 2).
In contrast, the owner of farm 18 stated that there were no problems
associated with crop and livestock production and, hence, that there was no
need to adopt alternative techniques of production. The farm supported few
dependents (i.e. two children under 3 years) and a family member was engaged
in permanent off-farm work (earning approximately KShsl,250 a month).
None of the innovations under examination had been adopted. The farm was
not terraced and trash lining on the contour was not used, with the result that
signs of severe gully erosion were evident on the cropland. Despite a
substantial build-up of manure in the boma, manure was not distributed to the
cropland because the farmer considered it inappropriate to apply manure to
unterraced cropland. Crops were plough-planted at onset, with an inter-
cropping arrangement of traditional varieties of maize, beans, cowpeas, and


Vol. 2, No. 1, 1991





OCKWELL ET AL.


green gram, which indicated that oxplough weeding was not under consider-
ation. There was no use made of improved crop varieties for the stated reason
that there was no advantage to be gained from buying seeds of improved
varieties when there was an ample supply of local varieties on-hand. Although
the farmer stated that severe cash and labor constraints mitigated against him
using improved techniques of production, these constraints did not appear to
be as binding as on farm 17. It was clear that his perception that there were
no problems on the farm was most influential in his not adopting any of the
improved practices under consideration. In response to the erosion evident
on his cropland, the farmer argued that if it became serious he would clear
more land for cropping.

Yields for the 1986 Short-Rain Season
The results presented in Table 3 are dry-weight yields for common crop
combinations for the case farms in relatively homogeneous areas. For
example, three of the four case-study farms in the locations of Mwala and
Wamunyu had grown a common intercrop combination of maize-cowpea-
pigeon pea over the short-rain season of 1986, so their ranges and means are
shown.
A striking feature of the results is the low crop yields obtained by farmers,
in what was considered by farmers and researchers alike to be a fairly average
season. Two factors that contributed to the low maize yields and which were
evident during the growing season were the low and highly variable plant
populations within terraces and low rates of cob set per plant. Overall, the best
sole-crop maize yield for the short-rain season was 2,000 kg per ha, which was
achieved by farm 7 for a dry-planted crop using farmyard manure. Maize yields
achieved at the NDFRC for the same season were between 4,000 and 6,000
kg per ha, depending on the input of fertilizer. For maize/legume intercrops,
the average yield of maize over 11 farms in the sample was about 300 kg per
ha (see Table 3).
Several other observations were made while documenting yields. First,
there appeared to be a strong yield response to the application of farmyard
manure. Although improved timeliness of planting and weeding often
confounded the effect of farmyard manure, marked differences in crop
performance were found across terraces where the only variable was manure
application. Second, manure was generally applied to the preferred crops,
such as maize and beans. These crops were usually planted as close to onset


Journal for Farming Systems Research-Extension






TECHNOLOGY ADOPTION IN EASTERN KENYA


Table 3. Summary of Crop Yields and Plant Populations for Case-Study Farms in
1986 Short Rains

Crop(s) Crop Population ('000/ha) Yield (kg/ha)
Range Mean Range Mean

Regional group I (Mwala/Wamunyu; N-3)


MZ/CWP/PIPa


MZ
CWP
PIP


4.8-34.2
5.4-33.2


0-1172
65-326


Regional group II (Makueni; N=2)


MZ/BN/PIP


MZ 21.0-22.0
BN 23.4-52.5
PIP 3.8-18.2


21.4
38.8
9.8


140-424
55-227


Regional group III (Mutomo; N=4)


MZ/CWP/PIP



MZ/GRG/PIP


MZ 11.0-25.2
CWP 0.4-12.0
PIP 0.8-3.4


MZ
GRG
PIP


5.2-25.4
6.4-68.2
1.6-2.4


Regional group IV (Embu-LM4; N-2)


MZ/BN


MZ 29.4-31.3
BN 28.3-60.3


30.4
44.3


202-392


Regional group V (Embu/Meru-LM5; N=2)


SOR/MLT SOR 42.6-43.9
MLT 62.7-69.4


43.2
66.1


473-780
366-431


a Abbreviations: MZ-maize, CWP-cowpea, PIP-pigeon pea, BN-bean, GRG-green
gram, SOR-sorghum, MLT-millet.

as possible and were often followed by a maize-cowpea crop in planting
sequence. Third, two weedings were common to all crops observed.

Additional Observations Relevant to Research and Development
Institutions
There were five additional features of the farming system that deserve
consideration for developing new technologies. First, timeliness of weeding


Vol. 2, No. 1, 1991


17.2
5.2
1.6

12.8
30.8
2.1


37-912
7-135


49-150
15-121





OCKWELL ET AL.


has an important impact on final crop yield through its effect on competition
for water and nutrients between the crop and the weeds. Timeliness of
weeding is affected by timeliness and method of planting, by the method of
weeding, and the amount of labor available. Despite the well-known
relationship between weeding and crop yields, there has been little new
weeding technology developed for smallholder farmers by the research-and-
development institutions during the last two decades. This is in sharp contrast
to the developments in the field of crop improvement during this period.
The most sophisticated weeding technology found on the farms studied
was the use of the ox-drawn mouldboard plough. Because this is an
implement designed for land preparation and planting, it is rather crude for
the control of weeds within an existing crop and can cause considerable
damage to the crop itself. Some work has been done by the NDFRC on
special-purpose weeding tools as well as detachable tools for mouldboard
ploughs (Figueroa and Mburu, 1984), but this does not seem to have resulted
in any technology being adopted by farmers in the region. It would seem,
therefore, that there is considerable scope for further research and develop-
ment in this field, perhaps based on the use of a single ox for draught power
rather than the usual pair of oxen.
An alternative is to develop other methods of herbicide application to
weeds within the crop, such as rope-wick technology (Gwynne and Murray,
1985). To date, this has received no attention from research-and-development
institutions in these semiarid regions. We realize, however, that farmers may
well prefer to make a single capital investment in a new tool or implement (as
many have done in purchasing ploughs) for mechanical control of weeds,
rather than bear the recurrent expenditure on herbicides.
A second issue concerns the preservation of available soil moisture.
Although this has been recognized in general as a target for research, new
approaches may be warranted. Erosion was still considered to be a problem
by 11 of the 14 case-study farmers who had terraced cropland. Dry ploughing
of this land was only implemented to a limited extent, suggesting that other
options should be investigated by the research institutions. One option is
stubble retention. However, there is a risk of stubble being consumed by
termites if it is not fed to livestock. Another is to conduct research into
alternatives to the existing "fanya-juu" terrace system for those farmers whose
capital or labor constraints prevent further construction of terraces. In this
regard an emerging technology is the formation of terrace mounds through
the planting of browse legumes, such as Leucaena, along the contours of


Journal for Farming Systems Research-Extension





TECHNOLOGY ADOPTION IN EASTERN KENYA


cropland. The establishment problem associated with Leucaena would be a
worthy subject for on-farm research.
A third problem area is the high rate of erosion on natural pasture land.
Research into methods ofpreventing degradation and rehabilitating such land
is needed. This is an extremely complex field, involving important technical,
social, and environmental interactions. The various roles that livestock play
in the farming system (Ockwell et al., 1990) result in the high stocking rates
that are a major contributor to the erosion problem. This probably means that
simply introducing a new technology will be insufficient. However, technol-
ogies such as forage legumes for the semiarid tropics are becoming available
(Menin et al., 1987), and the research challenge is one ofdeveloping methods
for establishing legumes on farmers' grazing lands.
Another challenge for both extension workers and researchers is to get
those farmers who use crop-protection chemicals to recognize the yield-
reducing effects of pests and diseases at a stage early enough in their respective
life cycles to take corrective action. In discussions with farmers it seemed that
some of their disenchantment with crop-protection chemicals was caused by
not applying them at the optimal time. This would seem to be an extension
issue. If there is a message for the research stations in the reluctance of farmers
to purchase these chemicals, it is perhaps that increased attention needs to be
given to incorporating resistance in plants at the breeding stage.
Fifth, although no immediate suggestions can be offered at this point,
recognition must be given to the fact that water collection for both livestock
and household needs is a major problem offarm households in this environment.
It represents a substantial drain on available labor supplies at critical times of
the year.


CONCLUSIONS

An important conclusion of this research is that a real understanding of the
level of adoption of new innovations by farmers of the semiarid tropics of
eastern Kenya requires the type of close examination embodied in the case-
study approach that was followed. Merely listing the technologies in use, in
the manner of a survey, might produce the mistaken impression that adoption
of innovations had proceeded far. However, by completing a detailed farm-
by-farm analysis, it was revealed that in many instances where an improved
technique is recorded as being in use its application is ineffective or only
available to a small proportion of the farm area. A more precise description


Vol. 2, No. 1, 1991





OCKWELL ET AL.


ofthe general situation on the case-studyfarms is that there is limited adoption
overall, with no farm having a complete package of the practices under review.
Within this general picture, some technologies had received more attention
by farmers than others. Those which were noncash using had widest use,
followed by low-risk cash using, and then high-risk cash using technologies.
This indicated that cash availability and risk-averse behavior were consider-
ations to be given serious attention by researchers in developing improved
farm practices. Higher-yielding crop varieties may do little to ameliorate
farmers' concerns about food security unless they have the capacity to reduce
the risk of crop failure in poor seasons. Even under average seasonal
conditions, the adoption of improved varieties that require ancillary inputs in
the form of fertilizer and crop-protection chemicals places severe strains on
farmers who are constrained in their access to cash. Indeed, it is possible to
dassify innovations in order to provide the research institutes with dearer
reasons for nonadoption. For example, crop-protection chemicals for use
against apion weevil on cowpeas are not applied at the required level because
of the cash constraint. That is, farmers growing improved cowpeas are
generally aware of the pesticide, but are constrained in its use by their cash
position. Conversely, the resistance of farmers to the use of inorganic
fertilizers would require their perception of the riskiness of the technology to
be altered.
Next, there was little evidence to support the conclusion that on-farm trials
conducted from 1980 to 1982 had been effective. Although on average the
16 farms that had been involved in on-farm trials had adopted more innovations
than the two farms that had not and seemed to be more aware of improved
practices even when not adopted, the differences between the groups was
slight. This was a consequence of the fact that, except for two PET farms, little
progress had been made down the innovation path. However, it would be
unreasonable to conclude that the trials had had no influence on farmers.
Because farm-level resource constraints were often acute, financial conditions
probably prevented the uptake of the improved practices.
The measurement of yields achieved in what was considered an average
season revealed very poor outcomes. This tended to confirm our other, more
informal observations of the ineffective, often partial way in which the new
techniques, when adopted, were generally applied.
The results of the study also showed that farmers vary markedly in their
resource endowments and in their abilities to respond to a risky production


Journal for Farming Systems Research-Extension






TECHNOLOGY ADOPTION IN EASTERN KENYA


environment. In view of these constraints and the fact that farmers differ in
terms of socio-economic considerations, not all technologies were equally
applicable to all farmers. This implies that there is a need to develop a range
of technologies to provide farmers with the option to select those that best suit
their particular circumstances, or to develop "generic technologies" that
farmers can adapt to their own farm situations.


ACKNOWLEDGEMENTS

The authors wish to acknowledge the assistance provided by the Australian
Centre for International Agricultural Research and the Government of Kenya.
Thanks are offered to Gunnar Harland, Jim Ryan, and Ken Menz, who
provided comments on various drafts of the paper.


REFERENCES

Anderberg, M.R. 1973. Cluster analysis for applications. New York: Academic Press.
Ashby, J.A. 1986. Methodology for the participation ofsmall farmers in the design ofon-
farm trials. Agricultural Administration 22(1):1-19.
Bakhtri, M.N., S. Gavotti, S. Odhiambo, and S. Nguluu. Undated. Farming systems re-
search at the NDFRC-Katumani. Project Field Document No. 2. U.N. Food and
Agriculture Organization/Ministry of Agriculture, Nairobi.
Byerlee,D., L. Harrington, and D.L. Winkleman. 1982. Farming systems research: Issues
in research strategy and technology design. American Journal of Agricultural Eco-
nomics 64(5):897-904.
Collinson, M.P. 1981. Alow-cost approach to understanding small farmers. Agricultural
Administration 8(6):433-50.
Collinson, M.P. 1982. Farming systems research in Eastern Africa: The experience of
CIMMTT and some national agricultural research services. MSU International De-
velopment Paper No. 3. Michigan State University, East Lansing.
Doorman, F. 1990. A social science contribution to applied agricultural research for the
small farm sector: The diagnostic case study as a tool for problem identification.
Agricultural Systems 32(2):273-290.
Everitt, B. 1980. Cluster analysis (2nd edition). London: Heinemann.
Farrington, J. 1988. Farmer participatory research: Editorial introduction. Experimental
Agriculture 24(3):269-79.
Figueroa, R.E., and J.K. Mburu. 1984. Development of the Bukura Mark II plough: A
multipurpose ox-drawn toolframe for small farmers. Proceedings of the Symposium
of Dryland Farming Research in Kenya. East African Agricultural and Forestry
Journal 44( 1):266-274.
Ghodake, R.D., and T.S. Walker. 1982. Yield gap analysis in dryland agriculture: Per-
spectives and implications for the eighties. ICRISAT, Patancheru, A.P., India.
Gwynne, D.C., and R.B. Murray. 1985. Weed biology in agriculture and horticulture.
London: Batsford Academic.


Vol. 2, No. 1, 1991





OCKWELL ET AL.


Harrington, L.W., and R Tripp. 1984. Recommendations domains:A frameworkfor on-
farm research. Economics Program Working Paper 02/84. International Maize and
Wheat Improvement Center (CIMMYT), Mexico City.
Jaetzold, R., and H. Schmidt. 1983. Farm management handbook of Kenya. Ministry of
Agriculture, Nairobi, volume 2.
Menin, L.K, W.M. Beattie, D.N. Njarui, BA. Keating, and R.K Jones. 1987. The
evaluation of forage and browse (alley crop) legumes for the semi-arid areas of Eastern
Kenya- an interim report. Pasture Network for Easter and Southern Africa (PANESA)
Workshop, Arusha, Tanzania, April, 1987.
Ockwell, A.P., KA. Parton, S. Nguluu, andL. Muhammad. 1990. Relationships between
the farm household and adoption of improved technologies in the semi-arid tropics
of Eastern Kenya. Agricultural Systems, submitted.
O'Leary, M. 1984. The Kitui Akamba Economic and social change in semi-arid Kenya.
Nairobi: Heinemann.
Tessema, S., E. Emojong, F.P. Wandera, and M. Nderito. 1985. Features of traditional
farmingsystemsas they affect livestockproduction. Project Field Document No. 6. U.N.
Food and Agriculture Organization/Ministry of Agriculture, Nairobi.
Wilkinson, L. 1987. SYSTAT: The system for statistics. Systat Inc., Evanston, Ill.


Journal for Farming Systems Research-Extension








Interaction Between Nitrogen Level and
Hessian Fly Protection of Wheat in Morocco

John Ryan, J.P. Shroyer, and M. Abdel Monem'



INTRODUCTION
The focus of current development efforts in international agriculture is to
stimulate increased food output in aid-recipient, lesser-developed countries.
Morocco typifies this situation. Cereals dominate the cultivated land area of
Morocco's semiarid rain-fed zone. Despite comprehensive approaches to
enhance cereal output, especially wheat (Triticum spp), per capital production
is decreasing because of high population growth. Of the obstacles to increased
cereal production described by Shroyer et al. (1990), the problem of the
Hessian fly (Mayetiola destructorSay) looms large. As a pest indigenous to the
rain-fed cereal belt of North Africa, its severity varies from year to year and with
location (Keith, 1988). Although chemical control is possible, its use is not
economically feasible for practical purposes in Moroccan agriculture. As a
consequence, research efforts have focused on developing resistant cultivars.
Research emanating from the Aridoculture Center in Settat, Morocco, and
from the United States has identified a Hessian fly-resistant bread-wheat
cultivar, Saada. Subsequently, it has been subjected to on-farm trials evaluat-
ing its agronomic traits compared to standard cultivars and farmer acceptance
(Riddle, 1989).
Although drought is an ever-present threat, nitrogen (N) deficiency is also
ubiquitous. Because N is the most limiting factor after drought for dryland
crops in Morocco, the response of any new cereal cultivar to N is of interest.
Consequently, N became the focus of the Aridoculture Center's soil-fertility
research program, with emphasis on soil type, previous crop, placement
methods, and residual availability (Abdel Monem et al., 1988a, 1988b,
1988c). Nitrogen responses of wheat were dramatic, but they were less so for
barley (Hordeum vulgare L.; Ryan et al., 1990a). Though these studies

1John Ryan and M. Abdel Monem, Mid-America International Agricultural Consor-
tium/Institute National de la Recherche Agronomique (MIAC/INRA) Aridoculture
Center, Settat, Morocco; J.P. Shroyer, Department of Agronomy, Kansas State Uni-
versity.





RYAN, SHROYER, AND MONEM


involved a diversity of soil types, they were confined to the same climatic zone
and used established cultivars. A soil-fertility survey of the various regions in
this semiarid zone suggests large variation in NO3 with changing environ-
ments (Abdel Monem et al., 1990b; Ryan et al., 1990b).
The first fertilizer trial with Saada highlighted both its Hessian fly resistance
and its response to N (Abdel Monem et al., 1990a). However, the rate of 60
kg of N per ha had a proportionately greater influence on Saada's straw
production than grain yields in comparison with the common bread-wheat
cultivar, Nesma, and a durum (Triticum turgidum var. durum L.) cultivar,
Cocorit. Indeed, Saada had a lowerN uptake and lower grain-protein content.
Although maximum N response was shown to vary with soil type in the
Settat area, ranging from 120 kg of N per ha for Vertisols to 40 kg of N per
ha for Rendolls (Abdel Monem et al., 1988c), little information was available
on N response in relation to climatic gradients. The recent euphoria about
Saada as a solution to the perennial scourge of Hessian fly underlines the
urgency of such efforts. The response of Saada to such high rates of N has not
been investigated. An additional consideration is the possible mitigating effect
of N where Hessian fly attack has occurred; it has been postulated that the
increased tillering caused by N fertilization compensates for fly damage.
This paper describes the results of a study to evaluate the interaction
between protection from Hessian fly (both chemical and genetic) and N
response of wheat in Morocco's semiarid zone. Saada was compared with the
most common cultivar, Nesma, using applications of 0, 40, 80, and 120 kg of
N per ha at five diverse locations, with varying soil and climatic conditions.
This paper also demonstrates that on-farm trials can be established successfully
and can provide useful and timely information for researchers and local
farmers.


MATERIALS AND METHODS
The five on-farm sites chosen for the study (Berrechid, Chamaia, Jemaa
Shaim, Khouribga, and Oulad Said) represent a diversity of environments,
particularly with respect to soils and rainfall, in the rain-fed semiarid zone of
central Morocco. Farmer cooperators were identified and contacted by local
extension personnel. Soils ranged from deep Vertisols at Oulad Said near
Settat and Jemaa Shaim to the south to shallow soils at Berrechid, Chamaia,
and Khouribga (Calcicxeroll). The sites also varied with respect to other
relevant properties (Table 1). Previous studies (Keith, 1988) showed a wide


Journal for Farming Systems Research-Extension






NITROGEN, HESSIAN FLIES, AND WHEAT IN MOROCCO 49

Table 1. Characteristics of the Trial Sites in Morocco
Rainfall Organic
Location and 1988-89 carbon NO, NaHCO,-P Exch-K
soil classification (mm) (g/kg) (mg/kg) (mg/kg) (mg/kg)
Chamaia
Petrocalcic Palexeroll 314 24.4 3.2 6.4 490
Oulad Said
Palexerollic Chromoxerert 373 15.1 2.0 3.4 355
Khouribga
Petrocalcc Palexeroll 416 29.6 5.4 18.4 375
Jemaa Shaim
Petrocalcic Palexeroll 434 21.5 2.8 2.0 200
Berrechid
Petrocalcic Palexeroll 443 21.5 2.0 22.4 185


range of severity of Hessian fly damage in this zone, with the southern areas
(i.e., Chamaia and Jemaa Shaim) usually hit the hardest.
The sites lie within the target area of the Aridoculture Center in Settat, a
facility central to the Dryland Agriculture Project of the Mid-America
International Agricultural Consortium (MIAC), in conjunction with Moroc-
co's Institute National de la Recherche Agronomique (INRA), and funded by
the U.S. Agency for International Development (USAID). Ryan et al. (1988)
describe the organization, development, and operation of this project.
The fields selected in each of the five cereal-growing areas were in cereals
the previous year, thus making a response to N likely. The sites were prepared
by cultivating with an offset disc in November 1988, after the onset of the fall
rains. Three types ofcultivars were tested: (1) the popular, but highly Hessian
fly-susceptible bread-wheat cultivar, Nesma; (2) Nesma treated with the
insecticide carbofuran; and (3) the newly introduced cultivar, Saada, which is
widely effective in Morocco (El Bouhssini et al., 1988) and carries the H5 gene
for Hessian fly resistance. Carbofuran, which is effective against the Hessian
fly (Morrill and Nelson, 1976; Regehr et al., 1986), was applied as Furadan
with the seed drilled into the plots at 1 kg per ha. Cultivars were compared at
four N fertilization rates (0, 40, 80, and 120 kg per ha). These treatments
covered a range of expected N response and the control (0 kg of N per ha)
represented local-farmer practices.
The experimental design was a strip block (Cochran and Cox, 1957) with
three replications. Nitrogen rates differed between strip blocks, and cultivars
were randomized within strips. The 10- by 3-m plots were drilled at 80 kg of
seed per ha in 30-cm rows. We used this design because (1) we were most
interested in comparisons between cultivars and the effect on those comparisons


Vol. 2, No. 1, 1991





RYAN, SHROYER, AND MONEM


of interaction with nitrogen fertility, and (2) the systematic arrangement of
fertility treatments made it much easier to show and explain the plots to
farmers. Nitrogen, at the specified rates, was subsequently hand broadcast as
urea, mimicking farmers' application. Because phosphorus (P) was not a
variable in the study, a blanket application of 80 kg of P2O per ha as super-
phosphate was drilled with the seed.
Whole plants from two 5-m inner rows were cut at ground level and
weighed to determine dr- matteryields. Bundles were threshed, and grain yields
were determined. Additionally, a 1-m row was taken from each plot to
estimate N fertilizer uptake and thus N-use efficiency, following grinding and
Kjeldahl digestion.
At all locations the site was lent by the farmer at no cost; he retained the bulk
of the straw and grain after subsampling at harvest. Farmer cooperators
identified sites near well-traveled roads, prepared the seedbed for planting,
and assisted with the harvest operation. The sites were protected on a 24-hour
basis by guards in order to prevent animal grazing, bird damage, and pilfering.
In all cases, the farmer cooperator or a member of his family acted as the site
guardian. The guards were briefed on the purpose of the trial and were given
a treatment sketch for the plot; they were thus able to serve as an on-site
"extension agent" to explain the trial to local farmers, as well as to farmer and
researcher groups visiting the sites.
Rainfall during the year was unusual and erratic. Although it normally
decreases to the south and east of Settat, rainfall during the study period at
Jemaa Shaim and Khouribga was, in fact, higher than at Oulad Said, near
Settat. The season was characterized by heavy initial rains in November,
followed by little or no rain during December and January, and normal rains
from February to April.


RESULTS

There were no cultivar versus N interactions compared for dry matter at any
location. Nesma produced lower dry-matter yields than Nesma + carbofuran
or Saada at four of the five locations (Table 2). Only at Khouribga did Nesma
+ carbofuran produce significantly higher dry-matter yields than Saada.
There were cultivar-versus-N trials (Figure 1) for grain yields at three of the
five locations (interaction at Chamaia was significant at P>0.08). At Jemaa
Shaim, with heavy Hessian fly infestation, Nesma produced significantly lower
grain yields over all N rates, whereas Nesma + carbofuran had lower yields than


Journal for Farming Systems Research-Extension






NITROGEN, HESSIAN FLIES, AND WHEAT IN MOROCCO 51

Table 2. Wheat Dry Matter (dm) and Grain Yieldsa of Wheat
at Five Moroccan Environments
Jemaa Shaim Chamaia Oulad Said Khouribga Berrechid
Cultivar dm grain dm grain dm grain dm grain dm grain

Nesma 2.0 0.3 3.3 1.3 3.3 1.0 2.7 0.9 4.0 1.3
Nesma + 5.1 1.6 3.6 1.4 4.2 1.4 3.5 1.3 5.9 1.7
carbofuran
Saada 5.8 1.6 3.7 1.3 4.2 1.0 3.0 0.7 6.3 1.6
LSD 0.05 0.8 0.2 NSb NSb 0.3 0.1 0.2 0.1 0.7 0.2

a Yields are in mg/ha.
NS = not significant.


Saada only at the 0 N rate. At Khouribga, Nesma + carbofuran produced
significantly higher grain yields than the other treatments over all N rates.
Also, Nesma outperformed Saada at 40 and 120 kg of N per ha. Nesma +
carbofuran at Oulad Said produced higher grain yields than the other
treatments over all N rates, except at 80 kg per ha, which showed no significant
differences in yields between Nesma and Nesma + carbofuran. At Chamaia,
Saada performed well at 40 kg of N per ha, but Nesma + carbofuran produced
significantly higher yields than Saada at 120 kg ofN per ha.
Cultivar grain-yield means varied significantly over N treatments at four of
the five locations (Table 2). At Jemaa Shaim and Berrechid, Nesma +
carbofuran and Saada had similar yields. At Oulad Said and Khouribga, Nesma
+ carbofuran had higher yields than either Saada or Nesma. At Khouribga,
Nesma had a higher yield than Saada. There was no difference in overall dry-
matter yields of Nesma + carbofuran and Saada at any N level (data not
shown). However, Nesma + carbofuran, which had a mean grain yield equal
to that of Saada with no N applied, produced significantly higher grain yields
than Saada when N fertilizer was applied. Nesma + carbofuran's harvest index
(0.33) advantage over Saada (0.27) became much more apparent with 40-kg-
of-N-per-ha fertilization. Yield loss from Hessian fly infestation, manifested
as the difference between grain yields of Nesma + carbofuran and Nesma, was
also greatest in the fertilized treatments. This indicates that farmers cannot
compensate for Hessian fly damage by applying N to induce greater tillering.
In order to illustrate N-use efficiency between sites, apparent N recovery
was determined for Saada. Although efficiency varied with N rates, all soils,
except the Vertisols at Oulad Said, had recovery values of less than 30 percent.
The Oulad Said values were 35 to 55 percent and coincided with maximum


Vol. 2, No. 1, 1991


























Nitrogen (kg/ha)


RYAN, SHROYER, AND MONEM


2.4,


2.2.
2.0-
1.8.
Id-
1.6-
1.4-

. 1.0-
0o.8
0.6-
0.4-
0.2-


SLSD (0.05) Chamaia

Nesma + Carbofuran




Saada


u0.0 ........ I
0 40 80 1
Nitrogen (kg/ha)

2.41


2.2
2.0o
1.8
" 1.6"
S1.4

S1.0
0.8
0.6.
0.4
0.2
0.0


0 40 80
Nitrogen (kg/ha)


0 40 80 120
Nitrogen (kg/ha)

Figure 1. Grain Yields of the Three Wheat Types at the Five Trial Sites in Morocco


Journal for Farming Systems Research-Extension


S LSD (0.05) Khouribga



Nesma + Carbofuran


Nesrna


Saada


I SD (0.05) Oulad Said

Ncsma + Carbofuran

Nesma

Saada





NITROGEN, HESSIAN FLIES, AND WHEAT IN MOROCCO 53

N response; the lowest average recovery occurred at Chamaia, where N
response was nonexistent.


DISCUSSION
This paper provides a basis for identifying technologies that have potential for
greatly expanding cereal output in Morocco, especiallyin the rain-fed semiarid
zone. The data demonstrate the serious impact ofHessian fly on cereal yields:
when plants are protected from this pest, yields are about 30 percent higher.
The development ofa Hessian fly-resistant wheat, Saada, which carries the H5
gene for resistance, is a major breakthrough in cereal breeding in Morocco.
Although it may not have the grain-yield potential of Nesma under favorable
conditions, Saada does produce high straw yields. This is an important factor
in an area where straw is a major animal feed. However, the quest continues
for Hessian fly-resistant cultivars with better grain-yield potential. Although
Hessian fly can also be chemically controlled, the material, carbofuran, is
exorbitantly expensive, and its effectiveness is dependent up on weather
conditions. Toxicity, limited supplies, the requirement of a grain drill in an
area where sowing is normally done by hand-broadcasting, and uncertain
yields are other factors that militate against farmer use of chemicals to control
Hessian fly.
Despite the fact that drought may limit N response in some areas, yield
increases from N application generally are dramatic. Such response could be
expected to vary with seasonal rainfall and its distribution pattern. In a region
where rainfall is so erratic (Watts and El Mourid, 1988), few seasons can be
said to be "normal." However, our data indicate that for a modest input of40
kg of N per ha, average yield increases of 30 to 40 percent are likely. The
implications for cereal output in the semiarid dryland zone of Morocco are
self-evident. The region, where 75 percent of the arable land is devoted to
cereals, is characterized by low N input; few, if any, farmers fertilize barley.
Although socio-economic factors impinge upon farmer acceptance and use of
fertilizers, the research emanating from the USAID-sponsored Dryland
Agriculture Project (Abdel Monem et al., 1988c; Shroyer et al., 1990) has
underscored N fertilizer benefits and catalyzed farmer adoption.
The lack of response to N of some sites or soils, coinciding with previous
observations from this dryland area, raises a number of questions about the
dynamics of N in such soils. Because many soils from this zone are relatively
high in organic matter, and because moisture and temperature during the


Vol. 2, No. 1, 1991






RYAN, SHROYER, AND MONEM


growing season are favorable, mineralization of soil organic N sources is a
probable cause of nonresponse (El Gharous et al., 1988). Indeed, the
relatively low N recovery from added fertilizer indicates possible loss mecha-
nisms such as volatilization. Because residual N carry-over is minimal in this
area (Abdel Monem et al., 1988b), such loss mechanisms need to be
investigated.
This study also highlights some positive aspects of conducting on-farm
trials in semiarid areas such as Morocco. Farmer cooperators were excited
about the availability of a new cultivar that resists the Hessian fly and has the
potential of producing excellent grain and straw yields. The cooperators
expressed amazement that nitrogen fertilization could give such a noticeable
response and increase grain and straw yields. Having the guards available, not
only to tend to the plots but also to talk with neighbors and visitors, proved
valuable in extending new information. The strip-block design proved simple
to install and an easy means of demonstrating major effects such as cultivar
differences and various N levels. Uncomplicated trials with visual impact are
required in traditional, low-input, illiterate, rural settings. The loss of some
trials to drought or unforeseen pest infestation must be accepted in conducting
diagnostic-demonstration trials in such marginal areas. Indeed, the severe
drought in the early 1980s rendered most on-farm trials in the semiarid zone
virtually valueless. However, on-farm trials, at all levels of farmer cooperation,
are necessary to extend new technologies while gathering new information.


REFERENCES
Abdel Monem, M., A. Azzaoui, M. El Gharous, J. Ryan, and P. Soltanpour. 1988a.
Fertilizer placement for dryland wheat in central Morocco. Pages 149-162 in J. Ryan
and A. Matar, eds., Proceedings of the Third Regional Soil-Test Calibration Workshop,
Amman, Jordan, Sept. 2-9, 1988. Aleppo, Syria: ICARDA, in press.
Abdel Monem, M., A. Azzaoui, M. El Gharous, J. Ryan, and P. Soltanpour. 1988b.
Residual nitrogen and phosphorus for dryland wheat in central Morocco. In J. Ryan
and A. Matar, eds., Proceedings ofthe Third Regional Soil-Test Calibration Workshop,
Amman, Jordan, Sept. 2-9, 1988. Aleppo, Syria: ICARDA, in press.
Abdel Monem, M., A. Azzaoui, M. El Gharous, J. Ryan, and P. Soltanpour. 1988c.
Response ofdryland wheat to nitrogen and phosphorus in some Moroccan soils. Pages
52-65 in J. Ryan and A. Matar, eds., Proceedings of the Third Regional Soil-Test
Calibration Workshop, Amman, Jordan, Sept. 2-9, 1988. Aleppo, Syria: ICARDA, in
press.
Abdel Monem, M., A. Azzaoui, M. El Gharous, J. Ryan, and P. Soltanpour. 1990a.
Demonstrating fertilizer response and Hessian fly resistance with wheat in Morocco.
Journal ofAgronomic Education 19:77-80.


Journalfor Farming Systems Research-Extension






NITROGEN, HESSIAN FLIES, AND WHEAT IN MOROCCO 55

Abdel Monem, M., J. Ryan, and M. El Gharous. 1990b. Preliminary assessment of soil
fertility in the mapped area of Chaouia. AlAwamia, in press.
Cochran, W.G., and G.M. Cox. 1957. Experimental design. New York, N.Y.: John Wiley
& Sons, Inc., p. 305-309.
El Bouhssini, M.,A. Amri, and J.H. Hatchett. 1988. Wheat genes conditioning resistance
to the Hessian fly in Morocco. Journal of Economic Entomology 81:709-712.
El Gharous, M., P.N. Soltanpour, and R.L. Westerman. 1988. Nitrogen mineralization
potential of some semi-arid soils of Morocco. In J. Ryan and A. Matar, eds.,
Proceedings ofthe Third RegionalSoil-Test Calibration Workshop, Amman, Jordan, Sept.
2-9, 1988. Aleppo, Syria: ICARDA, in press.
Keith, D. 1988. Saada wheat update. INRA-MIAC Report. Aridoculture Center, Settat,
Morocco.
Morrill, W.L., and L.R. Nelson. 1976. Hessian fly control with carbofuran. Journal of
Economic Entomology 69:123-124.
Regehr, D.L., J.H. Hatchett, D.L. Keith, and G.E. Wilde. 1986. Insecticides for Hessian
fly control in Morocco, 1985. Insecticide and Acaricide Tests 11:373-374.
Riddle, R. 1989. Farmer evaluations of a new bread wheat variety: A survey of Saada
promotion participants. MIAC/INRA Sociology Bulletin No. 5. Aridoculture Cen-
ter, Settat, Morocco.
Ryan, J., M. Abdel Monem, A. Amri, and K. El Mejahed. 1990a. Nitrogen fertilization
of improved barley cultivars in Morocco's dryland zone. Agronomic Abstracts- 62.
Ryan, J., M. Abdel Monem, and M. El Gharous. 1990b. Soil fertility assessment at
agricultural experiment stations in Chaouia, Abda, and Doukkala. AlAwamia 73, in
press.
Ryan, J., D. Keith, M. Abdel Monem, and S. Christiansen. 1989. Mid-America Inter-
national Agricultural Consortium's Dryland Agriculture Project in Morocco. Agro-
nomic Abstracts- 58.
Shroyer, J.P., J. Ryan, M. Abdel Monem, and M. El Mourid. 1990. Production of fall-
planted cereals in Morocco and technology forits improvement. JournalofAgronomic
Education 19:32-40.
Watts, D.G., and M. El Mourid. 1988. Rainfallpatterns and probabilities in the semi-arid
cereal production region of Morocco. Aridoculture Center, Settat, Morocco.


Vol. 2, No. 1, 1991








Farmer-First Qualitative Methods:
Farmers' Diagrams for Improving Methods of
Experimental Design in Integrated Farming
Systems1

Clive Lightfoot and D.R. Minnick



INTRODUCTION
Continued degradation in farmed environments and poverty in farm house-
holds challenge farming systems research-extension (FSRE) practitioners,
especially in tropical regions. Today's farming systems in South Asia have
salinized 59 million ha of irrigated land and desertified 150 million ha of rain-
fed cropland. In Sudano-Sahelian Africa, 5 million ha are salinized and 142
million desertified annually. Each year 3,349 ha in Africa and 1,571 ha in
Southeast Asia are deforested. In Africa and Asia roughly half the population
receives inadequate calories for an active working life, 30 percent of the
children are malnourished, and less than 10 percent enjoy access to sanitation
(World Resources Institute, 1988). FSRE practitioners must concentrate on
designing new farming systems that rehabilitate or regenerate farm environ-
ments and economies.
Regenerative farming systems for degraded farm land need to integrate
trees, fish, animals, vegetables, and crop enterprises. Trees serve many
purposes, including protecting land and providing investment. Likewise,
animals in an integrated farm provide inputs to other enterprises as well as
sources of meat, milk, hides, and investment. Fish provide valuable, high-
quality protein products and their production systems recycle important
nutrients and organic matter. For the farmer, integrating such an array of
enterprises on a farm not only offers economic security but also imparts
ecological protection. Use of diverse flora and fauna spreads income and


SPaper presented at the Tenth Annual Association for Farming Systems Research-
Extension Symposium, Michigan State University, East Lansing, October 14-17,1990.
International Center for Living Aquatic Resources Management, Inc. (ICLARM) and
International Rice Research Institute (IRRI), Manila, Philippines, respectively.





LIGHTFOOT AND MINNICK


inhibits pests and diseases. Waste and by-product flows between enterprises
help to cut external input bills and recycle nutrients (Lightfoot, 1990). The
examples in this paper illustrate farming systems that integrate trees, fish,
livestock, vegetable, and crop enterprises. A great deal more research needs
to be done if these kinds of environment-friendly systems are to become
commonplace in degraded tropical ecosystems.
New farming systems can only be generated with the participation of many
farmers. Most national, on-farm research teams can barely cope with a dozen
trials with five to ten cooperators each. There are many reasons for this. A
recent survey of farming-systems projects showed that current procedures
demand too much researcher control, input supply, and data collection
(Lightfoot and Barker, 1988). Researchers and their respondents concluded
that farmer participation should increase. A decade of designing integrated
farming systems on research stations for farmer demonstration has taught a
similar lesson (Edwards et al., 1988).
Farmers must participate in the design of new systems. Few scientists have
called for testing system-level hypotheses with very large numbers of farmers
(Sumberg and Okali, 1988). Those that have recognize the importance of
putting farmers knowledge, ideas, and wishes first. 'Farmer-First' procedures
provide researchers with tools for learning what farmers know and want to do
(Lightfoot et al., 1988; Lightfoot, 1989). Several documents detailing
qualitative procedures for encouraging farmer participation are now available
(Farrington and Martin, 1988; Chambers et al., 1989; Essers et al., 1989;
Waters-Bayer, 1989). Increasing interest in environmental issues and enterprises
such as agroforestry ensures continuing demand for farmer-first qualitative
methods.
We anticipate a demand for field procedures wherein very large numbers of
farmers can quickly and easily participate in designing environment-friendly
farming systems. Current methods of questionnaires and component tech-
nology trials limit farmer participation in quality and number. This paper
advances the idea that farmers' diagrams can improve field methods in on-farm
research. Our qualitative methods are directed at putting farmers' ideas first,
increasing the number of farmers able to participate, and orienting technology
testing to farm-system levels. We illustrate how farmers' diagrams can help put
farmers ideas about on-farm research priorities, experimental layouts, and
integrated-farming-system designs first.


Journal for Farming Systems Research-Extension





FARMER-FIRST METHODS: FARMERS' DIAGRAMS


FARMERS AND VISUALIZATION

The theory behind diagramming is that farmers, many of whom are illiterate,
visualize better than they verbalize. These farmers see their farms in pictures;
they can describe farms better by drawing than by talking. Research has shown
that different types of people have a proclivity to communicate with different
types ofmessages. We use Egan's definition of communication as the creation
of a common understanding by the exchange of messages (Egan, 1988).
Communication differences are related to the experiences and exposure of
individuals. Recent research suggests that individual memory references and
subsequent communication concepts are linked to two distinct memory
networks. One stores formal or semantic references, whereas the other stores
experiences or episodic references (Tulving, 1989). This research has
inferences for the process of communication between people with varying
levels of formal education (Minnick, 1990). Illiterate farmer communication
patterns rely primarily on concrete, visual, hands-on references, whereas
scientists rely more on abstract terms and symbols.
One technique that may be useful in promoting farmer participation is
cognitive psychology's notion of concept maps (Novak, 1977). Where maps
originate from farmer concepts they portray the realities and associations of
farm systems. One application of this methodology is to have groups of
farmers draw concept maps of their experimental designs and farming systems.
The concept map documents farmers' perceptions and is used by researchers
for future communication reference.


PROCEDURE FOR FARMER DIAGRAMMING

Farmer diagramming procedures are extremely flexible. However, they
require their proponents to have skills in eliciting information. The following
is a general outline of the process.
A group of farmers gathers in a comfortable place where there is plenty of
flat ground to draw on. Large sheets of paper can substitute for the ground
if the farmers feel comfortable with pens and paper. The researcher, who
should be accompanied by someone who will record and observe the process
and products, explains what information is sought. This explanation usually
requires the researcher to initiate the drawing, maybe even providing a stick
or seeds to start it off. As soon as farmers understand, however, they should
take over and start afresh. During the process, the researcher should be


Vol. 2, No. 1, 1991





LIGHTFOOT AND MINNICK


sensitive to who is participating and encourage the silent parties, especially
women, to join in. The whole process usually lasts up to about two hours. As
many as fifteen farmers can participate at the same time.
Although the process is continuous, it is best described in phases. The first
phase explains the method to farmers or asks them to help researchers
understand their concepts. This is an important and often difficult reversal for
researchers, who are used to telling farmers what to do. Preparing the ground
or materials to make the drawing is phase two. The third phase prompts
farmers to start the drawing. In the fourth phase, farmers take over as their
understanding of the exercise crystallizes. This may be signalled by farmers
enthusiastically searching for symbols to add to the drawing. In the final
phase, the full drawing or farmers' concept map is recorded by the research
team and verified by the group.
Eliciting knowledge from farmers is a skill that requires much practice.
Farmers must feel at ease. This can be fostered through humor, which may
encourage the self-conscious and less confident to draw. All attempts at
drawing should be incorporated; there is no such thing as a bad drawing or
someone who cannot draw. The drawing itself should be as large as possible
and uncluttered. Researchers should elicit all the information needed for one
element before going on to another. Clarity in the diagram is also preserved
if researchers are specific about each element. Symbols should represent the
element precisely, as confusion can quickly set in.


DIAGRAMS FOR DETERMINING WHAT
TECHNOLOGIES TO TEST

An example from eastern India shows how improved-technology diagrams
can influence farmers' decisions about which technologies to test first.
Groups of farmers were shown diagrams of improved technologies. Each
diagram, which had been prepared by researchers on large sheets ofpaper, was
explained to the group. They were then asked what changes they thought
were necessary for the technology to work on their farms. After much
discussion changes were identified, and the group was asked to select those
technologies they would like to test on their farms. All the technology
diagrams selected were then shuffled and handed back to the group for them
to sort in order of priority for on-farm testing.
One of the improved technologies entailed the stocking offish into flood-
prone areas called "chaurs." The researcher's diagram of this technology


Journal for Farming Systems Research-Extension






FARMER-FIRST METHODS: FARMERS' DIAGRAMS


(Figure la) shows the farmer putting fish fry into the chaur in July and netting
the fish in December and January. When the farmers saw this diagram they
pointed out that some of the fish were not caught. Indeed, they worried that
many, if not all, of the fry might escape during floods. The risks of loss were
high. The group suggested a modification to this technology (Figure Ib).
Farmers would like to see some kind ofnet enclosure in which the fish fry could
be placed and allowed to grow. They felt that this would ensure that the fry
put in would have a good chance of being taken out.



(A)


... h.r.'... in .........ar .


Fish harvested in December-January.


Fish stocked in July.


Fish placed in confined net in July. Fish collected in December.


Figure 1.
(A) Researchers' Design of Technology for Fish Culture in "Chaurs" Areas
(B) Farmers' Modified Design of Technology for Fish Culture in "Chaurs" Areas
Uttar Pradesh, India


Vol. 2, No. 1, 1991





LIGHTFOOT AND MINNICK


DIAGRAMS FOR DETERMINING HOW TO
TEST TECHNOLOGIES

Another example, also from eastern India, shows how diagrams ofexperimen-
tal layouts and activities put farmers' ideas of how to test technologies first.
A farmer, while showing researchers a researcher-designed, rice-fish exper-
iment on his farm, asked how fish growth could be improved. The farmer was
asked to draw diagrams of the experimental layout and a calendar of experi-
mental activities over the year. In the experimental field the farmer dug two
trenches about 2 m wide and 1 m deep (Figure 2a). He transplanted rice in
the first week of July and put in fingerlings ofrohu (Labeo rohita), catla (Catla


Trench


I I )t v y0 ,. V V ,' p 0 i .'
LI y r yV V P I y I ir I I't I 0 .I'
(A) Researchers Design for Rice-Fish Experimental Field Layout










(B) Farmers' Design for Rice-Fish Experimental Field Layout
I X _J V' U Mr p -P




tar Prades, Iia ce





SJournal for Farming Systems Research-Extension
pT Y Y VP I I V Y1 3' 3 'P A 1 0i P X '
)vr 0 I 0 V l~t / ff 4Jv .9 Ar Vt 1W It #

(B)
Figure 2.
(A) Researchers' Design for Rice-Fish Experimental Field Layout
(B) Farmers' Design for Rice-Fish Experimental Field Layout
Uttar Pradesh, India

Journalfor Farming Systems Research-Extension


' f 9,14- p Jj^ r* .j' y ur I y~y7 e ~
Iv^ YJ ^ 'I Y Y y y Jv y y; )
*II. fJ .v-S^ r fIf} f v: :I~u
f P-1 f ; *0 0 W f v It y i f
r H^ L *- C Rice V 1 V
Fish p Fish
Y ei~ I )vL~ p Y$ a
I yPg *o < r? y r P, a f~Y f yj K-ve
Y3P -r *& y~ux uI -J j^
YV.J v3 gW7 ~%~%L srwrP
_J A_ V V V_ _
y^? y.'^^^.~yy mib.


Trench






FARMER-FIRST METHODS: FARMERS' DIAGRAMS


catla) and bhakur (Cirrhinusmrigala) 15 days after transplanting (Figure 3a).
At transplanting and 45 days later he top-dressed the field with urea. The
farmer added cattle manure to the trench daily from transplanting until
October, after which he applied the manure once a week. The rice crop was
harvested in the last week of December. The fish were fed with rice bran and
mustard oilseed cake daily until November, and thereafter on alternate days.
In October, the farmer added grass carp (Ctenopharyngodon idella) fingerlings
because the rohu and catla were growing slowly.
When asked how he would improve the test field, the farmer drew another
set of diagrams. His experimental layout (Figure 2b) replaced the two deep
trenches with a single, square, shallow sump, 0.5 m deep, in the middle of the
field. He scheduled rice transplanting for June rather than July, harvesting
both rice and fish for late November, and stocking of grass carp and bhakur
fingerlings at 15 days after transplanting (Figure 3b). Manuring practices
remained the same but rice bran and oilseed cake applications were doubled.
After harvesting the fish, the farmer expected to dig out the sump and plough




mix zx xxnxxxx K xUxx i.. xxx XXX xxx Rice bran
0000 0000 00 00
Urca top dressing
VV 4 V Wh' .~ ~ Rice Fish

Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May
(A)

....... ....n.........x. 0oooooo o o ooo0 oo 00000 a ooo oo Cowdung
,6e rUrea top dressing
,c >, c I 5 CG ,Z I I I
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May
(B)



Figure 3.
(A) Researchers' Seasonal Calendar of Operations and Inputs for
Rice-Fish Experiments
(B) Farmers' Seasonal Calendar of Operations and Inputs for Rice-Fish Experiments
Uttar Pradesh, India


Vol. 2, No. 1, 1991






LIGHTFOOT AND MINNICK


the field in preparation fora winter wheat crop. This, he hypothesized, would
result in acceptable fish yields in a shorter time and maintain rice yields. The
additional wheat would also increase overall returns for the land and water
resources.


DIAGRAMS FOR DETERMINING INTEGRATED FARMING-
SYSTEM DESIGNS

The following examples from Vietnam, Bangladesh, and Malawi demonstrate
how diagrams of material flows between aquaculture, agroforestry, and
agricultural enterprises put farmers' knowledge of integrated farming-system
design first.

Vietnam
Farmers' knowledge about interactions in their rice-shrimp system was
elicited from four innovators who drew pictures, using colored pens and
newsprint paper, of land types, enterprises, and flows between them. The
process of explaining what was needed, giving examples, and drawing the
pictures took about one and a half hours. The diagrams depicted the flow of
water from the trenches to irrigate vegetables planted on dikes made from
trench mud and grown in rice-straw mulch. This illustrated to farmers the
concept of material flows between enterprises. These, along with more
complex transfers, are shown in Figure 4.



Trash fish
from canal
Straw Oil cake
Branches
~re 4~..~ l P. f and leaf i
Oil cake Grain ba, brn, seed
lour/chips Manure
M u l c h aD R i c eM



2.5 m 4.0 m Rice area
(0.2 ha) ] (1.3 ha)



Figure 4. Material Flows Between Enterprises of Rice-Dike Integrated Farming
Systems in Mekong Delta, Vietnam.
Systems in Mekong Delta, Vietnam.


Journal for Farming Systems Research-Extension





FARMER-FIRST METHODS: FARMERS' DIAGRAMS


Immediately after trenching, chicken and cattle manure from the home-
stead are moved into the rice-field trenches to induce phytoplankton blooms
for the fish and shrimp to feed on. Although shrimp diets are primarily natural,
they are fed during their first two months with other farm-grown by-products.
Germinated rice grain, cassava flour, and rice bran are typical shrimp feeds,
along with coconut and peanut oilcake, trash fish from the irrigation canals,
and duck or chicken carcasses. Also placed in the trenches are mango and
eucalyptus branches. These branches keep out cattle and poachers and
provide the undisturbed habitat that shrimps require.
Farmers learned that rice crops benefit from shrimps in two ways. First,
shrimps feed so heavily on weed seedlings that weeding expenses can be
reduced by one third. Second, chemical fertilizer application can be reduced
by 30 percent without reducing rice grain yield.

Bangladesh
A widow's homestead had bees, pigeons, a fish pond, and 24 broiler
chickens, as well as the typical trees, vegetables, and root crops found in
Bangladeshi home gardens. After researchers inspected each enterprise, she
and some ten neighbors formed a circle and drew out on the ground all the
enterprises and material flows between them. Using a stick the widow drew
boxes for each enterprise and then etched arrows between them to show
material flows (Figure 5).


Figure 5. Material Flows Between Homestead and Rice-Field Enterprises of
Integrated Farming Systems in Mymemsingh, Bangladesh


Vol. 2, No. 1, 1991





LIGHTFOOT AND MINNICK


Manure collected from animal pens is placed in vegetable plots, rice fields,
compost heaps, and the fish pond. Tree branches provide support for
vegetable crops and protection for the fish pond. The pond also provides
irrigation water for an adjacent vegetable plot. Feed and fertilizer inputs for
the fish pond are purchased whereas home-grown rice bran is given to the
animals. Similarly, other potential pond inputs such as tree leaf, ash, and
compost are used on other enterprises.
From the diagram, ideas emerged on better integration to reduce use of
some external inputs. The widow could reduce costs of external inputs and
better integrate her pond if, for example, household ash was substituted for
lime. Purchased fertilizer could be substituted by compost. Fish may feed on
tree or legume leaves just as well as on purchased rice bran.

Malawi
Four farmers drew a diagram on the ground detailing the connection offish
ponds to other enterprises on their farms. First they outlined the pond and
then named the materials that they put into the pond. The first drawing was
made by etching the ground with a stick. This diagram was abandoned
because the many linkages became confused. The farmers started again, this
time using ash and, wherever possible, objects to represent the products of
other enterprises that were going into or leaving the pond (Figure 6).
The farmers used fruits of guava, papaya, and avocado to represent the
input of rotten fruits into the pond. Leaves of leucaena, pumpkin, wild
vegetables, and cowpea represented the input of leaves into the pond. Some
maize bran was used to represent its use as a fish feed. Other inputs such as
cattle, sheep, and goat manure were represented by lumps of soil. The main
outputs from the pond were sediments for the vegetable gardens, water for
irrigating vegetables, and, of course, fish.
During the process, farmers learned of new inputs from each other. Some
did not know that guava and avocado fruits were good pond fertilizer, others
that fish would eat cocoyam leaves. That pond sediment could be used to
'fertilize' vegetable gardens was also news to some farmers.


CONCLUSIONS
We call for FSRE practitioners to concentrate on designing new integrated
farming systems that rehabilitate or regenerate farm environments and
economies. In this we anticipate a demand for field procedures wherein very


Journal for Farming Systems Research-Extension






FARMER-FIRST METHODS: FARMERS' DIAGRAMS


Figure 6. Material Flows Between Fish Pond and Other Enterprises of Integrated
Farming Systems in Zomba, Malawi


large numbers of farmers can quickly and easily participate in designing these
environment-friendly farming systems.
We advance the notion that farmers' diagrams can improve field methods
in on-farm research because visual images communicate to farmers better than
talk. In drawings everything can be seen at once, and mistakes can be
corrected easily and immediately. Drawings stay on the ground for passers-
by to join in and comment on, garnering wider views and greater consensus.
Farmers remember what they learn because the pictures stick in their mem-
ories. Researchers and extension workers learn about traditional knowledge.
The diagramming process builds self-reliance and confidence among the
farmers to incorporate new technologies from extension and become more
active partners in the development process. Thus we reverse the one-way flow


Vol. 2, No. 1, 1991





LIGHTFOOT AND MINNICK


of decisions from expert to farmer and put the farmer first.
This paper presents examples of how farmers' diagrams put farmers' ideas
and knowledge first in on-farm research-priority setting, experimental layouts,
and integrated farming-system design. Although our examples concerned the
integration of aquaculture and agriculture, these procedures are just as useful
for those wanting to integrate agroforestry, livestock, or any new enterprise
into the farming system.
These qualitative methods are now being tested and refined in several
projects in Africa, South Asia, and Southeast Asia (Maclean and Dizon, 1990).
A great deal of research still needs to be done before quick and easy methods
for many farmers to synthesize new integrated farming systems will be ready
for wide-scale use. A great deal more research needs to be done if environ-
ment-friendly systems are to replace today's destructive practices.


ACKNOWLEDGMENTS
We would like to acknowledge those scientists who helped us with the farmer
interviews. In Vietnam, from Cantho University, Drs. Nguyen Anh Tuan, Le
Thanh Duong, Duong Ngoc Thanh, and Nguyen Quang Tuyen and from
IRRI, Mr. O. Magistrado. In India, Dr.AjayKumarfrom RajendraAgricultural
University, Pusa; Dr. K. Tirkey from Birsa Agricultural University, Ranchi;
and Dr. V.P. Singh from IRRI. In Malawi, Dr. R.P. Noble and Mr. S.
Chimatiro-Phiri of Chancellor College, University of Malawi, Zomba, and
ICLARM, respectively. In Bangladesh, Dr. M.V. Gupta and Dr. M.A. Ahmed
from ICLARM. At ICLARM, we would like to thank Jay Maclean and Dr.
Roger Pullin for commenting on the manuscript. O.F. Espiritu, Jr., is to be
thanked for his fine artwork.


REFERENCES
Chambers, R., A. Pacey, and L.A. Thrupp, eds. 1989. Farmer first: Farmer innovation
and agricultural research. London, UK: Intermediate Technology Publications.
Edwards, P., R.S.V. Pullin, and J.A. Gartner. 1988. Research and education for the
development ofintegrated crop-livestock-fish farming systems in the tropics. ICLARM
Studies & Reviews 16. International Center for Living Aquatic Resources Manage-
ment, Manila, Philippines.
Egan, G.1988. Change agentskills: Assessing the designing ofexcellence. San Diego, Calif.:
University Association.
Essers, S., B. Haverkort, W. Hiemstra., and C. Reijntjes. 1989. Proceedings ofthe ILEIA
Workshop on Operational Approaches for Participatory Technology Development in


Journal for Farming Systems Research-Extension






FARMER-FIRST METHODS: FARMERS' DIAGRAMS


Sustainable Agriculture. Information Center for Low External Input Agriculture,
Leusden, The Netherlands.
Farrington, J., and A. Martin. 1988. Farmer participation in agricultural research: A
review of concepts and practices. ODI Agricultural Administration Unit, Occasional
Paper 9. Overseas Development Institute, London, UK.
Lightfoot, C. 1989. Farmer first approach. Pages 45-49 in J. van der Kamp and P.
Schuthof, eds., Methods of participatory technology development: Theoretical and
practical implications.. Information Center for Low External Input Agriculture,
Leusden, The Netherlands.
Lightfoot, C. 1990. Integration of aquaculture and agriculture: A route to sustainable
farming systems. NAGA, The ICLARM Quarterly 13 (1, January):9-12. Interna-
tional Center for Living Aquatic Resources Management, Manila, Philippines.
Lightfoot, C., and R. Barker. 1988. On-farm trials: A survey of methods. Agricultural
Administration e'Extension 30:15-23.
Lightfoot, C., O. de Guia, Jr., and F. Ocado. 1988. A participatory method for systems-
problem research: Rehabilitating marginal uplands in the Philippines. Experimental
Agriculture 24:301-309.
Maclean, J.L., and L.B. Dizon, eds. 1990. ICLARM Report 1989. International Center
for Living Aquatic Resources Management, Manila, Philippines.
Minnick, D. 1990. A case for courseware. International Brain Dominance Review
6(2):30-34.
Novak, J.D. 1977. A theory of education. Ithaca: Cornell University.
Sumberg, J., and C. Okali. 1988. Farmers, on-farm research and development of new
technology. Experimental Agriculture 24:333-342.
Tulving, E. 1989. Remembering and knowing the past. American Scientist 77(July/
August):361-367.
Waters-Bayer, A. 1989. Participatory technology development in ecologically-oriented
agriculture: Some approaches and tools. ODI Agricultural Administration (Research
and Extension) Network Paper 7. Overseas Development Institute, London, UK
World Resources Institute. 1988. A report by the World Resources Institute and the
International Institute for Environment and Development. New York: Basic Books,
Inc.


Vol. 2, No. 1, 1991








Sustainable Agriculture and Farming
Systems Research Teams in Semiarid West
Africa: A Fatal Attraction?1

J.L. Posner and Elon Gilbert



INTRODUCTION
In Africa, the current interest in sustainable agriculture (SA) represents a new
phase in efforts to improve agricultural production. Initially industrial prin-
ciples were applied to food crops in order to increase production. The "Green
Revolution" technologies met with little success in sub-Saharan Africa and the
farming systems research (FSR) approach began to gain favor in the mid-
1970s. This approach shifted the emphasis from high-input commodity
research to understanding the family farm-how it works, what its constraints
are, and what biological or economic "leverage points" exist that can be used
to increase food production. Understanding of farming systems increased
significantly in the process, but agricultural production per capital continued
to decline in most countries. More recently, researchers' attention has turned
to SA as a result of a continuing deterioration of the resource base in many
regions.
Research in SA seeks to improve long-term productivity and incomes
through a renewed emphasis on multiple cropping, integrated pest manage-
ment, genetic diversity, the natural maintenance of soil productivity (e.g.
nutrient cycling, organic matter management, pH maintenance, soil-struc-
ture amelioration), and improved water management (Francis et al., 1986;
Altieri, 1987; OTA, 1988). The semiarid tropics of West Africa present a
unique and difficult ecology for the application of this agenda. Rainfall is
unpredictable, and the low rural population density in the region results in
traditional production strategies that are extensive rather than intensive. The
population, however, is growing rapidly, and it is imperative that the productivity

1Paper presented at the Tenth Annual Association for Farming Systems Research-
Extension Symposium, Michigan State University, East Lansing, October 14-17,1990.
Agronomy Department, University of Wisconsin-Madison, and consultant, The
Gambia, West Africa.





POSNER AND GILBERT


of the system be increased. Concurrent with population growth, the economies
of the countries are deteriorating, which greatly restricts their ability to
support research necessary to develop sustainable technologies (Elliot, 1990).
This paper attempts to place the current sustainable agriculture program
within the context of the unique ecology of semiarid (600- to 1,200-mm
rainfall isohyets) West Africa. A brief review of the changing criteria for
evaluating agricultural performance and a discussion of the major technical
components of the SA agenda in West Africa highlight the complexity of the
task and dearth ofsolutions available. An overview oftrends in farming systems
will focus on the perplexing situation found in the region: one of population
increase without accompanying agricultural intensification. The authors
conclude that in the short run (10 years) there will be relatively few good
matches between actual production objectives of farmers and available SA
technologies.
The next section addresses the potentially fatal attraction between SA
research and the FSR teams in the field. The natural fit between SA concerns
and FSR methodology combined with increased donor interest in SA can lead
FSR teams to focus their efforts on long-term natural resource management.
SA research and technologies are unlikely to produce the short-term results
that farmers require and that FSR teams need to strengthen their status and
credibility. The final section of the paper deals with the question of what mix
of private, national, regional, and international organizations is best suited to
address SA issues so that they become part of the agricultural development
process in the semiarid tropics of West Africa.


EVOLUTION OF PERFORMANCE CRITERIA
Most agricultural research in West Africa has focused on increasing rural
welfare through increased production of basic food crops. Over the years, as
an understanding of the complexities of development has grown, the criteria
for evaluating new technology have become broader. Associated with these
new criteria are additional, increasingly demanding, field-monitoring require-
ments.
The initial commodity ("Green Revolution") programs took as their point
of departure increased productivity of the land and promoted the adoption of
a package of practices, including use of high-yielding varieties and fertilizers
and improved management. With this approach, major indicators of perfor-
mance were yield and maximum economic return per ha, usually measured on-


Journal for Farming Systems Research-Extension






SUSTAINABLE AGRICULTURE AND FSR IN AFRICA


station under researchers' management (see Table 1). The fundamental
philosophy was that the use of inputs would overcome environmental
differences and that basic packages would be sufficiently productive, so that
market forces would ensure the development of the necessary input delivery
systems. These accelerated food-crop production programs, however, have
had little impact on agricultural practices in the savanna of West Africa. There
are three reasons for their lack of success. (1) The main focus of the "Green
Revolution" technologies has been on irrigated rice and wheat and on hybrid
maize. Neither rice nor wheat are major crops in sub-Saharan Africa and
successes with maize have been mostly limited to eastern and southern Africa
(Lipton and Longhurst, 1989). (2) Work on African staples, such as millet,
sorghum, cowpeas, and cassava, is relatively recent (Wolf, 1986). (3) There
has been insufficient attention paid to the heterogeneity and unique charac-
teristics of the semiarid West African Sahel (Matlon and Spencer, 1984). Thus,
few introduced varieties have proven superior to local materials under farmers'
conditions (Matlon, 1985).
The introduction of a FSR approach shifted attention from returns to land
to the level of the farm family and its resources and aspirations (Harwood,
1979; Gilbert et al., 1980; Norman, 1980; Shaner et al., 1981). FSR

Table 1. Characteristics of Alternative Research Programs

Commodity Farming Systems Sustainable
Research Research Agriculture
Research

Objective increase production increase whole farm increase/maintain
through judicious use income production using
of purchased inputs renewable resources
and enhance future
productivity
Approach agro-industrial farm-management agro-ecology
principles
Level of field farm toposequence,
inquiry village, watershed
Time frame short-term short-term long-term
Performance *returns to land *farmer acceptance *resource
criteria *returns to labor conservation
*returns to capital *dependance on
*returns to land external inputs
*equity *genetic diversity
*risk analysis *returns to land
*equity


Vol. 2, No. 1, 1991





POSNER AND GILBERT


concentrates on better understanding local farming systems and their needs
in order to facilitate the introduction of new agricultural technology. A farm-
management approach is used, and social scientists join production biologists
to form "Farming Systems Research Teams." These teams conduct surveys
and trials, mainly on-farm. The FSR methodology brought several issues into
focus: (1) the importance of integrating crops and animals; (2) the aversion
that farmers have to financial risk; (3) the need to understand gender issues;
and (4) the rich knowledge base on which indigenous production systems are
founded. With this increased awareness, the evaluation of new technology
includes not onlyproduction and maximum-profit criteria, but also additional
economic indicators, such as multienterprise optimization, risk analysis,
calculation of transaction costs, returns to labor, and equity (Table 1). The key
evaluation criterion in this approach is farmer acceptance.
Growing concern for the impact of poverty and increasing population
growth on the tropical environment led to the establishment of the sustainable
agriculture movement (Brown and Wolf, 1985; Altieri and Anderson, 1986;
Dover and Talbot, 1987). This shifted the emphasis to an even higher level of
analysis. Within this framework, the time table is long-term and the focal point
is no longer the field or farm but the village or watershed (Table 1). The client
is no longer the individual farmer but the whole society. An agroecological,
resource-management approach to technology development is promoted, in
which foresters, water-management specialists, and ecologists are added to
the development team. Criteria for evaluating technology from the SA
perspective include determining the relative dependence on purchased inputs,
reducing run-offand erosion rates, minimizing leakages ofnutrients and water
in the agricultural cycle, and promoting increased genetic diversity. In
addition to production and profit data and household survey information, the
evaluation of technologies requires environmental monitoring over a number
of years and often over large areas.
The evolution of performance criteria has brought research systems a long
way from a simple one-dimensional focus on increasing the productivity of
land to a complex, multidimensional choice matrix that is tailored to individual
farming systems and takes both acceptability (individual and societal) and
sustainability into account. Conceptually this evolution is needed and
welcome, but the resulting increase in the scale and complexity of the research
agenda and methodologies dramatically increases the gap between what needs
to be done and what most national agricultural research systems (NARS) and
FSR teams are able to do.


Journal for Farming Systems Research-Extension





SUSTAINABLE AGRICULTURE AND FSR IN AFRICA


KEY COMPONENTS OF SUSTAINABLE AGRICULTURE IN
SEMIARID WEST AFRICA

In 1984 the Food and Agriculture Organization (FAO) of the United Nations
estimated that without adequate conservation measures, the area of rain-fed
cropland in Africa would decline by 16.5 percent by the year 2000 because of
land degradation. This loss of land, and of soil quality on the remaining
cropland, would lead to a 25 percent reduction in land productivity (OTA,
1988). The two main technical components of the SA agenda that respond to
this threat are improved water and nutrient management. A third SA component
is village-level organization; many SA technologies require community con-
cern and collective effort. This section describes the principal features of these
components. Table 2 divides the semiarid region of West Africa into four
ecologies and lists some of the key characteristics of each.
The essential principles in water management in the semiarid tropics are (1)
to manage rainfall so that there is maximum water infiltration (and nonerosive
removal of run-off) and (2) to optimize soil-moisture use by crops (Figure 1;
Barrow, 1988). Improving water management is difficult, not only because
rainfall is generally low and sporadic but intensities are often high and many
of the savanna soils have a surface crust when dry. It is the unpredictability of
the input, however, that makes it difficult to persuade farmers to invest in
improved water management. In wet years improvements may not pay for
themselves, whereas in very dry years often nothing will work. In addition,
Table 2. Characteristics of the Major Agroclimatic Zones
in the Semiarid Tropics of West Africa

Zone Annual Total Population Rural Production constraint
precipitation area population
density water soil fertility
(mm) (%) (%) (people/km2)
Sahelian < 350 24 16 7 little annual cropping
Sahelo-
Sudanian 350-600 30 19 20 ++++ +
Sudanian 600-800 21 59 19 +++ +++
Sudano-
Guinean 800-1100 24 6 9 + ++
a Plus sign indicates severeness of constraint. More plus signs mean a more severe
constraint.
Sources: Columns 1-5 based on World Bank data presented by P. Matlon (1989) at
the Summer Institute for African Agricultural Research, Madison Wisc. Columns 6
and 7 based on data from Sanders (1989).


Vol. 2, No. 1, 1991





POSNER AND GILBERT


Run-off'-- --------- ------ -< Run-on

Infiltration Uptake

Storage in soil in root zone



Drainage Capillary rise

Figure 1. Crop/Soil Water Balance (After Norman, 1984)
because of statistically rare heavy rainfall events, planning and construction for
huge volumes of water is necessary.
Following rainfall events, water on the ground follows well-understood
physical principles, and a number of improved management technologies have
been developed. These technologies are divided into those that can be
adopted by a single farm family and those that requirevillage-level coordination.
For example, the use of mulching or tied ridges will help improve infiltration
and reduce run-offon an individual field. Microcatchments can be used to
direct run-off towards individual trees or cropped areas and small rock
"diguettes" built across the slope can reduce the soil loss associated with run-
off. The introduction ofshorter cycle varieties, improved planting geometries,
new species, multiple cropping, improved weed control, and mulching can
help individual farmers optimize soil-moisture use by their crops. Construc-
tion of other interventions, such as contour berms, hedgerows, grass water-
ways, and integrated crop catchments, often require associations of farmers.
Physically, the water-management technologies work. The challenge for
researchers, however, is to make investments in water management sufficiently
attractive, either through enhanced yields or modest labor inputs, so that
farmers will undertake the additional work to adopt them. For example, tied
ridges take time to build and maintain, but, if they regularly increase yields of


Journal for Farming Systems Research-Extension





SUSTAINABLE AGRICULTURE AND FSR IN AFRICA


maize (Lal, 1987; Hulugalle, 1989) and cowpeas (Hulugalle, 1987), they may
become an accepted technology. By the same token, rock diguettes are an
attractive partial solution to reducing erosion because, even though they
require labor, they can be built during the dry season (Sanders, 1989).
Unlike the hydrologic cycle, inputs into traditional nutrient cycling are
generally stable and predictable. The main problem is that the biology and
chemistry of this cycling is not yet well enough understood to allow for
manipulations that readily enhance crop production. Biologically based
strategies to improve the fertility of the plateau soils have several components.
First and foremost, there is a focus on building up the organic content and
microbial populations in the soil, with the objective of reducing nutrient
leaching, synchronizing nitrogen and phosphorus availability with crop needs,
and increasing soil water-holding capacity (Figure 2; Rodale, 1989). The use
of grain legumes in rotations, green manuring, and planting of leguminous


Nutrient aching


Figure 2. Nutrient Cycling (After Rodale Institute, 1989)


Vol. 2, No. 1, 1991





POSNER AND GILBERT


trees are promoted to add biologically fixed nitrogen and readily decompos-
able organic matter to cropped fields. Additional inputs in the form of
compost and manure help to cycle nutrients from the homestead or rangeland
to the cropped fields. Another strategy is the use in crop rotations of deep-
rooted plants (e.g. pigeon pea and shrubs) that can serve as biological pumps,
returning nutrients leached below the root zone ofshallower-rooted crops to
the surface. Although all these techniques help to regenerate soil fertility, they
are generally labor-intensive. They may take six to eight years to stabilize a
degraded situation (Rodale, 1989) and another 15 to 20 years before
important yield increases might be realized in the semiarid tropics. In many
ways, the challenge of soil-fertility regeneration and maintenance in the
semiarid tropics is still in the hands of the station-based scientists, who must
discover how to manipulate the soil microflora and fauna in order to facilitate
substantial and faster increases in crop production.
Overall, rain-fed agriculture in the semiarid tropics is risky and degrades the
environment. Planting into a dry soil with little water-holding capacity and the
frequent occurrence of dry spells during the rainy season result in only modest
yields. Cropping leaves the soils exposed to erosion during the early rains and
the practices of shortening fallow periods and increasing animal populations
adversely affect soil fertility. Technologies exist that minimize the negative
effect of farming on the environment but they are generally labor-intensive
and their positive effects on crop production are less dramatic.
The need for collective action in addressing many SA concerns introduces
an additional level of complexity in the research and development process, yet
one that fortunately is not alien to West African rural societies. Development
efforts in the past have often served to undermine genuine indigenous
communal action at the village level by imposing organizations that are, at
best, conduits for subsidized inputs. These have tended to create dependency
and often fail when subsidized services are withdrawn. In fact, indigenous
village-level organizations may not be easily resurrected in the wake of the
failure of the externally promoted village-level institutions. Nevertheless, a
sense ofself-reliance byvillage-level organizations is essential to their successful
participation in sustainable development. Some agencies, notably nongov-
ernmental organizations (NGOs), have been sensitive to the need for self-
reliant village organizations, yet there continues to be strong pressures from
donor agencies and national governments to pursue development efforts at a
pace that often runs counter to healthy institutional development. Fortun-
ately, there is abundant information on participatory approaches to development


Journal for Farming Systems Research-Extension






SUSTAINABLE AGRICULTURE AND FSR IN AFRICA


(Esman and Uphoff, 1984; Biggs, 1989; Waters-Bayer, 1989) and a growing
number of promising examples (see ILEIA Newsletter; Amanor, 1989).
Together, they provide a strong indication that productive approaches to
village-level development exist.


TRENDS IN FARMING SYSTEMS IN
SEMIARID WEST AFRICA

Farming systems of West Africa are in flux due to both ecological and
economic reasons. Rainfall during the 1970s and early 1980s was particularly
poor and averaged one standard deviation below the long-term mean (Figure
3; Glantz, 1989). Because of the lower productive potential associated with
drought and transportation difficulties, fertilizer use on food crops in West
Africa was 1.4 kg of nitrogen, phosphorous, and potassium per ha in the late
1970s (cited in Matlon, 1987). Use has probably decreased since then,
because its relative cost has risen, further limiting yields. The unavailability of
fertilizer or unwillingness of farmers to buy it stagnates most production
systems, which reach a low-level equilibrium between nutrient removal as
yield and the occasional inputs from short fallows, manuring, or rotations.


50-
40-
30-
20-


-100-
0-


-20-
-30-
-40-
-5o II I I I I I I
1900 1910 1920 1930 1940 1950 1960 1970 1980
Time, in years

Figure 3. Rainfall Fluctuations in the Sahel and Sudan, Expressed as Percent
Departure From the Long-Term Mean, 1901-1984 (After Brown and Wolf, 1985)


Vol. 2, No. 1, 1991





POSNER AND GILBERT


In spite of these constraints, it is estimated that food production in West
Africa in the 1970s rose by an annual rate of 1.9 percent-mostly from
increasing the land under production of food crops (1.1%). Nevertheless, due
to the high rate of population growth (2.9%), per capital food production fell
by an annual rate of 2 percent (Paulino, 1987). Complicating the pressing
need to reverse this trend is the migration of many young adults away from the
rural sector and into urban centers. As a result, many farms are in the
anomalous position of having to produce more food with less labor.
Analysis, based on a population density/technology change model (Boserup,
1981), helps place the evolution of these farming systems in perspective. The
majority appear to be moving from lengthy bush-fallow rotations to a short-
fallow system. This transition means that land is often cropped six years in ten
rather than three years in ten. This shift is critical, because the natural
vegetation can no longer maintain soil fertility with shorter fallow periods.
Boserup (1981) found that these changes in cropping intensity are caused by
an increase in population density from approximately 10 persons per km2 in
the bush-fallow stage, to over 100 persons per km2 in the case of annual
cropping. Rural population densities in the region are only 7 to 20 persons per
km2 (see Table 2). These data suggest that land is fairly abundant in semiarid
West Africa.
Under abundant land conditions, Binswanger (1986) argues that farmers
need technologies that reduce their input requirement per unit of output.
Specifically, Binswanger concludes that this can occur with either the intro-
duction of stress-resistant varieties that result in higher yields or with the
adoption of labor-saving innovations. On the other hand, labor-intensive
technologies that improve soil fertility or reduce the probability of drought
stress will not be embraced enthusiastically. Sanders (1989) points out,
however, that plant breeding approaches have had only modest success in the
semiarid West African tropics over the past 25 years and are probably doomed
to continued failure until "environment-improving" techniques are in place,
which modify the intense water and fertility stresses of the area.
A second, logical, option is labor-saving innovations. In a review of 56
village studies in the semiarid tropics, Binswanger and Pingali (1987) found
that at population densities in the range of 16 to 50 persons per km2 (slightly
higher than current densities in West Africa), agricultural technologies, such
as animal traction, become common. According to this model, as population
increases, farmers are forced to adopt shorter and shorter fallow periods. This
results in less stumps in the fields and more time spent weeding. The use of


Journal for Farming Systems Research-Extension





SUSTAINABLE AGRICULTURE AND FSR IN AFRICA


oxen, donkeys, or horses becomes increasingly attractive because they permit
the cultivation of larger areas in a more timely fashion.
It is interesting to note that although animal traction was introduced in the
region shortly after World War II, it was only employed on about 15 percent
of the sown area by the mid-1980s (Spencer, 1985). In Burkina Faso, Jaeger
and Matlon (1990) found that the modest rate of adoption was because the
income gained by using animals was generally not high enough. They suggest
the three most common causes for the failure to adopt were: (1) the animals
were used for only one task (e.g. plowing) and were, therefore, underutilized;
(2) adoption did not result in increased production because neither yields nor
area cultivated increased; and (3) farmers, as well as credit and extension
services, did not appreciate the substantial cost of learning to use animal
traction (an average of 7 years), which resulted in unacceptable short-term
financial hardship. In parts of Senegal and The Gambia, where adoption is
almost universal, a variant of point two is highlighted. The use of animal
traction there does not appear to have resulted in increased production.
Rather, it has been driven by the desire to maintain production by speeding
up planting in the face of declining rainfall. Mechanization also permits two
weedings, despite the severe labor shortage in the area (Sumberg and Gilbert,
1988).
The generally low level of adoption in West Africa serves either to reinforce
the hypothesis that population densities are still low or that African farming
systems are experiencing growing populations but without the expected
evolution towards agricultural intensification. In either case, assuming that
Binswanger (1986) is correct, it can be predicted that labor-intensive solutions
to maintaining or increasing crop production will be met with little farmer
enthusiasm.
Just as the previous section raises questions about the current availability
of SA techniques that improve production, this section raises questions about
the priority that most Sahelian farmers accord to labor-intensive technologies.
In extensive production systems, resource preservation is not a priority. In
addition, in systems without animal traction manySA technologies (production
of forage legumes, composting, digging water catchments, building contour
berms) will be difficult to realize.
Most discussions of future trends assume that high population-growth
rates will shift the balance from labor to land scarcity in agriculture, which will
create conditions favorable to intensification along the lines that occurred in
Asia. This in turn will result in a recoupling of farming systems with the


Vol. 2, No. 1, 1991





POSNER AND GILBERT


principles of sustainable agriculture (ie. nutrient cycling, land stewardship,
multiple cropping) and higher levels of productivity will be achieved (Har-
wood, 1990). There are at least three reasons that dispute this scenario for
West Africa. First, migration rather than intensification is likely to continue to
be the primary strategy by which rural populations deal with labor surplus
(Colvin, 1981). Political and economic factors may progressively reduce the
attractiveness of migration, but not to the extent that flows will be stopped or
reversed. Second, terms of trade are likely to continue to run against the major
exports of the region groundnutss and cotton), reducing the attractiveness of
agriculture vis-a-vis other activities. Third, the environment will continue to
deteriorate, increasing the complexity and magnitude of the effort required to
reverse the trend.


THE FATAL ATTRACTION

In spite of its complexity, the SA agenda is compelling. The threatened loss
of productivity demands immediate attention. The approach is also scientif-
ically appealing for a number of reasons. It focuses on the landscape, rather
than individual farms or fields; its approach to soil fertility is in terms of
nutrient cycling, rather than response functions; it has the objective of
designing biologically sound production systems, rather than simply maxi-
mizing returns to land or labor; and it puts an emphasis on long-term
productivity, rather than immediate returns.
Farming systems research, with its emphasis on farmer participation,
multidisciplinarity, and location specificity is a seemingly logical home for
these sustainable agriculture concerns (Francis and Hildebrand, 1989). The
diagnostic methodology (rapid rural appraisals, recommendation domains,
formal surveys) can be used to characterize farming systems and their
interaction with the environment, a crucial step in sorting through the range
of possible SA interventions. The design and promotion of these technologies,
as is the case with FSR research, requires multidisciplinary teams of biological,
physical, and social scientists. Site specificity, a major concern in developing
SA technologies, already looms large in the FSR methodology because it
emphasizes on-farm trials. In addition, the farmer-participation model in FSR,
which encourages a two-way dialogue between farmers and researchers, is also
crucial to the SA approach.
The fit is not perfect, however, in terms of most NARS and their FSRteams.
A shift in focus from adapting production technologies to a resource-


Journal for Farming Systems Research-Extension





SUSTAINABLE AGRICULTURE AND FSR IN AFRICA


management outlook has several important drawbacks. The lack of proven
technologies, the scientific complexity ofSA research and monitoring, and the
frequently long lead times before results can be seen may only contribute to
the erosion of credibility that national-level FSR and agricultural research
institutions are now experiencing. The field testing and transfer of these
technologies will require a commitment of resources to a specific location for
several years that is difficult for most national programs to make or, more
importantly, deliver. The probable lack of farmer interest in labor-intensive
solutions without some incentives can marginalize the teams and separate
them from their farmer clientele. Equally important, a shift in focus will have
to be made at the expense of their agendas, which typically address farmers'
interest in raising net incomes through increased production and/or reduc-
tions in unit costs. The lure of donor funding may induce some to adopt the
SA agenda, but the comparative advantage of these teams, as currently
constituted, is primarily at the farm level, addressing farmer-identified pro-
duction constraints.
Many of these concerns apply to national research systems as well as
individual FSR teams. Shifts in donor interest away from commodity-based as
well as farming systems research toward sustainable agriculture issues can tilt
research agendas in directions that many NARS currently do not have the
capacity to deal with effectively.


IS THERE A PLACE FOR SUSTAINABLE AGRICULTURE
IN THE CURRENT WEST AFRICAN AGRICULTURAL
RESEARCH AGENDA?

The preceding comments suggest that FSR teams in the sub-Saharan region
are not well advised to invite the SA elephant into their fragile rowboat. This
does not mean the subject should be put aside or that FSR methods should
not be employed in moving toward more sustainable systems. SA research
should be undertaken now to generate the technologies that will fit into
farming systems in 10 to 15 years. This might take the form of special,
medium- to long-term projects in areas that are already representative of what
the situation will likely be by the turn of the century in terms of agricultural
intensification and/or environmental deterioration. Although resource con-
servation is needed for all of semiarid West Africa, only some areas (e.g.
Senegalese Peanut Basin, Mossi Plateau of Burkina Faso) are already experi-
encing the severe stress that threatens to become widespread in the future.


Vol. 2, No. 1, 1991





POSNER AND GILBERT


Increasing productivity through improved water- and nutrient-manage-
ment techniques will require both basic and applied research on-station, as
well as on-farm. In both arenas, the testing of technologies will involve a
commitment of resources (funds and staff) for several years without major
interruptions. Some combination of the larger national agricultural research
programs, international centers, and universities is probably best suited to
provide the needed leadership and continuity of effort that is required to yield
useful results. The effort should selectively incorporate FSR methodologies,
including farmer-participatory research, formal and informal surveys, and on-
farm trials. Smaller national institutions and FSR teams, in particular, can play
important supporting roles, but ones that are defined in a fashion that is
consistent with their own priorities and capacities. SA research should
contribute to the growth of research institutions and capacities at the national
level. In the near term, this means a primary focus on identifying, testing, and
disseminating productivity-increasing innovations that are attractive to farm-
ers and nearly ready to be used.
It would be a disservice to both the FSR teams and the overall SA agenda
to assign an undertaking of this complexity and magnitude to short-term,
donor-funded projects attached to FSRunits. The commitment to sustainable
agriculture needs to be much greater, if there is to be convergence within the
next decade between local farming systems, available technologies, govern-
mentpolicies, and national agricultural research capacities that will provide for
a widespread dissemination of resource-conserving agricultural practices.


REFERENCES
Altieri,MA. 1987.Agroecology, thescientificbasisofalternativeagriculture. Boulder, Colo.:
Westview Press. 227 pp.
Altieri, M.A., and M.K. Anderson. 1986. An ecological basis for the development of
alternative agricultural systems for small farmers in the third world. American Journal
ofAlternative Agriculture 1(1):30-39.
Amanor, K. 1989.340 abstracts onfarmer participatory research. Research and Extension
Network Paper No. 6. Agricultural Administration Unit, Overseas Development
Institute, London, UK
Barrow, C.J. 1988. The present position and future development of rain-fed agriculture
in the tropics. Outlook on Agriculture 17(3):112-119.
Biggs, S. 1989. Resource-poorfarmer participation in research: A synthesis of experiences.
OFCOR Comparative Study Paper No. 3. ISNAR, The Hague.
Binswanger, H.P. 1986. Evaluating research system performance and targeting research
in land abundant areas of sub-Saharan Africa. World Development 14(4):469-475.


Journalfor Farming Systems Research-Extension






SUSTAINABLE AGRICULTURE AND FSR IN AFRICA


Binswanger, H.P., and P.L. Pingali. 1987. The evolution of farming systems and
agricultural technology in sub-Saharan Africa. Pages 283-318 in V. Ruttan and C.
Pray, eds., Policy for agricultural research. Boulder, Colo.: Westview Press.
Boserup, E. 1981. Population and technology. Oxford, UK: Basil Blackwell.
Brown, L.R., and E.C. Wolf. 1985. ReversingAfrica's decline. Worldwatch Institute Paper
No. 65. Washington, D.C.
Colvin, L.G., ed. 1981. The uprooted of the western Sahel: Migrants' quest for cash in the
Senegambia. New York, N.Y.: Praeger Publishers.
Dover, M., and L.M. Tabot. 1987. To feed the earth: Agroecology for sustainable
development. Washington, D.C.: World Resources Institute.
Elliot, H. 1990. National agricultural research systems in developing countries: A report
on progress. Paper presented at the University of Tuscia, Viterbo, Italy, February 1-
2, 1990.
Esman, M.J., and N.T. Uphoff. 1984. Local organization: Intermediaries in rural de-
velopment. Ithaca: Comell University Press.
Francis, C.R., R. Harwood, and J. Parr. 1986. The potential for regenerative agriculture
in the developing world. American Journal ofAlternative Agriculture 1(2):65-74.
Francis, C.R., and P.E. Hildebrand. 1989. Farming systems research-extension and the
concept of sustainability. Farming Systems Research-Extension Newsletter No. 3:6-11.
Gilbert, E.H., D.W. Norman, and F.E. Winch. 1980. Farming systems research: A critical
appraisal. Rural Development Paper No. 6. Department of Agricultural Economics,
Michigan State University.
Glantz, M. 1989. Drought, famine, and the seasons in sub-Saharan Africa. In R. Huss-
Ashmore and S. Katz, eds., Anthropologicalperspectives on the African famine. New
York, N.Y.: Katz, Gordon, and Breach Scientific Publications.
Harwood, RR. 1979. Smallfarm development: Understanding and improving farming
systems in the humid tropics. Boulder, Colo.: Westview Press.
Harwood, R.R. 1990. The farming systems decoupling/integration cycle as a determinant
of research approach. Keynote address presented at the Tenth Annual Association for
Farming Systems Research-Extension Symposium, October 14-17, 1990, Michigan
State University, East Lansing.
Hulugalle, N.R 1987. Effect of tied ridges on soil water content, evapotranspiration,
root growth and yield ofcowpeas in the sudan savannah of Burkina Faso. Field Crops
Research 17:219-228.
Hulugalle, N.R. 1989. Effect of tied ridges and undersown Stylosanthishamata L. on soil
properties and growth of maize in the sudan savannah of Burkina Faso. Agriculture,
Ecosystems, and Environment 25:39-51.
Information Centre for Low-External-Input and Sustainable Agriculture (ILEIA).
Quarterly Newsletter. P.O. Box 64, 3830 AB Leusden, The Netherlands.
Jaeger, W.K., and P.J. Matlon. 1990. Utilization, profitability, and adoption of animal
draft power in WestAfrica. AmericanJournalofAgriculturalEconomics72():35-48.
Lal, R. 1987. Managing the soils of sub-Saharan Africa. Science 236:1069-1076.
Lipton, M., and R. Longhurst. 1989. New seeds and poor people. Baltimore, Md.: Johns
Hopkins University Press.
Matlon, P.J. 1985. A critical review of objectives, methods and progress to date in
sorghum and millet improvement: A case study of ICRISAT/Burkina Faso. Pages
154-179 in H. Ohm and J. Nagy, eds., Appropriate technologies for farmers in semi-
arid West Africa. West Lafayette, Ind.: Purdue University Press.


Vol. 2, No. 1, 1991






POSNER AND GILBERT


Matlon, P.J. 1987. The West African semiarid tropics. Pages 59-88 in J.W. Mellor, C.L.
Delgado, and M.J. Blackie, eds., Acceleratingfood production in sub-Saharan Africa.
Baltimore, Md.: Johns Hopkins University Press.
Matlon, P., and D. Spencer. 1984. Increasing food production in sub-Saharan Africa:
Environmental problems and inadequate technical solutions. American Journal of
Agricultural Economics 65(5):670-676.
Norman, D.W. 1980. The farmingsystems approach: relevancyfor the smallfarmer. Rural
Development Paper No. 5. Department of Agricultural Economics, Michigan State
University.
Norman, M.J.T.,etal. 1984. The ecology oftropicalfood crops. Cambridge University Press.
Office of Technology Assessment (OTA), U.S. Congress. 1988. Enhancing agriculture
inAfrica:A rolefor U.S. development assistance. OTA-F-356. Washington D.C.: U.S.
Government Printing Office.
Paulino, LA. 1987. The evolving food situation. Pages 23-38 in J.W. Mellor, C.L.
Delgado, and M.J. Blackie, eds., Acceleratingfood production in sub-Saharan Africa.
Baltimore, Md.: Johns Hopkins University Press.
Rodale Institute. 1989. Soil degradation and prospects for sustainable agriculture in the
peanut basin of Senegal. Report to USAID, Dakar, Senegal, August 15, 1989.
Sanders, J.H. 1989. Agricultural research and cereal technology introduction in Burkina
Faso and Niger. Agricultural Systems. 30:139-154.
Shaner, W.W., P.F. Philipp, and W.R. Schmehl. 1981. Farming systems research and
development: Guidelines for developing countries. Boulder, Colo.: Westview Press.
Spencer, D.S.C. 1985. A research strategy to develop appropriate agricultural technol-
ogies for small farm development in sub-Saharan Africa. Pages 308-326 in H. Ohm
and J. Nagy, eds., Appropriate technologiesforfarmers in semi-arid West Africa. West
Lafayette, Ind.: Purdue University Press.
Sumberg, J., and E. Gilbert. 1988. Draft animals and crop production in The Gambia.
Gambian Agricultural Research Service, Cape St. Mary, Banjul, The Gambia. Mimeo.
Waters-Bayer, A. 1989. Participatory technology development in ecologically oriented
agriculture: Some approaches and tools. Research and Extension Network Paper No.
7. Agricultural Administration Unit, Overseas Development Institute, London, UK.
Wolf, E.C. 1986. Beyond thegreen revolution: New approachesfor third world agriculture.
Worldwatch Institute Paper No. 73. Washington, D.C.


Journal for Farming Systems Research-Extension








Crop-Enterprise Selection of Farming
Systems in Eastern Gambia'

M.B. Kabay and Lydia Zepeda2



INTRODUCTION
In The Gambia, 90 percent of the country's population lives in rural areas
where crop production is the major source of food and income. Producer
prices for major crops are determined by a public agency, The Gambia
Produce Marketing Board, which uses price as an incentive for increasing
food- and cash-crop production. For such policy to be effective, it is important
to understand how farmers adjust production decisions given different
resource endowments and under different rainfall conditions.
The purpose of this paper is to model farmer response to different
production and rainfall conditions in order to assess the effects of agricultural
policies on food production and farmer income. The problem of determining
optimal crop combinations underdifferentscenarios, given resource constraints,
is one of constrained optimization: selecting among alternative crop com-
binations to maximize net farm returns. The specific objectives of the study are
(1) to identify optimal crop combinations at current Gambian price levels,
given available resources; (2) to identify farm constraints and to determine
their effect on food- and cash-crop production; and (3) to simulate farmer
response to different producer-pricing scenarios. The first two objectives are
addressed in this paper.
Differences in resource endowments and family size affect how much of an
input can be committed to a production process. Therefore, farms reflecting
different resource endowments and family size are modeled. Whether the
perspective of the analysis is the whole farm or a single crop also affects the
allocation of resources. Most economic and agronomic on-farm research in
The Gambia has focused on single-crop enterprises. Because farmers make
most of their decisions from a whole-farm perspective, it is appropriate to use

1Paper presented at the Tenth Annual Association for Farming Systems Research-
Extension Symposium, Michigan State University, East Lansing, October 14-17,1990.
The authors are, respectively, a graduate student and an assistant professor in the
Department of Agricultural Economics at the University of Wisconsin-Madison.





88 KABAY AND ZEPEDA

research designs that are consistent with farmers' thinking. Thus this study
uses a whole-farm approach. The erratic rainfall patterns of The Gambia
necessitate the incorporation of different rainfall scenarios into an analysis of
production decisions.
Background information for The Gambia study area and the farming
systems is discussed in the following section. The data are analyzed and used
to estimate a linear-programming model to represent farmer decisions given
their endowments. The model generates the cropping decisions for different
policy scenarios; the results are used to infer the aggregate effects of policies
on production. Implications of these policies are discussed in the final section.

The Study Area
With a land area of 4,000 square mi, The Gambia is one of the smallest
countries in Africa. The country is completely surrounded by Senegal on its
northern, southern, and eastern borders, and its western boundary is the
Atlantic Ocean. The River Gambia bisects the country and is navigable along
most of its length. According to the 1973 census, the country's population
was 493,555, with 90 percent living in rural areas. The 1983 census showed
a 41 percent increase in total population. This imposes considerable pressure
on available resources for agriculture and other sectors of the economy. The
country has a Sahelian climate with annual rainfall ranging from 800 mm in
the north to 1,000 mm in the south. Rains start in June/July and end in
November, the period during which upland crops are grown. The rainfall
pattern, however, is very erratic. Farms are diversified and emphasize some
level of food self-sufficiency. The major food crops in the country include
millet, sorghum, maize, and rice. Peanut, cotton, sesame, and vegetables are
the important cash crops. The gross domestic product for 1984-85 was
estimated at US$179 million, 75 percent of which was accounted for by the
agricultural sector.

The Farming Systems
This study focuses on the eastern Gambia because of its diverse resource
endowments and farming systems. In a soil fertility-management survey,
Boughton et al. (1987) classified the eastern Gambia into Riverine Mandinka,
Inland Wolof, and Extensive Upper River farming systems. This classification
was based on village location, rainfall amount and pattern, resource endow-
ment, and available fertility-management options. Because this study covers


Journal for Farming Systems Research-Extension





CROP-ENTERPRISE SELECTION IN THE GAMBIA


the same location, two of the farming systems identified by Boughton et al.
(1987) are used (Inland Wolofand Extensive Upper River).
Inland Wolofvillages are located on uplands away from the River Gambia.
Level of investment in mechanization (primarily animal traction) is relatively
high. Animal herds out-migrate in the dry season, limiting access to manure
for farms in the region. Also, the heavy soils in this region are unsuitable for
production of most crops. The short growing season requires cultivation of
early maturing crops such as early millet.
Extensive Upper River villages are located close to the River Gambia. The
shortage of upland area limits fallow. However, the availability of rice land
partially mitigates the pressure on land. Herds are resident in the dry season,
and their manure is an important component of the soil-fertility management
system. The level of investment in mechanization is relatively low in this area.
The growing season is longer than in the Inland Wolof region, allowing for
the cultivation of late-maturing crops such as late millet.


METHODS

The problem of choosing cropping enterprises subject to limits on land, labor,
and other resources can be formulated as a constrained optimization problem.
Linear-programming (LP) techniques are used to model and simulate farmer
response to different production, weather, and price scenarios. LP is appro-
priate for whole-farm analysis of alternative production choices, given resource
constraints.
The objective function is represented by the sum of net farm returns for
each possible production, consumption, purchase, and sale activity. Net
returns per unit of the activity are known. The objective is to determine the
level of each activity that will maximize the farm's net income. Land, labor,
cash, technological, and cultural constraints reflect the resources available to
the farmer. The optimal cropping system is one in which the input of resources
has the greatest return. Only if there are no binding constraints will the most
profitable activity be chosen, to the exclusion of all other activities. The
optimal solution indicates the crops to be grown, sold, and purchased, as well
as the resources to be used.


3 Mills et al. (1988) found investment in implements and draft animals was 7,492 dalasis
per farming unit for an Inland Wolof village and D3,817 for an Extensive Upper River
village.


Vol. 2, No. 1, 1991





KABAY AND ZEPEDA


Alternatives (millet, sorghum, maize, rice, peanut, cotton, and sesame)
available to producers in the Inland Wolofand Extensive Upper River areas
are modeled to compare producer-cropping decisions. The models are
constructed with secondary data from The Gambia and Senegal (Department
of Planning, The Gambia, 1987, 1988; Martin, 1988; Mills et al., 1988).
Alternative farm size and weather conditions are modeled to assess the impact
of resource endowments and weather on the cropping choices within each
region. Three farm categories are identified based on resource endowment
and family characteristics: small (2.5 ha), medium (5 ha), and large (10 ha).
Three rainfall states (good, normal, and bad)4 are incorporated to reflect the
erratic rainfall patterns in the study area.
Sensitivity analysis and shadow price determination are results of LP
analysis that are useful for policy analysis. Sensitivity analysis provides the
range in net returns per activity and the range in resource constraints over
which the solution is stable. This is used to determine how responsive farmers
are to changes in prices, and how resource availability limits their decision
making. Shadow prices are the imputed value of resources, or the value of
additional resources, to the farmer's net returns. Resources that are not
binding have a shadow price of zero. For instance, unused labor is worth
nothing to the farmer. If all the labor were used, the shadow price would
indicate how much net return would increase if the farmer had additional
labor. In other words, the shadow price indicates the maximum amount the
farmer would be willing to pay for additional labor. Shadow prices are useful
in guiding policy with respect to improving availability of resources.


RESULTS


Optimal Cropping Plan Under Current Prices
Optimal crop combinations under the various rainfall states for small,
medium, and large farms are shown in Table 1. The optimal cropping
strategies for medium and large farms tend to be different than those for small
farms. In a normal year, two thirds of the total acreage of both large and
medium farms is put into cash-crop production (peanut). The remaining


4Normal rainfall states are 873 mm and 941 mm, respectively, for Inland Wolof and
Extensive Upper River Farming Systems. Bad rainfall equals less than 70 percent of
normal, and good rainfall equals more than 100 percent of normal.


Journal for Farming Systems Research-Extension





CROP-ENTERPRISE SELECTION IN THE GAMBIA


Table 1. Optimal Crop Combination for Different Rainfall Statesa

Farm size Percentage of total acreage
Crop Good year Normal year Bad year
Large Peanuts 44.4 68.8 44.6
Millet 53.3 28.5 52.0
Sorghum and maize 2.3 2.7 3.4
Medium Peanuts 40.0 74.2 40.0
Millet 57.8 23.4 57.0
Sorghum and maize 2.2 2.4 3.0
Small Peanuts 0.0 65.6 0.0
Millet 99.0 31.2 99.0
Sorghum and maize 1.0 1.0 1.0
a Reflects 1989 price levels.

cultivated acreage goes to food-crop production (millet, sorghum, and
maize). If the rainfall situation is bad, the planting strategy changes in favor
of millet production. This reflects the farm household's concern for food self-
sufficiency. In a good year, millet occupies more than half the cultivated
acreage; however, the solution to this excess includes the sale of millet. This
suggests that at current price levels, millet can be a profitable crop in a good
rainfall year. Large and medium farms meet their food self-sufficiency re-
quirements in all three rainfall states.
Small farms follow a different planting strategy. The major cash crop,
peanut, is only cultivated during a normal rainfall year. This is probably due
to the higher labor requirement for peanut cultivation in a good year. Some
millet is sold during good years. During a bad year the small farm is under
tremendous pressure to meet the food requirements of the family, forcing the
entire cultivated acreage to be devoted to food-crop production.

Farm Profitability and Constraint Levels
In all three farm categories, output and net return per farm is greatly
influenced by rainfall. Because prices are fixed by the government, yield
variability translates directly into income variability. The results in Table 2a
show returns per farm decreasing as the rainfall declines. Smaller farms have
a higher net return per ha cultivated. The same trend holds true for returns per
labor unit. This suggests that the food requirements of larger households
more than offset the marginal benefits of the extra labor provided.
Land is a binding constraint for small and medium farms, with smaller farms
having higher shadow prices, from 1,000 to 1,370 dalasis (The Gambian


Vol. 2, No. 1, 1991





KABAY AND ZEPEDA


Table 2. Net Returnsa per Farm, Land Productivity Based onbReturnsa
per Ha Owned, and Net Returns? per Ha Cultivated

Farm size Rainfall state Expected
Good Normal Bad value
(A) NET RETURNS
Large 5,190.6 4,521.7 2,769.5 4,226.4
Medium 4,790.8 4,100.6 2,927.6 3,968.9
Small 2,675.0 2,319.2 1,761.7 2,264.1

(B) LAND PRODUCTIVITY
Large 519.1 452.2 276.9 425.0
Medium 958.2 820.1 585.5 793.8
Small 1,070.0 927.7 704.6 905.7

(C) CULTIVATION AND PRODUCTIVITY
Large 798.6 685.1 426.2 645.5
Medium 958.2 820.1 585.5 793.8
Small 1,070.0 927.7 704.6 905.7
SReturns are in the Gambian currency, dalasis.
Area cultivated: Good year 6.5 ha; normal year = 6.6 ha; and bad year 6.5 ha.



currency) per ha as opposed to the 630 to 995 dalasis per ha for medium farms.
This is contrary to the conventional wisdom that land is an abundant resource
in rural areas of The Gambia. What is frequently overlooked is that a high
proportion of available land is marginal land or unsuitable for crop production
with current technologies. The very small area that is suitable for maize
production reiterates this point. Shadow prices for maize land for medium and
large farms are greater than that for ordinary land (Table 3), indicating that
the return for maize production is very high. Technology and research that
could expand the area of land cultivated with maize would have a greater
return to the farmer than would expansion of any other crop.
The return generated per ha is dependent upon the size of the farm. As farm
size increases the return per ha owned or farmed declines (see Table 2b and
2c). The return per ha owned for small farms is twice that of large farms. This
is further justification for land redistribution: aggregate incomes and produc-
tion would increase if there were fewer large farms.
Labor for weeding is binding for small and medium farms. This implies that
new technologies that are labor intensive might not work for farms in these
categories. Labor during harvesting periods is also binding for all three farm
categories. Research to alleviate these bottlenecks in labor supply would be a


Journal for Farming Systems Research-Extension





CROP-ENTERPRISE SELECTION IN THE GAMBIA


Table 3. Labor Productivity: Net Returnsaper Adult Equivalent (AE)
Farm size AE Rainfall state Expected
Good Normal Bad value
Large 11.4 455.4 396.6 242.1 370.7
Medium 6.9 694.3 594.3 424.3 575.2
Small 4.8 557.3 483.2 367.0 471.7
L Returns are in the Gambian currency, dalasis.


desirable policy to improve the productive capacity and incomes of small and
medium farms. Returns to labor are greatest for medium-sized farms (Table
4).
Small farmers do not use all their available cash for production because they
use technologies that require low cash investment. Because small farmers have
no need for credit in their optimal cropping plan, subsidizing credit offers no
benefit to them. The Gambia Cooperative Union has been facing problems
recovering loans made to small farmers. This is probably because most of the
credit obtained is not used in income-generating agricultural activities. The
Union thus needs to reassess its current credit policy.
Medium and large farmers do use credit in their optimal-production plans.
Both medium and large farmers have adequate credit, and medium farmers do
not exceed their borrowing limit. Large farmers do. The shadow price on the
borrowing limit indicates that large farmers are willing to pay over 100 percent
in interest to exceed the limit. This is consistent with the actual rates charged
by moneylenders in the informal sector of The Gambia. Therefore, credit
availability is most profitable for the large farms. Credit subsidies would
benefit large farmers, who would be willing to pay much more for credit.

Sensitivity Analysis
The range in constraint values over which the solution is stable is deter-
mined by sensitivity analysis. Large farms have wider ranges in the land
constraint over which the optimal plan is stable than do small and medium
farms. This is because land is binding for small and medium farmers whereas
large farmers have land surpluses. Large farmers do not cultivate all of their
available land in the optimal solution. Medium and small farmers, however,
cultivate all of their available land, and decreases in the amount of land
available would change the solution. Small farmers have more flexibility in
terms of operating cash, however. This is because small farmers do not use all


Vol. 2, No. 1, 1991





KABAY AND ZEPEDA


Table 4. Binding Constraints
Farm size Variable Shadow price value
Good year Normal year Bad year
Large Maize land 4,441.2 4,013.1 3,154.8
Labor 23.7 22.6 20.1
Operating cash 1.4 1.3 1.0
Borrowing limit 1.3 1.2 0.9
Medium Land 995.0 803.2 629.6
Maize land 1,100.1 1,115.2 1,065.5
Labor 10.8 10.8 10.8
Operating cash 0.08 0.08 0.08
Small Land 1,368.0 1,259.4 1002.6
Maize land 992.9 1,065.1 987.9
Labor 10.0 10.0 10.0
Self-sufficiency 0.04 0.04


their cash for production. Because they do not need credit, large changes in
interest rates do not affect the production decisions of small farmers.
Range analysis of variable costs for production shows that solutions are
more stable in better rainfall years. This suggests that farmers would be less
willing to commit resources with cash requirements to agricultural production
in bad rainfall years. Personal discussions with farmers in the Inland Wolof
support this.
Larger farms have less stable solutions for cash crops than smaller farms,
suggesting that larger farms are more sensitive to changes in production costs
for cash crops. This is probably because smaller farms use technologies that
require lower cash investments. McIntire (1982) found that small farmers in
Upper Volta (now Burkina Faso) were unwilling to commit cash resources to
agricultural production. This is consistent with subsistence-production goals
of most farmers in the West African semi-arid tropics.


CONCLUSIONS AND IMPLICATIONS

The results of a linear-programming (LP) model are used to assess alternative
production activities available to farmers in the eastern Gambia. The objective
of the model is to choose activities that maximize farm returns subject to
resource constraints. The alternatives for small, medium, and large farms are
analyzed to compare optimal cropping strategies and binding resources for
each. Different rainfall scenarios are explored to determine how much impact
weather has on cropping decisions.


Journal for Farming Systems Research-Extension





CROP-ENTERPRISE SELECTION IN THE GAMBIA


The results of the analysis indicate that, contrary to popular opinion, land
shortages are a problem for small and medium farms. Large farmers are unable
to utilize all their land, however. This indicates that some form of land
redistribution would better utilize the resources available in the eastern
Gambia and would increase aggregate production and income. It should be
cautioned that no accounting for land quality has been made in this analysis,
and much of the land held by large farmers may be marginal land.
The apparent surplus of land on large farms may also mask the occurrence
of "strange farming." Strange farming occurs when landless people exchange
their labor to landowners for the right to use a plot of land. Thus, large farmers
can overcome their shortage of labor by using a resource they have plenty of,
land, rather than one they do not have enough of, cash. It is not known to what
extent strange farming exists in The Gambia, but the results of this research
indicate that it is likely to be prevalent among large farmers. Research is needed
to determine its prevalence, and what impact it has on aggregate agricultural
production and efficiency and in overcoming labor shortages.
Maize land has a large return for all farm sizes, under all weather conditions.
The limiting factor is that not much land is suitable for maize cultivation.
Research directed at improving technology to cultivate more maize or to bring
more land into maize production would have greater returns for medium and
large farmers than research for any other crop. The implication is that funds
should be diverted from cotton to maize research.
Labor is a binding resource on small and medium farms during the weeding
and harvesting seasons. Research on improving tools or machines to be used
for weeding and harvesting could alleviate this constraint and improve the
incomes of small and medium farmers.
Small farmers do not need extra cash under their optimal cropping plan.
Thus, they have no need for credit for production purposes. This would
indicate that the large number of defaults by small farmers on their loans is due
to using the loans for purchase of nonfarm goods. Medium and large farmers
do use credit in their optimal cropping plan. Medium farmers do not utilize
the full amount available, while large farmers do. The shadow price on the
borrowing limit of large farmers indicates that they would be willing to pay
over 100 percent interest to borrow more money. In other words, credit has
a larger payoff to large farmers, thus credit subsidies benefit them the most.
The implication is that credit subsidies are regressive and assist large farmers
to increase their agricultural production and incomes.


Vol. 2, No. 1, 1991





KABAY AND ZEPEDA


The implications for Gambian agricultural policy are that funds directed at
maize research and alleviation of seasonal labor shortages would have the
greatest impact on aggregate agricultural production and incomes in the
eastern Gambia under current prices. Land redistribution to small and
medium farmers would also increase production and aggregate farm income.
What remains to be evaluated are the pricing policies themselves and how they
influence producers' decisions.


SELECTED REFERENCES
Boughton, D., M. Kabay, and B. Mills. 1987. Findings of a rapid village appraisal of
farmers' soilfertility management strategies in the Kuntaur and Giroba cluster areas.
GARP Working Paper No. 4. Department of Agricultural Research, The Gambia.
Crawford, E.W. 1982. A simulation study ofconstraints on traditionalfarming systems in
northern Nigeria. MSU International Development Paper No. 2. Department of
Agricultural Economics, Michigan State University, East Lansing.
Dykstra, D.P. 1984. Mathematical programmingfor natural resource management. New
York: McGraw-Hill.
Department of Planning. 1987 and 1988. National agricultural sample survey. The
Gambia.
Eastman, C. 1986. Traditional Gambian land tenure and the requirements ofagricultural
development. Gambian Mixed Farming and Resource Management Project, The
Gambia.
Haydu et al. 1986. Mixed farming in the Gambia. Mixed Farming Technical Report No.
10. Gambian Mixed Farming and Resource Management Project, The Gambia.
Martin, F. 1988. Food security and comparative advantage in Senegal: A micro-macro
approach. Ph.D. dissertation, Department ofAgricultural Economics, Michigan State
University, East Lansing.
McIntire, J. 1982. Reconnaissance socioeconomic surveys in north and west Upper Volta.
ICRISAT West African Economics Program Progress Report No. 3. International
Crop Research Institute for the Semi-Arid Tropics, Ouagadougou, Upper Volta.
Mills, B., M. Kabay, and D. Boughton. 1988. Soilfertility managemetnstrategies in three
villages of eastern Gambia. Department of Agricultural Research, The Gambia.
Ministry of Agriculture. 1987. Department of Agricultural Research annual research
report. Republic of The Gambia.
Ramaswamy, S., andJ. Sanders. 1989. Populationpressure, new agricultural technologies,
and sustainability on the western plateau of Burkina Faso. Purdue University,
November.


Journal for Farming Systems Research-Extension




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs