• TABLE OF CONTENTS
HIDE
 Front Cover
 Title Page
 Preface
 Acknowledgement
 Table of Contents
 Introduction
 Description of the project...
 Organization of the OFAR progr...
 Description of target domains and...
 System constraints and opportunities...
 Design of on-farm trails
 Result of on-farm trails
 Farmers' opinions about interv...
 Early adopters
 Conclusion
 Reference
 Back Cover






Group Title: OFR bulletin; no. 1
Title: A case study of on-farm adaptive research in the Bida Agricultural Development Project--Nigeria
CITATION PAGE IMAGE ZOOMABLE PAGE TEXT
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00053844/00001
 Material Information
Title: A case study of on-farm adaptive research in the Bida Agricultural Development Project--Nigeria
Series Title: OFR bulletin
Physical Description: 29 p. : ill. ; 28 cm.
Language: English
Creator: Ashraf, M
Balogun, P
Jibrin, A
International Institute of Tropical Agriculture
Publisher: International Institute of Tropical Agriculture
Place of Publication: Ibadan Nigeria
Publication Date: <1985>
 Subjects
Subject: Agriculture -- Technology transfer -- Nigeria   ( lcsh )
Cropping systems -- Research -- Nigeria   ( lcsh )
Agriculture -- Research -- On-farm -- Nigeria   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Bibliography: Bibliography: p. 28-29.
Statement of Responsibility: M. Ashraf, P. Balogun, A. Jibrin.
 Record Information
Bibliographic ID: UF00053844
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: African Studies Collections in the Department of Special Collections and Area Studies, George A. Smathers Libraries, University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 001034581
oclc - 14547083
notis - AFB6912

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Title Page
    Preface
        Preface
    Acknowledgement
        Acknowledgement
    Table of Contents
        Table of Contents
    Introduction
        Page 1
    Description of the project area
        Page 1
        Page 2
        Page 3
    Organization of the OFAR program
        Page 4
    Description of target domains and system productivity
        Page 4
        Initial delineation of target domains
            Page 5
        Refinement of target domains
            Page 5
            Page 6
            Page 7
            Page 8
        Farm productivity
            Page 9
            Page 10
            Page 11
            Page 12
    System constraints and opportunities for experimentation
        Page 13
        The rice-based system
            Page 13
            Page 14
            Page 15
        Upland-based systems
            Page 16
    Design of on-farm trails
        Page 16
    Result of on-farm trails
        Page 17
        The rice based system
            Page 18
            Page 19
            Page 20
            Page 21
            Page 22
        Upland-based systems
            Page 23
    Farmers' opinions about interventions
        Page 24
    Early adopters
        Page 25
        Page 26
    Conclusion
        Page 27
    Reference
        Page 28
        Page 29
    Back Cover
        Back Cover
Full Text
/:. 24Z
6.- .-, Oc '(
n(





A Case Study of

On-Farm Adaptive Research
in the Bida Agricultural
Development Project Nigeria


M. Ashraf
P. Balogun
A. Jibrin


International Institute of Tropical Agriculture
PMB 5320, Oyo Road, Ibadan, Nigeria


OFR Bulletin No. 1











A Case Study of

On-Farm Adaptive Research
in the Bida Agricultural
Development Project- Nigeria




M. Ashraf, Agricultural Economist, IITA
P. Balogun, Research Assistant, IITA
A. Jibrin, Senior Agronomist, Bida ADP















International Institute of Tropical Agriculture
OFR Bulletin No. 1


MARCH 1985











Preface


On-farm research (OFR) has become a high priority of IITA's Farming

Systems Program. OFR includes identifying constraints and opportunities

for improvement; choice of improved technologies that fit the local farming

systems; their testing and evaluation under farmers' conditions and

dissemination of suitable technologies to farmers. OFR also provides

feedback to research stations for further research required. At every

phase of the process, biological and social scientists cooperate closely.

In cooperation with researchers in many national agricultural research

programs, IITA also organizes workshops for national scientists, trains

field researchers and develops procedures for socioeconomic surveys and

experimentation under African conditions. This work, initiated in

Nigeria, is now being extended to Benin, Cameroon, Ivory Coast, Tanzania,

Togo, and Zaire.

This publication is the first in a series of IITA OFR bulletins. It

describes work undertaken with the Agronomy Section of the Bida

Agricultural Development Project in Niger State, Nigeria. The study

emphasizes delineation of farmer target groups, on-farm testing, and

experimentation to develop recommendations to farmers.

A.S.R. Juo, Director
Farming Systems Program






















Acknowledgement


We wish to acknowledge the valuable contributions made by many
colleagues at various stages in the case study work reported in this
publication. C.H.H. ter Kuile, leader of the Farming Systems Program (FSP)
during 1981-84, arranged the collaborative research, and A.S.R. Juo,
current director of FSP, encouraged and supported the research program.
B.T. Kang, T. Kosaki, K. Alluri, 3.W. Gibbons, S.R. Singh, B.B. Singh and
M.A. Gowman were extremely helpful in making technical assessments of the
performance of IITA component technologies. M. Collinson, P. Ay, W. Vogel,
T. Gebremeskel, N.D. Hahn, M. Palada, and H.W. Mutsaers reviewed the report
and made valuable suggestions. Finally, we would like to thank 3.E. Rice,
E.L. Shiawoya and A.J. Daja, all of Bida Agricultural Development Project,
for their help in implementing the on-farm adaptive research program.
























Contents


Introduction . . .

Description of the Project Area .

Organization of the OFAR Program .

Description of Target Domains and System

Initial Delineation of Target Domains

Refinement of target Domains . .

Farm Productivity . . .

System Constraints and Opportunities for

The Rice-Based System . .

Upland-Based Systems . . .

Design of On-farm Trials . .

Results of On-Farm Trials . .

The Rice-Based System . .

Upland-Based Systems . . .

Farmers' Opinions About Interventions

Early Adopters . . ... .

Conclusions . . . .

References . . . .


. . . . 1

. . . . 1

. . . . 4

Productivity . . 4

. . . . 5

. . . . 5

. . .. . . 9

Experimentation . .. 13

. . . . 13

. . . . 16

. . . . 16

. . . . 17

. . . . 18

. . . . 23

. . . . 24

. . . . 25

. . . . 27










Introduction
The Bida Agricultural Development Project (BADP) was begun in 1980
with funding from the Federal Government of Nigeria (25.), Niger State
government (39Z) and World Bank (36.). The objectives of this five-year
project are to raise agricultural production by 25. and to increase farm
income. These are to be achieved by providing farm inputs such as
fertilizers, credit, tractor hiring services, and extension information, by
developing low-cost irrigation schemes, and by constructing feeder roads
for the transport of surplus farm products.
Late in 1981 BADP, in collaboration with IITA, initiated an agronomic
research program designed t6 identify major production constraints in the
local farming systems. The aim of the program was to identify intervention
points and develop technologies and recommendations which the project's
extension staff could pass on to local farmers. This case study is a
preliminary description of the resulting on-farm adaptive research (OFAR)
program and of the farmers' responses to new technologies.


Description of the Project Area
The project covers some 17,000 square km in the southern part of Niger
State and lies in the southern Guinea savanna zone. The topography is
characterized by gently undulating terrain underlaid by sandstone. One of
the major determinants of the cropping system is the network of rivers,
especially the Niger River and its tributary and the Kaduna River. These
rivers are characterized by large, swampy floodplains, which flood during
the rainy season and then gradually dry out during the dry season. The
floodplains and the complex of small river valleys and inland swamps
(fadamas) are usually referred to as lowlands, and rice is grown on them.
The major part of the project area, which does not flood, is referred to as
upland and is used for rainfed farming. Soils in the uplands are generally
sandy and acidic with low levels of organic matter and low cation exchange
capacity. They are also highly permeable and liable to erosion, especially
on steeper slopes. Soils in the lowlands are loamy and of higher quality,
allowing longer periods of cropping without a fallow period. Temperatures










Introduction
The Bida Agricultural Development Project (BADP) was begun in 1980
with funding from the Federal Government of Nigeria (25.), Niger State
government (39Z) and World Bank (36.). The objectives of this five-year
project are to raise agricultural production by 25. and to increase farm
income. These are to be achieved by providing farm inputs such as
fertilizers, credit, tractor hiring services, and extension information, by
developing low-cost irrigation schemes, and by constructing feeder roads
for the transport of surplus farm products.
Late in 1981 BADP, in collaboration with IITA, initiated an agronomic
research program designed t6 identify major production constraints in the
local farming systems. The aim of the program was to identify intervention
points and develop technologies and recommendations which the project's
extension staff could pass on to local farmers. This case study is a
preliminary description of the resulting on-farm adaptive research (OFAR)
program and of the farmers' responses to new technologies.


Description of the Project Area
The project covers some 17,000 square km in the southern part of Niger
State and lies in the southern Guinea savanna zone. The topography is
characterized by gently undulating terrain underlaid by sandstone. One of
the major determinants of the cropping system is the network of rivers,
especially the Niger River and its tributary and the Kaduna River. These
rivers are characterized by large, swampy floodplains, which flood during
the rainy season and then gradually dry out during the dry season. The
floodplains and the complex of small river valleys and inland swamps
(fadamas) are usually referred to as lowlands, and rice is grown on them.
The major part of the project area, which does not flood, is referred to as
upland and is used for rainfed farming. Soils in the uplands are generally
sandy and acidic with low levels of organic matter and low cation exchange
capacity. They are also highly permeable and liable to erosion, especially
on steeper slopes. Soils in the lowlands are loamy and of higher quality,
allowing longer periods of cropping without a fallow period. Temperatures









rarely fall below 20oC, except during the harmattan, and are not a limiting
factor to crop growth in the area. Rainfall is the major constraint and is
characterized by its seasonal nature, monomodal distribution and
variability from year to year. Seedbed preparation, germination and the
early stages of crop growth are entirely dependent on the amount and
frequency of precipitation, since the soil profile and particularly the
soil surface carry a large soil moisture deficit at the beginning of the
rainy season. Figure 1 shows the high variability in the amount of
rainfall during this period. Mean rainfall (1,182 mm/yr) exceeds evapo-
transpiration from 11 May until 10 October (152 days)*. However, rainfall
can only be relied on to exceed evapotranspiration from 1 July until 30
September (91 days). The period between early May and mid-June presents
considerable risk to farmers. The amount of moisture available may not be
sufficient for sustained crop growth.
An estimated 9Z of the total project area is under cultivation (BADP,
1982); most farmland is situated around the lowlands and the major urban
center, Bida. Cultivation in the lowlands is semipermanent to permanent
(following Ruthenberg's classification, 1980). Upland farming is based
upon the bush fallow or shifting cultivation system. Little of the lowland
area is cultivated in the dry season, although farmers do plant cassava
(which sustains them during the "hungry period" prior to the main harvest),
sugarcane (for chewing) and vegetables. Nomadic Fulani herdsmen migrate
into the area during the dry season with their cattle (estimated at 400,000
head per year) to find dry season grazing and drinking water.
The project area includes an estimated 65,000 farming families with an
average of 6.5 persons per family. Farm size ranges between one and three
hectares. Population is densest to the east of the Kaduna river,
especially around Bida. Farming is the full-time business of the rural
population and is their primary source of income. Other supplementary
sources of income are fuel wood, fishing, trading and employment in
government. The work of women in this predominantly Muslim area is limited
to processing and storage of farm products and other chores within the
household.




These data on rainfall were compiled at Bida airport, located at the
center of the project area, over a period of 23 years. Evapotranspiration
rate is adapted from Kowal and Knabe (1972).
2
















Rainfall, mm/decades ( 10 day period )
140

1Upper quartile
120

Mean
100 I
SLower quartile
80 -
Rainfall

60 Evapotranspiration


40


20 -_ J
ATO I I 1 1 0 I



J F M A M J J A S 0 N D
Months

Fig. I. Rainfall for IO0-day periods( decades) at Bida, Nigeria (1961-83).










Rice, sorghum and dried fish are the major surplus commodities in the
area, and a local marketing system has developed for these items. Most
villages hold a market every five days for sale of goods for local
consumption. Surpluses are sold through two marketing systems: major
markets, usually found on the main roads, and traders who visit the
villages during harvest to buy directly. Bida market, the dominant one,
is held daily. Bida and other large markets serve as the wholesale markets
for sorghum and rice. Traders from urban centers outside the project area
also come to villages each year to buy parboiled rice. It is estimated
that 16,000 tonnes of rice and 15,000 tonnes of sorghum leave the project
area every year (BADP 1982). On-farm storage and marketing of surplus farm
products are not considered to be major constraints at present.


Organization of the OFAR Program
The BADP was one of the first projects in Nigeria which sought to
organize agronomic research according to a farming systems perspective.
The project adopted this innovative approach for a number of reasons. The
project wanted (1) to identify new technologies which would become a
permanent feature of the local cropping systems and would help meet the
project's crop production goals; (2) to train the young and inexperienced
research staff in relating the research program to the needs of farmers;
and (3) to create a close relationship between research and the
decentralized extension service.
BADP research and extension staff helped design and implement the OFAR
program, which undertook the following five activities:
In the diagnostic phase, description of the target areas and
identification of opportunities for experimentation within
these areas.
Screening of on-station technology for use in on-farm trials.
On-farm testing and verification of selected technologies.
Survey of farmers reactions to selected interventions (this
yielded information that will be useful in designing programs
for subsequent years).
Recommendations to the project's extension service.


Description of Target Domains and System Productivity
The wide range of agroeconomic conditions of farmers in the project
area made it impossible to run a series of on-farm trials relevant to all










Rice, sorghum and dried fish are the major surplus commodities in the
area, and a local marketing system has developed for these items. Most
villages hold a market every five days for sale of goods for local
consumption. Surpluses are sold through two marketing systems: major
markets, usually found on the main roads, and traders who visit the
villages during harvest to buy directly. Bida market, the dominant one,
is held daily. Bida and other large markets serve as the wholesale markets
for sorghum and rice. Traders from urban centers outside the project area
also come to villages each year to buy parboiled rice. It is estimated
that 16,000 tonnes of rice and 15,000 tonnes of sorghum leave the project
area every year (BADP 1982). On-farm storage and marketing of surplus farm
products are not considered to be major constraints at present.


Organization of the OFAR Program
The BADP was one of the first projects in Nigeria which sought to
organize agronomic research according to a farming systems perspective.
The project adopted this innovative approach for a number of reasons. The
project wanted (1) to identify new technologies which would become a
permanent feature of the local cropping systems and would help meet the
project's crop production goals; (2) to train the young and inexperienced
research staff in relating the research program to the needs of farmers;
and (3) to create a close relationship between research and the
decentralized extension service.
BADP research and extension staff helped design and implement the OFAR
program, which undertook the following five activities:
In the diagnostic phase, description of the target areas and
identification of opportunities for experimentation within
these areas.
Screening of on-station technology for use in on-farm trials.
On-farm testing and verification of selected technologies.
Survey of farmers reactions to selected interventions (this
yielded information that will be useful in designing programs
for subsequent years).
Recommendations to the project's extension service.


Description of Target Domains and System Productivity
The wide range of agroeconomic conditions of farmers in the project
area made it impossible to run a series of on-farm trials relevant to all










farmers within the area. Therefore, it was necessary first to group

farmers with similar agroeconomic circumstances and then design a program

to produce recommendations relevant to each group or target domain
(CIMMYT Economics Program 1980).

Initial Delineation of Target Domains
The Project was visited by an IITA agricultural economist and an

agronomist shortly before the start of the planting season in 1982. After

conducting aerial reconnaissance to familiarize themselves with the general
features of the area, the team collected secondary data on population and
farm labor use patterns. Information about agronomic practices was also
gathered through short diagnostic surveys.
The obvious difference between farmers whose cropping system was based
upon lowland cultivation of rice and farmers whose cropping system was
based on upland crops led to the initial hypothesis that there are two
target domains: (1) farmers with a lowland cropping system and (2) those
with an upland cropping system.

Refinement of Target Domains
Before establishment of the first year's trials, there was not enough
time to refine the target domains. So, this was done concurrently with

those trials. Analysis of a yearly agronomic survey, referred to as FRADYS
(Field Records for Agronomic Details, Yields and Stands), and a series of

informal interviews formed the basis for verification. The data studied
were from the 1982-83 survey of 225 farmers in 15 villages. Farmers were
grouped by land area allocated to specific crops or crop mixtures to
discern dominant cropping patterns; it was assumed that aggregate area

reflects the interaction between the biophysical and socioeconomic circum-
stances of farmers and their priorities (Zandstra 1981; Norman 1981).
This analysis revealed that there are four target domains based on
cropping systems, one in the lowlands and three in the uplands:
Lowland rice based
Upland yam based
Upland cassava based
Upland cereals based
The major determinants of the four cropping systems were soil types
(including the varying level of soil fertility) and the availability of










farmers within the area. Therefore, it was necessary first to group

farmers with similar agroeconomic circumstances and then design a program

to produce recommendations relevant to each group or target domain
(CIMMYT Economics Program 1980).

Initial Delineation of Target Domains
The Project was visited by an IITA agricultural economist and an

agronomist shortly before the start of the planting season in 1982. After

conducting aerial reconnaissance to familiarize themselves with the general
features of the area, the team collected secondary data on population and
farm labor use patterns. Information about agronomic practices was also
gathered through short diagnostic surveys.
The obvious difference between farmers whose cropping system was based
upon lowland cultivation of rice and farmers whose cropping system was
based on upland crops led to the initial hypothesis that there are two
target domains: (1) farmers with a lowland cropping system and (2) those
with an upland cropping system.

Refinement of Target Domains
Before establishment of the first year's trials, there was not enough
time to refine the target domains. So, this was done concurrently with

those trials. Analysis of a yearly agronomic survey, referred to as FRADYS
(Field Records for Agronomic Details, Yields and Stands), and a series of

informal interviews formed the basis for verification. The data studied
were from the 1982-83 survey of 225 farmers in 15 villages. Farmers were
grouped by land area allocated to specific crops or crop mixtures to
discern dominant cropping patterns; it was assumed that aggregate area

reflects the interaction between the biophysical and socioeconomic circum-
stances of farmers and their priorities (Zandstra 1981; Norman 1981).
This analysis revealed that there are four target domains based on
cropping systems, one in the lowlands and three in the uplands:
Lowland rice based
Upland yam based
Upland cassava based
Upland cereals based
The major determinants of the four cropping systems were soil types
(including the varying level of soil fertility) and the availability of










moisture during the wet and dry seasons. In the lowlands the surface soils

are deep and loamy, and are subjected to flooding caused by surface runoff.

Water is supplied by seepage during the dry season. The upland soils are
of three types. The first is found on the lower edges of sloping fields

which have deep, rich surface soil and heavier soil texture, and experience

little erosion. In the local classification system, fields containing such

soils are called "yam fields" and farmers grow yams and maize on them.

These soils are of limited availability in the project area. The second

type has slightly loose soil texture and lower fertility, and lies on the
midslopes. These fields are known locally as "cassava fields" and are

usually planted to that crop. The third type is the top section of the
catena. These soils are either sandy or gravelly, highly eroded, leached

and very low in fertility. Most of the project land falls in this category.
These fields are planted to sorghum, millets, egusi melon, bambarra ground-
nut, groundnut, etc., crops which are best adapted to low-fertility soils.
The project area's Nupe population has settled in inland valleys where
they have access to drinking water during the dry season. Some farmers
have access both to lowlands and one or the other type of upland, while

many farmers have access only to different types of upland. We refer to
farmers cultivating both lowland and upland as having a rice-based cropping

system. Farmers on upland, some of which is suitable for yam cultivation,
are referred to as having a yam-based system. Farmers who grow both
cassava and upland cereals are referred to as having a cassava-based
system. Farmers who have access only to poor quality land, on which they

grow cereals and legumes, are within a cereals-based cropping system. The
definitions of target domains in this paper are therefore very much related

to biophysical factors, which in our view have a major influence over
farmers' cropping plans and management decisions, assuming that farmers

have a wide range of choices among a number of crops and intercropping
mixtures.

Although the dominant crop by area in all four target domains was
sorghum (Table 1), rice, yam and cassava were the major cash crops within
their respective domains. These crops also produced the highest yields
(Table 2) and failed least often (Table 3) in their respective domains.
Crop failure rates were highest in the upland cereals-based system. As
mentioned previously, this system is confined to the less fertile areas in










Table 1. Percentage of area grown to various crops and mixtures in target
domains of the Bida ADP, 1982

CroppDDing system domain
Rice Yam Cassava Cereals Over-
Crops and mixtures based based based based all
Rice 25 3 7
Yam 9 12 4 8
Yam + millets 5 2
Cassava 1 1 44 2 4
Sorghum 8 27 15 18 19
Egusi melon 3 1 3 2
Sorghum + melon 40 16 8 17 22
Sorghum + maize 5 14 3 6 9
Sorghum + millets 3 4 2
Sorghum + cowpea 4 2
Sorghum + groundnut 1 1 3 12 4
Sorghum + millets + melon 2 1 12 4
Cereals + others 1 8 3 5
Tubers + others 1 3 24 5 4
Others 4 4 3 11 6

Percent 100 100 100 100 100

Less than 0.5 percent.





Table 2. Crop yields in various target domains of the Bida ADP, Nigeria,
1982

Cropping system domain
Rice Yam Cassava Cereals Over-
Crop based based based based all
tonnes/ha

Rice 2.2 1.2 1.3 2.1
Yam 5.5 9.1 2.7 6.9 7.9
Cassava 2.2 3.9 6.6 1.6 5.4
Sorghum 1.3 1.2 1.1 0.9 1.1
Millets 0.5 0.8 0.9 0.8
Maize 1.4 1.1 1.2 0.5 1.0
Cowpea 0.3 0.5 0.5
Groundnut 0.3 1.0 1.3 0.8 0.8
Bambarra groundnut 0.6 0.5 0.5
Melon seed 0.1 0.3 0.2 0.2
Note: These figures are average yields for various inter- or mix cropping
combinations.
Less than 0.1 tonne/ha.
Less than 0.1 tonne/ha.










ADP area in which crops failed, 1982


CroDoina system domain
Rice Yam Cassava Cereals
Crop based based based based Overall
Yam 0 3 0 2 2
Cassava 13 0 2 15 8
Sorghum 4 1 3 5 3
Millets 0 *- 2 2
Maize 0 6 0 20 10
Cowpea 0 0 19 2
Groundnut 0 0 0 1 1
Egusi melon 6 3 0 2 4
Total 4 2 2 6 4

Less than 0.5 percent.

the uplands, on soils which are both highly erodible and shallow, and which
quickly show the effects of drought. The dominant cropping systems are
described below using the' "decision tree" method. This method has been
found useful in coming to an understanding of farmers' decision-making
criteria and in assessing opportunities for improving the existing cropping
syste.~s (,Shaane, et al. 1981; 8alcet and Candler 1982).

The rice-based system. Twenty-five percent of the farmers sampled use
this system; i.e.,' grow both lowland and upland crops. In late April or
May, upland fields are planted with tall indeterminate sorghum, usually in
a mixture with millet, maize or egusi melon (Figure 2). Since sorghum is
the major food grown for home consumption, priority is attached to ensuring
good establishment of this crop. Although the crop can compete well with
weeds once it is established, the fields must be kept clean during the
early stages of growth. Therefore, land preparation for the rice crop in
the lowlands occurs only after the first weeding of the upland crops in
late July and August, three months after the start of the rainy season.
Within the project, rice is grown mainly in two ecological
environments: on the floodplains and in inland valleys where much seepage
water comes from the surrounding uplands. In both environments ridging is
the major method of seedbed preparation, being dictated by water status and
position in the topographic sequence. Close to the bottom of the sequence,
where the water level is deep, rice is seeded directly on the ridges; while
in the upper part of the toposequence, ridges are made to conserve water
and facilitate weeding.

8


Table 3. Percentage of Bida










Farmers use ridging as a weed control measure at the establishment
stage and also during the first weeding, when weeds are pulled into the

furrow with a hoe and then buried. Herbicides are only rarely used, and

most farmers have no experience with them.

Harvesting of short-season upland crops such as millet and egusi melon
and weeding of the rice crop are the major activities until the dry season
in November and December, when sorghum and rice are harvested. The dry

season is a period of slack labor demand for most farmers in the rice-based

system, since only a small part of the lowlands is utilized for production
of dry-season crops on residual moisture and few opportunities exist for

off-farm employment.

Upland-based systems. Within the survey sample, 25% of farmers had
the yam-based system, 10% the cassava-based and 407 the upland cereals-
based system. Farmers in the yam-based system had the largest farms and
those in the upland cereals-based system the smallest. For both the yam-
and cassava-based systems, the labor input was high (365 and 239 man-days,
respectively), with peak labor demand occurring at the start of the rains
when sorghum is planted and after the rains when crops are harvested.

Farmers avoid potential conflicts in the allocation of scarce labor for
establishment of yams and sorghum by planting most of the yam area during

the dry season (Figure 3), thus spreading out the demand for labor over
time. Within the cassava-based system, cassava is planted after estab-
lishment of sorghum. Since cassava is a relatively drought-resistant crop
with a growth cycle of over 12 months, timeliness of planting is not as

important as it is for the rice crop; cassava can be established over
several months (Figure 4).
The upland cereals-based system (Figure 5) was the poorest domain;
crop yields were lower than in the other domains and farm size was smaller.
In the first half of the rainy season, farmers are usually busy planting
and weeding; during the second half many of them sell their labor to richer

farmers who have rice- and yam-based systems. Families in this domain sell
their surplus labor and hunt and fish to supplement their food and farm
incomes.

Farm Productivity
Farm productivity (measured in terms of caloric value of food
produced), total farm income and returns to farm inputs vary considerably
9





















OPENING PLANTING DECISIONS CROP PROBA-
STAGE STAGE I STAGE II STAGE III STAGE IV ENTERPRISES ABILITY
Mar Apr-May Jun- Jul Aug-Sep Oct-Dec
Rice Rice I. Fallow/Rice 247

O/ 2.Sorghum/ .390
S-- Melon
Sorghum 3.Sorghum .082

Sorfbu f- Maize 4 Sorghum/Maize .054
Rains start/ so ts 5. Sorghum/Melon .016
first planting -- /Millets
-eons Beas 6. Beans .012

SMaize 7 Fallow/Maize .007
Cassava Cassava Cassava/ 8.Cassava/Okra .020

Other 9. Other combi .172
nations
1.000

Fig. 2. Crop enterprise decision tree for the rice -based cropping system, Bida ADP (Avg. for 1982-83).


Fig. 3. Crop enterprise decision tree for the yam-based cropping system, Bida ADP( Avg. for 1982-83).




















OPENING PLANTING DECISIONS CROP PROBA-
STAGE STAGE I STAGE II STAGE III STAGE IV ENTERPRISES ABILITY
Mar Apr-May Jun-Jul Aug-Sep Oct-Nov
Cassava Cassava Cassava I. Cassava .440
o_ Sorghum/ 2*
Cassava Sorghum/ .163
Cassava Cassava .163
Cassava 3.

Raon te.e Cassava 4.Sorghum/Melon/ .064
Rain start Sor rv!--- Cassava
First planting Sorghum Sorghum 5. Sorghum .153

So rghuMelon 6.Sorghum/Melon .077
.i. Sorghum/ 7 Sorghum/
t Groundnut Groundnut .034
Other Other 8.Other
combinations .069

1.000


Fig. 4. Crop enterprise decision tree for the cassava-based cropping system,


Bida ADP( Avg. for 1982-83).


Fig. 5. Crop enterprise decision tree for the cereals-based cropping system, Bida ADP ( Avg. for 1982-83.











among the four target domain cropping systems. The root-crops-based

systems of yam and cassava yield greater quantities of food than the rice-

and cereals-based systems (Table 4). The farmers growing root crops not

only contribute more to the market (about 457 of the total amount of food

produced), but still have more food available per capital than the farmers

in the rice- and coarse cereals-based systems. When farm production of the

four target domains is converted to value terms, the root-crops-based

systems maintain their edge over the cereals-based systems.

The rice-based system, which requires access to good land and water

resources, is less productive mainly because the lowland fields are less

intensively cropped. Farmers' capital costs are mostly for seed, small
quantities of fertilizer and primitive hand tools. Farmers do not have

access to animal and mechanical farm power. Thus, the return to capital is
negligible. Nearly 85. of the farm costs are labor inputs (mostly from the
family), and returns to labor input are close to the rural wage rate. Af-
ter accounting for costs of labor and capital, return to land is negative


Table 4. Amount of food calories produced by the average farmer in Bida
ADP, 1982-83

Cropping system domain
Rice Yam Cassava Cereals
Crop based based based based Overall

1,000 calories

Rice 3,841 90 180 1,028
Yam 810 3,600 50 630 1,273
Cassava 123 62 10,134 48 2,592
Sorghum 4,802 8,918 3,773 2,950 5,111
Millets 60 1,189 730 495
Maize 576 2,340 241 252 852
Egusi melon 372 450 33 270 281
Groundnut 27 342 311 622 325
Bambarra groundnut 445 25 72 135
Cowpea 17 291 77
Vegetables 3 17 29 8 14
Sweet Potato 25 15 10

Total 11,076 17,324 14,596 5,777 12,194
Per hectare 5,483 5,397 6,950 4,377 5,891

Calories were measured on a fresh-weight basis and adjusted by 80% of
the total amount of produce to allow for storage and transportation
losses. Adjustments were also made for seed requirements. The calcula-
tions were based upon an FAO source (Schmitt 1979).










Table 5. Average, annual farm costs and income for the Bida ADP, 1982

Crooning system domain
Rice Yam Cassava Cereals
Item based based based based Overall
Average farm size, ha 2.02 3.21 2.10 1.32 2.07
Total labor input:
Man-days 248 365 239 147 250
Cost, $a 1,674 2,464 1,613 992 1,686
Seed cost, $ 153 285 66 86 147
Fertilizer cost, $ 16 3 18 9 12
Cost of farm tools, $c 92 79 79 77 82
Land rent, $ 27 43 28 18 29
Total farm cost, $ 1,962 2,874 1,804 1,182 1,956
Total farm income, $ 1,802 3,842 2,427 1,053 2,281
Net farm income, $ -160 968 623 -129 326
Measures of efficiency
Return to land, $/ha -66 315 310 -84 171
Return to capital, $ 0.4 3.6 4.8 0.3 2.3
Return to labor and
management, $/man-day 6.1 9.4 9.4 5.9 8.0

aLabor was valued at the current wage of 5 naira ($6.75)/man-day.

bFertilizer cost is based on the subsidized price of 2 naira ($2.70)/50-kg
bag.
CThis is the cost of two large and three small hoes, two cutlasses, one
axe, three sickles (rice farmers only), three baskets and two to five
sacks. The annual cost of cash capital is charged at a 25Z interest rate.
Land rent is assumed to be 10 naira ($13.5)/ha.


for the rice and coarse cereals-based (sorghum and millets) systems and
positive for the yam- and cassava-based systems (Table 5).


System Constraints and Opportunities for Experimentation
The Rice-Based System
As stated previously, farmers in the rice-based cropping system grow
rice in the lowlands and a mixture of coarse cereal and legume crops in the
uplands. Since the lowlands are rain fed and farmers practice little water
control, there is some risk of moisture stress at the end of the rainy
season. Rice planting is usually delayed because farmers first attend to
upland crops. Planting of rice is further delayed by the high labor input
necessary for ridging the land. Because of low paddy yields and less
intensive use of rice fields, the productivity of the lowlands is very low.










Table 5. Average, annual farm costs and income for the Bida ADP, 1982

Crooning system domain
Rice Yam Cassava Cereals
Item based based based based Overall
Average farm size, ha 2.02 3.21 2.10 1.32 2.07
Total labor input:
Man-days 248 365 239 147 250
Cost, $a 1,674 2,464 1,613 992 1,686
Seed cost, $ 153 285 66 86 147
Fertilizer cost, $ 16 3 18 9 12
Cost of farm tools, $c 92 79 79 77 82
Land rent, $ 27 43 28 18 29
Total farm cost, $ 1,962 2,874 1,804 1,182 1,956
Total farm income, $ 1,802 3,842 2,427 1,053 2,281
Net farm income, $ -160 968 623 -129 326
Measures of efficiency
Return to land, $/ha -66 315 310 -84 171
Return to capital, $ 0.4 3.6 4.8 0.3 2.3
Return to labor and
management, $/man-day 6.1 9.4 9.4 5.9 8.0

aLabor was valued at the current wage of 5 naira ($6.75)/man-day.

bFertilizer cost is based on the subsidized price of 2 naira ($2.70)/50-kg
bag.
CThis is the cost of two large and three small hoes, two cutlasses, one
axe, three sickles (rice farmers only), three baskets and two to five
sacks. The annual cost of cash capital is charged at a 25Z interest rate.
Land rent is assumed to be 10 naira ($13.5)/ha.


for the rice and coarse cereals-based (sorghum and millets) systems and
positive for the yam- and cassava-based systems (Table 5).


System Constraints and Opportunities for Experimentation
The Rice-Based System
As stated previously, farmers in the rice-based cropping system grow
rice in the lowlands and a mixture of coarse cereal and legume crops in the
uplands. Since the lowlands are rain fed and farmers practice little water
control, there is some risk of moisture stress at the end of the rainy
season. Rice planting is usually delayed because farmers first attend to
upland crops. Planting of rice is further delayed by the high labor input
necessary for ridging the land. Because of low paddy yields and less
intensive use of rice fields, the productivity of the lowlands is very low.










Since farmers attach great importance to the cash crop paddy, we therefore

sought to improve the system by changing the farmers' crop management
practices in the lowlands.
Easing the labor constraint in June and July is crucial for early rice

planting, which lessens the risk of crop failure, There are two possible
approaches to moving the rice planting date forward: (1) substituting flat
cultivation for the laborious ridge seedbed preparation and (2) planting a

portion of the area now planted to photoperiod-sensitive sorghum varieties
to a short-season modern variety that can be planted later in the season.

This would allow the paddy to be planted a few weeks earlier. A third
alternative for alleviating the labor constraint, saving labor for weeding
upland fields, was not considered since there was no known mechanical or
chemical weed control technology for the crop mixtures grown by the
farmers.
Getting farmers to plant rice on flat rather than ridged seedbeds
received high priority since this practice was expected not only to save

labor but to be compatible with the newly expanded tractor hire services
and chemical weed control methods provided by the project. It was also
thought that this practice would tie in with the project's work on informal
irrigation schemes in the lowlands and the use of contour bunding.

Other factors limiting rice yields in the area were low stand density
and iron toxicity. Stand density is closely related to the ridging of
seedbeds; wider spacing between ridges results in lower stand densities.
They could be increased either by reducing the gap between ridges or by
planting on the flat.
Iron toxicity is a more complex problem. Improved tolerant varieties
are one solution the farmers could easily adopt. Other expensive solutions

are the use of lime (1-2 tons/ha), drainage of seepage water and planting
on raised beds. Other problems such as irregular water supply and bird

damage were noted, but these were beyond the farmers' control and were
therefore not included in the short-term research program.
The movement of seepage water from the surrounding uplands into the
fadamas during the dry season offered another possibility for increasing
farm productivity. Since labor was not a constraint in the farming system
during the dry season (Figure 6), it was proposed that a "catch crop" could
be grown in the residual moisture without creating major problems with
resource allocation.
14




















Man-days

---I


Yam-based


Labor


I I I I I I I I I I l l


J F M A M J J A S N D


Months

Fig. 6. Average household labor input (man-days) for agricultural activities
in various cropping systems, Bida ADP, Nigeria 1983.


50-


40-


30 -


20-


10 -










Upland-Based Systems
The local cassava varieties were observed to be heavily infested with
green spider mites and cassava mosaic virus, both of which depressed
yields. The project decided to include in its research program screening
of improved varieties for tolerance to those problems. Although the
,project area is well suited to maize production, yields are limited by the
parasitic weed Striqa hermontheca. The project proposed to screen improved
varieties with some resistance to striga and find cultural control methods.
In the cassava- and cereals-based systems, the chances for improving
production are best in the midseason between July and November, when
surplus labor is available (see Figure 6). A short-season crop like
cowpeas can mature if farmers make good use of the remaining rainy season.
Project records and informal interviews have shown that the area planted to
cowpeas has declined rapidly within the last fifteen years because of
insect problems principally at flowering. Even so, cowpeas are still a
major component of the local diet, and 500 to 60.0 tons are imported
annually from northern Nigeria. Local cowpea varieties are indeterminate,
making the use of insecticides inefficient, since the crop flowers over a
period of several weeks. The availability of surplus labor in the
midseason and of new determinate varieties that require minimum insecticide
spraying, offered a chance to reestablish this crop in the project area.


Design of On-Farm Trials

The aim of the on-farm trials was to evaluate selected crop varieties
and management alternatives for overcoming crop production constraints and
to generate information confirming the significance of constraints
identified in the diagnostic phase. The results were used to assess the
agroeconomic feasibility of selected interventions, compared to the
farmers current practices, and to supply socioeconomic information which
was found to be lacking during the diagnostic phase. After discussions
with scientists and project staff, high priority was placed on research in
the lowland rice-based and upland coarse cereals-based domains. Both
needed urgent attention to improve their production. The rice-based
cropping system was found to have the most potential for improvement using
technologies available on-station. An on-farm experimental program was
designed to make better use of biophysical and human resources. This has










Upland-Based Systems
The local cassava varieties were observed to be heavily infested with
green spider mites and cassava mosaic virus, both of which depressed
yields. The project decided to include in its research program screening
of improved varieties for tolerance to those problems. Although the
,project area is well suited to maize production, yields are limited by the
parasitic weed Striqa hermontheca. The project proposed to screen improved
varieties with some resistance to striga and find cultural control methods.
In the cassava- and cereals-based systems, the chances for improving
production are best in the midseason between July and November, when
surplus labor is available (see Figure 6). A short-season crop like
cowpeas can mature if farmers make good use of the remaining rainy season.
Project records and informal interviews have shown that the area planted to
cowpeas has declined rapidly within the last fifteen years because of
insect problems principally at flowering. Even so, cowpeas are still a
major component of the local diet, and 500 to 60.0 tons are imported
annually from northern Nigeria. Local cowpea varieties are indeterminate,
making the use of insecticides inefficient, since the crop flowers over a
period of several weeks. The availability of surplus labor in the
midseason and of new determinate varieties that require minimum insecticide
spraying, offered a chance to reestablish this crop in the project area.


Design of On-Farm Trials

The aim of the on-farm trials was to evaluate selected crop varieties
and management alternatives for overcoming crop production constraints and
to generate information confirming the significance of constraints
identified in the diagnostic phase. The results were used to assess the
agroeconomic feasibility of selected interventions, compared to the
farmers current practices, and to supply socioeconomic information which
was found to be lacking during the diagnostic phase. After discussions
with scientists and project staff, high priority was placed on research in
the lowland rice-based and upland coarse cereals-based domains. Both
needed urgent attention to improve their production. The rice-based
cropping system was found to have the most potential for improvement using
technologies available on-station. An on-farm experimental program was
designed to make better use of biophysical and human resources. This has










involved the intensification of crop production by using slack-period land
and labor resources, adjusting the cropping calendar and changing the crop
management practices.
For the rice-based cropping system, the on-farm trials included rice,
cowpeas and sorghum. In addition, different types of sickles were
evaluated for their ability to save labor in paddy reaping and shorten the
period from paddy to dry-season cowpeas. Rice trials were conducted to
determine the effects of flat seedbed preparation, increased rice hill
density and modern rice varieties on the paddy yield and labor use pattern.
Cowpea trials were conducted mainly during the dry season to study the
possibility of growing short-season cowpea varieties on residual moisture
after rice. These experiments also explored the effect of planting dates,
slope gradients and tillage methods on crop establishment and cowpea
yields. In experiments on sorghum, short-season sorghum varieties were
planted later in August after completion of rice planting.
For the upland-based cropping systems, only a limited number of
experiments were conducted on cassava, maize and cowpeas. New cassava and
maize varieties were tested for resistance to common pests and diseases.
The cowpea trials determined the effectiveness of a minimum pest-management
regime and the possibility of planting in August, a relatively slack labor
period.
The number of on-farm agronomic experiments for any single crop was
limited to 8 and for any season never exceeded 25. However, after the
initial experiments had produced good results, the number of verification
trials was substantially increased with the help of extension staff.
Overall, the trial program consisted of a few centralized replicated trials
surrounded by a large number of simple unreplicated tests. The former were
managed by the research staff and the latter by the extension staff or by
farmers. The trial had a factorial design with two treatment levels, the
farmers' practice and the scientists' recommendation. The research team
which implemented the on-farm trial program included one agronomist with a
bachelor's degree, three supervisors with high school diplomas and nine
laborers.


Results of On-Farm Trials
In this section agroeconomic results are presented only for trials
which have produced somewhat conclusive evidence of the success of new
17










technologies and which have most attracted the farmers' attention. For the
short-duration crops--cowpeas, rice and maize--more agronomic response data
have been generated than for the annual crop cassava, for
which a longer period is required to produce sufficient data.

The Rice-Based System
Rice varieties. A total of six varieties were planted at different
sites. These varieties were selected based on their performance in a seed
multiplication plot and were the new varieties being recommended by the
extension service. Because the rains began late, rice planting by farmers
was delayed by two to three weeks. Planting of on-farm trials was also
delayed at all but one site, where the farmer's variety was photoperiod
sensitive and experienced little water stress after cessation of the rains.
Among the improved varieties, only one had a maturity time comparable to
that of the farmer's variety and did not experience moisture stress at
panicle filling. The local' indeterminate varieties, yielding 2.25 tons per
hectare, were thus better adapted to local physical and biological
conditions, with little or no water control, than the improved varieties.
At one site where water management was good and irrigation with
seepage water continued after the rains ceased, yields of the improved
varieties were 32Z to 40% higher than those of the local control.* These
results confirmed the need for an earlier rice planting date.

Seedbed preparation. We hypothesized that yield could be maintained
or even increased and that labor could be saved by planting rice on a flat
seedbed rather than on ridges. The trials produced conflicting agronomic
yield responses. Half the sites showed a positive yield response and the
other sites a negative response. Based on preliminary observations made
during the crop season and subsequent visits by a soil scientist, it was
concluded that the differing yield responses were caused by differing
levels of flooding and water control.




In this discussion no reference is made to level of statistical signifi-
cance for two reasons. First, most of the on-farm tests were superim-
posed on fields managed by the farmers and it was not possible to measure
site-to-site environmental variability. Second, farmers' assessments and
economic analysis were given great importance in determining the suita-
bility of the technology (Perrin, 1976). Farmers often accepted small
yield increments as long as they were not associated with peak labor demand
periods and when the treatment reduced labor input.
18










Table 6. Effect of management practices on labor productivity in lowland
rice, Bida ADP, 1982

Average Total la- Total Labor productivity
Management yield, bor input, revenue, kg/man- $/man-
practices kg/ha man-day/ha $/ha day day
Farmers' practice 2,203 118 1,190 19 10
High hill density (HD) 2,552 129 1,378 20 11
Flat cultivation (FC) 2,683 112 1,449 24 13
HD + FC 3,003 122 1,622 25 13



At sites whe:e there was less flooding with seepage water, flat
planting of rice (direct seeding) produced higher yields. However, at
sites where seepage water brought excessive flooding, the farmers' practice
of planting on ridges to raise plant beds produced better results.
Although iron toxicity was considered to be depressing yields, it required
further verification.
For sites with a positive yield response to flat seedbeds, yield gap
analysis showed that the switch from ridged to flat seedbeds contributed,
on average, an extra 514 kg/ha of paddy yield or about 237 above the
farmers' yield of 2.2 tons/ha. Table 6 shows that the management change
was highly beneficial, with the return to labor expressed as paddy yield,
and the gross return per man-day increased with little or no increase in
total labor use.

Stand density. The farmers' stand density of less than 100,000
stands/ha was considered too low to make full use of land and water
resources. As with the tillage methods, varying responses to increasing
stand density from 100,000 to 150,000 stands/ha were observed. At sites
having excessive seepage water, yield response to increased stand density
were insignificant. At sites having no flooding problem, increased stand
density produced an extra 300 kg/ha of paddy rice. The contribution of
increased stand density is expected to be much greater with moderate
fertilizer use.

Short-season sorghum. Two years of exploratory work on short-season
sorghum varieties has shown that under researcher management, ICRISAT
varieties are suitable and produce economic yields. This work has now
advanced to the on-farm testing phase, which includes a study of labor use

patterns and farmer assessments of the new varieties.










Exploratory cowpea trials. In the discussion of system constraints, we

mentioned that one reason for the low productivity of the rice-based system
is that it underutilizes the lowland fields. An informal survey of project

farmers indicated that after a paddy crop their lowland fields remained wet
for some weeks during the dry season. Since farmers have surplus family

labor during that period, they said they would welcome any innovation which
would increase their food supply by using slack period resources. The

survey further revealed that the average Nupe family consumes every week
two to three meals prepared from cowpea beans and spends about $250 per
year on cowpeas. A variety popularly known as "60-day cowpeas" was
therefore selected as a "catch crop" to be grown on residual moisture

during the dry season.
Exploratory trials were designed to determine whether cowpeas could be
grown after rice. The farmers were asked to manage the trials. The trial
design was simple including only two experimental factors, variety and
insect control. Four short-season varieties were selected, and the plot

was split into two sections, one section being sprayed to control insects
and the other section not.

Results from the two sites established in the first year were
encouraging. At the first site, yields were reasonable, especially those

of variety IT 82E-60, which had the shortest season length. Yields at the
second site, where crop establishment was delayed by two weeks, were
considerably depressed because of moisture stress and damage by goats.
Economic returns (Table 7) at the first site compared favorably with cash

costs, especially for a slack labor period. Total cash costs for seed and
fertilizer at the local market price and for spraying were calculated to be
$125/ha, while net return to labor and land inputs was $1,470/ha.
Cooperating farmers stated that demand for seed from other farmers was
high. Since the results of the first year's dry season cowpea trials were
favorable, the program was expanded in the next dry season to gain
information on factors affecting crop establishment, on suitable methods of
seedbed preparation, and on the effect of planting date and location of the
crop in the toposequence.
Zero tillage, strip tillage and conventional tillage with hand hoes
were all successful in germination of seed and good establishment of the

crop in paddy fields. Each method made it possible to reduce labor inputs










Table 7. Economic returns from exploratory short-season cowpea trials in
the Bida ADP

Yield, Labor prod., Total crop Net bene- Benefit/
Variety kg/ha kg/man-daya value, $/hab fit, $/hac cost ratio

1982-83
E-60 945 14.5 1,595 1,470 12.8
E-77 545 8.4 920 795 7.4
1983-84
E-60 608 9.4 1,026 901 8.2
TVx 3236 691 10.6 1,166 1,041 9.3

Based on 65 man-days/ha.
Based on a net field price of 1.25 naira/kg (1 naira = $1.35).
Assuming total cash costs of $1.25/ha, not including cost of family labor.


Table 8. Average yield of cowpeas during the dry season in Bida ADP,
1983-84

Topographic River flow Flood
position Fadama valley plain Overall

E-60 kq/ha
Middle 571 750 -597
Bottom 620 672 646
Average 578 711 -608
TVx.3236
Middle 707 750 712
Bottom 536 -830 634
Average 669 750 830 691
Note: Two insecticide sprayings were done at all sites. Data were col-
lected from a total of 20 sites. E-60 was planted at 9 and TVx 3236 at 11.
Sixteen sites were in fadamas, 3 in a river flow valley and 1 in a flood-
plain. Fifteen were located on middle positions and 5 on valley bottoms.

and permitted timely planting of larger areas to cowpeas. The performance
of crops planted along the slope/gradient depended upon the availability of
moisture. On the top sections of fadamas with 5. or steeper slopes,
moisture depletion was fast and so yields were uneconomical. Crops in the
middle and bottom parts of the fadamas matured successfully, giving
economic yields both in experimental and extension verification plots
(Table 8). Dry-season cowpea yields of 600 to 700 kg/ha with two sprayings
were quite attractive to farmers. They became so enthused about this new
crop enterprise that they refused to sell their produce to the project.
After two dry-season cowpea plots proved successful in the first year, more
than 300 project farmers voluntarily planted short-season cowpeas during
the next dry season. 1











Shift in system productivity. We had previously estimated that a

paddy yield increase of 800 kg/ha would result from changes in hill density
and tillage method, while requiring little extra labor input. Similarly,
the dry-season cowpea crop (E-60) produced 777 kg/ha yield with 60 days of
labor and a capital cost/ha of about $125 for insecticide spraying. We
have measured the effect of these extra costs and returns on overall farm
productivity (Table 9). Column one in the table represents the benchmark
level, while columns 2, 3 and 4 show the level of costs and returns
resulting from incremental changes in rice management practices,
cultivation of dry season cowpeas and adoption of both improvements,

respectively. These innovations have brought about substantial improvement
in net farm returns to labor and capital inputs. The net farm return of
all costs has become positive, having risen from its benchmark level of
negative income. Because of the low costs and high financial gains
associated with these innovations, many farmers have started increasing the
plant density of their paddy crop and growing dry-season cowpeas.


Table 9. Estimated gains in farm productivity with dry-season cowpeas in
the rice-based cropping system, Bida ADP

Improved
Bench- rice man- Dry-season
Item mark agement cowpeasa Both

Farm costs, $/ha
Labor 1,674 1,701 1,775 1,802
Capital 261 261 314 314
Land rent 27 27 27 27
Total 1,962 1,989 2,116 2,143

Farm returns, $/ha
Gross 1,802 2,018 2,263 2,479
Net after land and capital 1,514 1,730 1,775 2,138
Net after labor and land rent 101 290 461 650
Net after labor and capital costs -133 56 174 363
Net of all costs -160 29 147 336
Returns per unit of:
Land, $/ha -66 28 86 180
Capital, $ 0.4 1.1 1.5 2.1
Labor and management, S/man-day 6.1 6.9 6.9 7.6
Note: These figures are for an average farm size of 2.02 ha of cropped
land.
aDry-season labor cost during the slack farm period was valued at 507 of
the normal wage of $6.75/man-day.









Table 10. Relative efficiency of different sickles in rice harvesting

Area harvested (m2) in 20 min. Efficiency
Condition of paddy Local sickle Improved sickle increment, Z
Field condition
Wet 53.4 65.3 22
Dry 83.1 92.4 11
Lodging rate
Fully lodged 74.1 76.5 3
Semilodged 67.5 78.2 16
Erect 57.7 80.8 40
Overall 67.5 78.2 16

Note: Efficiency tests were carried out with 30 farmers in their fields.

Rice sickle. During the second year of OFAR trials, it was observed
that farmers preferred to harvest the paddy crop when it was overripened
and fully lodged, which not only caused loss of paddy through shattering,
but would also delay the planting of dry-season cowpeas. The farmers said
that lodged paddy was easier to harvest and saved labor time. A paddy
reaping experiment conducted with the farmer's unserrated sickle and an
improved serrated sickle confirmed the farmers' opinion (Table 10). This
experiment also revealed that before modern rice varieties (which are
comparatively short and do not lodge) will be accepted on a large scale,
farmers sickles must also be improved. Although farmers preferred the
serrated sickle (and a few rich farmers have acquired them from abroad),
the village and town blacksmith lacked equipment for making serrated
blades.

Upland-Based Systems
Some on-farm research was carried out in the upland-based systems
mainly to generate more information about the ability of improved crop
varieties to withstand attack by the most common insect pests and diseases
and to determine the suitability of introducing short-season cowpeas toward
the end of the rainy season.
The failure of the maize trials emphasized the need for a striga-
resistant variety and/or effective cultural controls. The cassava trial
was not harvested until the second year. Several of the improved varieties
appeared to be tolerant to cassava mosaic virus and cassava green spider
mites. Since these improved varieties had been successful elsewhere in
Nigeria, it was decided to set up cassava extension trials to assess the










agronomic performance of the new varieties under the care of extension

staff. Because of the success of dry-season cowpeas, farmers using the

upland-based system became very interested in growing main-season cowpeas.
The farmers were motivated by various cost economies related to the fixed

operational cost of spraying equipment, reductions in the on-farm storage
period and consequently in crop losses during storage, and improvement in

the cash flow. As stated previously, less labor is required during August/

September in the upland-based cropping systems.

Any crop which will mature within the last two months of the rainy
season would be of interest to the farmers, particularly those in the

cereals-based and cassava-based systems, where land is abundant. When
tested in upland systems, short-season cowpeas were found to produce yields

of 700 to 1,000 kg/ha under farmers' conditions with a minimum spraying
regime. Thus, for the main season, the cowpea enterprise has a benefit:
cost ratio of at least 8:1 for pest management coverage.


Farmers' Opinions About Interventions
Informal surveys of rice growers revealed the possibility of

recommending closer spacing and flat seedbed preparation to farmers in
areas with better water control. When asked about using flat seedbeds

instead of ridges, farmers said they wanted to try the idea on their own
plots and agreed that it might give as good or better yields and/or involve
less work than ridging. However, some farmers believed that planting rice
on flat seedbeds was either not feasible because of the variation in water

conditions during the season or would take more work to achieve good weed
control. Based on these reactions, BADP has undertaken to classify

different types of fadamas and to lay out plots demonstrating the value of
planting a denser crop on flat seedbeds.
Farmers' reactions to increasing stand density were mixed. Some
feared that increased stand density would reduce tillering and make it more

difficult and time-consuming to weed, since closer spacing would not leave
room for hoeing between stands. Adoption of closer spacing was observed in
one village, where farmers used smaller hoes. Survey results indicated
that smaller hoes were not readily available in the project area and that
it required a special effort to have the blacksmiths make small hoes.
Thus, wider spacing between hills was considered necessary for weeding with
large hoes.









Farmers liked the improved rice variety grown on "good fadama" for its

grain color, size and cooking quality. However, they indicated they would
plant a new variety only on part of their fadama land, where water control
is better, weeds are less of a problem and where they will apply chemical
fertilizers. When asked if keeping two types of seed is a problem, they
said they usually plant two to three varieties each year.
In growing dry-season cowpeas, farmers made a number of innovations.
Once farmers knew the moisture requirement of the crop, they proved fairly
accurate in selecting fields on the toposequence and using appropriate
methods of seedbed preparation. In wetter fields they preferred planting
on raised beds and in dryer fields on flat beds. To protect their cowpea
fields against rodents and monkeys, farmers used five different methods:
(1) fencing with tree branches, (2) fencing with fish nets, (3) tieing a
dog inside the plot, (4) erecting scare crows and (5) having young children
watch the crop at the podding stage. To protect the crop from Fulani
cattle, village chiefs alerted the herdsmen. In their cowpea fields
farmers erected small signboards which the herdsmen were eager to avoid
since these are traditionally used to label government property.


Eearly Adopters
During the three cycles of on-farm trials, a number of component
technologies were screened and tested by representative farmers. New
varieties of cowpeas, rice and cassava and improved management practices
for these crops have attracted the most attention from farmers. They have
taken full advantage of new information, improved planting material and
project support services which the OFAR program has brought closer to
farmers.
By bringing research closer to farmers, OFAR has proven effective in
producing useful recommendations, which have given the extension service
greater credibility. Even at this early stage of OFAR in the Bida project,
very good relationships between farmers and the project's extension service
have been developed. The extension staff is now exerting considerable in-
fluence through its advice and demonstration program for transfer of new
technologies.
After noting the success of short-season cowpeas in oh-farm testing
and verification trials and learning how easy it is to use the Electrodyn










sprayer, farmers showed a remarkable willingness to adopt the new cowpea
production package. During the 1984 main season, about 25. of all project
farmers had grown cowpeas, according to the project Monitoring and
Evaluation Unit.
Surveys undertaken in December 1984 in 24 villages where farmers were
practicing the rice-based system and had contact with the extension service
revealed widespread adoption of new cowpea varieties. Of the 85 farmers
surveyed, 907. grew main-season cowpeas in 1984, compared to only 77. in
1982. About 76X of farmers also planted dry-season cowpeas in 1984-85,
compared to 457. in the year before and none in 1982.
Rice farmers were not the only early adopters of new technology.
There were also many among farmers in the coarse cereal-based and
cassava-based cropping systems. Most farmers using the cereals-based
system planted a sole crop of cowpeas in July/August, while cassava growers
interplanted cowpeas in cassava fields during August.
Initially the farmers only source of cowpea seed was BADP, but
subsequently distribution of seed from farmer to farmer acquired greater
importance. Eventually, 10 to 15 farmers were obtaining their new cowpea
seed from each of the early adopters. Farmers preferred varieties E-60 and
TVx 3236.
During the main season of 1984, farmers reported an average seed yield
ratio of 1:24 (or about 600 kg/ha) for E-60 and 1:28 (or about 700 kg/ha)
for TVx 3236. As for grain quality, farmers preferred E-60, which has seed
of similar size and color to that of local cowpeas, but they liked the
taste and cooking quality of TVx 3236. Farmers were concerned about high
losses in storage. They tried to protect the grains by mixing them with
sand or wood ash and by sealing them in polythene bags, but these measures
were not very effective.
Among the 85 sample farmers, only 287 succeeded in purchasing an
Electrodyn sprayer. Another 457. realized that pests were their biggest
production problem and wanted to purchase the sprayer. The project's
entire stock of 100 Electrodyn sprayers and 300 "bozzles" containing
cymbush insecticide were sold at cost. Farmers placed orders for an
additional 100 sprayers and 1,000 "bozzles."
When asked why they were willing to invest their meager cash resources
in pest management equipment, farmers usually responded that this would









enable them to grow multiple crops of cowpeas. On-farm tests showed the

viability of growing two upland crops, the first planted in May and

harvested in July, and the second planted in August and harvested in
October. The third lowland crop can be planted between November and
January and harvested 8 to 10 weeks later. Commenting on the short season
length of TVX 3236 and E-60 cowpeas, many farmers remarked, "maybe it is a
crop that can stop hunger."


Conclusions

Three years of OFAR trials at 8ADP have shown that it is possible to
quickly identify and take advantage of opportunities for improvement in
local farming systems. Cultivation of short-season cowpeas as a dry-season
crop has proven so successful that demand for seed by farmers is far in
excess of the project's supply. In addition, many farmers have now
reintroduced cowpeas in the main season as an upland crop. Farmers in the
upland-based systems have responded enthusiastically to the new cowpea
technology, even though the two years of main-season trials had created
little interest because of the scarcity of good seed and high cost of
spraying equipment. The rapid adoption of new cowpea seed and pest
management technology has demonstrated the effectiveness of farmers in
disseminating new technology in Niger State.
Unless they stand to gain something, farmers are reluctant to offer
their fields for trials that require significant changes in crop
management. Their reluctance stems from the high cost of land and seedbed
preparation and from the risk involved in making changes. In such
instances farmers prefer to participate as a group in on-farm agronomic
tests. Farmer participation is essential to establishment and maintenance
of research and verification plots and for assessment of farmers' reactions
to the proposed changes in crop management.
The OFAR program at BADP has shown the potential benefits of
agroeconomic data collected by the project's Monitoring and Evaluation
Unit. The agroeconomic data gathered was sufficient for delineating the
important cropping systems and farmer's production practices. These data
facilitated the diagnostic part of the OFAR program and helped researchers
identify opportunities for agronomic experimentation. The project's
interest in OFAR enabled research institutes to test their component










technologies and helped BADP identify components which are suitable for
farmers. For example, the project discovered it could improve local
farming systems by introducing short-season cowpeas and the serrated sickle
and by showing farmers in good fadamas the value of increasing rice hill
density. The project also discovered the need' to undertake land
development engineering for control of surface runoff and seepage water to
permit better utilization of stream flow valley land.
The on-farm adaptive research program has been instrumental in the
development of close contact and cooperation between Bida ADP's commercial,
extension and agronomy sections. The division of project farmers into four
distinct cropping-systems-based target groups was helpful in making the
work of the different sections complementary. Identifying target groups
enhanced the researchers' understanding of farmers' circumstances and
helped them find opportunities for improvement. Technologies identified
through the on-farm adaptive research program were widely sought by the
farmers, giving rise to greater demand for support services.
The major impediment to success in the OFAR program has been lack of
trained personnel. Though the project did have a training program for its
field staff, the technical knowledge and ability of staff members remained
low. Since many of the junior staff were from urban areas, they had little
understanding or appreciation of traditional farming systems or of farmers'
reasons for carrying out certain operations. An OFAR program such as that
of BADP is dependant on field research staff carrying out their work
accurately and conscientiously. In West Africa as a whole, the lack of
trained research staff who can divide farmers into distinct target groups,
carry out diagnostic farm surveys and organize on-farm agronomic
experiments will be the major impediment to expansion of OFAR programs.
However, once the planning, organization and information gathering are
complete, local field staff can implement the OFAR program successfully.


References

1. Bida Agricultural Development Project (BADP). 1983. Mid-Term
implementation report, Vol. 2. Nigerian Federal Ministry of Agricul-
ture. Mimeo.

2. BADP. 1983. Field records for agronomic details, yield and stands:
preliminary report. Nigerian Federal Ministry of Agriculture. Mimeo.









3. Balcet, J.C., and W. Candler. 1982. Farm Technology Adoption in Nor-
thern Nigeria. 2 vols. Washington, D.C.: World Bank.

4. Collinson, M.P. 1982. Farming systems research in eastern Africa: the
experience of CIMMYT and some national agricultural research services,
Michigan State University International Development Paper no. 3. East
Lansing, Michigan.

5. CIMMYT Economics Program. 1980. Planning Technologies Appropriate to
Farmers: Concepts and Procedures. El Batan, Mexico: CIMMYT.

6. IITA. 1983. Minimum pest management in cowpeas. In Annual Report for
1982, 172. Ibadan, Nigeria.

7. IITA. 1984. Introducing a new component in a local farming system.
In Research Highlights for 1983, 7-10. Ibadan, Nigeria.

8. Kowal, J.M., and D.T. Knabe. 1972. An Agroclimatological Atlas of
the Northern States of Nigeria, With Explanatory Notes. Zaria,
Nigeria: Ahmadu Bello University.

9. Norman, D.W., M.D. Newman and I. Ouedsaogo. 1981. Farm and Village
Production Systems in the Semi-arid Tropics of West Africa: An Inter-
pretive Review of Research. ICRISAT Research Bulletin No. 4. Andhra
Pradesh, India.

10. Perrin, R.K. et al. 1976. Froi_! _rEqgRnmic .Da.t L Farmer Recommenda-
tions: An Economic Training Manual. El Batan, Mexico: CIMMYT.

11. Ruthenberg, H. 1980. Farming Systems in the Tropics. 3d ed. Oxford:
Clarendon Press.

12. Shaner, W.W., P.F. Philipp and W.R. Schmehl. Farming Systems Research
and Development: Guidelines for Developing Countries, Vol. 2.
Washington, D.C.: USAID.

13. Sinner, J., A. Jibrin and M. Ashraf. 1983. Dry-season cowpeas at Bida:
an initial assessment of exploratory trials. Farming Systems Program
Discussion Paper, IITA.

14. Zandstra, H.G., et al. 1981. A Methodologv for On-Farm Cropping Systems
Research. Los Banos: IRRI.




































































Published by
International Institute of Tropical Agriculture
PMB 5320, Oyo Road, Ibadan, Nigeria




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