Table of Contents
 Executive summary
 Overview of FSR
 Institutional issues
 Strengthening linkages
 Research issues
 FSR impact
 Appendix A: Agronomy
 Appendix B: Livestock
 Appendix C: Agricultural econo...
 Appendix D: Sociology/anthropo...
 Appendix E: Schedule and people...
 Appendix F: References
 Appendix G: Acronyms

Group Title: national farming systems research strategy for Tanzania
Title: A National farming systems research strategy for Tanzania
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00095073/00001
 Material Information
Title: A National farming systems research strategy for Tanzania report of a task force
Physical Description: 78 p. : ; 30 cm.
Language: English
Creator: Tanzania -- Wizara ya Kilimo
Publisher: National Coordination Unit for Farming Systems Research, Ministry of Agriculture
Place of Publication: Dar es Salaam, Tanzania
Publication Date: 1992
Copyright Date: 1992
Subject: Agricultural systems -- Research -- Tanzania   ( lcsh )
Genre: non-fiction   ( marcgt )
Spatial Coverage: Tanzania
Bibliography: Includes bibliographical references (p. 74-77).
General Note: "September 1992."
 Record Information
Bibliographic ID: UF00095073
Volume ID: VID00001
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 - 263030081

Table of Contents
        Cover 1
        Cover 2
    Table of Contents
        Page 1
        Page 2
        Page 3
        Page 4
    Executive summary
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
    Overview of FSR
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
    Institutional issues
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
    Strengthening linkages
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
    Research issues
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
    FSR impact
        Page 53
        Page 54
        Page 55
    Appendix A: Agronomy
        Page 56
        Page 57
        Page 58
    Appendix B: Livestock
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
    Appendix C: Agricultural economics
        Page 64
        Page 65
    Appendix D: Sociology/anthropology
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
    Appendix E: Schedule and people seen by the task force
        Page 72
        Page 73
    Appendix F: References
        Page 74
        Page 75
        Page 76
        Page 77
    Appendix G: Acronyms
        Page 78
Full Text


/ VO











PREFACE ............................................

EXECUTIVE SUMMARY ....................................

1 INTRODUCTION ...........................
1.1 BACKGROUND............
1.2 CONTENTS OF REPORT .................

2. OVERVIEW OF FSR ......................................
2.1 OBJECTIVES AND FUNCTIONS OF FSR ....................
2.2.1 D iagnostic Activities .............................
2.2.2 Planning and Priority Setting ........................
2.2.3 Design and Implementation of Experiments ...............
2.2.4 Evaluation .................. ...............
2.2.5 Recommendation and Diffusion ................. .....
2.3 FUTURE OF FSR IN TANZANIA .........................

3. INSTITUTIONAL ISSUES ...................................
3.1 FSR AS A SEPARATE PROGRAMME ......................
3.1.1 Research Advantages .............................
3.1.2 Organizational Advantages .........................
3.1.3 Division of Responsibilities: FSR and Commodity Research Teams
3.2.1 -National-Coordination Unit (NCU) ....................
3.2.2 Zonal Team s ..................................
3.3.1 Research Staff .................................
3.3.2 Field Staff ..................................
3.5 MANAGEMENT OF DONOR RELATIONS ...................

4. STRENGTHENING LINKAGES ..................
4.3.1 Research-Extension Liaison Officers (RELOs)
4.3.2 Zonal Committees ..................
4.3.3 FSR Linkage Activities ...............

5. RESEARCH ISSUES ...............

- II : I: II


5.3.1 System Interactions ...................
5.3.2 Agroforestry .......................
5.3.3 Social Dimensions ................... .
5.3.4 Sustainability and Natural Resource Management.
5.4.1 Strengthening Farmer Participation .........
5.4.2 Working with Farmer Groups .............
5.6.1 Clustering and Extrapolation .............
5.6.2 Use of Formal and Informal Surveys .......
5.6.3 Trial Management Differentiation ..........
5.6.4 Technology Evaluation .................

6. FSR IMPACT ....................................
6.1 INTRODUCTION ..............................
6.3 ADOPTION STUDIES ............................


A AGRO NO M Y ............................................
T E A M . . . . . . . . . . . . . . . . . . ..
RESEARCH TEAMS .............................
A3 NECESSARY SKILLS ..................................
PROGRAMMES IN AGRONOMY ....................

B. LIVESTOCK .............. .................... ..........
TO THE FSR TEAM ..............................

B3 NECESSARY SKILLS .........................
B5 LIVESTOCK TRIALS .........................
B5.1 On-Station Research .....................
B5.2 On-Farm Research ......................
B5.3 Types of Research to be Considered ............
B5.4 Animal Nutrition Trials ...................
B5.5 Clustering ........... .................
B5.6 To be Avoided .........................


C. AGRICULTURAL ECONOMICS ................................


C3 NECESSARY SKILLS ........ .......................... 65

D. SOCIOLOGY/ANTHROPOLOGY ......................... ..... 66
PROGRAMME .......... ...................... 66
DI.1 Basic Social Science Skills ............ ..... . ..... .. 66
D1.2 Key Social Science Subject Matter ............ .. ... ..... 67
D1.3 Social Science Tools ................... ........... ... 68
D2.1 Overall Training of FSR Teams in Basic Social Science Skills ..... 69
D2.2 Training of Economists in More Refined Social Science Skills . .. 69
D2.3 Recruitment/Training of Social Science Specialists ............ 69


F. REFERENCES ....................................... 74

G. ACRONYMS ........... ..... .. ............... ... ....... 78


A Task Force was convened in June 1992 to review the current situation of the national Farming Systems
Research (FSR) programme in Tanzania and to generate guidelines, priorities and strategies for its future
development. The time frame was for five years (1992-1997). Proposals have been made taking into
account the larger planning exercise already completed for the formulation of the National Agriculture
and Livestock Research Master Plan (NALRM). The principal clients for the report are the
Commissioner for Research and Training, the Assistant Commissioner for FSR (AC(FSR)), and the staff
of the FSR National Coordination Unit (NCU), and the Zonal FSR Coordinators. The exercise was
organized by the FSR NCU of the Department of Research and Training (DRT) of the Ministry of
Agriculture (MOA). The activities of the Task Force were funded by the Tanzania/Netherlands Project
for the National Development and Coordination of FSR in Tanzania.

The Task Force consisted of the AC(FSR), staff members of the NCU and external consultants, all of
whom have extensive experience with FSR in developing countries. The report itself arose from field
visits and discussions that took place during eight days between June 7th-16th, 1992. Because of limited
time, field visits were only possible to the Northern and Eastern Zones. Most of the Task Force
members, however, had had recent first-hand contact with on-farm and on-station research activities being
undertaken in other zones in the country. This report is based on these visits, reading available reports,
and in-depth discussions with senior administrative personnel, and commodity and FSR researchers in
the DRT, plus more limited discussions with other staff (e.g., extension personnel) in the MOA.

Members of the Task Force were as follows:

Staff of the NCU, MOA, Dar Es Salaam:

Mr. Timothy Kirway, Agricultural Economist, AC(FSR)
Ms. Zainab Semgalawe, Agricultural Economist
Dr. Ann Stroud, Agronomist, Technical Adviser, Tanzania/Netherlands Support Project

External consultants:

Dr. Geoffrey Heinrich, Agronomist, FSR Adviser, Department of Agricultural Research,
Ministry of Agriculture, Botswana
Dr. Deborah Merrill-Sands, Anthropologist, Senior Research Fellow, International Service for
National Agricultural Research (ISNAR), Netherlands
Dr. David Norman (Team Leader), Agricultural Economist, Professor, Kansas State University,
Dr. Sandra Russo, Animal Scientist, Assistant Director, International Studies and Programmes,
University of Florida, USA
Dr. Louise Sperling, Anthropologist, Bean Programme, International Centre for Tropical
Agriculture (CIAT), Rwanda



FSR is recognized in Tanzania as an effective tool for ensuring that the national agricultural research
system addresses the technological needs of resource-poor farmers in a direct and efficient manner.
Initiation of FSR-type activities at different times in different zones, and differences in the way it was
being applied provided part of the rationale for forming an NCU for FSR in 1989. The role of the NCU
is to develop and coordinate a strong, integrated FSR programme within the DRT in the MOA in

The terms of reference for this Task Force were developed by the NCU, and included reviewing the
current situation and assisting the NCU to develop a cohesive strategy for the continued development of
FSR in Tanzania over the next five years.

The Task Force was composed of the AC(FSR), staff members of the NCU, and five external consultants.
Disciplines represented by the consultants included agronomy, agricultural economics, animal science,
anthropology, and institutional management. This report is the product of discussions and activities that
took place in Tanzania in June, 1992. The schedule of meetings, field trips, workshop, and contacts is
listed in Annex E of the report.

The contents of this report are based on the terms of reference for the Task Force provided by the NCU.
Chapter 1 is an introduction and provides the rationale and terms of reference for the report. Chapter
2 describes the objectives of FSR in Tanzania, and implications for the future. Chapter 3 discusses
institutional issues related to the structure, organization, and management of FSR in the country. Chapter
4 discusses the crucial issues of linkages required to ensure positive impact of FSR on farmer production.
Research planning and priority setting, methodological issues, and management issues are specifically
dealt with in Chapter 5. Chapter 6 examines the importance of monitoring the impact of FSR on the
research system as a whole, and on farmer adoption of new technologies. While the document discusses
most issues in detail, it also provides specific recommendations on certain topics. A list of these
recommendations is provided below. The Task Force concluded, however, that the highest priority areas
for development over the next five years are:

Developing and strengthening the technical capacity of the zonal FSR teams.
Developing and strengthening the linkages between zonal FSR teams and commodity
research programmes.
Utilizing and adapting resource-efficient strategies and methodologies to allow more
direct farmer participation in the technology development process.
Applying methods for site selection which permit widespread extrapolation of results.
Using, whenever possible, methodologies which are efficient in the use of research

The Task Force has made a large number of suggestions which are embedded in the report itself. In
addition, a number of recommendations have been made. These are as follows the first number
indicates the chapter in which they are located while the second signifies their order within chapters:

3.1 FSR should be maintained as a separate national programme, but with priority being given to
improving linkages with commodity programmes.

3.2 The NCU should give priority to developing its capacity to provide technical back stopping and

scientific leadership to the zonal FSR teams.

3.3 High priority should be given to filling the two vacant positions for livestock and agronomy
research officers in the NCU with senior scientists who can provide scientific direction and
technical back stopping to the zonal teams.

3.4 Highest priority should be given to ensuring that strong team leaders are in place in all zones.
The essential qualities for these team leaders will be research experience, commitment to farmer
oriented research, aptitude for interdisciplinary research, and strong 'people management' skills.

3.5 Priority be given now to consolidation of capacity, rather than expansion of the FSR programme.
Training and development of the current cadre of young researchers will be the key to the
successful establishment of a national FSR capacity. To provide more scientific leadership, the
remaining posts should be filled by experienced scientists.

3.6 To ensure returns to investment in graduate training, all training scholarships should be given
with the stipulation that the individual will work for at least three years within the FSR
programme upon return from training. Similarly, whenever possible, research for the MSc theses
should be carried out in Tanzania or relate to issues relevant for FSR in Tanzania.

3.7 The balance between staffing and operating funds should be restored and that, at a minimum, the
1990-91 funding levels for recurrent costs per zonal team be reinstated. Also, there should be
no further expansion in FSR staffing without a concomitant increase in funds necessary to carry
out an effective research programme. Researchers should be encouraged to look for ways of
utilizing their extra research time (e.g., working with NGO's on FSR related activities).

3.8 Attention should be given to maximizing efficiency in research resource use through improved
planning of research activities and through the adoption of less expensive research methods.

3.9 The NCU under the leadership of the AC(FSR) and the CRT should take a proactive role in
dealing with donor agencies with respect to FSR matters. Initiatives should include not only
providing interested donors with papers reporting progress and achievements in FSR, but also a
briefing paper providing an overview of FSR in Tanzania, its strategy for programme
development, and its approach to FSR. Periodic meetings with donors of FSR projects should
be convened to improve awareness of overall national programme and how their respective
projects fit. The AC(FSR) should take an active role in contacting prospective donors to generate
interest in funding FSR initiatives in the zones.

4.1 For improved relevance and efficiency, the FSR programme should strive to increase farmer
participation in all phases of the technology development process: from on-station testing to final
verification trials. Adapting and applying methods for increased farmer participation should be
a priority area for attention in strengthening FSR in Tanzania in the next five years.

4.2 Farmer assessment should be built into the evaluation of every technology and should be a
prerequisite for releasing a variety or making a recommendation to extension.

4.3 Priority attention should be given to strengthening collaboration between FSR and commodity
research in the next five years. Agreement should be reached on the appropriate division of roles


and responsibilities between FSR and commodity research and appropriate linkage mechanisms
should be put in place. These linkages will be critical to the future success of FSR in Tanzania.

4.4 Every effort should be made to ensure that senior commodity scientists participate fully in the
design, execution, and analysis and interpretation of informal surveys organized by FSR teams
in the zones. Senior commodity scientists should also be consulted, when appropriate, in the
design of, and interpretation of results from, formal surveys. The ZDRT should take a lead role
in encouraging such types of participation and collaboration.

4.5 The RELO should be an experienced and respected agricultural officer, drawn from either
extension or research, who has sound technical competence, excellent communication and writing
skills, and strong personal initiative. The RELO should report to the ZDRT, but work more than
half time with the zonal FSR team. The appointment of a RELO should be approved by both
research and extension. The RELO should have guaranteed access to transport and his/her
activities should have a specific budget with contributions from both research and extension.

4.6 Additional funds need to be allocated to support the strengthening of FSR technology transfer
linkage activities either through a direct supplement to the FSR budget or channeled through a
budget allocation for the RELO.

4.7 A modest approach should be taken to developing links between FSR and planners and those
responsible for policy/support systems. Given limited resources available, FSR teams should not
be conducting separate policy studies. Rather a linkage should be formed to allow FSR teams
to communicate policy issues which derive from their work directly to planners and those
responsible for policy/support systems. Also planners should be encouraged to communicate to
FSR teams the types of micro-level data they need for policy planning purposes.

5.1 Because of their intimate knowledge of farmers' problems, FSR teams should participate actively
in research priority setting at the zonal level. They can also help through characterizing the major
zonal farming systems, and in selecting specific areas within the zone to work. As a result of
this priority setting exercise FSR teams should apportion their activities to arrive at a balanced
portfolio of researchable themes.

5.2 A system of research planning should be implemented at the zonal level, involving all parties
(i.e., FSR and commodity researchers, technology transfer agents, etc.), that will ensure joint
planning of research, good information sharing, communication and accountability during
implementation, and analysis, documentation and dissemination of the results. FSR teams should
give priority to synthesizing and writing up results from surveys and multi-season trial results as
key inputs into future zonal planning.

5.3 Current capacity of FSR teams predicates against initiating research which does not have a pay-
off in a reasonably short time period (e.g., sustainability issues which are not associated with
short-run gains in productivity). However, two areas where FSR has a comparative advantage,
and should give priority in the medium term are: crop-livestock interactions and social science
research, including gender analysis.

5.4 FSR teams should play a leading role in bringing the farmer perspective into the research system.
Also, where appropriate, teams should, in actively encouraging collaboration with the limited-

resource farmers, pay particular attention to collaborating directly with women, the major
agricultural workers on Tanzanian farms.

5.5 The FSR team members should develop communication skills for working with farmers, and have
the ability to collect, record and analyzes qualitative and non-conventional quantitative data.

5.6 FSR teams should become aware of, and increasingly use newer methods and tools for
differentiating users, while gender analysis should become a regularly used methodological tool.

5.7 To increase the return from limited research resources, greater attention should be given to
clustering experimental units and to site selection to permit broader extrapolation of results.

5.8 Informal surveys should serve as a prime survey method of the FSR team with attention focused
to improving their quality. Formal surveys should be conducted only to address highly
specialized subject topics, where quantification is particularly important; they are more costly and
are not necessary as a general verification tool.

6.1 Studies should be undertaken to document the impact of FSR-related activities on the technology
development process, and assess the degree, level and pattern of adoption of new technologies
as a result of FSR and other activities.




The 1991 National Agriculture and Livestock Research Master Plan (NALRM) identifies FSR as a first
priority programme. FSR, with its emphasis on problem identification and interaction with farmers, is
viewed as essential for the development of the country's agricultural sector. The NALRM indicates that
the national FSR programme is to be strengthened, although no specific resource allocations are made.
The report of the Task Force is designed to provide research managers in DRT with concrete guidelines,
priorities, and strategies for developing a sustainable and effective FSR capacity in Tanzania over the next
five years. Recommendations and proposals are made within the larger context of the NALRM.

FSR activities began in Tanzania in the mid-1970's in response to a growing realization that the
technological requirements of limited resource farmers needed to be addressed in a more efficient manner.
Donor agencies have played a significant role in supporting the development of FSR particularly from
1983 with a USAID-supported FSR project in the Eastern, Central and Northern Zones. Since then, FSR
activities have been implemented to varying degrees in all of the seven research zones in the country.
Initiation of FSR activities at varying times in the different zones, supported by different donors, and with
some differences in FSR methodology, provided part of the justification for the strengthening, in 1989,
of the NCU for FSR within the reorganized Department of Research and Training (DRT) in the recently
renamed Ministry of Agriculture (MOA). This has been facilitated with project funds provided by the
Netherlands Government [Merrill-Sands, Kirway and Semgalawe, 1990]. The Netherlands Government
is also helping to fund FSR work in the Lake Zone. The importance the Government of Tanzania (GOT)
gives to FSR is evident in the appointment of an AC(FSR), and the priority assigned to the FSR approach
in the NALRM.


The contents of this report are based on the terms of reference drawn up for the Task Force by the NCU.
To address the various points the report is structured as follows.

Chapter 2 provides an overview of the objectives and functions of FSR in general as well as its specific
application in Tanzania.

The next three chapters are devoted to discussing issues that reflect the way in which FSR is and will be
implemented in the Tanzanian context. Chapter 3 considers institutional issues relating to programme
structure and organization, resource management, training, and donor relations. Chapter 4 is devoted
to an examination of the various linkages that are critically important if FSR is to be effective and if its
contribution is to be maximized. The interactive linkages examined include FSR-farmer, FSR-commodity
research, FSR-technology transfer, and FSR-policy and support systems. Chapter 5 examines several
research issues identified as key to the future development of FSR in Tanzania, including planning and
programming of research, research gaps, and some methodological issues.

Chapter 6 considers the importance of paying greater attention to assessing the impact of FSR, both in
terms of helping to provide inputs that will improve the efficiency of the technology development process,
and in terms of adoption of improved technologies by farmers -- the credit for which is shared with


Short appendices are devoted to the major disciplines associated with FSR work (i.e., agronomy, animal
science, agricultural economics and rural sociology/anthropology). These indicate the contributions each
discipline makes, the skills that are required and suggestions for strengthening them over the next five
years. The final appendices provide a diary of activities and meetings, papers referred to in the text and
definitions of the acronyms.

- 10 -



One of the major goals of agricultural development is to improve productivity and the standard of living
of farming families. Two complementary strategies are important in achieving this, namely the:

Development and dissemination of relevant improved technologies.
Design and implementation of relevant improved policy/support systems.

In relation to this, the major objective of FSR is helping improve the efficiency, relevancy and impact
of the research and policy/support systems in accelerating this process of agricultural development.
Farmers (who are the users), technology transfer and development staff (who are the transmitters), and
researchers/planners (who provide the potential means) are all essential and important contributors to this
process. Although this breakdown indicates the major responsibilities of each of the actors, it is
important to emphasize that all the actors can, and should, contribute productively to the research process.
FSR plays an important role in bringing these different actors together and pooling their expertise in
improving the efficiency of the technology development process.

The facilitative role of FSR in encouraging the process of agricultural development can be disaggregated
into a number of functions.

Characterizing major farming systems and client groups, using agro-ecological and
socioeconomic criteria, to diagnose priority production problems and identify
opportunities for research.

Stimulating within research:
A problem-solving approach focused on farmers as primary clients of research.
An interdisciplinary systems perspective.

Undertaking on-farm experimentation, considering conditions of targeted groups of
farmers sharing common production problems and same potential solutions, by:
Adapting existing technologies to farmer conditions with inputs from farmers.
Focusing on specific research themes which involve system interactions.

Promoting farmer participation in research as collaborators and as evaluators of

Performing a linkage function through:
Feeding back information from farm-level research to commodity and
disciplinary programmes working on-station, to improve the efficiency of the
research process.
Providing integration across commodity/disciplinary groups.
Improving priority setting and planning of research.
Promoting collaboration with technology transfer and development agencies to
increase efficiency of technology transfer.
Feeding back farm-level information to planners and policy makers to create

- 11 -

more favourable conditions for agricultural development.

Analyzing and documenting impact of FSR-related activities:
On the technology development process.
In terms of technology adoption.

These functions are similar to those delineated and evaluated by Tanzanian FSR and commodity research
staff as presented in the planning document for the NCU [Merrill-Sands, Kirway and Semgalawe, 1990:

It is stressed that FSR is complementary to, and supportive of, commodity research. The two should be
viewed as interdependent, both making vital contributions towards the development of relevant
technologies for farmers.


FSR in Tanzania is defined as an approach designed to generate relevant technologies for specific clients,
namely resource limited households. To accomplish this primary objective a research process developed
by CIMMYT is followed: an interdisciplinary systems perspective is used involving farmers to diagnose
problems and needs which are prioritized, then potential solutions are designed, developed and evaluated:
This technology development process considers: existing aspects and relationships among various
components of the farming system (e.g., crops, livestock, trees, off-farm activities, etc.) and variations
existing among farmers (e.g., socio-economic differences, resource levels, farming practices, etc). FSR,
which emphasizes on-farm adaptive research and interactions among components of the farming system,
complements the technical commodity and disciplinary research which is applied in nature and conducted
on research stations. The process is iterative and involves five major activities: description/diagnosis,
priority setting and planning, experimentation, evaluation, and recommendation and diffusion.

The primary product of FSR is farmer-focussed technology that will provide a greater number of options
for small farmers to solve problems, to relieve constraints, and to exploit new opportunities in a
sustainable way (i.e., can be maintained or supported locally), will not further stress the environment,
and will make better use of or expand the use of existing resources.

To accomplish the above objective, researchers must develop, through observation and dialogue, an
understanding of the farmers' circumstances including among others how they perceive their situation,
develop production strategies, make decisions, set priorities and manage their resources. Research
emphasis is then given to building upon farmer's indigenous knowledge.

Four concepts are basic to the FSR approach:

Selection of research priorities is based on field-level diagnosis involving farmers.
Research is carried out on farm with farmers, particular attention being paid to selecting
research sites and farmer collaborators which are representative of larger target groups.
Emphasis is given to understanding system interactions.
Farmers are involved in technology development and assessment.

General principles for FSR arising from these basic concepts include:

- 12-

Recognizing that small farmer circumstances must be understood in order to find
acceptable solutions.
Targeting technological solutions to specific farmer groups operating within similar
socio-economic and agroecological conditions.
Appreciating that farmers have useful, well-founded indigenous knowledge.
Recognizing that farmers operate in systems where resources are limited and where risks
may be high, so they have multiple objectives.
Emphasizing greater participation by farmers in all stages of the research process.
Providing a more rigorous test of technologies under a more realistic human and
environmental background.
Providing linkages between farmers, station-based researchers, extension/development
staff and planners.

2.2.1 Diagnostic Activities

The objective of diagnosis is to understand the physical, social, economic and production features of the
farming systems leading to a preliminary definition of recommendation domains or client groups which
have similar problems, constraints and missed opportunities. To accomplish thisjollection and analysis
of secondary data is made followed by an informal or exploratory survey. This may be followed by a
formal or verification survey when quantifiable data are required. Single or multiple visits can be used
and both groups and individuals interviewed. Diagnosis continues during implementation of the
experimental work because specific data may be lacking. Through continual interaction with the farmer
and continued observation of the farmer's circumstances, the understanding of the farming system is
constantly going through revision, which can have implications on the research programme. The farming
system is dynamic and the farmer is continually reacting to outside forces (e.g., weather, population
expansion and retraction, policies affecting prices, input availability, etc.) which affect production

Hypotheses concerning problems and constraints and their causes are developed, tested, and adjusted as
the information becomes more complete. Examples of problems are those which directly reduce crop
yields or enterprise productivity, reduce the efficiency of inputs regardless of the effects on yields, affect
the farmer's use of resources, or those associated with sustainability.

2.2.2 Planning and Priority Setting

The researcher uses the data collected during diagnosis to decide on future activities based on the need
identified: plan experiments, further diagnosis, or increased extension activity. The planning of
experiments follows six steps [Tripp and Wooley, 1989]:

Identify problems limiting the productivity of the farming system;
Rank these problems;
Identify the causes of the problems;
Analyze the interrelationships among problems and causes;
Identify possible solutions to the problems;
Evaluate the possible solutions.

Farmers can collaborate in this process to ensure that problem identification, prioritization and solutions
are in line with their thinking. During the first year of work the diagnostic material is used extensively
in this process. Experimental results and further observations contribute to the iterative planning process.

2.2.3 Design and Implementation of Experiments

Based on the decisions made during planning, objectives are set and experiments designed and
implemented. Decisions concerning the type of experiment and type of management (i.e., the degree of
farmer involvement), its location (i.e., on-farm, on-station or both), data to be collected, provision for
farmer feedback, etc., must be made. Experiments can have a diagnostic function among others. Merely
interacting with farmers in their situation can help researchers to fine-tune their understanding of the
farmer's point of view and circumstances. Trials can range from verification of research station results
on farmers' fields to trials addressing innovative new research themes.

2.2.4 Evaluation

Evaluation can begin during the implementation of the experiment, when, for example, the researcher
may be forced to evaluate the validity of a treatment when the ease or difficulty of application becomes.
apparent. Farmer assessment occurs throughout and after the experiment. Evaluation must include a
combination of: statistical analyses, observations, farmer assessment, and economic analyses. The final
interpretation will be used to re-design future experiments or research programmes, or to finalize a
recommendation. Evaluation should also serve the function of redirecting research, both on-farm and
on-station in terms of adjusting selection criteria for technologies to redefining research priorities.

2.2.5 Recommendation and Diffusion

FSR researchers must work closely with commodity researchers and technology transfer agents in this
final stage of the FSR process when recommendations are formulated, verified, and disseminated to
specific target groups.

Although the methodology defined is based on the CIMMYT approach, it is being continually updated
to suit Tanzania's needs and the evolving capacity of its researchers. For example, stronger emphasis
is now being given to incorporation of more active farmer participation, understanding household food
security issues, gender issues, livestock systems, and crop/livestock interactions.

Efforts, coordinated by the NCU, have been made to develop a coherent approach and methodology for
FSR which is applied by teams in all seven zones. The emphasis on a common approach results from
the large number of inexperienced FSR staff. The assumption is that once a basic capacity is developed,
there will be more room for experimentation with new methods and techniques. Strong interactions with
commodity and disciplinary scientists are encouraged through diagnostic surveys, joint planning, priority
setting, and research monitoring and evaluation.

- 14-


The government of Tanzania has expressed a strong commitment towards FSR as a research approach
designed to produce technologies aimed at improving the productivity and well-being of limited resource
farmers. FSR has been institutionalized in Tanzania and a basic capacity for FSR is in place.
Nevertheless, many of the researchers are young and inexperienced and research quality is variable.

To realize its potential contribution to agricultural development in Tanzania, we believe that the national
FSR programme should focus on following five priority objectives in developing its programme over the
next five years:

Strengthen the technical capacity of FSR researchers in the zonal teams through
comprehensive training and technical support.
Build and strengthen linkages between zonal FSR teams and commodity and disciplinary
research programmes.
Increase farmer participation throughout the technology development process using cost
effective methods.
Apply methods for site selection which provide for broad extrapolation of results.
Emphasize cost saving techniques and methods wherever appropriate.

Specific suggestions and recommendations are presented in detail in the following chapters.

- 15-



The long term vision for FSR outlined in the NALRM foresees eventually disbanding the zonal FSR
teams when the 'farming systems perspective' has been sufficiently integrated within commodity and/or
disciplinary teams. The idea apparently is that the commodity or disciplinary programmes will assume
responsibility for on-farm trial work and that a separate unit of agricultural economists will be responsible
for surveys and special studies. We agree with one major portion of this proposal, that all researchers
need an FSR perspective (i.e., they need to be able to think in terms of designing and evaluating
technologies in a farming systems context). Nevertheless, we do not see the absorption of farming
systems research functions within commodity or disciplinary programmes as a viable option for
organizing FSR in the next five to seven years. Nor do we consider it advantageous to separate
agricultural economics as a separate programme at the zonal level. Moreover, we suggest caution in
considering this model even as a long-term goal for the evolution of FSR in Tanzania.

We believe that a strong case can be made, on both research and organizational grounds, for keeping FSR
as a separate programme. Given the relative youth of the FSR approach in Tanzania and the relative
inexperience of much of the FSR staff, an integrated national programme is needed to:

Provide the focus and concentrated effort required to develop a strong capacity in FSR.
Build up the research skills and competence of the FSR zonal teams.
Refine and adapt FSR methods appropriate for Tanzania.
Demonstrate the utility of the approach to improving the relevance of research to
farmers' needs.

Nevertheless, the advantages of this organizational model for developing research capacity an4 specialized
skills can only be realized if priority is given to developing strong linkages with commodity and support
programmes (see Section 4.2). The new research planning and programming processes to be instituted
in the zones as well as the clear assignment of responsibilities for establishing links between FSR and
commodity teams to the Zonal Director of Research and Training (ZDRT) and Zonal FSR Coordinator
provide the minimal conditions for ensuring stronger linkages across programmes.

3.1.1 Research Advantages

In addition to facilitating the development of specialized research skills, there are several other strong
research reasons for keeping FSR as a separate programme:

The systems orientation of FSR facilitates and provides a natural milieu for working on
research themes that cross a number of commodities and/or subject areas (e.g.,
crop/livestock interactions, sustainability, etc.). The FSR team approach, being oriented
to farming system complexities faced by farmers, is much more likely to ensure that such
issues are addressed than would be the case if only a commodity-based approach were
used to conduct research.

The systems orientation of FSR also helps to provide integration across commodity teams

- 16 -

(e.g., consideration of crop associations, such as cotton/cowpea mixtures in Eastern
Zone). The FSR team can help ensure that attention is paid to such interactions by one
of the commodity teams rather than being neglected by both.

The interdisciplinary, on-farm, operational approach of FSR has resulted in the
development of special skills and methodologies for ensuring incorporation of a socio-
economic, in addition to a technical perspective, in the development and assessment of
technology. The socio-economic perspective is particularly difficult to accommodate in
commodity or discipline oriented research and in straight forward, multi-locational testing
involving trials that are usually researcher managed and implemented.

The close interactions between agricultural economists and agronomists and other
technical specialists, promoted through FSR teams, fosters a stronger farmer and field
orientation to social science research. When agricultural economics is separated out it
tends to become more disciplinary in orientation and focus more on macro-level policy
and marketing analyses.

The farmer participatory orientation that underpins FSR helps ensure that: the voice of
the major client of agricultural research (i.e., the farmer and his/her household) is heard;
farmers participate directly in the research process; and that due recognition is given to
the fact that the farming community consists of different types of farmers for whom
technology needs may differ. FSR teams' intensive interactions with farmers help ensure
that problems articulated by farmers are fedback to the relevant commodity programmes
and, if considered of sufficient priority, are incorporated into the research programme.
They also help to ensure that a number of possible technological options are offered to
farmers rather than blanket (homogenous) technological packages that may fit the
circumstances of some farmers and not others.

Experiences from other countries, as well as from the Southern Highlands Zone in Tanzania, indicate that
both the interdisciplinary systems perspective and farmer participation tend to get marginalized when the
social sciences are separated out and on-farm work gets folded into the work of commodity and/or
disciplinary programs [Merrill-Sands, Kirway and Semgalawe, 1991]. Yet, it is precisely in these areas
where FSR can make its strongest contribution to enhancing research relevance.

3.1.2 Organizational Advantages

There are also important organizational reasons for keeping FSR as a separate programme, given the
Tanzanian context. The option of folding FSR into the commodity and disciplinary based programmes
appears attractive for integrating on-farm and on-station research, as well as for increasing efficiency in
resource use. Nevertheless, experiences from both within and outside of Tanzania, have shown that this
organizational model is difficult to implement successfully unless several basic conditions are met
[Merrill-Sands, Kirway and Semgalawe, 1991]:

Other institutions, such as marketing boards, strong extension services, non-governmental
organizations (NGOs), farmer organizations, or development agencies, place a strong
demand on applied research and development of appropriate technologies.

Commodity researchers have a strong client orientation and have received focused
training in problem-oriented and farmer-oriented research methods.

Value and incentive systems are in place which reward applied research aimed at impact
on farmers' fields rather than only the contribution to scientific knowledge.

Transport and operational funds are sufficient to permit numerous research programmes
and units pursue individual field activities.

Strong mechanisms are in place to ensure coordination across programmes in research
priority-setting, planning and review to ensure a system's perspective in problem
definition as well as responsiveness and accountability to clients' needs.

Since most of these conditions are not present in the Tanzanian context and are unlikely to be developed
in the next five years, we conclude that it would be both more effective and more cost efficient to retain
a separate FSR programme.

Recommendation 3.1 FSR should be maintained as a separate national programme, but with priority
being given to improving linkages with commodity programmes.

3.1.3 Division of Responsibilities: FSR and Commodity Research Teams

It must be emphasized that the recommendation that FSR remain a separate programme requires priority
attention be given to strengthening collaboration between FSR and commodity and disciplinary-based
programmes (see Section 4.2).

A basic condition for separate FSR and commodity teams to woik effectively together is a clear division
of responsibilities which builds on the comparative advantage of each and is agreed upon by both
[Merrill-Sands and McAllister, 1988]. There is no recipe for this; it depends on the types of research
problems identified and the relative capacities of the research teams.

Based on our understanding of the Tanzania situation, however, we propose the following guidelines for
collaboration in FSR: in broad brush strokes commodity teams should emphasize applied, component-
based, research while FSR should concentrate on adaptive research aimed at testing technological options
for diverse groups of farm households. More specifically, commodity teams should take the lead in:

Designing and implementing most trials carried out on-station, although FSR teams may
need to run controlled experiments involving multiple components on-station as well.

Conducting more standardized varietal testing (i.e., both on and off station) over wide
areas. FSR teams can assist them in developing skills for eliciting farmers' assessments
of varieties. If research is to be responsive to clients' needs, it is important for
commodity researchers to be exposed to farmers' problems and farming conditions.

On the other hand, FSR teams should take the lead in:

Testing varieties to fit specific niches in farming systems or to meet specific farmer

- 18 -

needs, such as early maturing maize.

Designing and implementing trials in which farmer participation is particularly important.
They have the comparative advantage in designing and analyzing trials which are
interpretable to farmers, where farmers are to have a major role in management and
implementation, and which strive to reflect real farm conditions.

Conducting on-farm (and in some cases on-station) agronomic trials which focus on
system interactions (e.g., crop-livestock) or crop associations (i.e., intercropping or relay
cropping). Commodity and FSR researchers should work together when complex, multi-
component, trials are to be carried out on farm.

Organizing informal surveys to characterize farming systems or explore specific issues
or problems. FSR teams should, however, always involve relevant commodity
researchers. Joint diagnosis of problems in the field improves the quality of problem
definition and facilities the subsequent prioritization of problems to be addressed in

Designing and implementing formal surveys, under the specific conditions where such
surveys are needed. Relevant commodity researchers should be consulted in the design
of the questionnaire and in the interpretation of results.

Encouraging commodity scientists to consult farmers early in the design of on-station
trials and in using levels of non-experimental variables which more realistically reflect
farmers' conditions.

Promoting linkages with extension and other partners (development agencies and NGOs)
which could assist-in carrying out a larger on-farm trial programme and in transferring_
technological alternatives to farmers.


The national FSR programme currently consists of a National Coordination Unit (NCU) based at the
headquarters of the DRT in Temeke and seven interdisciplinary teams based in each of the zones. The
AC(FSR) heads both the national FSR programme and the NCU. Each zonal FSR team is led by a Zonal
FSR Coordinator and has a minimum complement of an economist, agronomist, and livestock specialist.
The Zonal FSR Coordinator and his/her team report administratively to the ZDRT and technically to the
AC(FSR). Given this split reporting relationship, close cooperation is needed between the Zonal
Directors and the AC(FSR) to ensure the quality of FSR research, strong linkages between FSR and other
programmes, and effective deployment of FSR team members.

We concluded that this structure for the national FSR programme, with some fine tuning, would be
appropriate for the development of the programme over the next five years. Some recommendations and
suggestions for improving performance are outlined below.

- 19 -

3.2.1 National Coordination Unit (NCU)

We endorse the following functions for the NCU as agreed upon in the 1990 Senior Research and
Extension Administrators Workshop convened in Arusha [Semgalawe and Anandajayasekeram, 1990]:

Develop the overall strategy and framework for FSR in Tanzania.

Develop and institutionalize a dynamic national FSR programme focused on priority
problems confronting resource-limited farm households through:
Provision of technical leadership and training.
Consolidation and dissemination of research results.
Monitoring and supporting activities of zonal teams.
Assessing the impact and output of zonal teams.

Develop and manage a national cadre of trained and experienced researchers and
technicians for FSR.

Ensure coherence and coordination among zonal FSR activities.

Strengthen information exchange with FSR programmes in other countries and
international institutes as well as with other FSR efforts in Tanzania.

Solicit, rationalize, and coordinate donor support for FSR activities.

Develop strong zonal teams and ensure the sustainability of on-going FSR activities.

We believe that a strong FSR capacity in Tanzania can only be developed if the NCU is able to strengthen
its role in providing technical back stopping and scientific leadership to-the zonal FSR teams which are
currently staffed largely by young, inexperienced researchers. As a relatively new field of research, FSR
is undergoing continuous innovation and improvement in research methods and techniques. FSR's areas
of research specialization, such as surveys, on-farm trials, and farmer participatory methods, are still in
the process of development. The NCU, working closely with the zonal teams, has the comparative
advantage in stimulating the developing, testing, and evaluating such methods. It can also play a key role
in providing technical support to the teams, in reviewing research plans and monitoring the scientific
quality of research results, and in promoting dynamic and problem-focussed research programmes at the
zonal level.

Recommendation 3.2 The NCU should give priority to developing its capacity to provide technical
back stopping and scientific leadership to the zonal FSR teams.

To do this, in addition to the AC(FSR), the NCU will need to be fully staffed by a core multidisciplinary
team with expertise in agricultural economics, agronomy, and livestock research. If full-time staffing
cannot be put into effect in the near future, then the NCU might consider having one of its senior
scientists based in zonal teams. These persons would have to divide their responsibilities between
national coordination and technical back stopping and focused technical contributions to zonal FSR
activities. If outposted to teams, they should be placed in zones (such as Eastern and Central) where
regular communication with headquarters is feasible.

-20 -

Recommendation 3.3 High priority should be given to filling the two vacant positions for livestock and
agronomy research officers in the NCU with senior scientists who can provide
scientific direction and technical back stopping to the zonal teams.

To allow the NCU to devote more staff time and resources to technical back stopping of the zonal teams,
every effort should be made to streamline administrative procedures in the NCU. This is particularly
important for the technical advisor, currently funded through the Netherlands/Tanzania project supporting
the national development and coordination of FSR. Her skills and experience as a on-farm agronomist
should be harnessed to strengthen research in the zones and not dissipated on administrative or logistical
coordination functions. We suggest that, in addition to the hiring of a competent administrative secretary,
the recurrent administrative responsibilities should be clearly allocated among NCU staff members. The
administrative burden should be reduced once the full complement of NCU staff has been posted.

3.2.2 Zonal Teams

The NALRM indicates that the zonal teams will be staffed, at a minimum, by an agricultural economist,
and agronomist, and a livestock scientist. This staffing complement has been met in all zones but one,
albeit largely through the placement of junior scientists and new recruits. We feel that flexibility should
be maintained to ensure that the posting reflect the priority research issues of each zone. For example,
if in a particular zone livestock is not a major enterprise for small holders, zonal research would be better
served by placing an additional agronomist or applied social scientist rather than a livestock scientist, at
least in the near term. Two key areas identified for strengthening the zonal teams in the future are:

The need for stronger team work in planning.
Executing research and the need for stronger team leadership in most zones.

To strengthen their interdisciplinary research, we encourage each of the FSR :c.a= tG jointly plan and
review their research and share responsibilities for the execution of research. Economists should
participate in on-farm trials and agronomists should be actively involved in the design and implementation
of surveys and interpretation of results.

Developing strong FSR zonal team leadership will be a key step in building an effective and productive
FSR programme in the next five years. This should be a priority objective in the near term. The Zonal
FSR Coordinators have four key areas of responsibility that affect the overall effectiveness of FSR:

Providing scientific leadership for the team and research guidance for the younger more
inexperienced members.
Promotion of interdisciplinary and systems-oriented research.
Promotion of linkages with commodity scientists and with technology transfer agencies,
whether extension or NGOs.
Efficient use of scarce resources through effective team management and administration.

Their roles will become even more important as the emphasis in research planning shifts from
commodities to the zones.

We also suggest that the Zonal FSR Coordinators be given short course training in team management and
leadership, especially communication skills. Also, to improve the Zonal FSR Coordinators' ability to

-21 -

develop effective interdisciplinary teams, we suggest they are sent to short course training in areas outside
of their own disciplines. For example a team leader who is an agricultural economist could benefit
significantly from attending a course on data analysis for agronomists. This would allow them to provide
better guidance to the team as a whole as well as to the younger less experienced members.

Recommendation 3.4 Highest priority should be given to ensuring that strong team leaders are in place
in all zones. The essential qualities for these team leaders will be research
experience, commitment to farmer oriented research, aptitude for
interdisciplinary research, and strong team building and 'people management'


3.3.1 Research Staff

The approved staffing level for FSR in the NALRM is 39 national researchers. The national FSR
programme, currently with 37 national researchers and five expatriate technical advisors, is therefore
almost fully staffed. Staffing has expanded rapidly (i.e more than 35 percent since 1990) through the
recruitment of recent graduates. Almost half of the staff is inexperienced with only one year or less
working in FSR. Only a third of the staff have a research degree (i.e., MSc or above). The principal
challenge facing the FSR programme in the next five years will be to provide the formal and in-service
training required to develop these staff into a strong and effective research team of comparable standing
with the commodity teams.

Recommendation 3.5 Priority be given now to consolidation of capacity, rather than expansion of the
FSR programme. Training and development of the current cadre of young
researchers will be the key to the successful establishment of a national FSR
capacity. To provide more scientific leadership, the remaining pests should be
filled by experienced scientists.

We suggest several strategies for strengthening human resources within FSR:

The remaining posts should be filled by well qualified scientists with at least five years
research experience rather than by new recruits. Such cross programme transfers should
be feasible given the current reorganization being undertaken for the NALRM, but will
require the full support of the Commissioner for Research and Training.

Special attention needs to be given to in-service training of new recruits as well as more
senior commodity researchers transferred to FSR. This includes the in-country FSR
orientation course, short courses in FSR methodology, and internal study tours, as well
as trying to establish advisory relationships with senior scientists supportive of FSR in
the zones. Funds for much of this type of training are available through the
Tanzania/Netherlands project for strengthening the development and coordination of FSR.
Arranging such training is an important responsibility of the NCU, while the Zonal FSR
Coordinators should have explicit responsibility for providing research guidance and
support to young researchers on their teams.

No team should be formed without at least one experienced FSR researcher, even if this


is accomplished only through a temporary staff transfer of a year to 18 months.

The NCU should give priority to back stopping zonal teams which have least research
experience to ensure that they design and implement a sound programme of research.

There are ample funds for graduate training. Through various donor projects, 21 scholarships for
graduate training will be available over the next five years. We have several suggestions for the
management of graduate training:

The policy should be to send staff to two year MSc programmes which require
preparation of a thesis. Whenever possible the thesis should pertain to FSR work in

For staff who do not have the necessary academic qualifications for admission to such
programmes overseas, or who are not able to pass the TOEFL and GRE exams required
for admittance to American universities, one year masters programmes overseas (e.g.,
in the UK or Netherlands) should be explored. Such one year masters programmes could
also be used to give staff the additional training required for entrance into a full two year
MSc programme. Similarly, support should also be given for researchers to pursue an
MSc programme at Sokoine or other African universities.

At least two out of the 11 MSc training positions funded under the Tanzania/Netherlands
project for the national development and coordination of FSR should be converted to a
PhD scholarship position. Most other national research programmes have at least one
PhD researcher. Achieving more equality in the training and experience levels of FSR
and commodity researchers will significantly improve the possibilities for constructive
and collegial collaboration.

The NCU should make arrangements with a contracting agency, such as Winrock
International, to handle placement, administration, and support of at least some of their
staff in two year MSc programmes at American universities. NCU should also consider
hiring locally a training coordinator. Currently too much of time of NCU staff is being
spent on placement of students.

Training has to be carefully managed to ensure that the core research programmes of
zonal teams are not disrupted by sending staff for long-term training.

Recommendation 3.6 To ensure returns to investment in graduate training, all training scholarships
should be given with the stipulation that the individual will work for at least three
years within the FSR programme upon return from training. Similarly,
whenever possible, research for the MSc theses should be carried out in Tanzania
or relate to issues relevant for FSR in Tanzania.

3.3.2 Field Staff

Field officers play a very important role in implementing on-farm trials, working with farmers, and in
administering surveys. They should also receive training to upgrade their skills and competencies. It

-23 -

is strongly suggested that they receive specific training in farmer participatory methods and eliciting
farmers' evaluations of trials as well as in operating farmer managed trials.


Although national funding for research in DRT as a whole is low and a serious constraint to the conduct
of effective research, funding for the national FSR programme had been comparable to that of other
programmes until 1992. Currently teams are facing serious constraints in operating funds. While FSR
staffing increased by more than 25 percent in one year and the number of zones covered expanded from
four to six, operating funds were cut from 16 million to 7 million shillings in 1991. This represents a
reduction from approximately 590,000 Tsh per researcher to 200,000 Tsh per researcher. This
precipitous drop in funding has resulted in a significant decline in zonal research activities as well as
under utilization of programme staff.

Recommendation 3.7 The balance between staffing and operating funds should be restored and that, at
a minimum, the 1990-91 funding levels for recurrent costs per zonal team be
reinstated. Also, there should be no further expansion in FSR staffing without
a concomitant increase in funds necessary to carry out an effective research
programme. Researchers should be encouraged to look for ways of utilizing
their extra research time (e.g., working with NGO's on FSR related activities).

Given the existing and projected funding constraints for zonal level research, we suggest that some
reallocations be made within the Tanzania/Netherlands project for the national development and
coordination of FSR to permit increased funding of zonal research activities. Emphasis should be given
to collaborative commodity team/FSR research, to extension/development agency linkage activities, and
to research activities designed to increase farmer participation. FSR projects should be funded on a
competitive basis.

Recognizing that GOT research funds are likely to be constrained in the next five years, we encourage
the NCU and the zonal teams to seek creative means for increasing efficiency in resource use, primarily
with respect to operating funds and transport. While we did not have sufficient time to examine resource
management processes in depth, several strategies which have been used in other countries could be
considered. Some of these, which are discussed in more depth in Chapter 5, are as follows:

Focus attention on priority client groups.

Choose sites for on-farm work which will facilitate broad extrapolation of results.

Cluster on-farm trials to reduce transport costs and staff time spent in traveling. Locate
clusters as close to the station as possible without jeopardizing the relevance of research

Experiment with working with farmer groups as a means for expanding the on-farm trial
programme and as a cost effective means for involving farmers in trial design and
assessment of results.

Use more cost effective research methods such as informal, rather than formal surveys,


and farmer managed trials.

Institute regular team meetings to plan activities and coordinate transport and logistic

Develop collaborative research projects with commodity teams to share costs.

Develop partnerships with NGOs or extension/development agencies to implement a
larger trial programme or for executing surveys.

Recommendation 3.8 Attention should be given to maximizing efficiency in research resource use
through improved planning of research activities and through the adoption of less
expensive research methods.


Several donors are supporting FSR in Tanzania. The objective of the national programme is to increase
support by encouraging different donors to fund individual zones. This is a practical approach to securing
additional funds for research. Nevertheless, it will place a heavy responsibility on the AC(FSR) and the
NCU to closely monitor and coordinate donor activities in order to ensure a coherent research programme
focused on problems judged by the Tanzanians to be of highest priority. Different donors, operating
independently, each with their own vision of FSR and its role within research, and each focused on
project rather than national programme priorities, can quickly undermine the coherence and strength of
a national FSR programme. Senior research managers in DRT consulted in 1990 saw coordination of
donors as the most important responsibility of the NCU.

We have.several -suggestions for strengthening the coordinatiort of externally-funded FSR teams and
ensuring their effective integration into both the national FSR programme and zonal research activities:

Close collaboration between ZDRTs and the AC(FSR) will be required if the NCU is to
perform this critical function of coordinating donor support to diverse FSR teams.
ZDRTs should consult the AC(FSR).on projects being developed in their zones which
have an important FSR component.

AC(FSR) and his staff should be involved in the review at the national level of donor
projects with an FSR component and should be involved in the discussions with the donor

The AC(FSR) and NCU staff should be consulted on the recruitment of expatriate FSR
advisors or technical assistance staff to ensure that the proposed personnel have the
experience and qualifications required to contribute effectively to the development of FSR
in Tanzania and to zonal FSR research.

Expatriate technical assistance personnel working in FSR zonal teams should be fully
briefed on the functions of the national FSR programme and should be expected to
participate fully in the programme. They should report regularly to the AC(FSR)/NCU
and keep them informed of research developments; coordinate activities, such as in-

-25 -

service training, with the NCU; and contribute in training and technical support to the
national programme on a selective basis. These responsibilities, which may extend
beyond specific project duties, should be included in their job descriptions.

The AC(FSR) and his staff in the NCU should monitor the performance of externally
funded FSR projects to ensure their effective integration into the national FSR
programme as well as into the zonal research programme. NCU staff should seek regular
consultations with donor representatives at the project level.

A staff member from the NCU should participate as a full team member in formulation
or review missions of externally funded FSR projects, as occurred recently with the Lake
Zone FSR project.

In order to help donors better understand how their project fits within the overall national programme
and to generate broader support for national efforts, we suggest that the NCU invest more effort in
increasing donors' awareness of FSR in Tanzania and the role of the NCU. The NCU should prepare
a short paper providing an overview of FSR in Tanzania, its strategy for programme development, and
its approach to FSR, which can be circulated to donors interested in funding this type of research in the
region. They should also send working papers reporting achievements, annual progress reports, as well
as any significant publications emerging for FSR research to programme officers of donor agencies
supporting FSR in the region. We encourage the NCU, led by the AC(FSR), to hold an annual or
biannual meeting with donors of FSR projects to raise their awareness of how their projects fit within the
overall national programme. Funds for such meetings are available from the Tanzania/Netherlands
project for strengthening the national development and coordination of FSR.

When the NCU is fully staffed, the AC(FSR) should consider taking a more active role in contacting
prospective donors with the view to generating interest in funding FSR initiatives in the zones. If the
national staff have taken the lead in cultivating support for FSR, they will be able to work as active
partners with donors in formulating projects and tailoring them to the specific needs of Tanzania.

Recommendation 3.9 The NCU under the leadership of the AC(FSR) and the CRT should take a
proactive role in dealing with donor agencies with respect to FSR matters.
Initiatives should include not only providing interested donors with papers
reporting progress and achievements in FSR, but also a briefing paper providing
an overview of FSR in Tanzania, its strategy for programme development, and
its approach to FSR. Periodic meetings with donors of FSR projects should be
convened to improve awareness of overall national programme and how their
respective projects fit. The AC(FSR) should take an active role in contacting
prospective donors to generate interest in funding FSR initiatives in the zones.

-26 -



Farmers can and should be effectively involved in all stages of FSR research: from the first stages of
problem identification and priority setting, through on-farm trial design and experimentation, the
evaluation, and even dissemination, of technologies. Studies have also shown that they can play a
valuable role in on-station as well as on-farm research. Early in the research process, at researcher-
controlled sites (e.g., on central stations or at the many multilocational sites on farm), farmers help to
efficiently eliminate unacceptable options, identify or help modify the more promising ones; and aid
researchers to target technologies to specific ecological or socio-economic niches. Such involvement of
farmers is not only a matter of researcher's good will but is also necessary to ensure that research
products are practical and relevant to farmers. Farmer participation is an important tool for increasing
research efficiency through improved relevance in technology development. It can also expand the
capacity of the research programme at minimal cost to research. However, it should be emphasized that
strengthening farmer participation in research programme design and implementation involves developing
new research methods, techniques, and skills (see Sections 5.4 to 5.6).

Recommendation 4.1 For improved relevance and efficiency, the FSR programme should strive to
increase farmer participation in all phases of the technology development process:
from on-station testing to final verification trials. Adapting and applying
methods for increased farmer participation should be a priority area for attention
in strengthening FSR in Tanzania in the next five years.

To strengthen farmer-researcher linkages, we suggest that the FSR programme should give particular
attention to strengthening several mechanisms:

* On-farm trial designs should aim to meet both farmer and researcher needs. The design of
trials should be interpretable by both partners -- farmers and researchers. Only when the
objectives, assumptions, and form of research are understood by both actors can participation be
fully effective. Similarly, research on-farm should aim to reflect real farmer conditions. Only
with this realism are the results of research predictive of farmers' ability to use technologies
themselves. We therefore believe the FSR programme should place greater emphasis on
involving farmers in the management and implementation of on-farm trials.

* Direct researcher-farmer contact should encouraged. Both researchers as well as their field
officers should aim for direct dialogue with farmers. Direct contact, beginning in the
descriptive/diagnostic stage, is the best way to establish strong mutual understanding. Various
types of meetings can encourage such interactions, for example: informal descriptive/diagnostic
surveys, field days, both on-station and on-farm, where specific technologies or problems are
discussed and assessed; village meetings (for diagnosis, general information exchange,
evaluation); group tours, either exploratory or problem-focussed; or individual farmer-researcher
meetings to address specific research issues. Feedback of research results to farmers (e.g.,
results of informal or formal surveys or of a season's trials) has been weak in Tanzania.
Feedback should be strengthened as a means for building stronger collaboration and confidence
with farmers as well as for verifying the validity of research findings.

-27 -

* A procedure for qualitative assessment needs to be systematized and institutionalized.
Economic assessments themselves are important but insufficient for representing farmer's views
and describing the probability of technology adoption. Farmer-assessment helps to ensure that
research products are effective under future farmer management.

Survey methodology needs to be selected and modified to more fully capture in-depth local
knowledge and perceptions. The effectiveness of socio-economic tools the FSR teams currently
use for working with farmers needs to be re-examined, particularly the balance between informal
and formal survey techniques (see Section 5.6.2). Second, the range of techniques should be
expanded so as to be able to address socio-cultural issues that affect farmer decision-making as
well as farmers' ability to adopt technologies. Where possible, farmer participatory techniques
should be given priority (see Section 5.4.1).

Research should be designed to work with groups as well as individuals. We believe the FSR
programme should give major priority to working with farmer groups as a means for reducing
costs and increasing farmer participation. Studies show many types of groups are effective for
the process of technology development, for example: those self-selected who are interested in
particular technologies [Heinrich and Masikara, 1991]; those which are researcher-selected to
represent specific farmer expertise or interests [Sperling 1991]; and those selected by immediate
geographical proximity to facilitate clustering (see Section 5.6.1). Such farmer groups (usually
of 10-30 persons) have been shown to make research both more dynamic and cost-efficient (see
Section 5.4.2 for methodological appraisal of group work). At a later stage, the FSR teams,
might also consider issues of working with communities, that is, much larger numbers of
farmers. Communities are key in the management of the natural resource base and hence issues
of sustainability often need be addressed with a community focus.

* Researchers should strive to strengthen their links with intermediary organizations which
work with farmers. Given research resource constraints, the FSR teams might experiment with.
supplementing their direct work with farmers through intermediary institutions, such as NGOs
or churches. These latter groups often have strong links with farmers' organizations or
communities. With researcher guidance, these intermediary organizations can help conduct a
range of on-farm trials, surveys, and adoption studies. FSR teams might explore possibilities for
collaboration by first testing single component technologies, such as varieties, through
intermediary institutions. If the collaboration is fruitful in terms of broader contact with farmers,
collection of useful data, and cost reductions for FSR, collaboration in more elaborate
experiments might be considered. Since collaboration always requires an investment of resources
from both partners, the FSR teams should select partner organizations carefully and build up such
linkages incrementally. Wellard [1990] provides a useful overview of the advantages and
disadvantages of working with NGOs and Gilbert [1990] provides a constructive analysis of the
experiences of research and NGO collaboration in The Gambia.

Recommendation 4.2 Farmer assessment should be built into the evaluation of every technology and
should be a prerequisite for releasing a variety or making a recommendation to

- 28 -


Strong commodity and farming systems research are both necessary if research, as a whole, is to benefit
limited resource farmers. Collaboration between the two, based on their mutually supportive roles, must
be developed to achieve this objective. FSR depends on commodity and disciplinary research to provide
technology components to be adapted to location-specific needs as well as to provide specialized advice
in diagnosing and prioritizing farmers' problems and in designing potential solutions. Component
research depends on FSR for feedback on farmers' production conditions and constraints as well as to
adapt technology components and put them together into viable systems options which are both relevant
and feasible for farmers to adopt. FSR teams should ensure that farmers' multiple evaluation criteria
are applied during technology assessments and take a pro-active role in feeding back important
information to commodity research programmes. The nature of the links between FSR and
commodity/disciplinary research in Tanzania varies markedly across the zones, but in all cases these
critical links need to be strengthened.

Recommendation 4.3 Priority attention should be given to strengthening collaboration between FSR and
commodity research in the next five years. Agreement should be reached on the
appropriate division of roles and responsibilities between FSR and commodity
research and appropriate linkage mechanisms should be put in place. These
linkages will be critical to the future success of FSR in Tanzania.

Effective collaboration between commodity researchers and FSR teams depends on defining and agreeing
upon an appropriate division of responsibilities and establishing mechanisms for exchanging the
information and resources required for working together. In some zones, agreement on the division of
roles and responsibilities already exists, while in others there appears to be conflicting views. Where
there is no consensus, the development of understanding and agreement between both parties on the role
of FSR should be addressed on a priority basis. We have proposed a division of labor which we believe
builds on the comparative advantage of each type of research (see Section-3.1.3). Using this-as a basi_
for discussion, we suggest that the ZDRTs, in conjunction with NCU staff, assist the research teams in
each of the zones to review their current working relationships and to come to an agreement on an
acceptable division of labour between FSR and commodity and disciplinary research. Discussions should
include the ZDRT, the Zonal FSR Coordinator, commodity programme scientists, the Research-Extension
Liaison Officer (RELO) (i.e., if posted) and an NCU staff member. Ideally, this should be done prior
to the setting of zonal research priorities and apportioning of responsibilities.

Once agreement on the roles of all partners has been reached, then specific linkage mechanisms need to
be tested and put in place. In considering appropriate mechanisms, it must be remembered that all links
cost time and money. Given the resource constraints for research in Tanzania, it is unlikely that the ideal
configuration of linkage mechanisms can be put in place. Attention should be given to instituting a few
key mechanisms which have proven successful in promoting collaboration. They should be managed
actively, resources should be assigned to support them, and their effectiveness should be reviewed
periodically. Some suggestions for such priority mechanisms, drawn from experiences in other countries,
are made below:

* Joint problem diagnosis. Bringing groups together to jointly define priority problems and
propose solutions to be researched helps to build a strong basis and commitment for on-going
collaboration. These exercises are most effective when they take place in the field rather than
the conference room, and when they are conducted periodically rather than in a 'one off fashion.


Bringing commodity and FSR researchers together in informal diagnostic field surveys has proven
to be a particularly effective and feasible mechanism in national programmes. FSR teams would
typically take the lead in organizing such diagnostic exercises, but commodity researchers may
also wish to enlist FSR teams to work with them in prioritizing problems specific to individual
commodities. To facilitate participation, activities must be planned well in advance and resources
allocated. Also, the objectives must be clear and the exercises well managed so that senior
scientists' time is not wasted. The probability of success is greatly enhanced if such activities
are actively supported by senior research managers, such as the ZDRTs and commodity
programme leaders.

Recommendation 4.4 Every effort should be made to ensure that senior commodity scientists
participate fully in the design, execution, and analysis and interpretation of
informal surveys organized by FSR teams in the zones. Senior commodity
scientists should also be consulted, when appropriate, in the design of and
interpretation of results from formal surveys. The ZDRT should take a lead role
in encouraging such types of participation and collaboration.

* Joint planning, programming, and review meetings. It is important that FSR researchers
attend the annual commodity review meetings relevant to their zones and that commodity
researchers attend the annual FSR review meetings. Funds should be set aside to support such
participation. The new research planning and review processes proposed by the Overseas
Development Administration (ODA) Organization and Management Team in DRT for the zonal
level, have the potential to serve as a key linkage mechanism. Joint planning and review
meetings, if they are kept small, focused, and allow ample time for debate, have proven to be
one of the most effective mechanisms for strengthening linkages in other countries [Merrill-Sands
and McAllister, 1988]. Experience shows, however, that their effectiveness as a linkage
mechanism will be highly dependent upon the initiative and commitment of the ZDRT.

* Collaborative research. Collaboration in research activities, such as surveys or trials, promote
closer linkages because they provide an opportunity for each group to gain a better understanding
of each others' objectives, assumptions, research methods, and working conditions. Some
important considerations when promoting collaborative research are:

Collaborative research may be initiated by either FSR or the commodity research

On some topics, FSR may take the lead role (e.g., on-farm problem identification), and
on other topics, it may take a supportive role only (e.g., technology verification trials).

Designs of surveys or experiments should incorporate the objectives of both parties, and
not simply be dictated by one or the other.

To be able to collaborate successfully, researchers need to be able to plan activities
together well in advance, have the capacity and resources required to carry out their
respective responsibilities, and be able to communicate easily during the research process.
It should be noted that collaborative research activities tend to be quite resource and
management intensive. Ideally, such activities should be jointly funded. Funds within
the Netherlands/Tanzania project for strengthening FSR should also be used to support

- 30-

joint research activities.

* Field days/tours. Joint field visits to see on-farm trials are very effective mechanisms and
usually increase commodity scientists appreciation of the need for on-farm research. Field days,
held both on-farm and on-station, where both FSR and commodity researchers and farmers
attend, provide an opportunity to review and discuss not only current but also potential future
collaborative work.

* Informal consultations. The value of informal consultations among researchers should not be
underestimated as a linkage mechanism. Again, the strengthening of research at the zonal level
should promote more contact and consultations among researchers with common interests.
Researchers should be encouraged to consult one another on research design, analysis, and
interpretation this is the essence of the scientific process.

* Reporting. The preparation of brief, one or two page reports are useful in keeping other
researchers appraised of important findings from trials and surveys. A two-way flow of
information (i.e, from FSR to commodity researchers and vice versa) is very helpful. In addition,
these reports can be used by other agricultural staff (e.g., technology transfer agents). We
suggest that the practice used in Northern Zone of convening seminars to report and discuss the
key findings from surveys should be promoted in all zones.

In summary, good collaboration between FSR and commodity research teams is vital to ensuring the
impact of agricultural research. A necessary precondition for good collaboration is that both parties must
understand the different, but mutually supportive roles of each. In zones this understanding is lacking,
priority emphasis should be placed on its development. For good collaboration to occur, mechanisms
which promote regular communication and interaction throughout the technology development process
are also needed.


The linkage between research and technology transfer agencies is vital because new technology cannot
have an impact unless it is made available to, and adopted by, farmers. Technology transfer agents are
necessary for broad dissemination. Ideally there should be a two-way flow of information between
research and technology transfer agencies. Technology transfer agencies have broad coverage and work
directly with farmers. They may have important insights on farmer interests or constraints, or on the
performance of previously released technologies, which can sharpen the focus and potential impact of the
technology development process.

Additionally, technology transfer agencies are expected to understand new technologies and to assist
farmers in adapting and applying them. Technology transfer agents will be more confident and familiar
with technologies if they have been involved all the way through their development and testing.
Involvement with FSR in testing technologies can therefore lead to greater efficiency in the transfer

The term 'technology transfer' has been used here instead of the more common term 'extension agency'
because 'extension' has come to be associated with traditional public sector extension services and a one-
way top-down transfer of information from research to farmers. In the Tanzania context, it is important

-31 -

to stress both the two-way flow of technical information and a wider range of potential technology
transfer partners such as development agencies, private firms, NGOs, educational institutions, and
community organizations, to assist with testing, transfer and feedback (see Section 4.1). Technology
transfer also includes services of organizations responsible for supplying inputs and services to support
adoption of technologies.

The NALRM indicates that public sector extension is weak in Tanzania as are research-extension
linkages. To date, there has been considerable variation between zones regarding the development of
these links. In many cases, the linkages have been ad hoc and sporadic, at best. The NALRM indicates
that FSR should play a lead role in strengthening this linkage. It also identifies some specific linkages
mechanisms (i.e., the Zonal Advisory Committee (ZAC) and Research-Extension Liaison Officers
(RELOs) to strengthen interaction). We endorse these efforts, but we also caution that links with
technology transfer agents are the responsibility of research as a whole, not solely of FSR. We also
suggest that FSR consider linking with a wider range of partners to multiply its contacts with farmers as
noted above. Given the current status and resource constraints of public sector extension in Tanzania,
it may not be beneficial for research to invest a major amount of resources in developing this link alone.
In all cases, the government extension service should be appraised of the initiation and outcome of
collaborative activities with other partners at the zonal level.

Some potentially useful mechanisms for linking with technology transfer agencies at the zonal level are
reviewed below.

4.3.1 Research-Extension Liaison Officers (RELOs)

The NALRM recommends the appointment of RELOs at the zonal level reporting directly to the ZDRT.
Recent proposals on the reorganization of the DRT have recommended that the RELO be part of the FSR
team. RELOs have the potential to be a very useful linkage mechanism. However, experiences in other
countries (e.g., Botswana and Zambia) suggest that a very specific set of conditions are required for that
potential to be realized [Kean and Singogo, 1990]. Moreover, RELOs cannot be considered as a
substitute for other linkage mechanisms. Rather, their function should be to keep active a range of
linkage mechanisms which promote stronger communication and collaboration between research and
technology transfer agencies.

Some of the prerequisites for a RELO to function effectively are outlined below:

The RELO must be a senior officer, highly respected by both research and the technology
transfer agencies. This status is necessary because he/she will have major responsibility,
but no direct authority. He/she will only be able to function effectively if he/she has the
respect of both research and extension officers and the professional competence to be able
to persuade officers from the different agencies to work together. The RELO should also
have a high level of personal initiative.

Both research and extension should have a say in the selection of the RELO. Too often
secondments become a means for shifting 'deadwood' out of an organization. Similarly,
care should be exercised to avoid appointing junior officers for the sake of short-term
expediency. Such appointments will jeopardize linkages in the long run.

- 32 -

The RELO's job description should be defined jointly by research and extension and
endorsed at the highest levels. The job description should indicate the coordination
responsibilities of the RELO as well as the activities for which they are solely
responsible. Such activities may include: running on-farm tests of promising
technologies, organizing demonstrations, revising crop recommendations, incorporating
a farming systems perspective into extension training, producing newsletters, and
organizing the involvement of extension agents in research activities. The RELO's
performance of his/her assigned responsibilities should be periodically assessed by the
ZDRT and the Regional Agriculture and Livestock Development Officer (RALDO).

The RELO must be technically competent in order to review and draw out technology
recommendations from research results and make them available to technology transfer
agents. The RELO should also be responsible for feeding back information from
extension to zonal research.

The RELO must have excellent writing and communication skills; for the same reasons
indicated above.

Funds should be allocated specifically to support the coordination activities of the RELO.
Preferably these funds should come from both research and extension. The RELO should
also have guaranteed access to transport.

Given the high level of experience and technical competence required for the RELO position, we believe
the RELO should report directly to the ZDRT, but should be expected to work at least 50 percent of
his/her time with the zonal FSR team. We also believe that candidates from both research and extension
be considered for these positions. If the RELO is to work effectively with the FSR teams, it will be
important for him/her to receive in-service training in FSR. Such training should be the responsibility
of the NCU.

Recommendation 4.5 The RELO should be an experienced and respected agricultural officer, drawn
from either extension or research, who has sound technical competence, excellent
communication and writing skills, and strong personal initiative. The RELO
should report to the Zonal Director, but work more than half time with the zonal
FSR team. The appointment of a RELO should be approved by both research
and extension. The RELO should have guaranteed access to transport and
his/her activities should have a specific budget with contributions from both
research and extension.

In zones where RELOs have not been appointed, alternative linkage methods could be attempted in the
interim. For example, the ZDRT together with the Zonal FSR Coordinator, could apportion the RELO
responsibilities among the appropriate staff at the zonal research centers, and use the proposed Zonal
Advisory Committee (ZAC) and other fora for regular interaction with technology transfer agencies.
Funds allocated to RELO activities at the zonal level could still be used for linkage activities. Experience
in other countries has shown that sometimes links between FSR and technology transfer agencies can be
more direct and therefore more effective without a RELO.

- 33 -

4.3.2 Zonal Committees

The NALRM recommends the formation of both the ZAC and a Zonal Technical Committee (ZTC).
Priority should be given to establishing these committees and encouraging technology transfer agencies
to participate. The RELO should be appointed as member, since such committees will be key to his/her
effectiveness, as should the Zonal FSR Coordinator. Joint participation in meetings creates a channel for
regular communications, is relatively low-cost, and provides the basis for development of collaborative

4.3.3 FSR Linkage Activities

In addition to the posting of RELOs, the NALRM gives FSR a lead role in strengthening research-
technology transfer linkages. We support this recommendation, but believe that the cost of such activities
should be carefully considered. An increase in linkage activities is being recommended at a time when
the FSR budget has been reduced by 50 percent. Linkage activities require funds for travel and meetings,
field days, etc. If linkages are to be strengthened it will require either more funding for FSR teams, or
acceptance of a reduction in their research activities.

Recommendation 4.6 Additional funds need to be allocated to support the strengthening of FSR -
technology transfer linkage activities either through a direct supplement to the
FSR budget or channeled through a budget allocation for the RELO.

Some of the activities zonal FSR teams could carry out to strengthen links with technology transfer
agencies are listed below:

Technology transfer agents can be invited to participate in selected FSR activities;
particularly those involving identification of production constraints, but also in field days,
research presentations, and meetings organized with farmers to feedback results of each
year's trials and surveys.

Technology transfer agents can be invited to participate in workshops on FSR philosophy
and methodology. This will give them a common base of understanding for working
with the FSR teams. Such activities are supported under the Tanzania/Netherlands
project for strengthening the NCU and have already proven useful in the Lake and
Eastern Zones.

Where technology transfer agencies have the interest and resources, they can participate
in implementing parts of the on-farm trials programme; particularly testing and
demonstration activities. Collaboration with technology transfer agents in implementing
trials will be facilitated by the recommended clustering of trial sites (see Section 5.6.1.).
Field agents' participation in on-farm trials can only be effective, however, if supported
by the District Agriculture and Livestock Development Officer (DALDO). The RELO,
the DALDO and Zonal FSR Coordinator should work together to develop commitments
and arrange the logistics of collaborative activities.

In those zones where the Training and Visit (T and V) system is established, the Zonal
FSR Coordinator should attend the regular quarterly meetings.

- 34 -

When FSR teams have technological recommendations ready for wider testing or
dissemination, they should organize training workshops for extension agents on the new
technologies in coordination with senior zonal extension managers. Such training has
already been experimented with in the Eastern, Northern, and Lake Zones.

Finally, good links at the national level assist and support linkages at the zonal level. The AC(FSR)
should therefore maintain contacts with the Assistant Commissioner of Extension and, wherever useful
and possible, with relevant NGOs.

Thus, without strong linkages with diverse technology transfer agencies the impact of research may be
marginalized because findings never reach farmers. Direct contact between research and a limited group
of farmers, such as that provided through FSR, is not sufficient to ensure the broad transfer of promising
research results. At the zonal level, the first priority is to establish regular channels of communication
between all the potential partners through the establishment of the proposed ZAC and ZTC. Appointment
of RELOs should proceed with due regard for the importance and difficulty of the position. Where
RELOs are absent, alternative linkage mechanisms could be tested in the interim. Good communication
between the Assistant Commissioners of FSR and Extension at the national level will encourage and assist
zonal linkages. Ideally additional funding should be sought for supporting linkage mechanisms.


Linkages between research and policy/support systems are often vitally important in furthering
agricultural development. A recent study identified inadequacies in the policy/support systems as being
one of the major reasons for the low level of adoption of improved technologies in Africa, which resulted
in low returns to resources invested in agricultural research [Daniels et al, 1990].

Field level data generated by FSR can help planners in their design of appropriate policy/support systems.
FSR teams are uniquely placed to provide such a service. At the same time FSR teams often identify on-
farm production constraints that relate to policy/support systems. In addition some important production
technologies are difficult for farmers to adopt without external support or assistance, especially in more
marginal areas where research is attempting to overcome difficult agro-ecological constraints.

Although convincing arguments can be made for strong linkages between research and those responsible
for designing and implementing the policy/support systems, there are often some problems in ensuring
such linkages. Two important ones are as follows:

Specific linkage mechanisms are often not easy to establish since FSR teams generally
do not have a mandate for providing direct inputs on policy issues, and the proper
channels of communications between the two parties are often not well defined.

Research on policy/support issues could easily consume the limited research resources
of FSR teams, particularly of economists, at the expense of necessary micro-level
economic research directed at technology development and evaluation.

Thus the question of involving FSR teams in policy/support system issues must be approached with
caution. The following points highlight some potentially useful areas for constructive interaction between
FSR teams and planners or those responsible for the policy/support systems:

- 35 -

Research involving specific policy issues or support systems may be necessary to identify
real production constraints. For example, deficiencies in marketing systems may be
restricting production of certain products (e.g., milk, cotton, etc.). Such inadequacies
may be due to the unavailability of the necessary inputs (e.g., fertilizer, inadequate
transportation, etc.). These types of constraints need to be clearly identified and
communicated to those responsible for the policy/support systems.

In more marginal areas, farmer adoption of certain technologies may be extremely
difficult without some explicit help from the policy/support system. Such technologies
often involve 'lumpy' rather than divisible inputs and revolve around breaking constraints
in the farming system rather than avoiding the constraints through exploiting flexibility
Sin the farming system. This implies major changes on the part of farmers often requiring
credit to purchase these new 'lumpy' inputs (e.g., equipment) and training to ensure that
the management skills are provided which are necessary to reap the benefits from such
large changes. Obviously good linkages between research and those responsible for
policy/support systems are important to facilitate adoption of such types of technology.

Sometimes research may identify improved technologies that need to be moved through
several phases before adoption can take place. For example, an improved crop genotype
might be identified that potentially has a major impact on productivity. In such a case,
researchers can act as 'product champions' following the product through various stages
to ensure that seed production, seed distribution, technology transfer and possibly product
marketing systems, are put in place to ensure that farmers can adopt and benefit from the
improved crop genotype.

Because of the location specific, on-farm and farmer-based orientation of FSR, socio-
economic research is likely to develop useful and interesting findings for those
responsible for designing and implementing policy/support systems. Although these
findings may not result in new research issues to be addressed by FSR teams, it can still
be useful for the teams to synthesize their findings in brief reports and distribute them
to key officials.

Given the limited research resources available to FSR teams and the demands placed on those resources,
it is important to evaluate very carefully the type and amount of policy/support system type work that is
undertaken. As a result we suggest the following:

Appropriate channels for implementing the research to policy/support system link need
to be identified and established. The proposed ZAC as described in the NALRM appears
to be an appropriate entry point at the zonal level. At the national level there is
considerable merit in strengthening the recently initiated annual meetings between FSR
economists, NCU staff, and planners. The AC(FSR) may have to take the lead in
contacting agencies.

The main focus of FSR socio-economic research should be on micro-level issues relating
to technology development and evaluation. The amount of time and resources devoted
to policy analysis should be limited in scope over the next five years and confined to very
specific, critical issues and to those FSR teams where more experienced agricultural
economists are placed. Such work should only be undertaken with the NCU's

- 36 -

collaboration and consent. Such studies should only be undertaken if they will have a
useful input into the implementation of FSR team activities that have already been

During the next five years the primary roles of the FSR teams should be limited to
synthesizing findings relevant to planners and policy-makers evolving from their field
level work and to acting as product champions for promising technologies.

Recommendation 4.7 A modest approach should be taken to developing links between FSR and
planners and those responsible for policy/support systems. Given limited
resources available, FSR teams should not be conducting separate policy studies.
Rather a linkage should be formed to allow FSR teams to communicate policy
issues which derive from their work directly to planners and those responsible
for policy/support systems. Planners should also be encouraged to communicate
to FSR teams the types of micro-level data they need for policy planning

- 37 -



There is now general recognition in Tanzania that the priority setting and research planning processes
need to be strengthened. The FSR teams should participate actively at the zonal level and should carry
out more systematic planning of their research programmes. Methodology development in areas where
the farming systems holistic approach has a comparative advantage needs to be strengthened.

As stated in the NALRM, and recognized by all agricultural researchers in Tanzania, funding for research
is well below the minimum levels required to conduct research activities. Further, the nature of donor
funding and task specific projects has created a 'stop and go' pattern of research. It is necessary,
therefore, to set priorities that make the most economic use of resources for both the short term and long
term, and to concentrate on programmes most likely to pay off. Priority setting provides a framework
in which to structure, or place, the various donor funding efforts while maintaining the integrity of
national research priorities. Research priorities have been set in the NALRM [NALRM, 1992]. FSR
and agricultural economics fall into the first priority category and are considered essential services for
development of crop and livestock research programmes in Tanzania and for implementation of the

A distinction is made between national and zonal research priorities. FSR should play a role in
identifying and refining zonal research priorities [NALRM, 1992: 23]. FSR teams themselves should pay
particular attention to selecting relevant regions within zones (as they cannot do quality work throughout
an entire zone) and within these regions, to select sub-regions. The regions should reflect national and
zonal research priorities and the research sites should be carefully selected to ensure extrapolation to other
areas with similar farming systems. Mechanisms for doing this are set out in the next section (Section

Defining the client in the FSR process can cause a conundrum, however, as national, zonal, research
station, researcher, household food security, and even sustainability and conservation priorities may not
be in concurrence; demands on the FSR team can become overwhelming. The FSR team must be able
to set logical criteria for ranking the importance of thematic problems, again in the context of possibly
conflicting client needs. Defining 'who' is the client and making these multiple clients participants in
the FSR process is necessary; FSR researchers should apportion their activities amongst these clients,
making both short and long term plans based on ranked, integrative themes. Importantly, the FSR teams
must become proactive in defining these themes. The tendency in some zones to become a survey service
and in other zones to become a variety testing service is a reactive mode. Continuing to operate in such
a mode means that much of the power and unique contribution of FSR is likely to be lost.

FSR researchers work directly with farmers and are the key linkage in feedback of farmers' problems
to the research station, extension/development agency staff and policy makers. A balanced portfolio of
farmer priorities, zonal priorities, and national priorities keeps FSR relevant and efficient. Regular re-
evaluation of FSR's short and long term plans will be necessary to move the research process forward.

In other words, we suggest in research prioritization and planning (see Section 5.2), it would be useful
to think in terms of two flows. One is a more strategic and long-term approach (e.g., five years) for
which the main themes of research are determined. In this determination, representatives of all stake

- 38 -

holders (e.g., commodity and FSR researchers, technology transfer staff, etc.) would be involved to elicit
their commitment. Out of this would be developed the major research themes. The second flow would
be a short term or annual adjustment of research priorities, within the major themes, based largely on
results from the previous years' results, greater farmer involvement, and funding.

Recommendation 5.1 Because of their intimate knowledge of farmers' problems, FSR teams should
participate actively in research priority setting and planning at the zonal level.
They can also help through characterizing the major zonal farming systems, and
in selecting specific areas within the zone to work. As a result of this priority
setting exercise FSR teams should apportion their activities to arrive at a
balanced portfolio of researchable themes.


Given its pivotal role in identifying zonal research priorities, the FSR team needs to develop standardized,
routinized methods of research planning that include the concerns of the potential actors, or clients, in
the research programme. Research planning should therefore take into consideration the following:

* Joint planning. Zonal commodity specialists, in particular, should be important partners in the
research planning process at the zonal level. National priority commodity and livestock scientists
are stationed amongst various zones which have widely varying agroecosystems. Clearly, taking
these factors into account is important for research planning. The ZDRT is, of course, vital, in
encouraging and directing the planning process. It may also be useful to invite comment from
technology transfer agencies, where possible.

* Orientation towards the client. Knowledge and understanding of the needs of the client, and
determination of 'who' is.the client (i.e., -usually the resource poor farmer for FSR teams), in
large part defines the type of research that will be planned.

* Rational priority setting. As noted, FSR's role in defining researchable themes includes
national, zonal, and household food security issues. However, given resource constraints, special
efforts must be made to maintain a balanced portfolio of research activities. The FSR team
should not attempt to cover the entire zone nor solve every problem.

* Linkages with commodity and technology transfer agents. In planning on-farm research, the
strengths and weaknesses of these linkages must be considered (see Sections 4.2 and 4.3). Input
from commodity research teams and technology transfer agents is important.

We suggest that a specific planning process be set up, in consultation with the ZDRT and NCU, to
identify and rank research themes and to discuss the overall work agenda. We suggest mechanisms for
institutionalizing FSR research planning and implementation include the following components:

* Regular meetings at the zonal level. Regular FSR logistical staff meetings (i.e., possibly
weekly), monthly programmatic FSR team meetings, and the proposed ZTC and ZAC meetings
(see Section 4.3.2) should be instituted. These are in addition to the annual research planning
meetings. Regular meetings diffuse many problems before they start and are an important
mechanism for team building. The monthly meetings should be devoted more to informal

- 39 -

reporting of what is happening in the field. Joint meetings with commodity researchers and
technology transfer agents should be used for planning collaborative research. Depending on the
zone, these meetings may have to occur more often than quarterly (e.g., Eastern Zone has three
potential planting seasons while Central Zone has only one). The FSR team should continually
review and monitor their research programmes to remain up-to-date and to avoid unnecessary
confusion and duplication of efforts.

Writing and publication of results. We suggest that the FSR team make serious efforts to come
to closure on some of their experiments and surveys and write up the results. Several years'
worth of annual reports indicate the same trial being repeated yet results are not apparent. By
the same token, much useful and interesting information has been obtained through the numerous
surveys which would not only be useful for research priority setting by commodity researchers
but would advantageously publicize the efforts of the FSR teams. We suggest that job
descriptions for each FSR team member indicate time be devoted to writing up their research
results. Research work that is not documented is not complete. We suggest, NCU, whenever
possible with assistance from the DRT publications unit, should assist the zones in publicizing
surveys and trial results, through producing and distributing a publication series.

Recommendations. Within DRT a system needs putting in place that will improve the timely
approval, dissemination and, if necessary, subsequent modification of recommendations.

Recommendation 5.2 A system of research planning should be implemented at the zonal level,
involving all parties (i.e., FSR and commodity researchers, technology transfer
agents, etc.), that will ensure joint planning of research, good information
sharing, communication and accountability during implementation, and analysis,
documentation and dissemination of the results. FSR teams should give priority
to synthesizing and writing up results from surveys and multi-season trial results
as key inputs into future zonal planning.


5.3.1 System Interactions

Considerable diversity exists among the FSR teams in the types of research undertaken. Since its
inception, FSR in Tanzania has focused on describing and conducting research on farming systems but
has basically defined the farming system as the cropping system, and even more specifically, has
conducted adaptive research primarily on variety evaluations and various input levels (e.g., fertilizer,
manure, weeding labour, etc.). Several exceptions include some research on intercropping in several
zones (e.g., maize/beans, cotton/cowpeas, etc.), village-level livestock research at Mpwapwa involving
both livestock and crop researchers, and integrated pest management programmes (e.g., cotton scouting)
at Ilonga. Nevertheless, on the whole, the bulk of the research is, as stated, on variety and input
evaluations. The systems perspective of FSR indicates that FSR researchers have the ability, and the
comparative advantage, to conduct research on system interactions, and to be at the forefront of research
concerned with sustainability issues.

Crop-livestock interactions, such an integral component of most smallholder agricultural production
practices, have been essentially ignored by FSR, except in the Central Zone at Mpwapwa. In large part,


this has been due to the separation of agriculture and livestock research until 1989. In areas where
disease is not endemic or uncontrolled, livestock management practices, especially poor nutrition, are the
major constraint to improved productivity. The knowledge gap of agronomists and livestock researchers
as to what happens between the time when the food crop is harvested and when the livestock graze the
stubble in the field is not only a lacuna but one that few research programmes have addressed. Research
on crop residues and agricultural byproducts at the farm level is extremely limited, not only in Tanzania
but throughout Africa. However, this is targeted for research action in the NALRM. Research on animal
feed resources must be continually farmer and user oriented. For example, although the NALRM also
suggests promotion of pasture and forage crops, the land tenure situation in Tanzania likely predicates
against use of these crops to any great degree. High levels of research resources and researchers' time
has been spent on research stations testing forage varieties that in the foreseeable future do not fit into
the socioeconomic structure of limited resource agriculture with the possible exception of intensive peri-
urban dairying. This may also include the introduction of multipurpose trees through agroforestry
projects in areas where either there are few animals or adequate natural rangelands. FSR, in addressing
both livestock and crop issues can, if implemented properly, incorporate the 'user perspective' and ensure
an appropriate approach is used to address livestock and crop/livestock interaction issues.

5.3.2 Agroforestry

Agroforestry is another research 'gap' only recently being addressed in Tanzania. Because agroforestry
research has been technology and donor driven, it is often proposed as a solution to a problem that has
not been diagnosed. It is essential that agroforestry take an FSR perspective in Tanzania to maximize
its potential benefits most fully in meeting farmers' needs. To date, FSR teams have not been involved
in any of these projects despite the advantages FSR would bring to agroforestry research, particularly
in the initial diagnosis stage. Improvement of soil fertility, soil conservation, provision of fuel wood and
animal feed may all be solutions to problems that are not priorities for farmers. We suggest FSR teams
should establish linkages with agroforestry programmes through, for eamr'ple, assistance in diagnostic
survey and testing technologies on-farm, where appropriate.

5.3.3 Social Dimensions

Social issues at the small farmer level have been shown, throughout Africa, to have an important bearing
on technology use and adoption yet, while agricultural economics has been prioritized in Tanzania, social
science per se has not. All agricultural research has social implications. Research strategies must include
an awareness of farmers' social situations. Land tenure, farmer-herder conflicts, market accessibility,
processing, access and control of resources, and division of labour, are all important in the adaptive
research process. Social science can also bring the perspective of integrating farmers' own knowledge
to help solve research problems. Ignorance of this indigenous knowledge represents a missed opportunity.
The use of farmers' knowledge appears to be limited; many researchers appear to be either supplying a
technology (e.g., a new variety, a spraying programme) or testing a research-station developed
hypothesis. Because social issues have not been addressed, a poor understanding of client groups has
resulted in inadequately defined and targeted research. This needs to be rectified (see Section 5.5 and
Appendix D).

Social factors also include gender issues. For example, who has access to and use of land, who processes
the crop, who does what agricultural operations, how are decisions made in the household, etc. These

-41 -

are all important questions to ask and answer when designing technology with a user perspective (Section

5.3.4 Sustainabiity and Natural Resource Management

Sustainability is an underlying theme of many of these research 'gaps' in that, if they are ignored, the
research efforts are not sustainable. FSR teams should consider the sustainability of their research
activities but concentrate on those areas that have been set as zonal and regional priorities, keeping in
mind some of these key 'gaps', (i.e., crop-livestock interactions and social science research) in which
FSR has comparative advantage. We do not believe that the FSR teams, given their current capacity, will
be able to take a pioneering role in developing research on resource management and sustainability in the
next five years. We believe that the capacity to FSR teams to become involved in sustainability issues
should be limited for at least the next five years.

Recommendation 5.3 Current capacity of FSR teams predicates against initiating research which does
not have a pay-off in a reasonably short time period (e.g., sustainability issues
which are not associated with short-run gains in productivity). However, two
areas where FSR has a comparative advantage, and should give priority in the
medium term are: crop-livestock interactions and social science research,
including gender analysis.


5.4.1 Strengthening Farmer Participation

The farmer is the ultimate user of the products of agricultural research. Whether he/she adopts a
technology largely determines whether research will have an impact on-farm. In evaluating a technology,
the farmer generally considers and synthesizes two broad areas of information: whether the technology
solves the particular problem at hand; and whether the technology can be effectively integrated into the
farmer's ongoing activities with the resources he/she has available. Compatibility with the farmers'
systems, both in terms of agro-ecological and socio-economic conditions, determines whether technical
alternatives are 'adoptable'. A technical alternative emerging from the research system becomes a
'solution' only when it can be utilized and is adopted by farmers themselves. It is thus farmers who
determine the ultimate relevance of a technology.

However, farmers are not only clients of research, but also potential contributors to research. They
possess skills, unique from researchers, which can help increase the likelihood that research will identify
a successful solution. This complementarity of farmer and researcher skills and knowledge is suggested
in Table 5. 1. There is significant evidence to show that farmers, when integrated as partners, can help
to increase the effectiveness and efficiency of the formal research. More relevant technologies are
identified and may be identified more quickly [Ashby, 1987; Sperling, Loevinsohn and Ntambovura,
1992]. Potential technological failures are eliminated earlier in the research process [Sperling 1992].
Also, a wider range of technical options emerges from the collaboration [Heinrich and Masikara, 1991].

The NALRM has recognized the important role of farmers by recommending a more participatory
approach to their involvement in research [NALRM, 1992: 24]. FSR teams should have a comparative

-42 -

advantage in developing these roles of the farmer as first, the user of research products and, secondly,
an effective research partner. In the first instance, the teams have the unique skills to define who
particular users may be, to help target clientele.


Practical, local ideas New, exotic ideas

Knowledge of specific (agricultural history, social, Knowledge of fundamental (biological) mechanisms
biological diversity)

Knowledge of his/her own possibilities, constraints Knowledge of biological, macro-processes

Excellent observational capability Excellent analyzing techniques


Experimenters on a regular basis
Sources of ideas for potential solutions
Source: Modified from Scheidegger [no datel.

Recommendation 5.4

FSR teams should play a leading role in bringing the farmer perspective into the
research system. Also, where appropriate, teams should, in actively encouraging
collaboration with the limited-resource farmers, pay particular attention to
collaborating directly with women, the major agricultural workers on Tanzanian

In terms of the second issue, partnerships, we would like to emphasize that there are many ways of
-involving farmers in research- not all of which give the farmer a decision-making role. Within the
Tanzanian system, we identified three general ways in which research system currently works with
farmers. Scientists may simply contract farmers to provide land or services. They may consult farmers
about their problems and then semi-independently aim to develop solutions. This, in fact, seems to be
the dominant form of collaboration within FSR sites. There may also be rare cases, where FSR
researchers work with farmers in developing collaborative relationships as partners: that is, where farmers
have a strong say in setting the problem agenda as well as a leading role in actual trial design.

As a general rule, we again would emphasize that the more farmers have a decision-making role in
research, the more relevant the problem addressed and the more predictive the experimental results. FSR
teams have a mentor role within the research system in exploring/testing the benefits of more
collaborative ways of working with farmers. Part of FSR's challenge is to modify attitudes or stereotypes
about farmer skills and knowledge, and a large part is to develop the methodological capacity itself, as
a national programme, to do more participatory research (see Appendix D).

We suggest two broad areas where FSR researchers need more training and access to information so as
to effectively collaborate with farmers: basic social science techniques and those more specific to
participatory methods of research.

In terms of basic social science skills, the following need to be given priority:

-43 -

Communication skills for working with farmers.
Collection, recording, and analysis of qualitative data.
Collection, recording, and analysis of quantitative, non-conventional data.

More details on these needs are given in Appendix D.

Recommendation 5.5 The FSR team members should develop communication skills for working with
farmers, and have the ability to collect, record and analyses qualitative and non-
conventional quantitative data.

In terms of participatory research per se, we would give emphasis to two major themes: overall
awareness of the different types of farmer participation (e.g., where they can be integrated and when);
and second, specific methods for working with groups.

5.4.2 Working with Farmer Groups

Working with farmer groups requires some specific inter-personal skills, and flexibility in trial design
and analysis procedures (i.e., when the group format is used for conducting or trials programme) on the
part of researchers. However, these hurdles are not difficult to overcome, and farmers themselves can
be very helpful in developing operating systems. Once initiated, farmer groups have considerable
advantages for on-farm research programmes.

With farmer groups, researchers have the possibility of better addressing needs of client variability as well
as helping to develop a longer-term farmer research capacity. Groups also help to increase the efficiency
in limited-resource research systems. Table 5.2 summarizes some of the key advantages and
disadvantages of working with groups. Several FSR teams have already used groups to help diagnose
and prioritize problems; and in some cases groups have been used effectively to test, evaluate and
disseminate a wide range of technology options. We would like to encourage this cost-efficient trend.


Being able to differentiate among users is important at two levels: who will use the new technologies and
who will use the output of the improved production. Users can be differentiated along many axes, for
instance, by age, gender, wealth, skills, ethnicity, orientation of production (e.g., market versus
household consumption). FSR research is improved through the increased information that comes from
having a user perspective greater efficiency, better targeting, screening, greater likelihood of adoption,
etc. There have been a range of tools and methods developed to help scientists define user populations
(e.g., wealth ranking and participatory rural appraisal (PRA), etc.).

Several methods may help FSR teams to initially define user groups: a comprehensive literature review
(see Appendix D); specific ranking methods where the community itself defines groups, for example,
Grandin [1988], and an informal survey designed to explore varying socio-economic and agronomic
variables. We stress that differentiating users is a iterative process: initial categories always need to be
reviewed; users groups may be re-formed according to the technology being considered; and 'users'
change (i.e., their needs/wants and means evolve). Further details are given in Appendix D.



Group interaction stimulates discussion and particularly
highlights areas of conflict to be pursued in targeted
groups or individually

Group interaction can help build up farmers' interest
and commitment to collaborative research

Group interaction is especially useful for exploratory

Ratio of farmers to researchers changes the normal
dynamics of interaction giving farmers more clout to
lobby for specific needs (e.g., influence research agenda)

Ratio of staff time to farmer contact can be more


Logistically, farmer groups are probably the most efficient
format for handling farmer-managed trials

Working with farmer groups, and hence larger numbers of
farmers, enables many options to be tested simultaneously,
and allows the on-farm research programme to be more
responsive to the testing needs of station-based researchers

Group discussion is useful for technology evaluation and
can provide immediate feedback to station researchers

Groups can be used to increase interactions with types
of farmers under-represented in design and trial

Group discussion can be useful for efficiently collecting
general information on issues such as regional history
and indigenous technical knowledge where group
interactions may help to sharpen recall and/or insights

Source: Modified from Ashby [1990].

Participatory Rapid Appraisal (PRA) has been a method suggested for distinguishing among users in the
Tanzanian FSR context. We feel that this tool is relevant for sketching only the very broad categories.
With the PRA method, communities tend to unite themselves for action (i.e., rather than to show
differences), and the less powerful factions tend to be under-represented. PRA was developed principally
to work at the village and the prioritization of village needs (e.g., should the village concentrate on
terrace building or dam construction and who is most interested in doing what). Nevertheless, some of
the tools used in PRA may be useful, in learning to differentiate users, evaluating technologies, etc.
These could include mapping, transects, agroecosystems analysis, seasonal diagramming, matrix ranking
and scoring, etc. [Conway, 1985; Lightfoot et al., 1989 and 1990].

-45 -

Groups can be dominated or inhibited to produce a
false consensus or biased comments because of peer

Members will often withhold opinions on sensitive
subjects unlikely to be discussed openly in a group

Group activity must be culturally acceptable

Groups are sometimes less reliable for quantifying farmer
opinions because group members influence each other

Identifying or forming groups that represent user
populations or fit research purposes may be logistically
difficult or time-consuming when respondents are
geographically dispersed


One particularly effective method of user differentiation is gender analysis [Feldstein and Poats, 19911.
Gender analysis is a socioeconomic method used to analyze variables which are important to agricultural
technology development. Gender analysis starts with asking a series of questions about 'who' does what.
Several tools are used to answer these questions, the first being construction of an activities analysis or
agricultural calendar which contains all production activities, not just the major crops, including domestic
production, gathering activities, non-farm and off-farm activities, home processing, and household
maintenance activities such as house repair, fetching water and gathering fuel.

Gender analysis also examines the resources for agricultural production in terms of access and control.
Resources include land labour, capital, water, cash, seeds, fertilizer, credit, and implements. Access can
be gender specific or mixed. Control of a resource implies decision making about its use. By the same
token, the analysis of access and control of benefits disaggregated by gender can provide useful
information on whose capital, labor, and resources are being used and who will benefit from the end

Gender analysis benefits FSR and on-farm research when it is used to plan solutions to test, to identify
desirable characteristics of new technologies and evaluation criteria, to identify research areas of interest
to commodity researchers, to conduct ex ante analysis of possible solutions, and to identify the
appropriate user to be involved in on-farm experimentation. User differentiation has not been a priority
in Tanzanian FSR and should receive more attention in the next five years. Better targeting and farmer
participation leads to better design and testing of new technologies.

Recommendation 5.6 FSR teams should become aware of, and increasingly use newer methods and
tools for differentiating users. Given the importance of women in the farming
systems of Tanzania, gender analysis should become a regularly used
methodological tool.


5.6.1 Clustering and Extrapolation

The efficient use of resources (time, money, transport, personnel) is an important factor in many research
systems where these resources are limited. Tanzanian FSR teams are no exception. One useful way to
increase the efficiency of FSR and on-farm research is through 'clustering' of experimental units.

Clustering implies selecting experimental units to work with that are physically close to each other.
Whether the 'experimental unit' is a farm, a soil type, a household, or a village, clustering would imply
that they were selected, in part, because they were relatively close to each other. Clustering might also,
in the broadest sense, involve working with groups of farmers living in a community, rather than with
individual farmers across a region.

Clustering is a valid method for increasing on-farm research efficiency. The increased efficiency comes
from the facts that clustering:

Reduces travel costs.
Reduces travel time between units;
Allows each unit to be visited more often and therefore allows for the development of

-46 -

generally more accurate information;
Allows more time to be spent at individual units as opposed to time spent travelling
between units.

The primary concern, when using clustering, is that the experimental units should still be representative
of the population to which the results of the study will be applied (e.g., selected farms should represent
the population of farms to which the results of the study will be applied, the recommendation domain).
This requires that the relevant variables be assessed in the units selected, and that they are compared
against the same variables in the target population. However, this exercise is necessary whether
experimental units are clustered or not.

Given the resource constraints faced by FSR teams in Tanzania, clustering would appear to be an
important technique, and in fact, some FSR teams are already applying it.

Given the limited research resources, the returns from research need to be maximized. This means that
the clustering exercise must be undertaken in such a way that the research results can be extrapolated to
other areas in the same zone. Thus careful definition of recommendation domains is required to improve
the probability of successful extrapolation of the results. At the same time, the research prioritization
process needs to take into account factors such as the severity of the identified problem, the probability
of a successful pay-off to the research, the potential benefit to individual farmers, the total number of
farmers that will potentially benefit, etc.

Recommendation 5.7 To increase the return from limited research resources, greater attention should
be given to clustering experimental units and to site selection to permit broader
extrapolation of results.

5.6.2 Use of Formal and Informal Surveys

While the specific definition of formal versus informal surveys varies slightly within regions of Africa
[CIMMYT, 1985; Worman, Norman and Ware-Snyder, 1990], Table 5.3 highlights the two types with
respect to a number of research elements [Franzel and Crawford, 1987; Worman, Norman and Ware-
Snyder, 1990].

In the past, the informal survey may have been viewed as a 'pre-survey' whose primary function was to
refine the design of the formal survey [Franzel and Crawford, 1987]. We see this formal survey role of
'general verification' as little justified in terms of high costs with only marginal information gain. With
reference to the one large formal survey examined, researchers involved suggested that they had more
solidly confirmed findings of the informal phase but had learned nothing fundamentally new. Findings
of a methodological survey comparison in Kenya support this assessment [Franzel and Crawford, 1987].
We recognize the FSR researchers' dilemma: formal survey results are likely to be more convincing to
quantitatively-oriented clients -- one of the key clients here being commodity researchers. One way of
overcoming this bias is to encourage commodity researchers to participate actively in the survey itself:
in the design, execution, and even interpretation of the data analysis. In this way, commodity researchers
have a firm commitment to producing what they can consider a 'quality' product.

-47 -


Theme format Highly structured Use of guide questions or unstructured

Sampling techniques Formal random sampling techniques Informal random or purposive sampling
to ensure representative population purposive sampling procedures:
profile representativeness not assured

Question type Closed Open

Discipline interaction Less likely More likely

Interviewers Enumerators Researchers

Data collection process Relatively static Dynamic and reiterative

Potential quality of
Attitudinal Poorer Better
Qualitative Poorer Better
Quantitative Better Poorer

Total time allocation More Less

Certain strengths of the informal survey are particularly important in a developing FSR programme such
as that in Tanzania which emphasizes farmer participation in the research process as well as a user-
oriented perspective. The survey 'turnaround time' is rapid, and the more open format invites in-depth
discussion of farmer's values, knowledge and opinions. Done well, such conversations can not only
capture some of the socio-economic and technical complexities of farmer concerns, but also put the
farmer in the valued, respectful, collegiate relationship he/she deserves. The informal format further
encourages both FSR and commodity programme researchers themselves to go on-farm, to interact
directly with farmers, and to do so with an interdisciplinary team [Franzel and Crawford, 1987].

Several concerns frame the discussion of whether informal surveys are a sufficient tool for guiding the
research process. What is the quality of information obtained through informal surveys and to what
degree can survey results be extrapolated to a wider population? We feel that the usefulness of an
informal survey is tightly linked to the skills of the interviewer: his/her ability to probe farmer knowledge
and to guide the farmer towards identifying what prove to be critical issues. An informal survey has the
potential to produce comprehensive, cause-effect information but cannot inherently guarantee a quality
product. Further, the informal survey can adequately characterize the diversity of client populations, if
well conceived and tested. The FSR surveys reviewed by us varied greatly in quality; we make a
suggestion that FSR staff be further trained in basic social science research skills (see Section 5.4.1 and
Appendix D). Informal surveys should, we believe, serve as a prime research method of the FSR team.

When would a formal survey be appropriate? We suggest that formal surveys be conducted only to
address highly focused, specialized subject topics, where quantification may be particularly important,
for instance, impact studies. Formal surveys need be preceded by well-conducted informal surveys with
researchers aiming to get the various 'stakeholders' involved (e.g., commodity programmes, seed
companies, NGOs). Such stakeholder involvement may help to defray the costs but primarily assures that

- 48 -

the data are collected and the results presented in a form users find relevant and interpretable.

Formal surveys should be conducted with the understanding that researchers will aim for a quick
turnaround time, working for efficiency in data collection, analysis and publishing.

Recommendation 5.8 Informal surveys should serve as the primary survey method of the FSR team
with attention focused on improving their quality. Formal surveys should be
conducted only to address highly specialized subject topics, where quantification
is particularly important; they are more costly and are not necessary as a general
verification tool.

5.6.3 Trial Management Differentiation

Trials can be differentiated on the basis of who manages and who implements them (i.e., researcher
(technician) or farmer). A somewhat simplistic classification, that is often used and is particularly useful
for those not familiar with the principles of FSR experimentation, is to classify trials into three major
management types as follows:

Researcher managed.
Joint management between researcher and farmer.
Farmer managed.

Selection of how trials will be managed is related to research objectives, experimental design, types of
data to be collected, methods of analysis, and evaluation criteria.

Researcher managed trials on farmers' fields are similar to those conducted on the experiment stations.
Therefore, the level -of-testing achieved meets similar standards demanded, by experiment station-based
researchers. However, farmer managed trials are the most satisfactory for the farmer and provide the
most practical test of the technology. Cause-effect relationships and 'hard' (quantitative) objective-type
data are more easily obtained from researcher managed work, whereas farmer assessments and inputs into
the research process are more readily obtained from joint or farmer managed work undertaken on
farmers' fields. However, data collected under such formats, particularly at the farmer managed level,
are likely to be perceived by many commodity researchers as 'softer' (more qualitative) and more
subjective in nature.

Superimposed trials, a type of trial which has joint management, have considerable potential. The trials
generally involve superimposing a degree of research management on a plot or animal being managed
and worked on by the farmer. These trials tend to be single-factor experiments, such as looking at the
response of fertilizer superimposed directly on a farmer's own plot where he/she is providing both labor
and management of all other factors. We have seen examples of such types of trials in at least two zones
(i.e., Lake and Eastern). Superimposed trials can either be pre-planned or unplanned in the sense of
responding to a problem that has arisen and seeking a satisfactory solution.

All three types of trial management are vitally important in contributing to a well-rounded FSR
programme. Once researchers understand the purposes of the different types of trial management, it is
easier to recognize the complementarities that exist between them and, therefore, easier to adopt
appropriate criteria in evaluating their worth.

-49 -

Mainstream FSR has often, in spite of the iterative principle which underlies it, been implemented in
somewhat of a linear mode. For example, often technology design work often begins at the researcher
managed level; then when a technology is thought to be appropriate, it is passed on for joint testing with
the farmer, and then followed with complete farmer management.

We are uneasy about this linear connotation. Because of this and other convictions and considerations
such as the research capacity of farmers, the comparative advantage of FSR, and the need to improve the
efficiency of the technology development process, we suggest the following strategy for Tanzania:

Farmers should be involved much earlier in the research process than has often been the
case to date (e.g., in assessing varieties in on-station trials, in helping design and evaluate
treatments in researcher managed trials, etc.).

FSR teams have a comparative advantage in trials that maximize farmer participation and
assessment. Therefore they should emphasize work having a large farmer management

That methods that have been developed in recent years should be introduced to enable
the needs of more than one client (i.e., researchers and farmers) to be better satisfied
through farmer managed trials. Farmers' needs are likely to be satisfied as a result of
favourable assessment on their part in such trials. At the same time, researchers' needs
can now be better met through:

Using methods that can express farmers assessment in a more quantitative
manner which will more likely appeal to technically oriented scientists.

Utilizing statistical techniques such as regression analysis and modified stability

Thus increasingly the objective of ensuring farmer precision in the sense of ensuring that
the technology fits his/her needs does not have to be accomplished at the expense of some
degree of statistical precision.

Research resource efficient methods should be utilized in implementing on-farm trials.
Clustering work (Section 5.6.1) and using farmer groups (Section 5.4.2), especially in
farmer managed trials, are good examples of such approaches.

In implementing on-farm trials there should be clearly specification of the objectives,
which in turn will indicate what will be the most appropriate trial management type.

Use of superimposed trials should be encouraged not only because they provide a flexible
approach in responding to unexpected experimental opportunities or issues, but also
because the non-experimental variables reflect farmers' actual situation.

- 50 -

5.6.4 Technology Evaluation

Technology evaluation is often poorly understood and undertaken in the FSR teams. Certain aspects may
be routinely examined, such as the technical effectiveness of the innovation. Other features have been
less systematically pursued: particularly farmers' qualitative assessments and well-informed economic
analyses which highlight current and future financial pay-offs. While technology evaluation is addressed
at several points in this report (see Sections 4.1, 6.3 and Appendix D), we would like here to stress some
general principles by which FSR teams might better plan for their evaluations:

Technology should be evaluated by future users of the innovation. If a technology
is targeted for a range of farmers, for example, poor as well as wealthy, both should be
involved in its assessment. The two groups may have differing concerns which would
result in re-design. Similarly, within the household, varying technology use and expertise
should be recognized. Women usually harbour local knowledge of different cultivars and
manage seed selection, post harvest handling and storage. Therefore, it is they who most
often should be involved in on-farm varietal evaluations.

Technology evaluation should take place at different stages of the process of
technology development. Farmers can evaluate technologies at different stages and for
different purposes. At the diagnosis phase, technology screening can be used to elicit
farmers' criteria and understand their decision-making. During the planning and design
steps, farmers may help sort through prototypes, identifying the most promising to
promote for future testing. And, of course, during stages of on-farm experimentation and
validation, farmer assessments are critical for ensuring technology relevance and

Technology evaluation should be effected at different time intervals, once the
innovation has been released on farm (see also Section 6.3 and Appendix D1.2). FSR
teams should consider programming various types of follow-up studies to monitor:

Initial acceptance (e.g., perhaps at the end of one or two seasons).
Adoption (e.g., at the end of a three to four year period).
Impact (e.g., perhaps within five to ten years after the initial technology release).

Timing of studies will partly depend on the nature of the technology (i.e., varieties may
be accepted faster than alley cropping) and the manner in which it is being diffused (e.g.,
are the necessary inputs readily available?).

Researchers should consider the special roles of both ex ante and ex post analyses.
Technology evaluation may not always have clear-cut outcomes of
'promising/unpromising' or 'acceptable/unacceptable'. What is good for individuals may
not always be good for society. Many technically and economically successful
technologies favour some individuals at expense of others; for example, highly productive
cash crops (male-controlled resources) may push out lower-yielding consumption crops
(often women's domain). Income differences may be exacerbated both within and
between households. Similarly, increasing production in the short-run may result in
ecological degradation in the long-run (i.e., jeopardizing sustainability).

-51 -

Hence, technology evaluation, if it is to be useful, extends far beyond the criterion often
used (i.e., assessment of technical feasibility) to include assessments of the economic
viability, social acceptability and ecological sustainability. Further, these assessments
need to be evaluated both from the viewpoint of the individual farmer and of the society,
and through different periods of time.

Some standard evaluative criteria have been developed and are usually used particularly
with respect to technical feasibility and economic assessment. We note even here.
however, that often insufficient attention is paid to assessing the returns to the most
limiting factor (especially labour allocated during the peak labour demand period) and to
assessing the dependability of the technology.

Methods are also being developed to synthesize farmer qualitative assessments [Ashby,

It is in the realm of ecological assessments (ex ante or ex post) that knowledge of simple
analytical methods remains particularly incomplete. Obviously responsibility for
developing such measures goes beyond the mandate of FSR teams.

In conclusion, we believe it is important that a more systematic and holistic approach to technology
evaluation needs to be taken than has often been the case to date. Finally, these ex ante evaluative
criteria discussed in this section need to be supplemented by adoption type studies. Such studies (see
Chapter 6) help in providing information on impact and on whether and what adjustments need to be
made as the result of the adoption process.




There is increasing pressure on maximizing the return from limited research resources and for justifying
the budgets that are allocated. Thus, there is a need to adopt cost efficient methods for undertaking FSR
and also to document and publicize the value of FSR, not only to the various clients in the agricultural
development process but also, to those responsible for providing the resources required for its
implementation. In the Tanzanian context these originate from both domestic and foreign sources. FSR
activities have two types of impacts. The two, which are discussed in the following two sections, are as

They can and should have a positive impact on improving the efficiency of the technology
development process through constructive interaction with other actors in the technology
development process. These are particularly with farmers and commodity researchers,
but also, to a lesser extent, to those responsible for the technology transfer process and
for designing and implementing the policy support systems.

Along with other actors, particularly extension, they can help in accelerating the process
of adoption of relevant improved technologies.

If adoption of improved technologies is viewed as the final product, the outputs resulting from activities
relating to the first point above can be viewed as intermediate products. We believe much more emphasis
needs to be given to careful documentation and dissemination of information concerning the impact of
specific FSR activities.


The process of achieving a constructive impact on the technology development process requires specific
attention to:

Making sure that the research agenda is action oriented and focused on identified farmer

Ensuring that FSR activities are undertaken with cost (resource) efficient techniques (i.e.,
in as cheap a manner as possible) while at the same time maintaining quality. Using and
strengthening the research capacity of farmers is an important component of this.

Packaging and disseminating research information in a manner appropriate for the
different clients involved in the agricultural development process.

Examples of outputs from FSR activities that can help contribute to the technology development process
and which will facilitate the development and adoption of relevant improved technologies include the

Providing commodity researchers with: on-farm and farmer-based evaluations of

- 53 -

technologies they develop; information on possible new researchable topics based on
research at the farm level; inputs with respect to possible prioritization of research
problems based on farm and farmer-based research; information on farmers' use and
preferences regarding specific commodities and their management practices and
constraints, etc.

Helping technology-transfer agents through provision of relevant information,
recommendations and collaborating in joint activities, training sessions, etc.

Making available to all interested parties, including planning/policy, information on
technologies that have been tested on farm and that appear to be relevant according to
both research and farmer-based evaluative criteria. Such technologies are likely to
include a number of options and, in doing so, to include both:

Targeting information -- indicating under what type of technical and socio-
economic conditions the proposed technology is likely to work.

Conditional information -- indicating what farmers should do in order to at least
partially adopt it under sub-optimal conditions.

When deemed necessary or desirable, providing information to those responsible for the
policy/support systems on matters relating to the current effectiveness and efficiency of
institutions and programmes facilitating or hindering the agricultural development
process. This is particularly important when changes are required to encourage adoption
of proposed improved technologies that have been earlier been evaluated as being suitable
for dissemination.

Unfortunately information on this type of impact is rarely col!ected and documented, something we
believe needs to be rectified. Such documentation is particularly important since many FSR practitioners
argue that FSR facilitates a process rather than producing a unique product in its own right. There is a
lot of such information in Tanzania. If collected we believe such information can provide a useful idea
of FSR impact. We believe that such documentation is particularly important in Tanzania where both
internal and external support has been unstable and potentially wavering.


Because the desirable final outcome of the technology development process is adoption of relevant
improved technologies many actors, other than just FSR teams, need to receive credit for the success.
FSR teams are in a good position to undertake adoption studies to provide an objective measure of the
impact resulting from collaboration and interaction amongst the various actors in the agricultural
development process. Rarely have such studies been undertaken; we believe greater emphasis on
adoption studies is justified in order to better 'sell' the importance of farmer-focussed commodity and
FSR research in the technology development process.

We believe that in the quest for resource efficient research, technology adoption studies can and should
serve more than one purpose. Adoption studies should be conducted with the following objectives in

-54 -

An assessment of the degree, level and pattern of adoption. Fulfilling this objective
obviously provides the most tangible measure of impact.

The level of benefit to an individual household from adoption of the technology.

An assessment of the possible reasons for rejection or poor levels of adoption, or
unexpected types and levels of adoption, as possible inputs into:
Identifying further priorities for research designed to improve technology to
encourage greater adoption.
Proposing possible adjustments in the policy/support systems to encourage greater
Redefining the recommendation domain.

Adoption studies can address more than one technology at a time. It is also useful to conduct adoption
studies at different time intervals (Section 5.6.4) for any given technology. Through conducting an
adoption study at regular intervals (e.g., every second year) and covering a range of technologies in each
study, an FSR could develop, over time, a good database on the adoption and impact of new technologies
in their region.

Recommendation 6.1 Studies should be undertaken to document the impact of FSR-related activities on
the technology development process, and assess the degree, level and pattern of
adoption of new technologies as a result of FSR and other activities.

-55 -


This appendix involving agronomy deals specifically with: describing the contributions agronomy should
be making to the FSR programme and the contribution FSR agronomists should be making to other
commodity and disciplinary research programmes; the skills which are necessary to make these
contributions; and comments on on-going and planned agronomic activities in the FSR programmes.
Lastly some potentially useful linkage mechanisms are listed.


Contributions by agronomists to the FSR team should include the following:

Assist in developing and administering surveys (formal and informal) to describe
important agronomic aspects of zonal farming systems, and to help select 'target'
populations or systems.

Assist in developing and administering surveys, and conduct on-farm experiments to
accurately diagnose and quantify, or rank, agronomic constraints to production, and
identify areas of opportunity, within the zones. Assist actively in prioritizing research

Be prepared to assist with the development of non-agronomic surveys, and to screen their
agronomy-related content.

Work with farmers, other team members, and commodity team members to design, test
and adapt technology options for addressing identified production constraints and
opportunities. Also they should be prepared, when necessary,.to adjust trial designs and
analytical methods to incorporate the interests of farmers, and scientists in other
disciplines (e.g., economists).

Work to develop strong, two-way, interactive links with other team members, farmers,
commodity programme scientists and technology transfer agencies. These linkages will
greatly strengthen the contributions agronomists can make.


Contributions FSR agronomists can make to commodity research teams include the following:

Provide commodity researchers with descriptive/diagnostic data to assist them in targeting
their research on real farm problems, in the design of practical technology options, and
in prioritizing research issues. (Note: It is best if the same commodity scientists can
participate in descriptive/diagnostic activities).

Assist commodity researchers with the verification and adaptation of technology options.
They can also help to introduce farmers and farmers' perspectives, early in the

- 56 -

technology development process.

Take the lead in integrating technology components into effective, composite systems
options. They can also take the lead in studying cross-commodity research topics such
as intercropping or crop/livestock interactions.

Champion the farmers' perspective in the research system as a whole.


Skills required by FSR agronomists in order for them to undertake their duties satisfactorily, include the

A strong and broad-based grounding in the many facets of their science, and good
observation skills in the field.

Good inter-personal skills which allow them to work well with farmers, other team
members, commodity programme scientists, technology transfer agents, etc. They also
need to learn, modify and apply up-to-date research methods that allow for farmers'
systematic participation in the technology development and dissemination process.

Sound understanding of statistics to develop and apply innovative and resource-efficient
trial designs, while maintaining scientific rigour and credibility.

A systems perspective, and a willingness to learn about and integrate factors outside their
immediate discipline.

Good writing skills to communicate with a wide range of client groups including farmers,
commodity research scientists, extensionists, policy makers, etc.

Specific implications of the above are that at the undergraduate level, training should cover a wide range
of topics including soils, crop physiology, ecology, weed science, integrated pest management, etc. and
possibly also include some sociology courses. At the post graduate level, training is necessarily more
focused, but a solid grounding in statistics is important. Regular in-service training courses can provide
training in communication skills and farmer participation methodology.


Some basic observations concerning the agronomic component of FSR programmes in Tanzania include
the following:

In most zones, further work on describing local farming systems and on problem
diagnosis and quantification would appear to be justified, particularly in light of the
importance attached by the NALRM to zonal research priority setting. This work should
be done in conjunction with other FSR team members as well as commodity programme

- 57 -

* Some zonal FSR programmes are well integrated with commodity research teams, but
most are still doing basically researcher managed trials on-farm. Most FSR programmes
would benefit from developing and applying methods that encourage much more farmer
participation in research programs (see Section 5.4). They could also benefit from the
use of more varied and innovative trial designs that allow the perspective of farmers and
other disciplines (e.g., economics) to be included. In most cases there is a strong need
for the use of more systematic data collection procedures as well (e.g., data collection
forms for each trial that are prepared before the season and regularly updated).

* More resource-efficient research systems need to be developed and applied. Standard,
researcher managed trials are perhaps the least resource efficient designs for on-farm
research. They require more inputs, more field visits and more researcher time than
most other designs. With research resources being as limited as they are in most zones
in Tanzania, there is an urgent need in general for the application of resource efficient
techniques. These include appropriate trial designs (i.e., usually more farmer managed
types), clustering of research sites and experimental units, and the use of farmer groups.

* More innovative trial designs and research approaches are required in most zones to
facilitate integration of FSR with commodity research programmes. If FSR continues to
use standard, researcher managed trial formats, they will be very limited in the number
of research topics they can address on-farm, due to the demands these types of trials
make on a researcher's time. Less time consuming, more resource-efficient trial designs
are required if FSR agronomists are to address a wide range of research topics, and
thereby accommodate commodity research interests on-farm.

* Though there are some clear exceptions, most zonal FSR teams need to give greater
priority to developing linkages with both farmers and commodity research programmes
in their zones.

- 58 -


This appendix deals with the contributions the discipline, in this case, animal science, should be making
to the team and to other commodity researchers and what necessary skills are needed for effective
contribution to FSR. The following issues are also addressed:

How to handle common livestock/crop interaction areas (e.g., ox cultivation, crop residue
use, etc.).
Modes of interaction between key station and FSR livestock researchers.
How livestock trials can be integrated into team work.
What types of trials should be avoided by the FSR teams.


Livestock are an important component of smallholder farming systems in many areas of Tanzania yet,
to date, have been somewhat neglected and even left out of the FSR teams' research programmes. In
part, this is due to a number of institutional and researcher biases:

The complexity of the animal situation.
The traditional viewpoint of livestock researchers.
The multiple nature of livestock use.
The dominance of crop commodity research in all zones.
In most instances the youth and inexperience of the FSR livestock team member.
The lack of interaction between FSR team members and between the FSR livestock
person and other zonal researchers.
The 'definition' of farming systems by the FSR t-'-s doirg diag-osti, work ."d surveys
as being only cropping systems.
The perceived difficulties of working with livestock on-farm.

All of these issues need to be sorted out so that recommendations can be made for collaborative research

Given all of the above constraints and complexities, therefore, the FSR livestock researchers can best
assist the FSR team by:

Playing a leadership role in encouraging interactions between themselves and the other
FSR team members and between themselves and commodity researchers.

Assisting in all survey work to ensure that the livestock components are addressed.

Working with farmers, other team members, and commodity researchers to design, test,
and adapt technology options which address key, identified production constraints
especially related to crop-livestock interactions.

-59 -


Livestock interact in multiple fashion with other components of the farming system. The diversity of
livestock production systems in Tanzania ranges from pastoralism to intensive, smallholder dairy
production, includes cattle, sheep, goats, poultry, rabbits, donkeys, and swine, and occurs in every
agroecosystem in the country. Even factors as apparently simple as ownership can become extremely
complex as traditional usage may dictate who can have which products from the animals, who can own,
who can borrow, etc. Further, most 'farming systems' surveys done since the early 1970s have
interpreted farming systems as being equivalent to cropping systems and have gathered either cursory or
no data on livestock. It is also difficult, at times, to conduct economic analyses of livestock production
because it frequently produces products which are not readily marketable (e.g., draft, manure, fuel, etc.).

Therefore, the contributions which the FSR livestock researchers can bring to commodity research in
general include:

Making other researchers aware of the systems interactions between crops and livestock
by providing them with descriptive and diagnostic data.

Assisting commodity researchers, even other livestock researchers, with verification and
adaptation of technology options.


The skills needed by a livestock researcher on an FSR team are analogous to those listed for an FSR
agronomist in Appendix A. These are:

Strong training and understanding of the many facets of animal science, and good
observation skills in the field which will improve that understanding.

Good interpersonal skills to enable them to work with farmers, other team members,
commodity researchers, technology transfer agents, etc., and the ability to learn, modify,
and apply research methods that allow for farmers' systematic participation in technology
development and dissemination.

Basic understanding of statistics and experimental design, including innovative methods
used for livestock on-farm trials.

A systems perspective, and the willingness and humility necessary to learn about and
integrate knowledge from outside the animal science discipline.

Good writing skills.

Additionally, the livestock researcher and, ideally, everyone on the FSR team, should understand the
contributions and interactions which come about from the integration of livestock into farming systems.
The team should be able to identify within system linkages and relationships: availability of grazing/feed

-60 -

resources; water availability; labor availability; land tenure systems; management practices, stock wealth,
and production goals; animal diseases and animal genetic potential; and socio-economic conditions.
Furthermore, the team should analyze the relationship of these factors to animal performance. Stocking
rate and grazing pressure over the seasons, the size and structure of the herd or flock, the prevalence of
disease, and the animal breed all affect animal performance (e.g., calving rate and mortality, weight gain,
adult mortality rates, milk production, reproductive performance, etc.).


With the exception of staff at Mpwapwa and Uyole, we sense that the FSR livestock officers are fairly
young and inexperienced, both in their profession and as FSR team members. Their real or perceived
junior status makes them unwilling or unable to be forceful in their interactions with the team, to pursue
and persuade other team members to conduct collaborative research, to participate in team activities.
There are exceptions, of course, but by and large, this scenario holds true for most of the FSR livestock
researchers. The FSR coordinator should play a key role in fostering interactions to the extent, perhaps,
of insisting that the livestock officer accompany other team members on field visits. Some of the FSR
teams are heavily in involved in survey work, both for their own research and as contractors. As it
would be difficult to imagine a farming system in Tanzania with no livestock, it is important that the
livestock officer participate in survey work -- helping design the survey instrument, conducting
interviews, and analyzing the data.

As for interactions with other livestock researchers, given the current state of uncertainty about the
closing down of many livestock programmes and institutes, it is difficult to make recommendations that
could be implementable.

Off-station, the FSR livestock officer should be aware of, and make contact with, other agencies and
organizations that are working with livestock. As a few examples, the Mbeya Or.ensation Project, the
Swiss Dairy Project (Mbeya), the PANET (Tanzania pastoralist network), the various LITIs (Livestock
Training Institutes), and so on, are efforts to foster livestock production and offer potential opportunities
for productive collaboration.


B5.1 On-Station Research

Livestock systems research, especially on-farm, faces more difficulties than crop research due to livestock
mobility, producer attitudes, multiple outputs, communal land tenure systems, life-cycle duration, and
life-cycle synchronization. Because of these difficulties, it may be necessary to conduct some livestock
trials on the research station when facilities are available. The important role of on-station research for
livestock must be stressed. Some technologies are ready to be adopted without on-farm testing (e.g.,
zero-grazing of dairy cattle). Other, more broadly applicable technologies should be developed before
going on-farm, then refined more specifically for clusters or groups of farmers. Trials should always be
designed for a well-defined group of farmers who are able to obtain the experimental variables while non-
experimental variables should be held at farm level. There must be a clear understanding of technical
relationships (e.g., refining technologies). On-station research can also be used to develop technologies
relevant to specific target areas (e.g., breed improvement and forage evaluation for specific ecosystems).

-61 -

This is probably the most common use of on-station research. Finally, insufficient resources of the FSR
team may prevent them from travelling and supervising on-farm trials.

B5.2 On-Farm Research

In justifying the need for doing livestock on-farm trials, the FSR team should ask if the technology is
relevant to the problems identified, are on-farm trials appropriate for testing the technology, and what
is the practicality of doing on-farm trials, keeping in mind the above mentioned concerns related to on-
station research. Two types of livestock on-farm trials can be chosen: statistical trials, which require high
levels of researcher management and supervision, and monitoring trials, where the researcher observes
and monitors farmers' reactions to the technology being tested. In livestock systems research, it may be
more appropriate to conduct trials aimed at monitoring farmers' reactions. Results are usually not
statistically analyzable by ANOVA but other methods of analysis can be used.

B5.3 Types of Research to be Considered

The FSR teams could consider conducting livestock research on animal nutrition, maintaining or
improving soil fertility with farm yard manure and multipurpose trees and shrubs, and possibly on animal
traction if implements have already been adequately tested on-station and farmers have a history of
keeping livestock. Introducing and implementing the use of animal traction is a long process involving
high cost, high price of the package (i.e., needing a cash crop), high cost of animals, development of
implements, farmer and animal training, and varied use of tillage techniques, It is not a task to be taken
on by inexperienced researchers, especially if the farmers have zero experience with draft animals.

A traditional livestock researcher bias seems to underlay some of the FSR teams in that livestock is
equated with cattle. They should consider working with animals other.than cattle, especially sheep and

BS.4 Animal Nutrition Trials

The highest degree of relationship between crops and livestock is realized through animal nutrition trials.
For such trials, technological solutions to improving animal nutrition could include crop improvement,
changes in livestock management, pasture improvement, feed supplementation. Crop improvement
strategies with crop/livestock interactions could include selection of crop varieties which yield residues
with higher nutritive value or quantity of residues, changing crop combinations to produce residues
preferred by livestock, and changing time of planting. Feed supplementation strategies most definitely
have to take into account farmer circumstances and could include the use of fodder banks, fodder trees
(e.g., multipurpose trees), agricultural by-products, and urea-molasses licks.

B5.5 Clustering

In the report, we recommend that the FSR teams begin to cluster farmers in order to improve their
resource efficiency. By the same token, any trials done with livestock should also be clustered, perhaps
in more than one way. For example, to conduct a livestock trial within a cluster of farmers already

-62 -

determined by the FSR team, it may be necessary to cluster, or block, animals as well such factors as
age and calving/kidding history.

BS.6 To be Avoided

The difficulties of doing livestock research have been repeatedly stressed, cost and complexity being two
of the myriad of factors. Complicated or complex animal trials should be avoided (e.g., involvement in
design and manufacture of animal traction equipment, introduction of animal power to new users,
elaborate fertilization trials not realistic of farmer practices, etc.).

-63 -


Economics is one of the disciplines currently represented on all the zonal FSR teams. It is also the
discipline specialization of two of the three current professional staff of the NCU. Therefore there is no
need to argue the case for agricultural economics representation on the FSR teams. Therefore, this
somewhat abbreviated appendix summarizes the contributions agricultural economics can make and the
skills necessary to make those contributions possible. These are discussed in the context of the severe
research resource constraints currently existing in Tanzania.


FSR economists can and should make contributions at all stages of the FSR process. Their degree of
contribution and the level of responsibility will, of course, depend on the specific activity. What is
important to try and ensure at all stages of the FSR research process is that a genuine interdisciplinary
approach is always used -- something that has been implicitly recognized throughout the report. FSR
economists contributions should include, but are not necessarily confined to, the following:

Playing a leadership role, along with inputs from other disciplines, in conceptualizing,
developing, implementing, analyzing and writing up results of both informal and formal
surveys. Purposes of such surveys can differ greatly, for example, descriptive,
diagnostic, evaluative, assessment of impact, adoption, etc. All are important in a well-
rounded FSR programme.

Playing a supportive role to other disciplines (i.e., particularly agronomists and animal
scientists) in designing, implementing, evaluating, analyzing and writing up results on
FSR team trials. The greater the degree of farmer management (involvement) the more
significant is likely to be the role of FSR economists.

Be prepared to provide, whenever appropriate, the professional skills economists have
to contribute in terms of research resource efficient data collection techniques, statistical
skills (e.g., regression analysis), economic evaluation analysis, etc.

In the absence of a professional sociology/anthropology presence, particularly provide:
an even more prominent social science presence in the form of strong communication
skills (i.e., particularly with farmers); helping in organizing and running farmer groups;
providing expertise in qualitative data collection and analytical techniques, etc. (see
Appendix D).

Work to develop strong, two-way, interactive links with other team members; farmers;
commodity programme scientists; technology transfer agents; and, planners responsible
for developing policy/support programmes.


The contributions that agricultural economists can make to other actors in the agricultural development

-64 -

process, particularly commodity programme scientists and those responsible for designing policy/support
systems, include all those just mentioned above. In addition, FSR economists through relevant marketing,
adoption studies, etc., can provide information to those responsible for designing policy/support systems
that will help improve initial definition of relevant policy/support studies or, in the case of studies relating
to technology adoption, help in proposing adjustments that will encourage greater or better adoption.


There are a number of skills required of agricultural economists that will facilitate them operating
effectively in FSR teams. These include the following:

Preferably to the BSc level, a strong grounding in technical agriculture to help them in
interaction with technical scientists -- both on FSR teams and in commodity research

Training, at a minimum, to the MSc level, in agricultural economics with specific
attention being paid to:

Production/farm management including analytical techniques such as marginal
and risk analysis, budgeting and sensitivity analysis, gross margin and break-even
analysis, etc. [CIMMYT, 1988; Amir and Knipscheer, 1989; Dillon and
Hardaker; Worman, Norman and Ware-Snyder, 1990].

Good grounding in statistical techniques with ability to handle parametric but
also, preferably, non-parametric methods.

Knowledge uf-data coll-ection techniques (i.e., both informal and formaL
methods), and relevant microcomputer skills including software relating to word
processing, database storage, spreadsheets, statistical and graphical analysis, etc.

Knowledge on how to undertake technology impact/adoption studies and some
appreciation and understanding of macro-economic relationships.

Inculcation of a true systems perspective and an understanding of FSR

We recognize that it may be somewhat difficult to obtain training in all these areas in a
formalized agricultural economics MSc level course. Where this is not possible, short
courses and 'peer' individuals can help overcome deficiencies (see Appendix D2.2).

Perhaps most important is the need for inter-personal and communication skills which
will permit FSR economists to work effectively with other FSR team members and other
actors in the agricultural development process. Related to this is also the need for good
writing skills.

-65 -


Although there is considerable empathy in Tanzania for the potential role sociology/anthropology can play
in FSR, that potential still has to be realized. This appendix outlines the contributions sociology could
make in the Tanzanian FSR programme, the specific social science training needs for this perspective to
be included and finally, proposes a strategy for compiling existing social science information.


Social science has key contributions to make to FSR programmes, which are separate from, but
complementary to, the role of economics. There are several areas in which the Tanzanian FSR
programme needs to strengthen its present sociological/anthropological capacity.

DI.1 Basic Social Science Skills

Working with farmers implies being able to build an easy, trusting relationship with rural men, women
and children, and having the know-how to collect their insights in a format which can feed into the formal
research system. All FSR teams should be encouraged to refresh or develop skills in the following:

* General communication. This refers to the ability of the interviewer to elicit farmers'
viewpoints and reflections. Interviewers need to establish a quick rapport with farmers, develop
techniques to listen well, and need skills to probe for in-depth information [Ashby, 1990; IAC,
1991]. Techniques for working with groups differ from those of the individual interview: FSR
teams need to sharpen both sets of skills. For group methods, see Krueger [1988], Kumar
[1987], Morgan [1988].

* Collection and recording of qualitative data. Various types of qualitative information are
important for guiding research: for example, farmer evaluations of specific technologies, their
discussions of crop histories, and farmer analyses of their own constraints. Researchers need to
be trained to collect and verify such qualitative data in ways which offer maximum insight to
research. 'How' or 'what to record' is far from a natural skill.

* Analysis of qualitative data: both qualitatively and quantitatively. In order for qualitative
data to feed into the formal research system and help guide priority setting, technology design
and development, researchers must learn to manage complex and often large qualitative data sets.
In some cases, qualitative data might best be synthesized descriptively, that is, analyzed along
major themes and inter-relationships. In other cases, qualitative records may need to be
quantified, both to highlight trends as well as to be more interpretable by a range of scientists.

To date, in-depth qualitative data has been largely ignored within the FSR system precisely
because researchers lack expertise to collect and manage it in a rigorous manner. We note that
it is often qualitative data which is richest in farmers' insights and explanation of causes and
effects of current practices. For overview of qualitative research methods, see Bernard [1988],
Patton [1980].

* Analysis of non-conventional quantitative data. Farmer involvement in research, and

-66 -

particularly in trial design and management, often results in data sets which are heterogenous
within and among locations; different farmers have differing trial priorities and methods of
experimenting (e.g., varying planting densities, intercropping choices). Such data, while realistic
of actual farming practices, can be difficult for researchers to analyze and interpret. We suggest
that researchers be equipped to handle this farmer-generated 'non-conventional' quantitative data.
In some cases, systematic qualitative assessments can be quantified, in others, non-parametric
statistics may be appropriate, and in still others, statistical tests are being refined to address
heterogeneity [e.g., Heinrich and Masikara 1991, Stroup, Hildebrand and Francis, 1991]. The
more technical flexibility scientists have in handling 'non-conventional' data, the more liberty
they may be willing to give farmers in all phases of on-farm research.

Only with improved communication data collection, and management skills can FSR teams:

Adequately represent farmers' viewpoint to researchers (as a 'cultural broker') and
Efficiently involve farmers themselves in research.

The four skills discussed in this section should be regarded as sine qua non of FSR.

DI.2 Key Social Science Subject Matter

FSR takes place in a complex socio-economic as well as physical environment. In some cases,
researchers may take the lead in describing the needs, priorities, and opinions, of their farming clientele;
in other cases, it is farmers' knowledge itself which needs to be systematically processed so as to guide
research. There are several subject matter areas where FSR teams should consider putting more research

* User-differentiation. The need to distinguish among different farmer clients has been discussed
elsewhere (see Section 5.5). Two points requiring emphasis are::

Gender, or specifically, differences between male and female roles, is one of the major
variables shaping agricultural production in Africa. For instance, men and women have
different work roles, technical knowledge, and access to resources. The two groups may
also be affected differentially by technology development. For example, in switching
toward commercial crops, the household overall may increase its cash income. However,
this cash income is often controlled by men which means women and children may be
adversely affected by lack of actual food for consumption or the cash to buy food. FSR
researchers must become both more gender-sensitive as well as develop specific
techniques for working across what may be gender boundaries. However, gender is but
one variable describing a farmer. Age, ethnic group, class, skill, etc., are combined with
gender to create unique user profiles (e.g., a young woman may have very different
concerns from an older woman, a poor grandmother certainly has different constraints
from a richer one). FSR teams in general must be better trained to distinguish among
these 'composite' rather than 'unique variable' user groups to help create technologies
which reach various clientele.

The issue of different user groups is one which should pervade all stages of technology
development. At the beginning, researchers may work to identify distinct user groups

-67 -

and describe their varying wants/needs. Different user groups, however, should also be
involved in the design and testing of technologies, and in their evaluation. Further,
adoption and impact studies should be conceived so as to anticipate and monitor the
effects of technologies among varied farming clientele.

* Farmer evaluation of technologies, adoption and impact studies. We have made several
recommendations for putting more emphasis on farmer evaluations and monitoring the adoption
of technologies (see Sections 4.1, 5.6.4, 6.3). Here we emphasize the need to systematically
programme monitoring activities at different stages of the research process. For example, in
terms of varieties, farmer evaluations may be scheduled for the end of a season, acceptance
studies within one to two years, adoption analyses with a three to four year period, and impact
studies taken five years after the initial release. Timing of the latter studies partly depends on
the speed with which the technology is diffused and adopted.

* Local knowledge systems: 'indigenous knowledge'. Local knowledge can be a particularly rich
source for collecting and synthesizing technical information. Analyses of local knowledge may
identify farmer innovations to be built upon, may highlight the rationale for a range of farmer
practices, and suggest how agricultural practices are evolving. Such knowledge is variable within
communities (i.e., differing by age, gender, work experience, skill, etc.), highly dynamic, and
takes considerable skill to elicit effectively. We suggest that specific studies should examine how
farmers decide on crop allocations and how communities themselves guide the management of
their natural resources.

* Informal systems of experimentation and technology exchange. Farmers themselves are
constant experimenters and, indeed, have been responsible for much of the technological
innovation of the last 5,000 years. Many have systematic ways of experimenting with new inputs
and cultural practices. Building on farmers' own methods of experimentation may help the
research system to develop more realistic and predictive testing procedures (see Section 4.1). In
addition, studies of informal systems of farmer input exchange and information sharing may also
provide insights into designing improved technology support systems (e.g., seed dissemination

* Analyses of differential resource control. Related to user differentiation above, this refers to
the formal legal systems and local rules which determine who has access to what (i.e., land,
capital, labour). For example, researcher familiarity with overall land tenure will be particularly
important in designing soil management practices and livestock production strategies.

While some of these proposed themes can be partially pursued by economists or agronomists well versed
in basic social science methodology (i.e particularly the first and second points above) others will
probably require the expertise of someone with more advanced sociological/anthropological training.

DI.3 Social Science Tools

FSR teams can cover much of their qualitative information needs if they master the use of three basic
tools: the in-depth interview (sometimes with a key informant), the informal survey, and focus groups.
Participatory tools per se are plentiful. For example, see Waters-Bayer [1989] and Kamp and Schuthof
[19891, for descriptions of methods. They presuppose basic communication and qualitative data analysis

- 68 -

skills. FSR teams might benefit from having an overview of the possible participatory methods currently
used, with emphasis on those which can be used in many situations (e.g., brainstorming and participatory
diagramming). Rapid techniques in general (i.e., participatory rural appraisal or sondeo methods) are
most useful for researchers doing highly site-specific work, in which they can confirm the quick
community sketch with repeated visits to the site(s) in question. They may also be justified when one
wants to build up team spirit as well as community commitment to research. Good informal surveys can
usually achieve the same ends -- but with more in-depth information.


We suggest several strategies for incorporating social science skills into the ongoing Tanzanian FSR

D2.1 Overall training of FSR teams in basic social science skills

FSR coordinators, FSR researchers and field officers should be able to communicate effectively with
farmers and collect basic qualitative data (see Appendix D. 1). All should also be exposed to the range
of possible farmer participatory approaches, an overview of possible participatory tools (see Appendix
DI.1), and skills for building gender awareness/analysis. Short-term courses (two to four weeks) can
provide the base for such training, if instruction is oriented towards practical application within the
Tanzanian context. We suggest that a consultant familiar with African agriculture be brought in for such

D2.2 Training of economists in more specific social science skills

In addition to the above (see Appendix D2.1), we recommend that economists, already integrated into
FSR teams, be trained to collect and analyze in-depth qualitative data (see Appendix D1.1). They can
be helped to develop comprehensive interviewing skills, to consider a range of socio-cultural concerns
(i.e., user differentiation), and particularly, to take the lead in analyzing and synthesizing qualitative
information. While short-term courses (i.e., two to four weeks) can help introduce or reinforce in-depth
qualitative skills, we feel that economists should have the opportunity to interact more regularly with a
sociological/anthropological specialist. For this purpose, we suggest that the NCU try to develop an
ongoing relationship with a social science consultant within the region, preferably in Tanzania. Such a
consultant would be expected to deliver initial training, make periodic visits to field sites (perhaps two
to three times a year) as well as to help economists design and follow through on specific research

D2.3 Recruitment/Training of Social Science Specialists

There is a need for more specific expertise to look at key subject matter (i.e., particularly the last three
points in Appendix D1.2). We suggest putting emphasis on recruiting and training for applied social
scientists with a focus of agriculture and interdisciplinary investigation. There are several means for
building an applied social science component within the FSR teams. We suggest that:

-69 -

The current social science designates within the FSR teams be given training appropriate
to their duties. Neither designate currently has a rural sociology or applied anthropology
background and both will need ongoing support from more experienced social scientists.
We envision three training possibilities:

One or both designates develop a mentor relationship with an applied social
scientist within the region. For example, Zambia would seem a good possibility.
The FSR sociologist might spend a month at the mentor's home site, then return
to Tanzania to develop parallel activities. The sociologist could keep in contact
with his/her mentor through correspondence and periodic exchange visits.

One or both designates might develop a mentor relationship with an applied
social scientist working within Tanzania but not specifically within an FSR
programme. The proposed CIDA funded project in the Northern Zone might
provide one such an opportunity. In such a case, the mentor responsibilities of
the project scientist might have to be specified within the donor contract.

One or both designates be aided by a visiting social science consultant who would
be willing to make repeated visits to Tanzania and help FSR teams plan and
execute ongoing research programmes. The arrangement would parallel the
current livestock consultant format.

The position of gender-specialist, slotted for the Lake Zone, be modified so as to recruit
an applied social scientist with considerable gender expertise. Gender, a key variable in
agricultural research, needs to be combined with other variables so as to define user
groups (Appendix DI.2). Such a social scientist could be encouraged to:

Investigate key subject matters.
Rotate among FSR teams imparting basic social skills.

The proposed training for a rural sociologist/applied anthropologist [Merrill-Sands,
Kirway and Semgalawe, 1990: 41] be upgraded to include one slot for a Ph.D. degree.
Such a scientist could prove to be a longer-term resource person for the FSR teams.


We sense that a good deal of social science information already exists which could be valuable for aiding
FSR teams to understand their farming communities -- particularly in terms of the themes of user
differentiation, indigenous knowledge, and differential resource control. The University of Dar Es
Salaam has a well-known social science tradition and has been among the Africa leaders in gender
analysis. Further, Tanzania has welcomed a good number of expatriate social science researchers to
conduct thesis research in rural areas. The problem with this information, much of it academic, is that
it is not readily accessible to FSR practitioners. We suggest that the NCU employ a local consultant to:

Write an annotated bibliography of social science studies which could be used in applied
research -- specifying subject matter, quality of information, dates and methods of

- 70 -


Collect, through photocopying and/or ordering, key Tanzanian social science research

Compile a list of current addresses of important social science researchers (i.e., applied
or academic) who might be sources of continuing information on rural Tanzanian
systems, both in and outside Tanzania.

Such information will help contextualize informal surveys, give research an historical perspective, and
may possibly lead to important areas of FSR investigation.

-71 -


6 June 1992 Arrival of S. Russo, G. Heinrich

7 June Arrival of D. Norman, D. Merrill-Sands, L. Sperling

8 June Meeting of Task Force with NCU staff: Assistant Commissioner for FSR, Mr.
T. Kirway, Dr. A. Stroud, and Mrs. Z. Semgalawe at Temeke

Brief meeting with FSR team members from Southern Zone: Mr. H. Temu, Miss
A. Gongwe, and Mr. A. Mtambuki

9 June Further meetings with NCU staff at Temeke

Meeting with Assistant Commissioner for Livestock Research, Mr. D. Mpiri

10 June D. Merrill-Sands, G. Heinrich, L. Sperling, and T. Kirway travelled to Northern

D. Norman, S. Russo, and A. Stroud travelled to Eastern Zone

11-12 June Northern Zone contacts: Dr. J. Haki, Zonal Director; Dr. A. Mgonja, Wheat
and Barley Research Coordinator; Mr. W. Modestus, Agronomy; Mr. S. Lyimo,
FSR Zonal Coordinator; Mr. F. Nkonya, Economist, FSR Team; Dr. O. Edgy,
CIAT Regional Bean Programme.

Eastern Zone contacts: FSR team: Mr. A. Chilangane, Acting FSR Zonal
Coordinator; Mr. V. Akulumuka, Agronomy; Mr. O. Ringia, Economics; Mr.
N. Kaden'guka, Livestock; Mr. G. Mamkwe, Mrs. E. Hoya, and Miss D.
Mende, Field Officers. Zonal researchers: Dr. J. Assenga, Acting Zonal
Director, Grain Legume Breeder; Dr. A. Moshi, National Coordinator, Maize;
Mr. J. Kabissa, National Coordinator, Cotton; Mr. Mligo, National Coordinator,
Grain Legumes; Dr. Saadan, National Coordinator, Sorghum and Millet; Dr.
Mbwaga, Plant Pathologist; Mr. Mkangwa, Soils; Mr. Baijukya, Soils; Mr.
Mvuugi, Rice.

12 June Travel to Dar Es Salaam

13-14 June Work at hotel

15 June FSR Strategy Workshop, Embassy Hotel. In addition to Task Force and NCU
staff, participants included: Mr. D. Mpiri, AC(LR); Dr. S. Lyimo, Mr. R.
Ndondi, Acting AC(CR); Mr. L. Ngatunga, AC(SS); Mr. S. Shomari, ZDRT,
Southern; Dr G. Madata, Acting ZDRT, Southern Highlands; Dr. S. Das,
ZDRT, Central; Mr. D. Shirima, Acting FSRC, Western; Mr. J. Mende, Acting
AC(T); Mr. A. Kaliba, Acting FSRC, Central; Mr. G. Kajiru, FSRC

- 72 -

Representative, Lake Zone; Mr J. Kabissa; Mr. A. Mussai, FSRC, Southern
Highlands; Mr. J. Assenga; Mr. V. Akulumuka; Dr. C. Mushi, Bean Research
Coordinator, Lyamugu; Mr. K. Brown, O&M; Mr. P. Agaard, O&M.

16 June Continue writing
D. Merrill-Sands and L. Sperling departed

17 June D. Norman departed

18 June G. Heinrich departed, S. Russo left for Arusha and Mbeya. First draft left with
NCU and taken by other Task Force members

2 July Comments from NCU and other Task Force members and corrected drafts sent
to D. Merrill-Sands in USA.

28 July D. Norman and D. Merrill-Sands sent second draft to NCU and other Task Force

30 August Comments sent back to D. Norman by this date

Mid-September Final draft sent to NCU and copies sent to other Task Force members

- 73 -


Although all the references cited in the body of the text are listed in the following, the list also includes
others that might be of use and relate to subject material discussed in the report.

Amir, P., and H. Knipscheer, 1989. Conducting On-farm Animal Research: Procedures and Economic
Analysis. Morrilton and Ottawa, Canada: Winrock International and International Development
Research Center.

Andrew, C.O., and P.E. Hildebrand, 1982. Planning and Conducting Applied Agriculture Research.
Boulder, Colorado: Westview Press.

Ashby, J., 1987. The Effects of Different Types of Farmer Participation on the Management of On-Farm
Trials. Agricultural Administration and Extension, 25:235-252.

Ashby, J.A., 1990. Evaluating Technologies with Farmers: A Handbook. Cali, Colombia: IPRA
Projects, International Centre for Tropical Agriculture.

Bernard, H.R., 1988. Research Methods in Cultural Anthropology. Newbury Park, CA: Sage

Center for International Development and Environment (World Resources Institute), 1991. Participatory
Rural Appraisal Handbook. Washington, D.C.

CIMMYT, 1985. Diagnostic Phase Training Manual. Nairobi: CIMMYT. (Manuscript)

CIMMYT, 1988. From Agronomic Data to Farmer Recommendations: An Economics Training Manual.
Completely Revised Edition. Londres, Mexico: CIMMYT.

Conway, G., 1985. Agroecosystem Analysis. Agricultural Administration, 20:31-55.

Daniels, L., J. Howard, M. Maredia, J. Oehmke and R. Bernstein, 1990. The Impact of Agricultural
Research: A Review of the Ex-Post Assessment Literature with Implications for Africa. East
Lansing: Michigan State University. (Draft).

Dillon, J.I., and J.B. Hardaker, 1980. Farm Management Research for Small Farmer Development.
FAO Agricultural Services Bulletin No. 41. Rome: Food and Agriculture Organization of the
United Nations.

Feldstein, H.S., and S.V. Poats (Eds.), 1991. Working Together: Gender Analysis in Agriculture. Two
Volumes. West Hartford, Connecticut: Kumarian Press.

Feldstein, H.S., C.B. Flora and S.V. Poats, n.d. The Gender Variable in Agricultural Research.
Ottawa, Canada: International Development Research Centre.

Finan, T.J., and J. van Willigen, 1991. The Pursuit of Social Knowledge: Methodology and the Practice
of Anthropology. In Willigen, J. van and T. L. Finan (Eds.), Soundings: Rapid and Reliable

- 74 -

Research Methods for Practicing Anthropologists. National Association for the Practice of
Anthropology (NAPA) Bulletin 10.

Franzel, S., and E. Crawford, 1987. Comparing Formal and Informal Survey Techniques for Farming
Systems Research: A Case Study from Kenya. Agricultural Administration 27:13-33.

Gilbert, E., 1990. Non-Governmental Organizations and Agricultural Research: The Experience of The
Gambia. Agricultural Administration (Research and Extension) Network Paper 12. London:
Overseas Development Administration.

Goldman, A.E., and S. Schwartz-McDonald, 1987. The Group Depth Interview: Principles and
Practices. Englewood Cliffs, New Jersey: Prentice Hall.

Grandin, B.E., 1988. Wealth Ranking in Smallholder Communities: A Field Manual. London, U.K.:
Intermediate Technology Publications.

Heinrich, G.M., and S. Masikara, 1991. Trial Designs and Logistics for Farmer Implemented
Technology Assessments, with Large Numbers of Farmers: Some Approaches used in Botswana.
Paper presented at l th Annual Farming Systems Symposium, Michigan State University, East
Lansing, Michigan.

Herpen, D. van and J. A. Ashby, 1991. Gender Analysis in Agricultural Research. Cali, Colombia:
International Center for Tropical Agriculture.

International Agricultural Centre, 1991. Communication Manual of the International Course on Rural
Extension. Wageningen, The Netherlands: IAC.

Kanip, J. van de -and -P.- Schuthof, 1989. Methods of Participatory Technology Devclopment:
Theoretical and Practical Implications. Leusden, Holland: Information Centre for Low External
Input Agriculture (ILEIA).

Kean, S., and L. Singogo, 1990. Research-Extension Liaison Officers in Zambia: Bridging the Gap
between Research and Extension. OFCOR Discussion Paper No. 1. The Hague: ISNAR.

Krueger, R.A., 1988. Focus Groups: A Practical Guide for Applied Research. Newbury Park,
California: Sage Publications.

Kumar, K., 1987. Conducting Group Interviews in Developing Countries. A.I.D. Program Design and
Evaluation Methodology Report No. 8. Washington, DC: Agency for International

Lightfoot, C., V.P. Singh, T. Paris, P. Mishra, and A. Salman, 1990. Training Resource Book for
Farming Systems Diagnosis. Manila, Philippines:IRRI and ICLARM.

Lightfoot, C., N. Axinn, P. Singh, a. Bottrall, and G. Conway. 1989. Training Resource Book for
Agro-Ecosystem Mapping. Manilla and Dehli: IRRI and the Ford Foundation.

Merrill-Sands, D., and J. McAllister, 1988. Strengthening the Integration of On-Farm Client-Oriented

-75 -

Research and Experiment Station Research in National Agricultural Research Systems:
Management Lessons from Nine Country Case Studies. OFCOR Comparative Study Paper No.
1. The Hague: ISNAR.

Merrill-Sands, D., S. Biggs, R. Bingen, P. Ewell, J. McAllister, and S. Poats, 1991. Institutional
Considerations in Strengthening On-Farm Client-Oriented Research in National Agricultural
Research Systems: Lessons From A Nine Country Study. Experimental Agriculture, 27:343-373.

Merrill-Sands, D., T. Kirway and Z. Semgalawe, 1990. Project for Strengthening the National
Development and Coordination of Farming Systems Research in Tanzania. Dar Es Salaam:

Merton, R.K., 1989. Focussed Interview. New York: Macmillan.

Morgan, D.L., 1988. Focus Groups as Qualitative Research. Oualitative Research Methods. Volume
16. Newbury Park, California: Sage Publications.

NALRM, 1992. National Agricultural and Livestock Research Masterplan. Tanzania Department of
Research and Training. The Hague: ISNAR.

Patton, M.Q., 1980. Oualitative Evaluation Methods. Beverly Hills, California: Sage Publication.

Poats, S.V., M. Schmink and A. Spring, 1988. Gender Issues in Farming Systems Research and
Extension. Special Studies in Agriculture Science and Policy. Boulder, Colorado: Westview

Rhoades, R.E., 1982. The Art of the Informal Agricultural Survey. Lima, Peru: International Potato

Scheidegger, U. (n.d.). The Complementarity of Farmer and Research Roles in Research. Butare: Great
Lakes Regional Programme. (Manuscript).

Semgalawe, Z.M., and P. Anandajayayasekeram, 1990. Summary Recommendations for Senior
Research and Extension Administrators Workshop, AICC, Arusha. Dar Es Salaam: DRT.

Sperling, L. 1991. Farmer Participation and the Development of Bean Varieties in Rwanda. In Moock,
J.and R. Rhoades (Eds.), Diversity. Farmer Knowledge and Sustainability. Ithaca, New York:
Cornell University Press.

Sperling, L., M. Loevinsohn, and B. Ntambovura, 1992. Rethinking the Farmer's Role in Plant
Breeding: Local Bean Experts and On-Station Selection in Rwanda. Paper submitted to
Experimental Agriculture.

Stroup, W., P.E. Hildebrand and C.A. Francis, 1991. Farmer Participation for More Effective Research
in Sustainable Agriculture. Paper proposed for American Society of Agronomy Special
Publication: Technologies for Sustainable Agriculture in the Tropics.

Tripp, R. and J. Wooley, 1989. The Planning Stage of On-Farm Research: Identifying Factors for

- 76 -

Experimentation. Mexico, D.F. and Cali, Colombia: CIMMYT and CIAT.

Waters-Bayer, A., 1989. Participatory Technology Development in Ecologically Oriented Agriculture:
Some Approaches and Tools. Network Paper 7. London: Agricultural Administration Unit,
Overseas Development Institute.

Wellard, K., J. Farrington, and P. Davies, 1990. The State, Voluntary Agencies and Agricultural
Technology in Marginal Areas. Agricultural Administration (Research and Extension) Network
Paper 15. London: Overseas Development Administration.

Worman, F., D. Norman, and J. Ware-Snyder (Eds.), 1990. Farming Systems Research Handbook for
Botswana. ATIP RP No. 3. Gaborone: Agricultural Technology Improvement Project,
Department of Agricultural Research, Ministry of Agriculture. have reference).

-77 -


AC(FSR) Assistant Commissioner for Farming Systems Research
ANOVA Analysis of Variance

CIMMYT International Maize and Wheat Centre
CRT Commissioner for Research and Training

DALDO District Agricultural and Livestock Development Officer
DRT Department of Research and Training

FSR Farming Systems Research
FSRC Farming Systems Research Coordinator

GRE Graduate Record Exam

LITI Livestock Training Institute

MALD Ministry of Agriculture and Livestock Development
MOA Ministry of Agriculture

NALRM National Agriculture and Livestock Research Master Plan
NCU National Coordination Unit
NGO Non-Governmental Organization

ODA Overseas Development Administration

PANET Tanzanian Pastoralist Network
PRA Participatory Rural Appraisal

RALDO Regional Agricultural and Livestock Development Officer
P ELO Research Extension Liaison Officer

T and V Training and Visit Method
TOEFL Teaching of English as a Foreign Language

UK United Kingdom
USA United States of America

ZAC Zonal Advisory Committee
ZDRT Zonal Director of Research and Training
ZTC Zonal Technical Committee

- 78 -

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