Understanding Farmers

Explaining soil and water conservation in

Konso,Wolaita and Wello, Ethiopia

Tesfaye Beshah

rropical Resource Management Papers

Documents sur la Gestion des Ressources Tropicales
.tesfaye.thesis 29-01-2003 15:12 Pagina 1

Understanding Farmers

Explaining Soil and Water Conservation

in Konso, Wolaita and Wello, Ethiopia

Tesfaye Beshah

2003
.tesfaye.thesis 29-01-2003 15:12 Pagina 5

To Asnakech, Wondimagegn, Gebriel and Tsega

.tesfaye.thesis 29-01-2003 15:12 Pagina 7

Preface
This doctoral dissertation deals with soil and water conservation (SWC) practices that are vital
to Ethiopia’s agriculture and economy. Since the 1980s, many SWC projects have been imple-
mented in the highlands of Ethiopia for which millions of dollars were spent. However, the
legacy of these projects did not leave the country with the outcomes promised two decades
ago, regarding sustainable land use, food security, and natural resource management. In view
of these observations I decided to devote my doctoral work to finding out at least some of the
reasons for this lamentable outcome.
Two interrelated observations prompted me to focus on this issue. First, the gap between
farmers’ practices and the claims of research organisations and second, the variation among
farmers’ land management practices in specific localities in different parts of the country.
The information available on these issues was less satisfactory when I started looking to gain
an understanding of why farmers are doing what they are doing on their farms. In spite of the
inadequate knowledge on this issue, project after project is carried out without learning
about the pros and cons of the previous ones. Thus, there is very little improvement in the per-
formance of research and development interventions in the country in spite of the growing
gaps.
Understanding farmers’ soil and water conservation practices therefore became the main
theme of my research. This study provides insights on how farmers have reacted to externally
introduced SWC technologies within and across farming systems. Their responses were
analysed in terms of their knowledge and attitudes on soil erosion and soil and water conser-
vation. These analyses were carried out at the watershed, household, farm and plot levels, as
deemed appropriate. The relationships between knowledge, attitudes, and practices were
assessed. In this process, determinants of farmers’ SWC practices were analysed within and
across farming systems. Suggestions for improvements are provided based on these findings.
It is my hope that this study will challenge the R&D and extension professionals in the
country, particularly with respect to natural resource management, to work towards under-
standing farmers as a basis for promoting a genuine development partnership with them.
My PhD journey has brought me in contact with many people both in my country and abro-
ad. These contacts have helped me to cope with academic tasks and to achieve my goal in
many ways.
Among these people, my utmost words of thanks go to Prof. Niels Röling to whom I remain
very much indebted. His unique resourcefulness and guidance in science and kindness were
sources of energy for my long and demanding research processes. I met Niels when I did my
MSc degree in MAKS at Wageningen. He was my teacher and the first supervisor for my MSc
thesis. Years later he gave me the opportunity to be his PhD student. His supervision and gui-
dance during my PhD study was not limited to Wageningen. He also visited me in the field in
1999, which was very inspiring and useful in filling the gaps in my research. His contribution
to my career is unprecedented and always remembered. Besides this, I have been invited to his
house a number of times during my stay in Wageningen. I would also like to thank Niels and
his wife, Dr. Janice Jiggins, for their hospitality throughout these years.
The topical area of my research, soil and water conservation requires supervision and sup-
port from a scientist in that area. Prof. Leo Stroosnijder kindly accepted my request to become
my promoter. Leo supervised me with patience in my ‘cross-border’ study, as many of the tech-
nical concepts in SWC were new to me, coming from a social science background. I benefited
a lot from his critical and stimulating comments at all stages of my study. Thank you Leo for
your guidance and encouragement that helped me to browse the vast field of soil and water

7
.tesfaye.thesis 29-01-2003 15:12 Pagina 8

conservation so that I could marry my knowledge of social science with technical science.
Many solutions to the puzzles in my research came to light with the insights provided by
the farmers of Konso, Wolaita and Wello. Listening to them as they spoke from their life expe-
riences is addictive. With three years of intermittent field visits to each of these areas, I was
privileged to enjoy hospitality from several farm families. Without their unreserved co-opera-
tion and support, my PhD journey would not have come to a successful end. I thank them all
for showing me their world.
My communication with farmers was facilitated by my interpreters and assistants. In
Konso, I thank Mr Orkaido Olte and Mr Kora Gera. Orkaido’s contributions ranged from day-
to-day activities to stimulating professional discussion on Konso’s life and agriculture. The
assistance of Mr Kora Gera, who is well informed about the Konso culture, at a later stage of
the study was very useful. In Wolaita I had the assistance of Mr Birega Uka. I thank Birega for
his genuine concern and contribution to my project. At Maybar, South Wello, I met a team of
three persons working at the Maybar Soil Conservation Research Station. These are Ali Ahmed,
Seid Belay and Seid Hussen. This team was a great asset to my integration into the communi-
ty and in assisting me in a range of research activities, at times under great pressure. Thanks
to them, I felt very much at home at Maybar, in spite of its pseudo-remoteness, as it is poorly
connected with the nearby towns. My stay in the three research villages and meeting waves of
farmers coming to the centre is an unforgettable experience. I would also like to thank Dr Eva
Ludi, for her kind and collegial co-operation during our joint stay at Maybar station in 1999.
In addition, I want to thank many high school graduates in Wolaita and Wello who assisted
me in the survey work. I thank friends that I met in Konso, Wolaita and Wello over the years
for their support and for providing a friendly environment while I was away from my family.
The bureau’s of the Ministry of Agriculture in the SNNPR and the Amhara Regions, Zonal
and Wereda Office of Agriculture, Wereda Administration and Peasant Associations have facili-
tated my access to the communities in the study area. Offices of agriculture at various levels
also provided the secondary data. I thank the officials and experts in these offices for their
time and interest in my research. Their professional participation is highly appreciated. My
thanks also goes to Mr Amare Mengiste and Michael Assefa of the FARM-Africa for their sup-
port during my work at Konso.
Owing to the requirement of my research, I carried our case studies in areas that are sub-
stantially distant from each other. The major limitation in this connection was transportation
problem. Thanks to their understanding and kindness, Dr Marco A. Quinones and Mr Takele
Gebre of SG-2000, provided me with transportation for the entire research period. Without
their support, the cost for rental cars would have made my research operation impossible. In
addition, they gave me access to their office facilities, whenever I needed them. Everyone at
SG-2000 was friendly and kind to me. I thank them all.
My other source of logistic support during the early stage of my research came from the
Natural Resource Management and Regulatory Department of the Ministry of Agriculture. I
thank the co-operation of Mr Tamiru Habte and Mr Diribu Jemal, who accommodated my
requests from their limited resources in the department. The staff of the department was very
helpful in providing information on SCRP sites and professional discussions on SWC in
Ethiopia. I thank them all.
My study has benefited from the excellent academic environments of Wageningen
University and Research Centre. It is indeed a ‘knowledge centre’ where I could meet a num-
ber of people during courses, seminars and workshops from whom I obtained valuable sug-
gestions and stimulating questions that helped me to think further on my subject of study.
Some of them were kind enough to take time to read one or more of my case study chapters

8
.tesfaye.thesis 29-01-2003 15:12 Pagina 9

with useful comments followed by discussion. In this regard, I would like to very much thank
Dr A.W. Van den Ban who read all of my case study chapters and gave me invaluable com-
ments. I also thank him for access to his library and lively discussions on my topic and a range
of other issues on extension in developing countries. Dr Ruerd Ruben, Dr Niek Koning and
Prof. Cees Leeuwis also helped with the earlier version of chapter 6. I thank them all for their
enlightening comments and suggestions to deepen the chapter. I thank Dr Loes Maas for a
vivid discussion on the survey design and several tips on statistical analysis. She was always
ready to help me. I also thank Dr Ivo Van der Lans who provided me with suggestions during
the early stage of my data organisation and analysis. I thank Dr Geert Sterk for translating my
summary into Dutch. Whenever my computer was stuck because of the huge load of data, I
could always get the help of the experts from Systems Management at the Leeuwenborch. I
thank all of them for their instant help during my entire study period. I also thank the libra-
rians of the WUR, specially those at the Leeuwenborch for their friendly and professional assi-
stance.
My friends and colleagues in the Communication and Innovation Studies Group, former
and current PhD students are my family circle to whom I returned to for all my needs and sup-
port. Among these I would specially like to mention my office-mate Mr Jasper Eshuis for his
collegial and friendly treatment and Mrs Joke Janssen for all the support she gave me. I thank
the staff members of the Soil Erosion and Soil and Water Conservation Group for their hospi-
tality and encouragement during my study. The other source of social support has come from
my Ethiopian friends with whom I play ‘Little Ethiopia’ in Wageningen. Moreover, many of my
friends and family members in Ethiopia have supported and comforted my family in my
absence. I thank you all...
Great thanks go to AUA-Larnestein Project that very generously financed my study. I would
particularly like to thank Mr Kleis Oenema, Co-ordinator of the Ethiopia Project, for his con-
tinuous support, encouragement and concern for the progress of my study right from the
beginning. He has been very understanding and flexible when it came to making my progress
possible. My words of thanks are also due to the former CTA of the project, the late Dr M. de
Boer and his successor, Dr R. Baars for their facilitation of my project. I thank the Alemaya
University for granting me a study leave.
I thank Mrs Mundie Salm who edited the language. I particularly appreciate her timely
feedback and readiness to help me. I thank Mr Luc Dinnissen for designing this volume with
a professional engagement.
Being brought up in a village near the then Alemaya College of Agriculture, now University,
one of my early wishes was to obtain a Doctoral degree. Thank God my dream has come true.
For this I thank my parents who brought me to the school gate, while they remained illitera-
te. I thank them for such a gratifying and lasting gift.
My last but not least words of thanks go to my wife Asnakech Estifanos, our children,
Wondimagegn, Gebriel and Tsega for all the inconvenience you have encountered during my
long years of absence and long-distance support. I specially appreciate and thank Asnakech for
her endurance in shouldering all family responsibilities by herself. Tsega in particular is too
young to understand my extended field trips and travel abroad. I dedicate my dissertation to
you all with honour and pleasure!

Tesfaye Beshah

17 February, 2003

9
.tesfaye.thesis 29-01-2003 15:12 Pagina 11

Table of Contents

Preface 7

List of illustrations 16

List of tables 17

1 Introduction 19
1.1 Background of the research 19
1.2 Introduction to the country 20
1.3 Problematic situation 20
1.4 Statement of the problem 23
1.5 Objectives of the study 24
1.6 Research questions 25
1.7 Relevance of the study 25
1.8 Organisation of the thesis 26

2 Context of the Study 27

2.1 Introduction 27
2.2 Physical environment of Ethiopia 27
2.2.1 Topography 27
2.2.2 Climate 28
2.2.3 Soils 29
2.2.4 Vegetation 29
2.2.5 Agro-ecological zones 29
2.3 Population pressure 30
2.4 Rural organisations 31
2.5 Farming systems and technology 33
2.5.1 Farming systems of Ethiopia 33
2.5.2 The technology 35
2.6 Land tenure 35
2.6.1 Pre-incorporation period (before 1875) 37
2.6.2 Post-incorporation until the downfall of the Monarchy (1875 - 1974) 38
2.6.3 During the Derg Regime 39
2.6.4 After the Derg 40
2.7 An overview of agricultural research 40
2.8 An overview of agricultural extension 41
2.9 Environmental policies 42

3 Theoretical and Analytical Frameworks 45

3.1 Introduction 45
.tesfaye.thesis 29-01-2003 15:12 Pagina 12

3.2 Perspectives and approaches to land degradation and intensification 46

3.2.1 Land degradation: assumption or reality? 46
3.2.2 The path to agricultural intensification 47
3.3 Planned change and development intervention 49
3.4 The farming systems perspective 52
3.5 Indigenous knowledge 54
3.5.1 Scientists’ vs farmers’ knowledge 55
3.5.2 The nature of soil and water conservation by farmers 56
3.6 Social learning 57
3.7 The attitude - behaviour model 58
3.8 Empirical studies on determinants of soil and water conservation 61
3.8.1 Household variables 61
3.8.2 Farming system 63
3.8.3 Characteristics of technology 64
3.8.4 External institutional factors 65
3.9 Analytical framework 66

4 Research Methodology 73
4.1 Introduction 73
4.2 Overall design 73
4.3 Preparatory activities: exploratory visits, meeting communities and field staff 75
4.4 Qualitative research 75
4.4.1 Group interviews and selection of individuals 75
4.4.2 Methods of data collection 76
4.4.3 Methods of data analysis 77
4.4.4 Field assistants 78
4.5 Survey research 78
4.5.1 Secondary data 78
4.5.2 Design of the survey 78
4.6 Compensation of the community 79

5 Society and Indigenous Soil and Water Conservation in Konso, Southern Ethiopia 81
Abstract 81
5.1 Introduction 82
5.2 The study area 83
5.2.1 Location and population 83
5.2.2 Climate 83
5.2.3 Soil 84
5.2.4 Research and development interventions 84
5.3 Methodology of my study 84
5.4 Socio-economic situation 85
5.5 Overview of the production systems 87
5.5.1 Crop production 87
.tesfaye.thesis 29-01-2003 15:12 Pagina 13

5.5.2 Farm tools 88

5.5.3 Livestock production 88
5.5.4 Forest production 89
5.6 Origin and identity 90
5.7 Basis of social differentiation 91
5.8 Land tenure on a timeline 92
5.8.1 Prior to the incorporation 92
5.8.2 From incorporation until the Derg Regime (1897-1974) 93
5.8.3 From 1974 to date 93
5.9 Local institutions: indigenous governance, its structure and foundations 95
5.9.1 The origin of indigenous governance 95
5.9.2 Institutional structures 99
5.9.3 Foundation of indigenous institutions 100
5.10 Labour organisation 101
5.11 Beliefs, institutions and social learning in Konso’s agriculture 103
5.11.1 The role of belief systems and institutions 103
5.11.2 Mechanisms for social learning 104
5.12 Farmers’ knowledge, land management practices and problems 104
5.12.1 Farmers’ knowledge and land management practices 104
5.12.2 Replicability of soil and water conservation practices in Konso 112
5.13 Problems and prospects of agriculture in Konso 113
5.14 Conclusions 114

6 Farmers’ Response to Interventions in Soil and Water Conservation in Wolaita, 117

Ethiopia
Abstract 117
6.1 Introduction 118
6.2 The study area 118
6.2.1 Location of the study area 118
6.2.2 Historical background 119
6.2.3 Local institutions 120
6.2.4 Description of the farming systems 120
6.3 History of external interventions 121
6.3.1 The Wolaita Agricultural Development Unit (1970-1982) 122
6.3.2 Soil Conservation Research Project (1982-1995) 122
6.4 Methodology of my study 126
6.4.1 Unit of analysis 126
6.4.2 Sampling 127
6.4.3 Methods of data collection 127
6.4.4 Methods of data analysis 128
6.5 Results and discussion 129
6.5.1 Socio-economic situation of farmers in Wolaita 129
6.5.2 Farmers’ knowledge, perceptions and practices in soil and water conservation 136
.tesfaye.thesis 29-01-2003 15:12 Pagina 14

6.5.3 Soil and water conservation at Gunno and surrounding areas revisited 149
6.6 Conclusions 156

7 Farmers’ Response to External Intervention in Land Management, Maybar, South 159

Wello, Ethiopia
Abstract 159
7.1 Introduction 159
7.2 The study area 160
7.2.1 Location and population 160
7.2.2 Historical background 161
7.2.3 Local institutions 162
7.2.4 Physical setting of the area 163
7.3 The Soil Conservation Research Unit at Maybar 164
7.3.1 Inception and termination 164
7.3.2 The Maybar research site 164
7.3.3 Description of technical interventions 165
7.3.4 Runoff and soil loss results 167
7.4 Methodology of my study 168
7.4.1 Unit of analysis 168
7.4.2 Sampling 168
7.4.3 Methods of data collection 169
7.4.4 Methods of data analysis 169
7.5 Results and discussion 169
7.5.1 Socio-economic situation of farmers in Wello 169
7.5.2 Farmers’ knowledge, perceptions and practices of soil and water conservation 175
7.5.3 External interventions at Maybar: achievements and constraints 185
7.6 Conclusions 193

8 Discussion 197
8.1 Introduction 197
8.2 Ethiopia: a country prone to erosion that has needed to take measures to redress it 198
8.3 Research question 1: what are farmers’ responses to the introduction of SWC 199
measures?
8.3.1 Soil erosion vs soil fertility 200
8.3.2 Side effects of SWC practices 201
8.3.3 Differential reaction to bunds 201
8.3.4 Conclusion 202
8.4 Lessons from Konso: the role of the ‘soft’ side of land 203
8.5 Research question 2: what are farmers’ knowledge and attitudes with respect to soil 204
erosion and SWC
8.6 Research question 3: what do farmers do to conserve soil and water on their farms? 204
8.6.1 The relationship between farmers’ knowledge, attitude and practice in SWC 205
8.7 Research question 4: what are the determinants of soil and water conservation 206
.tesfaye.thesis 29-01-2003 15:12 Pagina 15

practices among small farmers within and across socio-economic and agro-ecological
environments?
8.7.1 Land tenure and SWC 206
8.7.2 Incentives for SWC 207
8.7.3 The role of local institutions in SWC 207
8.7.4 Household’s socio-economic characteristics and SWC 209
8.7.5 Characteristics of SWC technology 211
8.7.6 Physical factors 212
8.8 Research question 5: how can constraints to promote soil and water conservation be 214
overcome?
8.8.1 Biological vs physical conservation 215
8.8.2 Indigenous vs formal knowledge 215
8.8.3 Policy implications of the study 216
8.9 Issues for further research 218
8.9.1 Action research into a Maybar Irrigation System 218
8.9.2 Testing the status of soil fertility 219
8.9.3 Replication of the Konso study in Ankober 219
8.9.4 Farmers’ motivation in mass mobilisation for SWC 219
8.9.5 Periodic production and consumption study 220
8.9.6 Comparative study of gender factors in SWC 220

References 223

Summary 237

Samenvatting 239

Appendix 1: Methods of soil loss and runoff data collection 241

Soil loss and runoff assessment within plots
Soil loss and runoff assessment under experimental plots - Soil conservation
impact assessment

Appendix 2: Correlation results of Wolaita and Wello 242

Acronyms 243

List of local terms 244

About the Author 245

.tesfaye.thesis 29-01-2003 15:12 Pagina 16

List of Illustrations n

List of Boxes
Box 5.1 Example of how Fereyuma works 98
Box 6.1 Farmers’ rationale about bund construction and maintenance 154
Box 7.1 An account of indigenous knowledge about soil moisture for sowing 174
Box 7.2 Office back to office on irrigation in Maybar, South Wello, Ethiopia 192

List of Figures
Figure 3.1 Analytical framework with the determinants of soil and water conservation 67
Figure 5.1 Mean monthly rainfall in Konso (1971-2000) 83
Figure 6.1 Trends of soil fertility from 1990-2000 in Wolaita area 144
Figure 7.1 Trends of soil fertility in selected areas of the Maybar catchment, South 180
Wello, Ethiopia
Figure 7.2 Farmers’ reasons for decline of soil fertility, Maybar, South Wello, Ethiopia 181

Map 30

List of Plates
Plate 5.1 A typical terraced field in Konso 87
Plate 5.2 Farm tools used in indigenous Konso agriculture 88
Plate 5.3 Terraced fields irrigated by flood rivers in Konso 112
Plate 5.4 The growing scars in the once- glorious Konso stone terrace network 113
Plate 6.1 A typical land use pattern in Wolaita 121
Plate 6.2 Hydrometeric station at Gunno, Wolaita 125
Plate 6.3 The view of Waja catchment in the Damota Mountain, Wolaita 156
Plate 7.1 Stone bund moved for utilisation of the accumulated soils 187
Plate 7.2 A moved bund in the lower slope 187
Plate 7.3 Fields with selective moving stone bunds 188
.tesfaye.thesis 29-01-2003 15:12 Pagina 17

List of Tables

Table 2.1 Slope and area coverage in highland agro-ecological zones 28

Table 5.1 Generation grading system in the eastern villages, Konso, SNNPR, Ethiopia 96
Table 5.2 Sequences of Timba circulation in Doketu between Kalkusa and Herpa 97
Xeltas, Konso, SNNPR, Ethiopia
Table 5.3 Summaries of problems of decling soil fertility, Konso, SNNPR, Ethiopia 111
Table 6.1 Summary of experimental results at Gunno, Wolaita, Ethiopia 126
Table 6.2 Number of sampled farmers per Wereda, village and location in relation to 128
Soil Conservation Research Project
Table 6.3 Household characteristics and access to resources in Wolaita, Ethiopia 130
Table 6.4 Slope gradient of farm plots, Wolaita, Ethiopia 133
Table 6.5 Farmers’ prioritisation of problems in crop production 134
Table 6.6 Farmers’ response of agreement to statements on knowledge of soil erosion 137
and soil fertility processes, Wolaita, Ethiopia
Table 6.7 Farmers’ perceptions of soil and water conservation and crop production, 139
Wolaita, Ethiopia
Table 6.8 Farmers’ reasons for soil erosion, Wolaita, Ethiopia 142
Table 6.9 Farmers’ reasons for decline of soil fertility, Wolaita, Ethiopia 145
Table 6.10 Soil and water conservation practices, Wolaita, Ethiopia 147
Table 6.11 Decision on soil bunds, Wolaita, Ethiopia 150
Table 6.12a Farmers’ reasons for not constructing and maintaining soil bunds, Wolaita, 152
Ethiopia
Table 6.12b Farmers’ reasons for not contructing and maintaining soil bunds, Wolaita, 155
Ethiopia
Table 7.1 Average results of erosion experiments (1982-93), Maybar, South Wello, 167
Ethiopia
Table 7.2 Household characteristics and access to resources, Maybar, South 170
Wello, Ethiopia
Table 7.3 Rankings of farmers’ crop production problems, Maybar, Ethiopia 173
Table 7.4 Farmers’ prioritisation of livestock problems, Maybar, Ethiopia 175
Table 7.5 Farmers’ responses to statements on knowledge of soil erosion and soil 176
fertility, Maybar, South Wello, Ethiopia
Table 7.6 Farmers’ perception of soil and water conservation and crop production, 178
Maybar, South Wello, Ethiopia
Table 7.7 Farmers’ reasons for soil erosion Maybar, South Wello, Ethiopia 180
Table 7.8 Farmers’ reasons for decline of soil fertility, Maybar, South Wello, Ethiopia 181
Table 7.9 Soil and water conservation practices, Maybar, South Wello, Ethiopia 183
Table 7.10 Farmers’ comparison of Dib/Weber and the introduced stone bund, Maybar, 183
South Wello, Ethiopia 183
Table 8.1 Major soil and water conservation practices in Konso, Wolaita and Wello, 205
Ethiopia
Table 8.2 Summary of determinants of SWC in Konso, Wolaita and Wello, Ethiopia 214
.tesfaye.thesis 29-01-2003 15:12 Pagina 19

1 Introduction

1.1 Background of the research

The initial idea for this study was born during a long period of observation of my neighbour-
hood where ponds and lakes have dried up and continue to dry up partly because of siltation.
The cases in point are Lakes Adele, Langay and Alemaya, located in eastern Ethiopia where I
grew up. The last lake has lost a considerable part of its surface area and depth. In addition, I
know of gullies that have claimed long and wide tracks of land in my neighbourhood, but that
have not seen any remedial action.
Travel opportunities in 1995 and 1996 to the north and south have allowed me to compare
the land features in many parts of the country. During these trips, I witnessed extensive land
degradation in South Gondar, Wello and Tigray. My travels to southern Ethiopia helped me to
appreciate the diversities of Ethiopia’s agriculture. Information from workshops and litera-
ture have added to my concerns about the soil erosion issue in my country. One report that
particularly rose my interest spoke of 1,900 million tonnes of soil eroding annually from
Ethiopia (equivalent to 100 t ha -1 yr -1).
In the past, I wondered why farmers were not adopting soil and water conservation prac-
tices that were recommended by researchers. After my observations of various farming sys-
tems in the country, I began to think that farmers should have alternative options of soil and
water conservation to protect their soil from erosion and that I needed to understand what
the issues are that influence their actions. Owing to my educational background in
‘Management of Agricultural Knowledge Systems’ (MAKS), I have been following the extension
approaches and strategies in my country. Farmers’ responses to the introduced soil and water
conservation (SWC) practices became one of my preoccupations. When I was given an oppor-
tunity to pursue a PhD degree, I geared my attention to these issues. The outcome is the
dissertation that lies before you.

19
.tesfaye.thesis 29-01-2003 15:12 Pagina 20

My research interest then focused on gaining an understanding of what farmers know and
what they do to protect their land from water erosion. This query was pursued for two wide-
ly different systems. The first system did not receive any significant external intervention,
whereas the second system experienced intensive soil and water conservation (SWC) interven-
tions. The purpose of studying the first system was to learn what farmers do to manage their
land without external intervention, while the examination of the second system was to under-
stand both farmers’ land-management strategies and responses to the interventions.
In order to find contexts in which to carry out my study, I began a literature review and
consultation with experts in the field. Finally, I selected Konso for the first case (an indigenous
system) and Wolaita and Wello for the second case (intervention systems). Wolaita and Wello
hosted the former Soil Conservation Research Project (1981-98).

1.2 Introduction to the country

Ethiopia is a country typified by a predominantly subsistence agrarian economy. The major

contribution of agriculture to the economy is evident from its share of the GDP that has var-
ied between 46% and 58% between 1980/81 and 1997/98, (Multat, 1999). In addition, it provides
a livelihood for 85% of the population, generates over 90 percent of the export revenue main-
ly from coffee (60 percent), and produces raw materials for the industries and food needed by
its fast-growing population, though hardly meeting these needs in the last three decades or
so. The per capita income level of the country in 2000 was US$100 (World Bank, 2002), from
only around $120 in the late 1980s.
Ethiopia has a total area of 1.223 million square kilometres (Ethiopia, 1992). Of the total
area of the country, 60% is said to be suitable for cultivation. A substantial proportion of this
land is rainfed, whereas irrigated agriculture covers only 4.6% (FDRE, 1997).
According to the 1994 population and housing census, Ethiopia’s human population num-
bered 53.5 million (CSA, 1998). In 2000, the population projection was 64.3 million, with an
average annual growth rate of 2.7 for the period of 1980-2000 (World Bank, 2002).
Ethiopia’s livestock population is the highest in Africa and the tenth highest in the world
(Togoe and Dejene, 1983; Jahnke and Getachew, 1983). The country’s livestock population
based on the 1985 survey was 30.6 million TLU (MoA, 1984; cited in FDRE, 1997).
Though land provides a means of livelihood for the majority of the population of the coun-
try, land resources are facing increasing degradation mainly due to water erosion in the form
of sheet and rill erosion (Constable, 1985; Hurni, 1993).

1.3 Problematic situation

Being one of the oldest civilisations in the world, Ethiopia has an agricultural tradition that
is over 2,500 years old. However, the land was brought into cultivation at different times in
history. Generally, the northern part of the country has experienced extensive agriculture for
a long time, whereas the southwestern highlands, which show relatively less soil degradation,
were brought into agriculture in the last couple of centuries (Hurni, 1994). Generally, farming
is said to have been intensified gradually since the fourth millennium B.C., leading to subtle
soil erosion that had remained unnoticed for centuries (Galperin, 1981; cited in Mulugetta,
1988). Soil degradation due to extensive agriculture in the northern part of the country is pos-
tulated to have caused the collapse of the Axumite Kingdom in the seventh century A.D.,

20
.tesfaye.thesis 29-01-2003 15:12 Pagina 21

which was one of the earlier civilisations in the country. Consequently, the centre of domi-
nant civilisation gradually moved to areas with better soils. Lalibela between the 11th and
14th centuries, and Gonder and Menz, between the 16th and 17th centuries experienced the
same shift. The Gonder and Menz civilisations are also believed to have declined due to soil
degradation (Hurni, 1988).
The roles of the agricultural sector remain unsurpassed despite ill-conceived policy mea-
sures and development strategies that have crippled the sector over the centuries.
Prior to unification of the northern kingdoms, which took place in 1855-68, there was
widespread instability between tribes, religious groups and regional kingdoms. During the
monarchy (before 1974) internal instability and exploitation of the peasantry by landed classes
were notable agrarian problems, among others. During the military government (1974-91),
land tenure insecurity, war and the shadows of the centralised economy were the bottlenecks
of the agrarian economy. The economic policies pursued at that time included: a production
quotum based on fixed below-market prices, misguided co-operative movements, poor sup-
port services such as in extension, input supply and credit for smallholders, and forced vil-
lagisation, to mention a few. The 1990s came with the decentralisation of the state without
adequate pillars to readdress the cumulative negative consequences of the past. During these
historical epochs, there were human diseases, drought, livestock and crop diseases and pests
that tested the country’s development.
The extent of land degradation in Ethiopia was studied by different agencies at national,
watershed and plot level using standard technical procedures (EHRS; SCRP; EFAP –Vol. II, 1994;
World Bank, 1996; PDRE, 1997). Among these agencies, Ethiopia Highlands Reclamation
Studies (EHRS), financed by the FAO and the Swiss-financed Soil Conservation Research Project
(SCRP) are widely cited sources.
The EHRS estimated that half of the highland area (27 million ha) was significantly eroded,
and another 14 million ha were seriously eroded and left with relatively shallow soils. The
study designated over 2 million ha of land to suffer from irreversible erosion, which is said to
be unlikely to sustain farming in the future (Constable, 1985).
In the mid-1980s, the EHRS estimated soil loss in the Ethiopian highlands at 1,900 million
tonnes annually. Excluding about ten percent of the sediment that is carried away by
perennial rivers and adjusting for deposition within the highland, the soil loss from cropped
land is estimated at around 100 t ha-1 every year. About 80% of highlands erosion occurs on
croplands. The resulting on-site effect of erosion was estimated at about 2 percent annual
reduction in grain production. Agricultural cost of degradation from the mid-1980s to 2010
was estimated at a total of 15 billion birr, which is equivalent to a present value of birr 4.2
billion, at a 9 percent discount rate (Constable, 1985).
Hurni estimated rates of soil loss on slopes in Ethiopia at approximately 1493 million t yr-1
and on average 12 t ha-1yr-1 (Hurni, 1987; cited in Hurni, 1993). An estimated annual rate of soil
loss on croplands is 42 t ha-1yr-1. At this rate, the total soil of the present cropland will be
removed within 100-150 years. The resulting annual production loss due to soil erosion is
between 1 and 2%. An additional 1% loss to annual production due to biological soil deterio-
ration is estimated to take place (ibid.).
From the above discussion, the two widely cited sources on soil erosion in Ethiopia show a
substantial difference with respect to annual erosion rates. However, their annual loss to crop
production shows comparable trends.
In spite of the above-mentioned observations, a study by Bojö and Cassells (1995) argues
that previous analyses of soil erosion in Ethiopia have not made allowances for the re-deposit-
ed eroded soil within and between areas of productive land. A response to this argument from

21
.tesfaye.thesis 29-01-2003 15:12 Pagina 22

researchers at the University of Berne, Centre of Development and Environment (1997) dis-
agrees with their findings. They contend that the database used by Bojö and Cassells has lim-
itations. Dessaglegn (1998) also dismisses the claims of both the EHRS and that of the SCRP on
the grounds that the estimates provided by these sources are simplistic and misleading, and
have had disastrous policy implications.
Apart from the result of scientific studies on specific areas, the prevalence of erosion in the
country is directly visible in major watersheds and basins. In the north, the hills and moun-
tains have suffered from loss of vegetation cover and fertile topsoil, leaving bare stones
behind. This is visible from the thick mass of soil taken away by major rivers such as Abay (the
Blue Nile). The soil carried away by the river is partly the result of people’s style of resource
management in the upper catchments. The damage is clearly observable during the rainy sea-
son as the Abay river crosses over Lake Tana. The lake is coloured (brownish) with soils brought
down by Abay and its tributaries, showing the action of human agency along the catchment
like a mirror. In the south, the Rift Valley lakes with closed basins are threatened by siltation
due to the cultivation of hills and mountains in the catchments. One good example is Lake
Abaya with a total area of 1,070 square kilometres that is virtually coloured by reddish soil
from the drainage area of 14,487 square kilometres (FDRE, 1997). Lake Chamo that is separat-
ed from Lake Abaya by a small piece of land commonly called a ‘natural bridge’ does not show
a similar level of colour change 1. Gullies and the total removal of the soil body are widely
observed phenomena on roadsides and riverbanks in many parts of the country. The other
clear evidence of soil erosion, if not the rate, is the siltation of Koka Dam, an important source
of hydroelectric power in the country. The dam was built in 1960 and now faces difficulties to
operate regularly due to siltation (Ministry of Water Resources, 1999). Power interruptions
during lower rainfall periods force the electric power authority to ration electric supplies by
zones.
Cereal production in Ethiopia averaged 5.7 million tonnes over the period 1980 to 1997. This
production was obtained from 4.8 million ha of land (with a yield of 1.19 tonnes per hectare).
Over the same period, cereal production increased by only 0.9% with a 0.3% increase in yield
per annum after 1994, due to an improved policy environment, increased availability of inputs
such as fertiliser and relatively good weather. However, the achievement of the mid-1990s was
not maintained due to bad weather in the subsequent years (Mulat, 1999).
While the country has a very large livestock population, for which it is said to rank first in
Africa and tenth in the World, performance of the sector remains poor. For instance, beef and
milk yields are 120 and 350 kg per animal per year, respectively (Jabbar, et al., 2000).
Performance of the sector is limited by a lack of feed and by animal diseases.
Forested land is said to be declining in Ethiopia at a rate of 6% per annum, i.e., between
150,000 to 200,000 ha per annum. At this rate the country’s forest resource will vanish with-
in 15 to 20 years (EFAP, Vol. II; 1994). An extensive crop production system is increasingly
threatening hillsides, mountains and grazing areas, aggravating the decline of vegetation
cover under different land-use systems. For instance, a study by Tolcha (1991) in the eastern
part of Ethiopia shows that 81.3 percent of visible erosion was due to human factors, whereas
the remainder was attributed to natural causes. Deforestation is evident from a shift of pri-
mary forest to secondary vegetation that is seen in the remaining hillsides and mountains,
with the exception of limited protected forests. Cultivated areas in the steep slopes are quick-
ly exposed to water erosion. This is even more serious in the areas where relevant soil and
water conservation practices are missing.
According to recent studies, centuries of dependence on agriculture and forestry for food

1 These differences are interesting to investigate in case there are any other reasons.

22
.tesfaye.thesis 29-01-2003 15:12 Pagina 23

and energy supplies has exposed the country to severe soil erosion. The country’s forested
land base has fallen from the reported 40 percent to less than 3.5 percent within about a cen-
tury (Aklog Laike, 1990; cited in FDRE, 1997). The estimate of 40% is far from being true accord-
ing to Ritler (1997) who based his argument mainly on traveller’s accounts from 1699-1865.
Agricultural and rural development in Ethiopia, although claiming that it includes
people’s participation, remain delivery-oriented in terms of its extension service rather than
encouraging farmers’ innovations. Extension systems formulate and promote ‘on-the-shelf’
package technologies that result from past research. The operation of the national extension
approach is modelled after the Sasakawa-Global 2000 extension approach. The overall organi-
sational framework of the system is designed on the basis of a locally developed extension
approach called the Participatory Demonstration and Training Extension System (PADETES)
(Tesfaye, 1999). Even though much is still desired on the technical side, methodological short-
comings can explain the limited success of one of the largest conservation programmes in
Africa over the last two decades (Bekele-Tesemma, 1997).
With respect to the country’s research system, the research agenda has been identified by
researchers and approved through peer reviews. The farmers’ roles have been and still are very
passive in technology generation. A new direction of change in this regard is found in a series
of research strategies prepared by the newly reorganised Ethiopian Agricultural Research
Organisation (EARO) over the last five years. In addition to disciplinary areas of agriculture,
EARO’s research strategies include Farmer-Research-Extension Linkages. However, how farm-
ers participate in technology generation is still under consideration.
In the 1980s, a massive soil conservation programme was launched in Ethiopia using the
food-for-work incentive. During that period, 15% of the Ethiopian highlands that required
conservation efforts was covered (Hurni, 1988). In terms of physical works:
• about 600 km of earth and stone bunds were constructed on cultivated lands;
• about 300,000 km of hillside bunds were built for the afforestation of steep slopes;
• about 100,000 ha of hilly land were closed for regeneration of natural vegetation; and
• thousands of tree seedlings were raised in nurseries and transplanted on the afforestation
sites (ibid.).

1.4 Statement of the problem

Different explanations have been given to the low performance of agriculture in Ethiopia.
Commonly mentioned problems are drought, war, pests, insecurity of land tenure, population
pressure, soil erosion, overgrazing, deforestation, lack of efficient rural organisations, stag-
nant technology, distorted economic policies, weak institutional supports, etc.
These explanations often lead to solutions coming from outside the very community that
is facing the multitude of problems. The community’s indigenous knowledge on resource
management, local institutions and coping mechanisms were not given any attention.
Instead, the methodological approach used is the transfer of technology that suits research
and extension agencies. The introduction of a soil and water conservation programme is no
exception to this.
In the early 1980s, the Soil Conservation Research Project (SCRP) initiated soil erosion con-
trolled experiments in representative agro-ecological locations in the country. The overall
objective of the SCRP was to support soil and water conservation activities in the country. Its
specific objectives were:
• to monitor soil erosion damage, soil loss and runoff, catchment sediment loss and runoff,

23
.tesfaye.thesis 29-01-2003 15:12 Pagina 24

land use and production, and human and livestock parameters in seven research units rep-
resenting different agro-ecological zones 2;
• to develop viable models of soil loss, runoff, catchment runoff, and sediment and produc-
tivity losses for the research areas, and to test their applicability for larger areas;
• to develop ecologically sound, economically viable, and socially acceptable conservation
measures and approaches in different research regions; and
• to train project personnel and to have research fellows to improve the country’s research
and implementation capability in soil conservation.

The research activities of the project were more or less similar in most of the research units.
The SCRP catchments and surrounding areas were conserved through mass mobilisation with
food-for-work and other incentives. The SCRP project was decentralised in 1996 and finally
phased out in 1998. The research sites are now managed by regional research systems.
This study revisits the way the project introduced its conservation practices, assesses
changes in farmers’ land-management practices both within the SCRP catchments and
surrounding areas, notes the time taken for any observable changes in land-management
practices, and draws lessons at the national level for initiating technical interventions.
Insights into these issues are of paramount importance for the country’s research and exten-
sion systems in promoting sustainable land-management practices.
The other focus of this study is on indigenous soil and water conservation methods.
Farmers in Ethiopia have developed a variety of soil and water conservation practices over
time. Unfortunately, adequate studies have not been made on indigenous soil and water con-
servation methods in order to design sustainable land-management practices, particularly
with respect to the role of social institutions in land management. This study deals with pre-
viously overlooked issues in indigenous soil and water conservation in Ethiopia. The focuses
of this study are therefore to understand the foundation of indigenous soil and water conser-
vation and to assess the current situation and bottlenecks. This aspect of the study provides
insights into the shortcomings of the technical interventions that have characterised formal
soil and water conservation initiatives in Ethiopia over the last three decades.
This research deals with a range of factors that are likely to influence farmers’ decisions on
soil and water conservation. These include: social, economic, institutional, biophysical and
technological issues. The study employs different levels of analysis, as deemed necessary;
these are the plot, farm, household and watershed levels.
In the end, active soil and water conservation is the outcome of farmers’ decisions and
actions. Change in the current rather disastrous trends in soil erosion in Ethiopia can only
come about from changes in land use by farmers. This land use is not easy to affect by regu-
lations, subsidies and other impositions. Land use can only be changed on a voluntary basis.
Farmers must understand and profit from the changes. Therefore, it is of paramount impor-
tance to understand farmers’ soil and water conservation practices.

1.5 Objectives of the study

The overall objective of this study is to examine the land management behaviour of Ethiopian
farmers with specific reference to soil and water conservation, by analysing individual and
collective reasons for using indigenous technical knowledge and/or technologies from formal
research institutions.

2 Six out of seven research units are located in the present Ethiopia while one of the units is located in Eritrea.

24
.tesfaye.thesis 29-01-2003 15:12 Pagina 25

The specific objectives are:

1. to assess farmers’ response to SWC introduced by the SCRP;
2. to understand farmers’ knowledge and attitudes with respect to soil erosion and SWC;
3. to identify farmers’ SWC practices;
4. to identify determinants of SWC among small farmers within and across agro-ecological
and socio-economic environments; and
5. to identify alternative approaches and strategies to promote improved soil and water con-
servation practices.

1.6 Research questions

An understanding of farmers’ land management practices in general and that of soil and
water conservation in particular requires a holistic approach, rather than a reductionistic dis-
cipline-oriented research study. This approach is combined with survey research to develop an
understanding of farmers’ land-management practices at the household and watershed levels.
Each farm household has its objectives and opportunities. Attempts to achieve the objectives
involve a balance between opportunities and constraints. Thus, farmers’ perception of
resources, natural and social environments, and their specific land-management knowledge
and skills need to be examined to understand their management decisions. The present land-
scape is the result of the past pattern of resource utilisation that was found to be socially
desirable and appropriate according to the prevailing level of know-how. Repairing the dam-
age done certainly requires an understanding of the root causes of the problem, their dynam-
ics and determinants.
With this in mind, this study tries to reveal farmers’ soil and water conservation practices
in a broader land-management perspective. The analysis is made in a historical context to help
understand the current land-management practices and their problems. It deals with both
indigenous and introduced soil and water conservation technologies. This approach was
adopted first because of variations of farmers’ practices in both indigenous and introduced
practices. And secondly, such an approach gives the opportunity to examine the dynamics of
farming systems.
The specific research questions dealt with are:
1. What are farmers’ responses to the SWC initiatives introduced by the SCRP?
2. What are the farmers’ knowledge and attitudes on soil erosion and SWC?
3. What do farmers do to conserve soil and water on their farms?
4. What are the determinants of SWC among small farmers within and across socio-economic
and agro-ecological environments?
5. How can constraints to promote SWC practices be overcome?

1.7 Relevance of the study

Formal and organised intervention in soil and water conservation in Ethiopia dates back to
the 1970s. This was mainly triggered by the famine relief operation from 1973 onwards, which
was further intensified by another famine in 1984. The focus of intervention over the last
three decades was mainly technical without due consideration of social, economic and insti-
tutional factors. This study, which focused on an understanding of farmers’ views, is of
paramount importance to the country’s natural resource management.

25
.tesfaye.thesis 29-01-2003 15:12 Pagina 26

With respect to research and extension, among others, this study comes up with a suggestion
for a strong farmers’ participatory framework that will guarantee a technology generation and
utilisation process that takes into account farmers’ knowledge, attitudes and practices that
ultimately determine the kind of land-use techniques that will be adopted. Thus, understand-
ing farmers’ current land-management practices is a good entry point.

1.8 Organisation of the thesis

This thesis argues that farmers’ land management, as a concrete behaviour, is a result of the
interplay of several socio-economic, institutional and technical variables, which are also affect-
ed by emergent behaviour from these interactions. Thus, a better understanding of land man-
agement can only be achieved if one tries to understand the interrelationships among these
variables. Furthermore, this study argues that entry points and actions for improvement
become clearer with such a holistic approach.
Chapter two provides the physical setting for this study. It includes a discussion of the rela-
tionship between the physical environment of Ethiopia such as its land features and rainfall
that are closely related to forces of erosion, farming systems, social and institutional factors
such as population, rural institutions, land tenure, research and extension organisations and
environmental policies.
Chapter three deals with the theoretical and analytical framework of this study. The determi-
nants of land management are complex and manifest themselves at various levels and among
various actors, often with different responses. Owing to this, the study draws on various theo-
retical perspectives and concepts to explain field observations reported in Chapters 5 to 7. The
theoretical perspectives are land degradation and intensification, planned change and devel-
opment intervention, farming systems, indigenous knowledge, social learning, and an attitude-
behaviour model. Following that, the chapter presents empirical studies and an analytical
framework.
Chapter four presents the general methodology employed in the study, while each case study
also has its own detailed methodology section. Issues dealt in this chapter include; the selec-
tion of study units, methods of data collection and analysis.
Chapter five presents a case study of indigenous soil and water conservation practices from
Konso, in southwestern Ethiopia. A discussion about indigenous SWC methods is by no means
unique to Konso and has taken place in other case study areas as well. The Konso case is how-
ever here given special emphasis due to its widespread use of indigenous SWC methods that
have existed for over four centuries. This thesis uncovers the historical background of the sys-
tem, its knowledge basis, institutions that facilitate a process of innovation and mechanisms of
social learning. Furthermore, critical arguments as to whether it is possible to adapt Konso’s
case to other areas, as well as a description of its current situation and future prospect are dis-
cussed.
Chapter six and seven are intervention cases that deal with the Soil Conservation Research
Project (SCRP) at Wolaita and Wello, in southern and northern Ethiopia, respectively. Here the
focus is on how farmers have responded to the interventions introduced by the SCRP. The chap-
ters provide a biophysical and socio-economic background of the area, followed by an in-depth
analysis of farmers’ knowledge about, perceptions on and practices for SWC. The chapters end
by discussing farmers’ responses to the SCRP intervention.
Chapter eight presents a synthesis of the case studies. It addresses the major lessons learned
and identifies the determinants of SWC at the micro and macro levels. Finally, the chapter clos-
es the thesis by presenting issues for further research.

26
.tesfaye.thesis 29-01-2003 15:12 Pagina 27

2 Context of the Study

2.1 Introduction

The purpose of this chapter is to provide a contextual setting for the present study. Section 2.2
deals with the physical environments of Ethiopia where its topography, climate, soils, vegeta-
tion and agro-ecological zones are briefly discussed. Sections 2.3 and 2.4 respectively deal with
population pressure and rural organisations. Following that, the chapter dwells on farming
systems and technology in section 2.5. Thereafter, section 2.6 presents a historical profile of
land tenure in the country. Sections 2.7 and 2.8 present an overview of agricultural research
and extension in Ethiopia. The final section (2.9) discusses environmental policies in Ethiopia.

2.2 Physical environment of Ethiopia

2.2.1 Topography

Ethiopia covers a geographic area of about 1,223 million square kilometres (Ethiopia, 1992).
The country is characterised by diverse topographic features. It is noted for its high moun-
tains, a few of which exceed 4,000 masl. For instance, Ras Dashen is 4,620 masl. The largest
share of the country’s land falls under what is often called the ‘lowlands’ (55%), whereas the
remaining area falls under the ‘highlands’. The demarcation line between the lowlands and
highlands is 1,500 masl (FDRE, 1997a). The highlands are divided into two parts by the Rift
Valley, which runs from the northeast towards the southwest. More than 80% of Ethiopia’s
human population and about 60% of its livestock population depend on the highlands, which
are suitable for agriculture and human settlement.

27
.tesfaye.thesis 29-01-2003 15:12 Pagina 28

Four topographic regions are recognised in Ethiopia (WCMC, 1991). These are Wurch (2,900 to
3,500), Dega (2,500 to 2,900 masl), Weyandega (1,800 to 2,400 masl) and Kola (below 1,800 masl).
The terrain is characterised by an undulating landscape, which is extremely vulnerable to
water erosion, particularly when ecological disturbance occurs due to removal of vegetation.
According to Constable (1985), 23% of the highlands falls within the slope category of less than
8%, while 18% of the highlands is estimated to have slopes of between 8 and 16%. About 60%
of the highlands have slopes that exceed 16%, while 34% of the area exhibits slopes that exceed
30%. The distribution of slopes in the three highland agro-ecological zones is given in table 2.1
below.

Table 2.1: Slope and area coverage in highland agro-ecological zones

Zone Slope (%) (‘000) km2 Percent of total

High Potential Perennial 0-8 15 10.4

8-16 31 21.5
16-30 39 27.1
>30 59 41.9

Total 144 100

High Potential Cereal 0-8 44 33.5

8-16 22 16.8
16-30 26 19.8
>30 39 29.8

Total 131 100

Low Potential Cereal 0-8 38 25.3

8-16 25 16.7
16-30 42 28.0
>30 45 30.0

Total 150 100

Source: Constable, M. (1985). Ethiopian Highlands Reclamation Study (summary). Working paper 24. Land Use Planning and
Regulatory Department, Ministry of Agriculture and FAO, Addis Ababa.

2.2.2 Climate

Climate is a crucial physical factor that determines land-management practices for agricul-
ture. Farming systems that prevail in Ethiopia are partly the result of climatic situations that
affect the soil fertility, vegetation and length of the growing period. The climate in Ethiopia
is highly determined by the topographic conditions of the country (Westphal, 1975; McCann,
1995). This is reflected in its temperature and rainfall quantity and distribution. The average
annual temperature varies between 20 and 29 degrees centigrade. Rainfall ranges from 200
mm per year up to 2,700 mm for the southwestern part of the country (FDRE, 1997a). The aver-
age rainfall decreases from the southwestern margins of the Ethiopian highlands eastwards
to the Horn of Africa (Westphal, 1974).

28
.tesfaye.thesis 29-01-2003 15:12 Pagina 29

Following rainfall and temperature patterns, four traditional seasons are known in the
country. These are: Kiremt, Bega, Belg and Tsedey. Kiremt is the wet season (June-August/
September), while Bega is the dry season (October-February). March to May (Belg) is the short
rainy season in the north and the central highlands, while it is the main season for some areas
in the south. The distribution of rain between Kiremt and Belg creates a bi-modal rainfall pat-
tern in the country. Unlike the temperature, its rainfall shows high variability. Rainfall vari-
ability in both quantitative and temporal distribution is one of the primary agricultural prob-
lems reported in all corners of the country.

2.2.3 Soils

Ethiopia’s soils are formed from three types of parent rock (Last, 1962; cited in Westphal,
1974):
• the granites of the crystalline basement which tend to form shallow, sandy soils;
• the volcanic rocks, such as basalts, which tend to produce fertile loams, generally red in
colour, but sometimes black; and
• the limestones and sandstones which form shallow, poor, sandy soils.

According to the assessment so far, there are 18 soil type associates in Ethiopian. The major
ones are: Lithosols, Nitosols, Cambisols, Regosols, Vertisols, Fluvisols, Xerosols, and Acrisols.
Among these Nitosols, Cambisols and Vertisols account for 23%, 19% and 18% of arable lands
respectively. It is noted that the exact characteristics and distribution of the Ethiopian soils
are still not fully documented and mapped (EARO, 2000a).

2.2.4 Vegetation

Vegetation is one of the biological features resulting from the integrated forces of topography
and climate. Speaking from a biophysical point of view, the Kola zone is characterised by
steppe, savanna and woodlands, while dense forests occupy the Weynadega zone. Savanna and
steppe formation occurs in the lower part of Dega too. The Wurch zone is covered by montane
forests and alpines (WCMC, 1991). Apart from the Wurch zone, which has massive, high moun-
tains and a few extremely inaccessible areas, the rest of the country has changed to a cultur-
al landscape. This is because of human interference in the form of parks, livestock grazing,
bush fires, and farming (commercial or smallholder) that has greatly changed the natural veg-
etation.

2.2.5 Agro-ecological zones

The interaction between biophysical and socio-economic factors has resulted in a classifica-
tion of five different agro-ecological zones in Ethiopia (Constable, 1985). These are: the High
Potential Perennial (HPP), the High Potential Cereal (HPC), the Low Potential Cereal (LPC),
Agro-pastoral (AP) and the Pastoral (P) zones. Zones with perennial and cereal crops fall in the
highlands, whereas, the two forms of pastoral systems fall in the lowlands. It should be noted
that this classification is based on an assessment in the early 1980s.

29
.tesfaye.thesis 29-01-2003 15:12 Pagina 30

This study covers three of the above agro-ecological zones; Wello, in the LPC zone of the
northern highlands; Wolaita, the HPP of the southern highlands; and Konso in the AP area
(see Map 1). This last area is better described as seed-hoe-terrace, after Westphal (1975).

Map1: Location of the study areas in Ethiopia

South Wello site

Addis Ababa

Wolaita site

Konso site

2.3 Population pressure

A record of the Ethiopian population first appeared in the early 1960s. Before the first nation-
al population and housing census in May 1984, population estimates were based on sample
surveys. The second national population and housing survey was carried out in 1994. The
existing data show rapid population growth during the last three decades.
In 1960, Ethiopia’s population was estimated at 23.5 million, with a 2.0% annual rate of
growth. In 1984, the census data revealed that Ethiopia’s population numbered 42.2 million,
with an annual growth rate of 2.9%. Since then the growing rate of Ethiopia’s population
increase has attracted the attention of policy-makers and the international community. In
1994, the population reached 53.5 million at an annual growth rate in excess of 3.0% (CSA,
1998). The country’s population is expected to double after 25 years, based on the 1992 rate of
increase (about 3%). On the same basis, the population of the country is projected to reach 94
million and 140 million in the years 2010 and 2025, respectively (Jones, 1990).
During the time of the monarchy (before 1974), population pressure was not explicitly asso-
ciated with resource degradation. However, there were considerable initiatives in environ-

30
.tesfaye.thesis 29-01-2003 15:12 Pagina 31

mental conservation and protection. These were achieved through the establishment of dif-
ferent authorities for forestry and wildlife. This has by and large served to increase
government revenue at the cost of rural people who are denied access to the resources in the
national parks and state forest. The issues of soil erosion and food security are not included
in the primary agenda for development. During the reign of Haile Sellassie, tax incentives
were given to bring more land under cultivation (Pankhurst, 1966).
An appreciation of a large human and livestock population is deeply-rooted in the value sys-
tem of Ethiopians. The previous acceptance of a large human population can be ascribed to
continuous regional tribal conflicts, which more often than not led to bloody conflicts. In
addition, high human mortality has also contributed to the attitude favouring large numbers
of children who are often seen as both assets and insurance during old age and times of
calamity. Besides being one of the essential resource bases for survival, livestock is also seen
as a sign of social status. Animals are used in cultural and religious ceremonies, and as a
means of exchange and settlement of social and state commitments. These value systems have
been reinforced because of risks of animal disease, in addition to man-made and natural dis-
asters.
The image of population among state and scholars began to change in the 1970s and 1980s.
Population growth came to be seen as one of the major causes of land degradation in Ethiopia.
This theory subscribes to the Malthusian school of thought and claims that population
growth causes an increased demand for fuelwood, crop- and grazing lands, which cause clear-
ing of trees for cultivation (see 3.2.2 for details). This problem is said to be reinforced by aggra-
vated erosion, which is created by the removal of the vegetative cover and crop residues, and
the diversion of dung that apparently reduces land productivity. This view is widespread in
the government reports (EFPA, 1994; FDRE, 1997a). For instance, EFAP (1994) states that ‘… pop-
ulation density in the HPC zone will increase from a 1985 level of 62 to 166 persons/km2 in
2015. The livestock population is also growing at 1.1 percent per annum, resulting in serious
overgrazing…’
Paradoxically, this conviction did not lead to a family planning campaign to tackle the per-
ceived root cause of resource degradation; i.e., population pressure. Instead, a massive physi-
cal soil and water conservation and tree-planting programme was carried out in the 1980s,
which is the technical option. However, a state-led resettlement programme that involved
200,000 families, with an average of three people per family of settlers took place after the
mid-1980s (Pankhurst, 1990). In 1994, the government issued a population policy, which
targets a decrease in human fertility rates. According to the policy, it is envisaged to reduce
fertility from 7.5 to 4%. It is worth noting that the existing population is enough to intensify
the resource pressure for a few decades to come.

2.4 Rural organisations

Discourse on the impacts of rural organisations dominates the post-monarchical era of

Ethiopia. These works revolve around organisations created by the Derg in its pursuit of social-
ist transformation, i.e., the Peasant Association (PA), Service Co-operatives (SC) and Producers’
Co-operatives (PC). There are several studies, which have focused on the evolution of these
institutions in the country and their impacts on the agrarian economy (Dessalegn; 1984, 1990;
Mengistu, 1986; Aster, 1982). Thus, it suffices to mention only the most relevant issues to the
present study.

31
.tesfaye.thesis 29-01-2003 15:12 Pagina 32

Peasant Associations (PAs) were established soon after land reforms in March 1975. One PA
was organised on a territory of approximately 800 ha. PAs were charged with administrative
matters under their jurisdiction, including that of land distribution. Staff of the Ministry of
Agriculture provided technical backstopping to PAs. Generally, PAs were given the authority
to reinforce law and order at the grassroots level.
During the Derg period, even though members elected PA leaders, the latter maintained
their loyalty to the government bodies rather than to the farmers. As a result, they continued
to implement unpopular government programmes such as villagisation, military services,
forced resettlement, tax levies and contributions from the peasants. Thus, they turned into a
puppet organisation instead of a means to empower farmers and ensure self-governance. This
gap still persists, leaving much to be desired for popular participation.
With the intention of increasing access to household merchandise and farm inputs to the
rural people, the government also established Service Co-operatives (SC), which encompass a
minimum of four adjacent PAs. Service co-operatives have benefited the rural people through
the provision of small commodities at relatively accessible locations. However, they later
began to serve the desires of the government too. For instance they implemented an unpopu-
lar, quota-based grain purchasing system that was operated by a government organisation
called the Agricultural Marketing Co-operation (AMC). Apart from that, they embezzled mil-
lions of poor peasants’ capital in the process. The AMC has long been dissolved as a result of
market liberalisation measures.
As the Derg Regime consolidated its power, it entered into the mainstream socialist econo-
my. In 1979, it began what it called the Arenguade Yemirt Zemech, which means, ‘campaign for
the Green Revolution’. According to the appraisal of the Derg policy-makers, their dream of
social transformation could not be realised using the peasant economy, which uses ‘ineffi-
cient technologies.’ As a result, the state was committed to initiate Producers’ Co-operatives
(PC) and to support them with modern technology, credit and extension. Thereafter, PCs
began to emerge here and there in the country after a series of agitation measures by the
experts of the MoA and a special cadre who were trained in politics and technical fields, name-
ly the Political and Production Cadre. PCs were established under the jurisdiction of the PAs.
A minimum of five farmers could form a PC. Members joined all their means of production,
including land and oxen. Distribution was based on work points according to the socialist
principle, ‘from each according to his ability, to each according to his work’. For this purpose,
activities were classified and work points assigned, which later served as a basis for distribu-
tion of benefits.
Inputs such as fertiliser, improved seeds and machinery were channelled into the co-oper-
atives. Paradoxically, good land in particular irrigable land was automatically allocated to the
co-operative organisation. This created deep resentment among the peasants who did not see
the benefit of co-operatives and were not willing to join them. This development began to cast
doubt and uncertainty that replaced the jubilation and joy of the early days of the land
reforms.
PCs were disbanded soon after the May 1990 reforms. The Service Co-operatives met the
same fate almost within a year after the downfall of the military regime in May 1991.
The PAs were reorganised immediately after the new government maintained law and
order. Reorganisation of the PAs in the rural areas was part and parcel of the peace and sta-
bility maintenance plan of the new government.
A similar effort was made to reorganise Service Co-operatives, but the impact was not
uniform like that of the PAs. New legislation to reintroduce Producers’ Co-operatives was pro-
mulgated in January 1994. Recently, a separate body that organises co-operatives has been

32
.tesfaye.thesis 29-01-2003 15:12 Pagina 33

established. However, progress in this regard is negligible and continues to be so until farmers
see the benefit of joining hands for their own development. It should be noted that the rural
people are not against co-operation per se, which they have known throughout their lives. The
resistance emanates from the manner in which these rural organisations were politicised and
how they served the state at the expense of the rural people.
From the foregoing discussion, it is clear that there is a lack of genuine popular participa-
tion in the development processes. Rural people were plagued by a multitude of institutional
problems that retarded, the installment of more adaptive resource management.

2.5 Farming systems and technology

The development of agriculture in Ethiopia has followed different patterns. In some regions,
mixed crop and livestock production is followed while in others pastoralism dominates.
The northern highlands of Ethiopia had developed into a mixed crop-livestock production
system well before the Axumite Kingdom (100BC-1000AD). Cattle production preceded crop
cultivation (de Contesion, 1981, cited in Tewelde Brehan, 1989). Development of the ox-plow is
central to the mixed crop and livestock production systems in Ethiopia. According to archae-
ological evidence (McCann, 1995), the likely time for the introduction of ox-plough based agri-
culture in northern Ethiopia is the first millennium B.C.
For centuries, Ethiopian farmers have conserved (in situ) several crop varieties that they
identified and domesticated. Owing to their efforts, the country is now one of the 12 Vavilov
centres of crop genetic diversity in the world (Dole, 1985; cited in Towelde Brehan, 1989). This
successful history includes a few crops that are used as food crops only in Ethiopia; i.e., teff
(Ergrostis tef), enset (Ensete ventricosum) and the root crop, anchote (Coccinia abyssinica).
The present landscape of Ethiopia is the result of farming systems developed in different
parts of the country. A farming system and how people manage their natural resources are
determined by: their principal means of livelihood; biophysical conditions; the degree of inte-
gration between crop and livestock production systems; the level of technology in crop
production; types of crops grown; species of animals raised; customs and culture of people;
settlement pattern; values and belief systems (religious/spiritual); social status and stratifica-
tion; political system, etc. These factors may have positive or negative impacts on the natural
resources that need close examination based on historical events. The major features of the
farming systems in Ethiopia are discussed below.

2.5.1 Farming systems of Ethiopia

According to Westphal (1975), the agricultural system of Ethiopia can be broadly classified
into four types: the seed-farming complex, the enset-planting complex, shifting cultivation,
and the pastoral complex. Agro-ecological distributions, main crops, tools and farming tech-
niques in each of these systems are briefly discussed below.

The seed-farming complex

The seed-farming complex covers the eastern, central and northern highlands, as well as
Konso and adjacent Weredas. After the work of the Ethiopian Highlands Reclamation Study
(EHRS), the seed-farming complex was classified into two zones: the low-potential cereal zone
(LPC) and the high-potential cereal zone (HPC). Most crops are grown from seed. Cereals, puls-

33
.tesfaye.thesis 29-01-2003 15:12 Pagina 34

es, oil crops and, to a much lesser extent, tubers are important crops. The northern and
central highlands are ox-plough dominated, whereas the eastern highlands and Konso use a
combination of plough and hoe cultivation. The use of oxen is limited to the gently sloping
area of Konso.
One remarkable issue in this farming system is a lack of attention to tuber crops, vegeta-
bles, trees and shrubs, including fruit trees on the farm, mainly in the central and northern
highlands (e.g., Wello, Gojjam, Gondar, Shoa and Tigray). As compared to the time of Westphal
(1975), vegetable production is expanding into regions that have access to markets for both
inputs and outputs. In the Eastern and Konso-sorghum complex, trees are well integrated with
cereal production. In addition, root crops came to be widespread in Konso, whereas vegetables
are a typical feature of eastern Harargie. Owing to this, the farming system in this sub-com-
plex evolved to controlled livestock production. Thus, livestock either grazes on a controlled
basis or is stall-feed. This stage of crop-livestock integration is very favourable for soil and
water conservation measures. Unlike other parts of the farming system, biological or physical
conservation structures in these systems are not ruined by roaming animals during aftermath
grazing. The north and central highlands lack fencing in the outer fields. After harvesting, all
farmlands turn into common property as far as grazing and collection of dung is concerned.

Enset-planting complex
The Enset-planting (Ensete ventricosum) complex is located in the southwest of the country,
between 1,600-3,000 masl (Westphal, 1975). Nowadays, coverage of enset is increasing its geo-
graphical range. This farming system includes perennial crops such as coffee and chat (Catha
edulis) in addition to root crops. Enset is cultivated as a staple food or as a supplement to cere-
als and tuber crops. In some regions, it occurs in dense forest where tuber crops dominate,
with of secondary importance, cereal production practised as shifting cultivation. According
to Constable (1985), the HPP zone and the HPP zone-forest areas are located in this complex.
Hoes and similar hand tools are common in this farming system. However, the ox-plough is
also widespread in this zone, used in cereal and pulse plots. Exclusion of livestock from the
crop field is stricter in this system.

Shifting cultivation
Westphal (1975) classified a few areas of the country under shifting cultivation. These areas
are now located in the Benshangul-Gumuz Regional State in the west and in the Southern
Region. Nowadays, shifting cultivation is practised in the Gambella Region and a few pockets
in the southwestern enset-planting complex. Apart from this, there is hardly any room for
shifting cultivation in the country. Crops grown vary slightly from region to region, depend-
ing on food habits and the climate. In general, cereals, pulses, tubers, oil crops and spices are
grown in this system. The major tools are stick and hoe. According to Constable (1985), shift-
ing cultivators practise an agro-pastoral production system.

The pastoral complex

Livestock is essential to all farming systems. In the seed-farming complex, cattle are raised
mainly to ensure the availability of draught power. In the enset-planting complex, cattle are
raised to ensure the availability of manure.
In some communities combinations of socio-political and agro-ecological factors forced
them to add crop production. Those who see their comparative advantage in livestock have
continued with a pastoralist mode of production. Thus, both agro-pastoral and pastoral sys-
tems dominate different altitude zones of the country, mainly in the lowlands. Expanding

34
.tesfaye.thesis 29-01-2003 15:12 Pagina 35

croplands have a limited mobility of livestock between different altitudinal pastoral zones
(Westphal, 1974). Socio-political, institutional and technological changes that took place dur-
ing the last three decades, since the study by Westphal (1974,75) only mean the worsening of
the mobility situation for pastoralists, which intensifies degradation of grazing lands.

2.5.2 The technology

Technology is not always a product of scientific institutions. Human beings are inherently
capable of modifying their environment in the process of adaptation, whereby technology is
created and subsequently utilised. The struggle between the environment and people never
stops, though under some circumstances, a long time may pass before intended changes are
realised. For various reasons, some societies adhere to certain technology for centuries where-
as others pass a comparable level of technology in a relatively shorter period of time. For
instance, the revolution of farming tools for different operations in developed countries such
as The Netherlands and the stagnation of the same in a developing country such as Ethiopia
explains this observation.
Livestock and crop production techniques of Ethiopian peasants within the four farming
systems exhibit an enormous amount of technology. Key features of each farming system are
the result of continuous adaptation by the people. The selected tools and techniques of farm-
ing are in accordance with the biophysical, socio-economic and institutional environments in
which they operate (topography, climate, land size, market, and land tenure).
Livestock production has not shown any appreciable degree of innovation. However, con-
tinuous change was observed in crop production systems in both north and south. The major
breakthrough was the expansion of the ox-plough over 2,000 years ago. The conversion of pas-
ture and forested lands to crop production, induced by population growth, resulted in a
decline of wetlands and natural forest.
Agriculture in Ethiopia is a rainfed system. Water harvesting techniques are limited to the
use of ponds for livestock drinking during part of the dry period. Crop failures in some areas
occur in spite of very accessible rivers or lakes. It is paradoxical that there is very limited
irrigated agriculture in a country with abundant water resources. Irrigation is limited to river-
banks and on a very small scale. However limited that may be, irrigation structures in the
Hararghie highlands and Konso are very impressive. After the 1960s, expansion of small-scale
commercial farms in the Rift Valley made the extensive utilisation of the Awash River as
compared to other rivers of the country, possible. Nowadays, better practices of irrigated agri-
culture have developed in many parts of the country due to the increasing opportunities for
fruit and vegetable cultivation.
In summary, this study attempts to provide the dynamics of the farming system with a
focus on the current land-management practices, more specifically with respect to soil and
water conservation.

2.6 Land tenure

Land tenure can be conceptualised as a mode of institutionalisation of claims by human agen-

cy - in terms of private, common and public - on land and its embodied resources (Sfeir-Younis
and Dragun, 1993). A land claim witnesses far-reaching consequences at various levels, rang-
ing from family to country to continental levels.

35
.tesfaye.thesis 29-01-2003 15:12 Pagina 36

In Ethiopia, land tenure has been a contested issue for successive generations under differ-
ent political systems. While the land question is as old as the country’s civilisation itself, the
predominant analytical approach adopted in studying land tenure is an economic one. The
underpinning theory is the structure of property right regimes, which falls within the
domain of institutional economics. According to this theory, property rights create ranges of
incentives for efficient land resource utilisation (Feder and Feeny, 1991; Eggertsson 1990:125-
128). This debate fails to appreciate the dynamics of the issue that has evolved from social, eco-
nomic, institutional and political environments. Land-tenure systems should be viewed with-
in the historical context of each society, showing specific stages of societal development.
In Ethiopia, land continues to play a crucial role in the society’s spiritual, economic, polit-
ical and institutional development. During the time of the monarchy, land was a centre for
power relations between the citizens and their leaders, governors and the kings. In short, it
was a vital political instrument. Different nationalities in Ethiopia had their own land-tenure
system. It can be said that land was and still is both the foundation and the pillar of the econ-
omy. It has served as a means of transaction in power relations in the community. During the
Derg Regime, the use of land as a political instrument has by and large shifted to its use as an
instrument of economic development based on socialist principles. Thus, the tenant-landlord
relationship was replaced by a state-peasant relationship. Peasants were then confronted with
socialist economic principles and their auxiliary problems. Most of the inhibiting policies
that prevailed during the Derg period were removed after 1991, when a democratic
government came into power, even though state-peasant relations continued.
Thus, an understanding of the land-tenure issue in Ethiopia and its role in the economy
requires a retrospective analysis. Assuming the same role for land tenure during the monar-
chy, the Derg and post-Derg Regimes, imposes a barrier to a better understanding of the con-
cept. Examining this perspective is increasingly important in view of the disproportionate sig-
nificance given to land tenure in agricultural production and resource degradation in
Ethiopia by many scholars (Merid, 1986; Amare, 1977; Dessaglegn, 1994; Zemenfes, 1995;
Yigremew, 1999; Dejene, 1999). Among them, Merid and Zemenfes in particular make a direct
link between land tenure and land degradation. For instance, Zemenfes (1995) writes:

‘… Land degradation in Ethiopia, therefore, has its roots in the power structure and political economic pro-
cesses that created exploitative forms of property relations, governed the distribution of produce, and reg-
ulated access to resources, specially land.’

The unique way in which the Ethiopian Empire was built by itself leaves much to be desired
in understanding the nature of farming systems in different parts of the present Ethiopia.
This entails an assessment of the indigenous farming system in the north and among differ-
ent nationalities in the south before formation of present-day Ethiopia. With the exception of
a few (e.g., Sandford, 1994; Takele, et al., 1994; Watson, 1998; Terefe, 2001), many studies on
land tenure do not give enough attention to the fact that there were independent land-tenure
systems in different parts of the country.
In what follows, an overview of land-tenure systems in the country is presented in four dif-
ferent periods. These are: pre-incorporation of the south; post-incorporation; during the Derg
Regime; and post-Derg. These periods are used in the case studies to show the dynamics of
tenure relations in Ethiopia. Moreover, it is intended to examine the constraints and oppor-
tunities of intensive land management in relation to the general political climate and the
main features of the land-tenure system.

36
.tesfaye.thesis 29-01-2003 15:12 Pagina 37

2.6.1 Pre-incorporation period (before 1875)

During the pre-incorporation period, land tenure in northern Ethiopia and the newly incor-
porated regions was different. Therefore, the pre-incorporation phase is discussed separately.

Land tenure in the north during the pre-incorporation period

During the Axumite kingdom (100 BC-1000AD), the tradition was to subjugate any encoun-
tered society by war and to impose a tribute on their resources or other valuable items of the
time such as ivory, incense, cattle, etc.
Even though there was no evidence for a clear land-tenure system during the Axumite king-
dom, after the adoption of Christianity in the fourth century A.D., different monasteries were
granted land from the state (Pankhurst, 1966). This form of land grant which continued until
the 20th century, marked the association of state and church, and thereby religion and
people. Kings and rulers of different times have exploited their association with the church
through grants and gifts to legitimise their kingdom. Thus, people who believed in the doc-
trine earnestly accepted the absolute power of kings over the land and people. Owing to this,
the church and more specifically, the ecclesiastics formed an important land-holding class
until 1974.
In the course of societal development, consistent structures of property rights emerged in
the north. These are the Rist and Gult land-tenure systems. The traditional form of Rist was
established through the pioneer cultivation of certain territory. It was institutionalised as a
result of handing down of land from father to son on a successive basis. The second form of
Rist was based on a grant from the sovereign. This necessarily involved some kind of specific
reciprocation to the sovereign, either on a temporary or regular basis. This commitment could
also occur in the original form of Rist if deemed necessary by the sovereign, but in most cases,
it was a matter of outlook; i.e., a position given to the kings in relation to land over which he
is said to hold the ultimate power.
Traditional Rist land is usually associated with graveyards of one’s ancestors (Atseme rist).
This hints that the term Rist was in use well before the introduction of Christianity in the
north, where every family used its own land as a graveyard as does the non-Christian south to
date. Under this system, kinship members hold the ownership of land while individuals have
independent user rights. Thus, in terms of property rights, individuals have exclusive rights
to their farms. However, exchange and transfer to non-members was restricted. When a need
to sell arises, selling to non-members may only be accepted if anyone in the group fails to buy
or satisfy the needs of the person who intends to sell the land. The final sale of the land, how-
ever, needs approval by the state or its representatives. Under the Rist system, there is securi-
ty of land within the group, but there is no individual land transfer right. In this way, the tra-
ditional system of land tenure protects against total landlessness, while also being responsive
to members’ needs.
The other form of land tenure that developed with more differentiation of state functions
was the Gult system. The state found it increasingly necessary to compensate the citizens it
had mobilised for peace or war efforts. The recipients were local rulers, members of the royal
family, the nobility or priests. To that extent, Gult is an individual right as opposed to Rist,
which is a collective right. With the approval of the sovereign, Gult rights can be converted to
permanent and thereby, hereditary rights. This type of Gult is called Ristegult.
The Gult system gives the holders the right to appropriate dues, which would otherwise
have been collected for the state. The Gult holder may or may not directly participate in the
production process. This distinction is important when the form of due involved could be part

37
.tesfaye.thesis 29-01-2003 15:12 Pagina 38

of the primary products of the land. No physical removal of the land operator is involved,
though this phenomenon was observed in the second half of the twentieth century with the
improvement of markets. The right of the Gult holder was temporary, lasting as long as he
maintained his relationship with the ruling class. Peasant-military, peasant-priests, and a
range of other people receive land from the government in return for different forms of obli-
gations directly related to working the land. These forms of land-based relationships include:
gindebel meret, milimil meret, bernos or zitet meret, etc., (Gebr-wold, 1962) which are differentiat-
ed according to the forms of payments involved.
The Rist and Gult systems in the north were applicable only to the cultivated land (Merid,
1986; Tewolde Brehan, 1989). Mountains, valleys and rivers were therefore open access prop-
erty. Also, trees could be chopped for fuelwood and construction purposes from the allocated
areas, irrespective of their status.
In spite of their potential or actual shortcomings, the Rist, Gult and their modified form of
Ristegult systems all exhibit how society adjusted to the emerging social, political and institu-
tional environments. This is a good indicator for the need to adopt a flexible land-tenure
arrangement.

Land tenure in the south during the pre-incorporation period

There is a lack of a recorded history on the economic, social and political life of people in the
south. However, conflicts over resources for crop production, grazing, hunting, fishing and
trade routes were common. Like in the north, the accumulation of wealth in the form of land
and livestock replaced hunting and gathering practices. In the process, the original settlers
who came to claim the land evolved as spiritual, political and economic figures. Land and live-
stock were the two common properties that were used in reward and punishment systems of
the different communities.
In some societies, land in totality belongs to the ruling body. Use of land has ‘costs’ that
must be borne by the users. For example, in the old Gibe State of the Oromo people, all land
belonged to the king. Similarly, in the Kingdom of Keffa, the king owned the land. In both
cases, the members of the society had access to land following different patterns of relations
with the king and his administration. They could use the tribal or village grazing land and
own livestock (Huntingford, 1955; Cerulli, 1956). In his recent work, Terefe (2001) indicates
that the available information on pre-class Oromo society shows that land was controlled by
the Qomo (clan) during the Gadaa system where equality was pursued. As Terefe also indicated,
there is every reason to expect diversity of land tenure among Oromo people who live in a vast
geographic area, involving a sedentary agriculture and pastoralism.
In Konso, the early settlers who established ownership by lighting fires emerged as politi-
cal figures, landlords and religious leaders (Hallpike, 1972; Ambron, 1984; Watson, 1998). A
detailed discussion on land tenure in Konso is presented in Chapter 5.
Similarly, the Wolita people also had a very strong kingdom, which had its own system of
governance and land administration. In Wolita, the land belonged to the king and the citizens
had access to land through inheritance, different social arrangements and military service,
which was very rewarding in the kingdom (see also Data, 1998).

2.6.2 Post-incorporation until the downfall of the monarchy (1875-1974)

This period, though it covers only one century of a country of people with a long history,
brought far-reaching changes to the country’s social, economic, political and institutional

38
.tesfaye.thesis 29-01-2003 15:12 Pagina 39

aspects. The previous centuries were very important in shaping the country up until this peri-
od. Land tenure itself is the interplay of these complex phenomena during the monarchy.
At the beginning of this period, Rist was the dominant land-tenure system in the north,
whereas Gult dominated in the south. These trends created a dual land-tenure system in the
two major regions, which caused a heated debate during the monarchy.
Several operational names were given to Gult in the south, where three forms of obligations
were applied. These were Siso arash (one-third), Erbo arash (one-fourth) and Ekul arash (one-half).
These calculations were made after the tithe (one-tenth) was deducted. In the case of Ekul
arash, the tenant received oxen and seed from the Gult holder (Pankhurst, 1966; Cohen and
Weintraub, 1975). In addition, people in the south with the exception of local power holders
(chiefs, kings and their associates), were reduced to Gabbar (tax) payers. It should be noted that
the same name, Gabbar, was applied to all taxpayers in the north. However, the forms of rights
held by the two Gabbars were different. In the north, a Gabbar had both physical property
(land) and rights, with the exception of those on Gult land, whose fate was technically similar
to the south, but who socially identified themselves with landlords or the church. In the
south, a Gabbar was patronised because he was made to pay a tribute on what had been his
own land.
All in all, before 1974, land tenure in Ethiopia consisted of semi-feudal tenant-landlord rela-
tions. Tenancy was, however, a typical feature in the south after incorporation into the north.
Lack of land security was one of the major bottlenecks to improving land management and
production during the monarchy. In this connection, Terefe (2001:300) reports that farmers in
the western Shoa zone of the Oromia Region refused to engage in soil conservation efforts
such as contour ploughing, drainage, crop rotation, applying manure/ash on farmland and
leaving crop residues/stubble after the harvest, even though the landlords were pressing ten-
ants for such conservation measures. Such location-specific evidence is useful to understand
the link between land-tenure security and land-management practices among farmers at dif-
ferent periods in history.

2.6.3 During the ‘Derg’ Regime

One of the radical steps taken by the Derg Regime upon seizing power from the monarchy was
to institute land reforms (Negarit Gazette, 75). The intellegentsia had been struggling for this
movement under the slogan Meret larashu! (‘land for the tillers!’).
Land reforms broke the relationship between the tenants and the landlords. The land
reforms proclaimed that a maximum of ten hectares of land should be provided to each farm-
ing family. Any person who owned more than ten hectares was obliged to surrender the extra
land to the PAs for redistribution to other peasants who were short of land.
The proclamation made it clear that land belonged to the state. Peasants were entitled to a
usufruct right. One could not transfer the land in the form of lease or sale. But, one’s family
members could continue to use the land under the same regulation.
The fact that land was designated as state property, the state machinery at the grassroots
level, comprising the PAs and the Wereda administration, was redistributing land as it saw fit.
Furthermore, the PAs had been allocating land for public facilities without any compensation.
Any attempt to compensate one farmer hardly materialised without affecting others. At best,
grazing areas or hillsides were allocated, still violating a claim over the land, collective or indi-
vidual. In some regions, no less than four land distributions were carried out. The cumulative
effects eroded the peasants’ emotional acceptance of the government’s land reforms. This

39
.tesfaye.thesis 29-01-2003 15:12 Pagina 40

phenomenon may have affected the farmers’ land-improvement practices. But, what was the
nature of farmers’ responses as a result of this disincentive regarding specific land-manage-
ment practices? And how does this relate to the farming systems in each region? These ques-
tions and related issues will be examined in this study.
Shortly before its downfall, the Derg Regime introduced socio-economic reforms in May
1990. The reform, among others, included: the use of hired labour, removal of the grain pur-
chase quota, cancellation of villagisation, which inflicted a serious loss of assets among peas-
ants, and the dissolution of Producers’ Co-operatives if members so wished.

2.6.4 After the ‘Derg’

The Derg Regime was overthrown in May 1991 and replaced by a Tigrian-led coalition force,
called the Ethiopian People’s Revolutionary Democratic Front (EPRDF). The constitution of the
new government stipulated that ‘the right to ownership of land is exclusively vested in the
state and people of Ethiopia and shall not be subject to sale or other means of exchange.’
(Ethiopia, 1995). As far as the fundamental land policy is concerned, no change has occurred
since 1974.
Unlike that of the Derg Regime, however, the new constitution acknowledged the right of
compensation for loss of immovable building property and permanent improvements made
on the land. In addition, unlike the 1975 land reforms, the new land policy allowed for the
lease of land and use of hired labour. It was stated that citizens had the right to obtain land
freely for farming purposes, with the right to remain on their holding without eviction or dis-
placement. This study examines to what extent land tenure enters into the farmers’ decision-
making equation, in tacit or explicit ways and the relationship between land tenure and other
policy environments. The fact that the study was carried out in different agro-ecological and
socio-economic environments makes the results even more relevant for policy information in
this regard.

2.7 An overview of agricultural research

Agricultural research in Ethiopia was initiated and institutionalised within educational agri-
cultural institutions. The pioneering institutions were Ambo and Jimma, established in 1947
and 1952, respectively. When the College of Agriculture at Alemaya began its operation in
1953, based on the USA’s Land-Grant College model, it became the national co-ordinator for
research and extension. In 1963, the college administration was transferred from the Ministry
of Agriculture (MoA) to the Addis Ababa University. Consequently, the administration of the
national extension service was transferred to the MoA. On this development, the Institute of
Agricultural Research (IAR) was established in 1966, being mandated with agricultural
research. All in all by the mid-1960s, agricultural education, research and extension that were
under one umbrella were separated (Nichola, 1985). This institutional modality still prevails.
IAR was organised as a semi-autonomous organisation accountable to the Board of
Directors. The main research areas of the institute were crop and livestock production, with
an emphasis on the former. In the course of its periodic restructuring, additional mandate
areas were included. These include: agricultural mechanisation, socio-economics and farming
systems research. Soil and water conservation research is one of the less developed research
areas (EARO, 2000a). The bulk of the research in SWC was carried out under the SCRP. The IAR

40
.tesfaye.thesis 29-01-2003 15:12 Pagina 41

has been operational for three decades, executing its responsibilities in these areas.
Besides the IAR, there were and still are many organisations that contributed to the
National Agricultural Research System (NARS) that carries out research in agriculture and nat-
ural resources as its auxiliary or major task. These include: universities and colleges, various
agencies and authorities (the National Seed Industry Agency, Tea and Coffee Authority,
Environmental Protection Agency) and the CGIAR Centres (ILRI, CIMMYT).
The IAR has also been working towards ensuring the utilisation of the technologies it devel-
oped. Apart from its collaborative work with the MoA that began in the mid-1970s (see below),
the IAR has created an internal mechanism to pursue technology validation and dissemina-
tion. These efforts began with the initiation of farming systems research in the late 1970s. In
1985, the Research and Extension Division (RED) was established within the IAR to further
improve the relevance of its research programmes.
The establishment of the federal system of administration in 1991 resulted in the decen-
tralisation of the research system as well. Thus, the regional agricultural research institutes
were established, which so far number five. This necessitated the transfer of the IAR’s former
stations in respective regions to the newly promulgated regional research institutes. This pro-
cess has created duality in the country’s research system, the federal and the regional
research centres. The IAR with its federal centres was reorganised as the Ethiopian
Agricultural Research Organisation (EARO) according to proclamation No.79.1997 (FDRE,
1997b). This proclamation enforced a transfer of research functions from other organisations
in part or in full. These include: the Debre-ziet Research Centre, the National Soil Research
Laboratory, Fishery Research Institute, National Animal Health Research Institute, Wood
Utilisation Research and Forestry Research Centre.
The EARO’s strategy to bridge the gap between research, extension and farmers is through
research-extension advisory councils at various levels, following the federal system of admin-
istration. These are: federal, regional and zonal levels, each with their specific organisation
and management functions (EARO, 2000b). The effectiveness of this strategy remains to be
seen after it is adequately tested in the real-life situation. In the next section, an overview of
the extension system is presented.

2.8 An overview of agricultural extension

Like that of agricultural research, the birth of the formal agricultural extension system in
Ethiopia was in the educational institutions that began in 1953 with the Point Four
Programme at the then Imperial Ethiopian College of Agriculture and Mechanical Arts, now
Alemaya University. As mentioned earlier, the College coordinated the national extension sys-
tem until 1963.
The extension system has passed through different approaches, to some extent influenced
by the rural development approaches followed in the country (for details see Tesfaye, 1997,
1999). These are:
1930-50 Limited activities by religious organisations and national institutions with
mandates in the agricultural sector.
1953 The Land-Grant College approach based at the then College of Agriculture.
1958 Community Development approach that used the general agricultural exten-
sion model.
1968 Maximum Package such as the Chilalo Agricultural Development Unit, Ada
District Development Project, Wolaita Agricultural Development Unit, all

41
.tesfaye.thesis 29-01-2003 15:12 Pagina 42

using the general agricultural extension model.

1971 The Minimum Package Projects that have covered considerable parts of the
country. These also followed the general agricultural extension model.
1986 Peasant Agricultural Development Project, introducing the Training & Visit
system.
1993 Peasant Agricultural Development Project, using the Sasakawa-Global 2000
extension approach.
1994 Peasant Agricultural Development Approach, following the Participatory
Demonstration and Training Extension System (PADETES).
1995 to date Rural Centreed Development, using the PADETES and modified Sasakawa-
Global 2000 approaches, commonly called the New Extension Intervention
Programme.

Since its initiation and to the present day, the focus of the extension system in the country
was to introduce new technologies to the farming communities, mainly fertilisers, but also
improved seeds and other agrochemicals. The pace of the extension service since 1994 has
shown a great leap in the provision of these inputs to farmers. For instance, ‘annual sales of
chemical fertilisers have grown from 32,000 tonnes in 1982 to 280,000 tonnes during 1998’
(Setotaw et al., 2000). Even then, fertiliser coverage of arable land in the 1994/95 cropping sea-
son was only 28% (CSA, 1995; cited in Setotaw et al., 2000).
The extension system of the country operates without a concrete extension policy. The
focus of periodic restructuring was on approaches and methods. One important development
in extension was the effort made to collaborate with research in the country. In this connec-
tion, the outreach programme initiated in 1974 between the IAR and the MoA’s Extension
Projects’ Implementation Department (EPID), commonly known as IAR/EPID was the turning
point in research and extension linkage following their separation in the early 1960s. This pro-
gramme was discontinued in 1977 and reinitiated in 1980/81. The new linkage mechanism
was created between the IAR and the Agricultural Development Department (ADD) of the
MoA, thus known as IAR/ADD. These experiences paved way for a more articulated Research
and Extension Liaison Committee (RELC) in 1986 (EARO, 2000b). The RELC was initiated with
a Training and Visit system under the PADEP approach. The replacement of the rural devel-
opment approach and adopted extension approaches shown above coupled with the lack of
defined responsibilities and logistical support of the RELC weakened its operation after 1993.
The reorganisation of the IAR into the EARO lead to the establishment of Research and
Extension Advisory Councils at various levels. The MoA is an active partner in these councils.
The ups and downs of the research and extension linkages in Ethiopia have raised many con-
cerns and curiosity as to how these councils would fare in this matter, given the poor training
opportunity of the extension staff, the composition of the council and its working modalities,
to mention a few.

2.9 Environmental policies

The environmental movement in Ethiopia began during the monarchy and was intensified
during Derg periods and continued since then. The state’s focus during the Derg periods was
on wildlife protection. This policy direction led to the establishment of a number of parks and
sanctuaries in different parts of the country and the delineation of hundreds of hectares of
forestland. The drive for this initiative was to improve the tourist industry that also involves

42
.tesfaye.thesis 29-01-2003 15:12 Pagina 43

protection of historical places with their cultural heritage. To that end, much legislation con-
cerning natural resources has been issued since the mid-1960s (Dessalegn, 1998).
The turning point in the environmental movement took place during the Derg Regime,
prompted by the famine of 1973/74 and the socialist principles adopted by the Derg. The
degraded landscapes of the northern (particularly, Wello), central and eastern provinces
became the testimony for the expatriate experts who have long been ringing the warning bell
for the environmental crisis in Ethiopia. Thus, unlike the pre-Derg period where the state was
rather reluctant to address environmental problems, and thus limited its activities to forestry
and tourist developments, the Derg government entered mainstream natural resource man-
agement that included the introduction of soil and water conservation practices on arable
and non-arable lands. In addition, socialist oriented collectivisation and politically motivated
resettlements took place in an attempt to overcome technological gaps and to ease land
pressures. The environmental discourse followed by the state was founded on the Malthusian
concept of crisis: ‘…land degradation … was caused by backward agricultural practices, a
‘primitive’ system of land use, and high population pressure (RRC, 1985; cited in Dessaglegn,
1998).
The soil and water conservation division was established within the Ministry of Agriculture
in 1970 with one senior and six junior Ethiopians, supported by an expatriate expert (Berhe,
1996). The work of this division began with routine activities in an around the capital. The out-
break of famine in 1973 led to the joint operation of the then Relief and Rehabilitation
Commission. As a result of this collaboration field activities in the famine affected regions
began, where the food-for-work (FFW) scheme was initiated in the country. FFW based conser-
vation activities increased over the years after the 1973 famine, which was followed by that of
the 1984/85 famine. This phenomenon intensified the state’s environmental activities. In the
process, the work of the Ministry of Agriculture was split into two. The drought-prone regions
(Tigray, Wello, Gonder, Shoa, Hararghie, Gamugofa and Sidamo) were put under FFW project,
while the rest of the country was left under the public extension system that followed the
path outlined earlier. The watershed approach was followed in the entire conservation activi-
ties in the FFW regions (ibid.). According to Berhe, the former senior MoA staff indicated that
the watershed approach failed to include the socio-economic factors and operated on techni-
cal parameters only. Technical support of conservation activities was given by the former Soil
Conservation Research Project (SCRP) that was jointly financed by the Ethiopian and Swiss
governments. The national research system did not take part in these activities. In addition to
the SCRP, the Ethiopian Highlands Reclamation Study has substantially influenced the
government actions.
The financial estimate of environmental protection in Ethiopia from 1974-1993 was 900
million US dollars (Dessalegn, 1998). The largest donor of this fund was the World Food
Programme.
Conservation movements during the Derg Regime were based on mass mobilisation with-
out legislative backing (Dessaglegn, 1998). A formal conservation policy in recent times came
into the picture after the project agreement signed between the World Conservation Union
(IUCN) and the government on March 15th, 1990 (FDRE, 1997a). The result of the agreement
was what became the Conservation Strategy of Ethiopia (CSE) that was conceived during the
Derg and realised during the tenure of the new government that fully endorsed the initiatives.
The strategy documents were released in early 1997 and have since then come into force.
Following the federal conservation policy, few regions have prepared a regional conservation
policy, while others are in the process of being written.

43
.tesfaye.thesis 29-01-2003 15:12 Pagina 44
.tesfaye.thesis 29-01-2003 15:12 Pagina 45

3 Theoretical and Analytical Frameworks

3.1 Introduction

This study is about understanding farmers’ land-management practices. Its major aims are to
understand how farmers make decisions and to identify the determinants of their decisions.
It draws its conceptual framework from the theoretical perspectives on land degradation and
intensification, planned change and development intervention, farming systems, indigenous
knowledge, social learning, and an attitude-behaviour model.
The study was motivated by field observations of symptoms of land degradation. Therefore,
an examination of the debate on land degradation and on its alternative, intensification, cre-
ates an appropriate entry point into the discourse (section 3.2). Following that, section 3.3 pre-
sents a brief discussion of planned change and development interventions that underpin
interventionism and its drawbacks. From there, a discussion along theoretical lines leads to
the need to understand farmers’ decision-making processes, with special reference to soil and
water conservation. At this juncture (section 3.4), a discussion of the farming systems per-
spective provides the setting for analysing why farmers’ practices differ in different agro-eco-
logical and socio-economic situations. Unlike the traditional farming systems that focus on
the analysis of the components of the farming system from a ‘hard’ systems perspective, this
study carries out a farming systems analysis in line with ‘soft’ systems thinking. This helps us
to compare the similarities and differences of each farming system in relation to land-man-
agement practices and related subsystems. Farming systems are the making of people as they
try to interact with their natural environment. The inevitable result of such an interaction is
learning, which is the basis of all indigenous knowledge. With this logic, the discussion of the
farming system naturally progresses into outlooks on indigenous knowledge, farmers’ and
scientists’ knowledge and the nature of farmers’ soil and water conservation (as discussed in
section 3.5).

45
.tesfaye.thesis 29-01-2003 15:12 Pagina 46

The learning that takes place as a result of the human environment is predominantly
social. Social learning underpins the knowledge and practices people develop, forming the
basis of indigenous knowledge. Without learning, the adaptive capacity of the community is
finite. Therefore, a brief review on social learning underlining its importance in this study is
presented in section 3.6. Notwithstanding the importance of social learning, individual deci-
sion-making processes are also important. Therefore, a typical model dealing with individual
decision-making is reviewed; that is, the ‘attitude-behaviour’ model (in section 3.7). This
model is used for its ability to provide structure and for its theoretical rationale rather than
its measurement approach. Thus, the conceptual framework of the model is used to analyse
the relationship between farmers’ practices and potential determinants from within the
household and the environment. Following these theoretical lines, the chapter goes on to pre-
sent empirical studies on the determinants of soil and water conservation in section 3.8.
Finally, the overall analytical framework of the study is presented in section 3.9.

3.2 Perspectives and approaches to land degradation and intensification

Land degradation is defined as ‘the temporary or permanent lowering of the productivity of

land.’ (Young, 1997). Young indicates different types of land degradation that include, soil ero-
sion, soil fertility decline, salinisation, water-resource degradation, forest clearance and
degradation, pasture-resource degradation, and loss of biodiversity.
Land degradation studies in developing countries have used at least three approaches: the
classic (technical), populist and neo-liberal approaches. The classic approach takes for granted
that the extent of and solutions to the problems of land degradation are well known, but the
problem is to get people to implement them. Contrary to the classic approach, the populist
approach argues that the nature and extent of land degradation are imperfectly understood,
that local peoples often reject conservation technologies for good reason and resort to their
own practices and adaptations. The neo-liberal approach shares some views with the classic
approach, with regard to soil and water conservation technology. While assuming problem
definition to be unproblematic, this approach claims that incentive structures motivate farm-
ers to adopt these technologies, through extension (Biot, et al., 1995).
These approaches have influenced a number of projects in several countries, including
Ethiopia, by laying the foundation for soil and water conservation interventions (ibid.).
Nowadays, rural development projects and soil and water conservation projects are mostly
guided by the populist approach, with some elements of the neo-liberal approach appearing
in the process.
The following section presents an overview of the salient issues of the debates on land
degradation and intensification. Thereafter, the historical and epistemological background of
interventionist policies in the developing world is given, along with relevant critiques.

3.2.1 Land degradation: assumption or reality?

In the 1990s, assumptions about land degradation widely held by the classic and neo-liberal
approaches faced challenges on-the-ground stating that degradation cannot be defined and
measured unequivocally, nor taken to be straightforward. The definition of land degradation
cited above also indicates the difficulty it poses to measuring land degradation.

46
.tesfaye.thesis 29-01-2003 15:12 Pagina 47

The complexity of land degradation leaves much to be desired when specifying the rela-
tionship between soil erosion and decline in productivity on one hand, and soil conservation
and increase in productivity, on the other (see Mazzucato and Niemeijer, 2000; Sonneveld,
2002; Eswaran, et al., 2001). In this connection, Bie (1990, cited in Biot, et al., 1995) suggests that
the views of different actors in land management should be considered in the definition for
land degradation as all have their own perceptions on land degradation and the criteria to be
used for it. This view is consistent with the ‘social learning’ perspective which claims that dif-
ferent actors perceive different things according to their engagement with their immediate
environment. Therefore, resource-users have their own reasons for what they do with their
resources, their perception of the process and whether they see any problem or not. Without
the participation of the resource-users themselves, the perception of outsiders on degradation
will not guarantee that conservation measures will be implemented (see chapters 6 and 7).
The lack of clarity on land degradation has created two schools of thought on this subject.
One school of thought acknowledges the occurrence of land degradation as threatening the
ecosystem and the future of people on planet Earth. This group may be called the ‘land degra-
dation’ school. This school follows the Malthusian thesis of population and environment
which posits that high population growth is detrimental to the environment due to excessive
pressure on the natural resources to meet diverse human needs, food, shelter, energy, infra-
structure, etc. This viewpoint is subscribed to by several scientists and international
organisations (Hurni, 1993; Smaling, et al., 1997, Eswaran, et al., 2001; Pinstrup-Andersen and
Pandya-Lorch, 2001; World Bank, 1992). Among these, the World Bank (1992) considers popu-
lation growth as both the cause and effect of environmental problems. This is assumed under
similar land–management practices. The vicious circle of environmental crisis is explained in
Box 3.4 (World Bank, 1991:61), which states that: ‘rapid population growth, agricultural stag-
nation, and environmental degradation are closely interrelated and mutually reinforcing.’
Productivity decline due to soil erosion and desertification is estimated to have reached up
to 50%. Yield decline in African soils due to past soil erosion is estimated to range from 2 to
40%, with a mean loss of 8.2% for the continent (Eswaran, et al., 2001). Recent studies scale
down the extent of land degradation and its environmental costs to a nation on the grounds
that much of the removed soil is deposited further down the slope, valley or plain, and that
the yield loss in one area could be compensated by yield gains elsewhere (FAO cited in
Lomborg, 2001; Bojo, 1995).
The second school of thought describes the ongoing process of natural resource develop-
ment and utilisation as a normal phase within the dynamic process through which people
adapt their environment without necessarily degrading it. This group can be called the ‘Eco-
dynamics’ school, represented by the thesis of Ester Boserup (1965), and includes: Tiffen 1994;
Leach and Mearns, 1996; Dessalegn, 1998; Mazzucato and Niemeijer, 2000; and Rasmussen, et
al., 2001) (see Section 3.2.2 below).
Some scholars view the two schools of thought as complementary rather than contradicto-
ry (Netting, 1993).
In view of its recent popularity, the next section deals with the arguments of the ‘Eco-
dynamics’ school, with a focus on agricultural intensification.

3.2.2 The path to agricultural intensification

Agricultural intensification involves the technological progression from previous methods of

production to another method. According to Boserup (1965), agricultural intensification is

47
.tesfaye.thesis 29-01-2003 15:12 Pagina 48

considered to be an outcome of the stage within agricultural production processes triggered

by population growth. An increase in population density leads to, first, expanding the area
under cultivation, or when that is no longer possible, shortening fallow periods. Shortening
the fallow period initially leads to a decline in soil fertility. Being rational 3 enough, the soci-
ety changes its farming practices by using the relatively abundant resources, often labour, to
satisfy the higher demand for food. The change in farming practices may involve an increased
frequency of cropping, a shift from hoe to plough, application of manure, irrigation and pos-
sibly industrial inputs such as fertiliser and machinery. Thus, factor substitutions and tech-
nological changes play an important role in the Boserupian theory.
Hayami and Ruttan (1985) advance a similar view in their ‘theory of induced innovation’.
They advocate that the ability of a country to take on rapid growth in agricultural productiv-
ity and output seems to hinge on its ability to make choices among the alternatives 4 to solve
the limiting factor of endowment. According to these authors, the market 5 plays a decisive
role in achieving intended growth (See also Scherr and Hazell, 1994).
On the whole, the Eco-dynamics school of thought promotes the assumption that intensive
labour investment will generate a surplus to be reinvested, which would offset the opportu-
nity cost of a long fallow period, and also contribute to growth in other sectors through link-
age effects and induced institutions. The driving forces behind these institutions and linkage
mechanisms are markets. However, Boserup (1965:118) warns that the positive contribution of
population to economic development may not be fulfilled in densely populated communities
if rates of population growth are high. She does not go on to provide an indication of how high
a ‘high population growth rate’ 6 is, that will deter agricultural intensification, however.
Evidently, providing a standard cut-off point would be complex, and hence room is left, for the
variability of outcomes and the relationships among the variables involved.
What this school seems to overlook is the time span between the appearance of the nega-
tive effects due to overpopulation and the time when the expected intensification takes effect.
In developing countries where there is a massive dependence on biomass by a large majority
of the population, the expected effects that trigger a high population towards intensification
are likely to create even further negative effects on the environment.
Tiffen, et al. (1994) came up with a study that backs up the Boserupian theory in Africa, by
providing evidence that negates the Malthusian crisis relating to increased population in the
21st century. The study shows how more people can lead to more output and less erosion.
These conclusions were drawn from changes in land use in the Machakos District (Kenya) from
1930-1990. The authors argue that the success in Machakos is mainly due to a conducive poli-
cy environment in Kenya rather than to other unique situations. To that extent, they expect a
replication of the Machakos trajectory elsewhere. However, this optimism cannot be gener-
alised even for Kenya itself. For instance, a study by Ovuka in the nearby District of Murang’a
claims the opposite; i.e., ‘more people, more erosion …’. His study covers the period between
1960 to 1996. The author ascribes the difference between the two districts to different enforce-
ment of SWC practices, steeper slopes, and legal land consolidation in Murang’a District
(Ovuka, 2000:122). Scoones and Toulmin (1999) also doubt that high population growth in

3 Note that this rationality explains the major difference of this school with that of classic Malthusian theory that consid-
ers a doomsday scenario as an inevitable outcome of population pressure.
4 Biological technology offsets the limitations of land while mechanical technology offsets shortage of labour.
5 They hypothesise that technical change is guided along an efficient pathway, by price signals in the market, provided
that the prices efficiently reflect changes in the demand and supply of products and factors, and that there exists effec-
tive interaction among farmers, public research institutions, and private agricultural supply firms (Hayami and Ruttan,
1985:88).
6 Note that population growth, besides increasing land productivity through management, also gives incentives to agricul-
ture through the rural-urban migration that is expected to follow, due to stern working conditions in agriculture by rais-
ing the price of food (Boserup, 1965:118-119). Of course, cheap food imports could nullify this effect.

48
.tesfaye.thesis 29-01-2003 15:12 Pagina 49

some African countries would lead to further intensification under the current market con-
ditions.
Recent findings in Burkina Faso take the argument of ‘more people, less erosion’ even
further by illustrating how African farmers are capable of making more adjustments than pre-
sented by the population pressure theory (Mazzucato and Niemeijer, 2000). This study empha-
sises that changes in agriculture occur not only because of population pressure, but also in
response to changes in the social, economic and environmental contexts in which agriculture
takes place. It therefore rings a bell to re-think soil and water conservation according to
changes in the society. The message of this study, in a nutshell, is that we have to be aware of
the fact that farmers never sit still, twiddling their thumbs, while the context of their agri-
culture changes. The study further shows that productivity increases have sustained a rapidly
growing population while maintaining environmental sustainability. These authors, while
endorsing Boserup’s warning that a high population does not always coincide with techno-
logical change, strongly advise that imposing a particular path will not help to achieve con-
ditions of productivity growth and environmental sustainability, but rather make these paths
the subject of scientific inquiry (op.cit., p. 305).
Development policy in developing countries is highly influenced by what Lele and Stone
(1989) call ‘policy-led intervention’. This policy direction goes against the Boserupian devel-
opment model, or as they call it, ‘autonomous intensification’. Their recommendation of a
policy-led intervention model is based on inconsistencies 7 with the assumptions of the
autonomous–intensification model, particularly referring to Africa.
In connection to agricultural intensification, it is worth noting that a positive relationship
between intensification and environmental quality is not always expected, as intensification
processes 8 are likely to bring some negative effects on environmental quality. Hence, the ques-
tion of tradeoffs or synergy comes into the picture (Lee, et al., 2001). This is so because the dif-
ferent views on what is meant by environmental quality are value-laden (Pretty, 1995).
What is the implication of the foregoing discussion for the Ethiopian case studies? Do they
reflect a Malthusian or a Boserupian footprint? Chapter 8 addresses this question based on the
three case studies from Ethiopia. For now, let us look at an overview of interventions in a
broader development perspective.

3.3 Planned change and development intervention

The concept of intervention found its birth in a world outlook that originates from the dom-
inant ‘western’ culture and the economic standardisation of western and non-western cul-
tures. The economic gaps perceived by western countries’ institutions and multilateral organ-
isations such as the World Bank between the West, and the poor countries in Africa, Asia and
Latin America, have led to an interventionist policy in the poor continents by western insti-
tutions that triumphs to date (Escobar, 1995). The motto for intervention was coined as ‘pover-
ty alleviation’, which is nowadays addressed as ‘poverty reduction’. The start of this movement
was the Marshall Plan for Europe after World War II. The interventionism was ushered in by
multitudes of agencies, consultancy firms, programmes and projects commissioned through

7 These are: 1) the negative effects of extremely rapid population growth as compared to the slowly rising densities envis-
aged in the hypothesis; 2) the substantial concentration of population, even in land-abundant countries; 3) the conflict
between the social and private gain of a large family size at low levels of labour productivity for poor households; 4) the
tendency to ‘mine’ the land for immediate survival versus the social need to protect soils as a productive resources; and
5) unequal access to land and even expropriation from smallholders as land value increases.
8 These include: depletion of water resources, salinisation and micro-nutrient depletion in soils, chemical contamination
of groundwater, and reduced genetic diversity.

49
.tesfaye.thesis 29-01-2003 15:12 Pagina 50

the foreign agencies of the western countries and scientific organisations, and the United
Nations. Replicas of these agencies formed in developing countries, including their state
machinery, promoting the same principles of interventionism at the grassroots level.
Born out of a paternalistic view, their interventionism aimed at creating more and more
‘consumers’ by turning them into objects of knowledge and management (Escobar, 1995).
Thus, interventions took place in a range of sectors in the economy, from infrastructural devel-
opments, and policies of education, agriculture, health, etc., These support measures often
result in what is commonly called a ‘cargo’ image among the recipient community (Röling,
1981; Long, 2001). The interventions are assumed to be blessed and appropriate for the recipi-
ent cultures, as they had proved to be appropriate in the donor cultures. In the words of
Escobar (1995), they are assumed to be neutral, desirable, and universally applicable. In addi-
tion, the solution to the identified, universal problem of poverty/underdevelopment is always
a western-defined ‘development’. This is in turn believed to be achievable through financial,
scientific and technological interventions 9. Since the 1980s, the concept of development inter-
vention has shifted towards ‘sustainable development’, by marrying the two old enemies,
growth and environment. Development in this equation also targets poverty as both a cause
and an effect of environmental problems (Ferguson, 1990; Escobar, 1995).
An analysis of the architecture of development projects is clearly shown in the studies of
Ferguson (1990) and Escobar (1995) that focused on the World Bank, in Lesotho (in 1975) and
Colombia (in 1949), respectively. The rules of the game for such interventions were created by
top international institutions and subsequently picked up and reproduced by others. The fate
of most projects at least with respect to the objective of alleviating poverty, whether it was in
the 1950s or in the 1990s was the same: i.e., failure. The chief reason for this result, among
others, is the basic assumption that it would be possible to reduce and solve a problematic sit-
uation in the developing world with a technical fix. In this connection, Ferguson (1990:256)
writes that, ‘by uncompromisingly reducing poverty to a technical problem, and by promising
technical solutions to the sufferings of powerless and oppressed people, the hegemonic prob-
lematic of ‘development’ is the principal means through which the question of poverty is de-
politicized in the world today.’
Interventionist policy in the developing world was not limited to the ‘American Dream.’
The Eastern Bloc’s communism has also littered the socio-economic and political landscape of
many countries including Ethiopia, where the present study was carried out. For instance, in
Ethiopia, socialist interventionism was reflected in state-owned enterprises, centralised plan-
ning, collectivisation, villagisation, and the destruction of indigenous institutions, to men-
tion the major ones. Some black spots of the socialist era have not yet been cleared away even
a decade after the socialist government was overthrown. These features of socialism have been
experienced with varying degrees of severity in many parts of the former Eastern Bloc (Scott,
1998).
Long (2001:30) posits intervention as ‘a ‘multiple reality’ made up of differing cultural per-
ceptions and social interests, and constituted by the ongoing social and political struggles
that take place between the various social actors involved.’ In this process, the ‘recipient’ cul-
ture or group continuously interacts with the intervening party and the subject of
intervention, even long after the formal termination of the project. ‘Social learning’ in the
community not only passes down what has happened in the past, but also helps people to
construct knowledge and to store information. In this manner, social learning not only deter-

9 However, the general trend so far shows that the modified Marshall Plan waged against poverty in the developing coun-
tries seems to be unsuccessful, unlike its original form in Europe. Is this because of a lack of understanding to do the
right thing, or is it a lack of will to do so after half a century? Of course, this question equally applies to the global and
local elites, institutions, nation-states, professionals, etc., and the answer also includes non-intervention.

50
.tesfaye.thesis 29-01-2003 15:12 Pagina 51

mines the outcome of the present intervention, but also similar interventions in the future,
making people’s participation crucial for the success of development projects.
In view of the multi-faceted nature of poverty-reduction strategies, the term intervention is
used differently in different fields. In this study, however, its use is limited to technological
change in agriculture and natural resource management. Interventions in these areas are
commonly known as ‘Green Revolution’ technologies or its successor, ‘integrated rural devel-
opment’ projects. The epistemological grounds of the Green Revolution are vividly portrayed
by Elizabeth Bird (1984, cited in Escobar, 1995:159):

‘The messages [in the Green Revolution literature] are, first, that these development planners know what
‘the people’ in the ‘developing countries’ want; second, that what they want is what ‘we’ have; third, that
‘they’ are not yet advanced enough to be able to fully indulge themselves without repercussions; and fourth,
that discipline, prudence and forbearance are some of the qualities necessary to success.’

The assumptions of the Green Revolution and its offshoots are reproduced overtly or covertly
in development projects and programmes even to date, when participation and democratisa-
tion are loudly pronounced.
In agriculture and natural resource management, interventionism was realised largely
through extension services or integrated rural development projects where extension plays an
umbrella role in deploying an intervention. In view of this, huge financial and material
resources have been invested in extension in both developed and developing countries. The
driving force for its universal acceptance is the theoretical background of extension that
attracted policy-makers the world over. The most influential theory underlying extension
[intervention] is ‘diffusion theory’ with its offshoot, ‘Transfer of Technology’ (ToT), advanced
by E.M. Rogers.
Even though ToT contributed substantially to materialising the Green Revolution aims in
many countries (mainly in Asia and a few Latin American countries), it has been criticised for
its major shortcomings (Chambers and Jiggins, 1987; Röling, 1988, 1992; Rogers, 1995; Leeuwis
with Van Den Ban, forthcoming). The key shortcomings of the model are given below (for a
detailed discussion on this issue, see Leeuwis with Van Den Ban [forthcoming]:
• It is a sequentially linked linear model. Here science is seen as the major source of new
ideas that lead to the neglect of local knowledge;
• It has a pro-innovation bias. Here, what science produces is assumed to be beneficial for
users so that its adoption is worthwhile (non-adoption and farmers’ own innovation are
often considered as negative responses rather than as a conscious choice, a learning pro-
cess, or otherwise);
• Technology is often perceived as a ‘product’ and therefore, packaged by research and exten-
sion for delivery to farmers. That is in turn assumed to trickle down from model farmers to
the followers. In a worst case scenario, which occurs more often than not, technologies are
developed under different conditions from those of the end users, and yet their specificity
is often neglected; and
• The delivery model proved successful for resource-rich farmers in areas with developed
infrastructure, but was highly irrelevant for resource-poor farmers.

The extent to which the interventions assessed in this study reflect the drawbacks indicated
in this section will be examined in Chapter 8 when the research questions of the study are
revisited.

51
.tesfaye.thesis 29-01-2003 15:12 Pagina 52

Notwithstanding the negative side of interventions presented above, the purpose of this dis-
cussion is not to pronounce anti-interventionism. Interventions are condemned when they
are confrontational, manipulative and abusive. Intervention may be feasible when it is of a
catalytic and supportive nature (Wilson and Morren, 1990). These kinds of interventions are
well in line with the participatory school tradition that flourished after the ‘farmer-first’ idea
coined by Robert Chambers.

3.4 The farming systems perspective

The notion of a system helps us to operationalise the systems concept in improving system per-
formance, assuming a system in which human interest is directly or indirectly involved. In my
thesis, I have in mind a system of sustainable use of natural resources, more specifically, soil
and water conservation.
A system is a mentally constructed entity, whose existence can be configured so far that it
is made discernible to human minds, as an abstract or concrete system. The common elements
in the conceptualisation of a system are shown below (Hurtubise, 1984; Checkland, 1985;
Checkland and Scholes, 1990; Röling, 1994; Engel, 1997):
• A set of elements or subsystems within a system which are related in some organised way,
not arbitrarily;
• These elements interact in a dynamic way, due to certain patterns evoked by inherent
and/or induced interrelationships among them, which may lead to adaptive action;
• A system is more than the sum of its parts;
• The entity we mentally constructed has a certain mentally constructed boundary, which
not only distinguishes it from its environment (also a mental map), but also shows the
extent of change processes in the form of transformation within our system itself and its rela-
tionships with other systems;
• The environment, by and large, sets the pace for the adaptive action of our system, while it
is also affected by this action and/or at times partially guided by a system;
• Our specific system interacts with one or more systems within this environment in their
own contexts; and
• Relationships between and among elements or subsystems of a system show certain pat-
terns in energy/information flow, in the form of input-output, which indicates a kind of
hierarchy of relationships and that this emergent property/synergy is not only discernible with-
in a system (e.g. organism), but also observable between systems within an environment
(e.g. food chain, national, regional or international geopolitical relationships).

The very existence of an entity depends on its emergent properties, which change according
to the context, to ensure its adaptation to the system. This phenomenon holds true for the
mapping in human minds, which is triggered by the experiences in the world to which we are
directly or indirectly exposed (Checkland and Scholes, 1990:23; Maturana and Varela, 1992:95-
96). Had it not been for the triggering elements in the context that facilitate learning, human
adaptation to environment would hardly take place, which would be detrimental to the very
existence of people.
The farming systems perspective entered into the arena of systems thinking after the mid-
1970s, when researchers began to appreciate the holistic nature of the farming problem that
is full of highly interwoven factors. Taken as an entity, farming systems have all the system
properties discussed above. The state of a given farming system is shaped by the interaction

52
.tesfaye.thesis 29-01-2003 15:12 Pagina 53

between the natural and social environments internal and external to each farming system.
Natural environments refer to climate, soils, topography, vegetation cover, etc., while the
social environments include markets, land tenure, political stability, family size, cash avail-
ability, land and livestock owned, labour supply, management capacity, credit and extension
services, etc. (Ruthenberg, 1980). Meeting an objective in a given farming system depends on
the proper matching of relevant elements from each domain.
The farming systems perspective was initially, however, constructed using a ‘hard’ systems
methodology that very much focused on a system as an aggregation of biophysical factors
rather than on the social aspects dealt with by the entities (farming) (Checkland, 1985; 1989).
As a result, farming systems research passed through its years of hegemony without much
impact on changing the reductionist approach of research and extension organisations in the
developing world (van Eijk, 1998). It focused on the understanding of researchers of the farm-
ing system rather than on the learning processes of farmers themselves (Lynam, 2002).
However, its landmark contribution as a forerunner to the participatory approach goes down
in the history of institutional innovations in R&D in agriculture (Farrington, 2000; Collinson
and Lightfoot, 2000). The essence of a farming system is not the mere analysis, but the tracing
and triggering of the synergy of the system. This gap was created due to a lack of appreciation
for ‘soft’ systems thinking. However, this does not mean that it is not possible to redress the
shortcomings of the farming systems approach and to exploit the systemic vision and its
appreciation for location-specific knowledge that farmers developed, and the need to fit tech-
nological changes into a system (Collinson and Lightfoot, 2000). Soil and water conservation,
that seems to be a typical ‘hard’ systems domain, becomes a ‘soft’ systems concern when the
multitude of social factors that affect a farming system is taken into account. This issue is not
only about soil or water, i.e., purely biophysical factors, but also about all other social, eco-
nomic and institutional aspects. The interests of people and human agency signify the role of
the soft systems approach to solving the problem. Had it not been for these elements, only
engineers and agronomists would have been enough to solve the problems of soil erosion the
world over, which is not the case. The negative consequences of missing this view are illus-
trated by this study.
The hierarchy of farming systems shows that soil and water conservation falls within the
domain of the cropping subsystem (Ruthenberg, 1980). By studying the hierarchy of sub-
systems of farming systems, we can improve our understanding of the functioning of the
farmer’s household and land-management practices such as SWC. Households link SWC and
the cropping system with several other subsystems to fulfil their needs, rather than focusing
only on SWC as outsiders often do. Farmers base their decisions on specific characteristics of
their farm plots. To understand the decisions of farmers, one needs to make an analysis at the
farm-plot level. Farmers look at specific plot characteristics such as slope, distance from
house, soil types, crops grown, tenure, etc. In view of this, key factors related to soil erosion,
soil fertility and soil and water conservation were analysed at the plot level in this study.
The farming systems perspective is used in this study to develop a better understanding of
land-management practices in the three case study areas. This perspective helps to under-
stand why farmers are doing what they are doing. The study argues that the current farming
practices are rooted in the agrarian history of each case study area 10. While farming systems
could be comparable at a higher level, a detailed analysis of their components and functions
reveals a multitude of differences that need to be considered for the understanding of land

10 Technologies arise out of particular sets of historical and social circumstances; different people have different attitudes
and commitments to them and, because of the dynamic influences over their origin and maintenance, they evolve and
change continuously (Scoones, et al., 1996:11).

53
.tesfaye.thesis 29-01-2003 15:12 Pagina 54

management (Westphal, 1975; Ruthenberg, 1980). Technical experts tend to focus on crop or
livestock systems, even on a single crop or animal species, such as a wheat or dairy production
system. These approaches conceal the central role of people as resource managers and
decision-makers. The most crucial element of any farming system is the people who shape and
in turn change their system (learn) as a result of their interaction with the immediate
environment. Tools used for various purposes that play crucial roles in farming systems are
developed, adapted or adopted by the people (Netting, 1993).
The next section deals with the knowledge that helps people to adapt their environments.

3.5 Indigenous knowledge

Thanks to the work of many scholars and development experts, the concept of indigenous
knowledge has obtained currency in the development literature (Richards, 1985; Warren,
1988, 1991). The term is often used interchangeably with local knowledge, traditional
knowledge and indigenous technical knowledge (see below).
Indigenous knowledge has a wide range of roles in a society, which ensures the
achievement of livelihoods, including not only technical knowledge in production but also
knowledge with respect to institutions, health and security. It is unique to a given communi-
ty, culture or society inasmuch as the scientific knowledge is unique to the formal institutions
such as research institutions, universities and private companies, which are also communities
with their own identity and value system (Warren, 1991; Marsden, 1994).
Indigenous knowledge is the result of social learning (see section 3.6 below). It is generated
through a social interaction as a person tries to make his/her environment suitable for living.
A piece of knowledge on a given social phenomenon is developed, tested, improved upon and
stored through utilisation in the community of origin; thus, the knowledge is socially con-
structed (Scoones and Thompson, 1994; Leeuwis with Van Den Ban, forthcoming). This makes
knowledge indigenous to the locality and to its descendants. When knowledge developed in
this manner is handed down from generation to generation, it is called traditional knowledge.
Nowadays, due to the ease of communication, people’s knowledge tends to blend useful com-
ponents from different sources, changing its indigenous nature. For this reason, some practi-
tioners prefer to use ‘rural people’s’ knowledge instead of ‘indigenous’ or ‘traditional’ knowl-
edge (Brouwers, 1993).
There are ample opportunities for cross-fertilisation of indigenous and formal knowledge
in development. It is this characteristic that drew the attention of development practitioners
to the use of indigenous knowledge in development. Lack of this focus in the past has led to
the failure of numerous projects (Richards, 1985; Warren, 1988, 1991; et al., 1996; Critchley,
2000). During the last two decades, particularly in the 1990s, the role of indigenous
knowledge is recognised in the International Agricultural Research Centres and multilateral
agencies such as the World Bank and the United Nations Agencies (Warren, 1991).
However, how to articulate indigenous knowledge and to use it in promoting sustainable
development is not well understood by the formal systems. What has often proved difficult is
when and how to integrate the formal and the indigenous knowledge systems and even when
to employ only one of the two sources of knowledge. A gap emerges on two sides. On the one
hand, some practitioners over-romanticise indigenous knowledge with the view that it should
continue without change, and to even support it in spite of the changing internal and exter-
nal contexts. At the other extreme, indigenous knowledge is totally ignored or at best pas-
sively treated in research and extension processes, and other development arenas. Both

54
.tesfaye.thesis 29-01-2003 15:12 Pagina 55

extremes have detrimental effects on development as seen in many parts of the world.
Shortcomings in the utilisation of indigenous knowledge could be overcome if the ways in
which local people learn were well understood and appreciated by the scholars of the modern
knowledge system. Local people learn through concrete experience rather than through
theory formation and testing. Their theory originates from previous experiences that are sub-
sequently tested through practice rather than from divergence or assimilation, which focus
on abstract conceptualisation (Kolb, 1984). Richards (1985), when referring to farmers’ experi-
mentation, writes that farmers make their points on the ground, not on paper. The starting
point is, therefore, to examine the practices of actors in a given domain of indigenous
knowledge. For instance, in the area of soil and water conservation, outsiders need to closely
understand the rationale of the decision-making process of farmers before they try to improve
their methods.
Farmers have their own way of perceiving things in their farm and community. As they per-
ceive, they shape their practices using their evaluative frames of reference. The way farmers’
perceive the agricultural landscape, its problems and solutions sometimes differs, and at
times conflicts, with that of outsiders such as scientists and policy-makers. The next section
discusses this issue to gain a better understanding and to take necessary precautions in work-
ing with farmers.

3.5.1 Scientists’ vs farmers’ knowledge

The dichotomy between scientists’ and farmers’ knowledge is linked with the difference
between so-called western culture and non-western culture, which are said to be driven by dif-
ferent forms of knowledge (Marglin, 1990): the ‘episteme’ and ‘techne’ knowledge bases, respec-
tively. Unlike Marglin, Leeuwis (1993:85-86) sees a symbiotic relationship between the two
forms of knowledge. Marglin (1990:24) writes that ‘in the west, the knowledge system of man-
agement (episteme), particularly ‘scientific management’, is characterized by impersonality,
logically deduced, analytical, articulate, universal, cerebral, theoretical, verifiable and egali-
tarian’ processes. Techne, by contrast is characterised by ‘personal, intuitive, non-analytical,
implicit, contextual, tactile and emotional, practical and discovery oriented’ 11 processes. Even
though Marglin like Scott (1998) claims to have based his ideas on the Greek terms, his usage
of the terms differs from that of Scott 12. Scott uses the term ‘metis’ in the same way Marglin
use the term techne, and uses techne in the same way Marglin uses the term episteme. Unlike
Marglin, who views a sharp distinction between episteme and techne, Scott considers techne and
episteme to be knowledge of the same order and treats them differently from metis, which falls
on another continuum of knowledge from the other two. Metis develops in situations which
are ‘transient, shifting, disconcerting and ambiguous, none amenable to precise measure-
ment, exact calculation, or rigorous logic.’ (op. cit., 320). According to Scott, in a broader
sense, metis represents a wide array of practical skills and acquired intelligence in responding
to a constantly changing natural and human environment. In view of this, he even tends to
prefer metis to indigenous technical knowledge and its variations.
Over time, this differentiation has narrowed down to scientific and non-scientific
knowledge. In this study, non-scientific knowledge is referred to as farmers’ knowledge, which
Richards (1985) calls the ‘people’s science’. In this work, Richards has underlined that

11 See also Marglin’s (1991) work cited by Leeuwis (1993:85) in which more or less similar explanations and definitions are
given for episteme and techne.
12 Scott himself has clearly shown their differences in his footnote 17 for chapter 9, page 425.

55
.tesfaye.thesis 29-01-2003 15:12 Pagina 56

scientists consistently undermine rural people’s discovery and insight owing to their ‘formu-
laic approaches’ that worship literature and formalised experiments (Richards, 1985:155-156).
He further stresses that farmers are involved in a multitude of experiments in their own ways,
to adapt their environment. These processes are enriched through social learning, as was dis-
cussed in section 3.6. The other vivid observation about rural people’s knowledge is found in
the work of Brouwers (1993). Brouwers argues that rural people’s knowledge unlike that of
scientists focuses on a specific location and functionality, rather than on a universal applica-
tion. The work of Song (1998) shows that the farmers’ breeding strategy is holistic, location-
specific, and adaptive to the changing context, unlike that of breeders who follow specialised,
wider adaptation and standard approaches. Strategies of rural people’s knowledge cannot be
exaggerated, as they need to solve the problem they encountered in ways allowed by the social
resources, because they act for survival rather than to know, like scientists.
The boundary of a given knowledge base is limited by the scope of the social learning that
is behind it. Farmers explain what they perceive, in a similar way to scientists’ ‘objective’ find-
ings. The problem of modern agriculture is the lack of the ‘laboratory’ situation in real life
that has resulted in a mismatch between what scientists produce and what farmers accept.
The magnitude of the gap increases with the involvement of human action in the process and
less so when it is purely physical or biological, as in the hard system. Owing to the adherence
of the scientists to their original design, adaptation by farmers is considered as a rejection or
non-adoption whereby farmers are to blame (Ogunfiditimi, 1993; Scoones and Thompson,
1994; Yohannes, 1999; Moser and Barrett, 2002).
Millar and Curtis (1999) argue that the diversity of knowledge in the society is due to dif-
ferences in the environment such as socio-political-cultural contexts, rather than epistemic
differences that create a division between scientific and local knowledge. Stuiver et al., (forth-
coming) recognise the ‘specificity’ of scientific knowledge, which needs specific ‘laboratory’
conditions to be manifested. In that respect, scientific knowledge does not differ from farm-
ers’ knowledge. Thus, they adopt a notion of contextual knowledge rather than ‘scientific’ and
‘local’ knowledge as it is conventionally distinguished. Kothari (2002: 229-230), however,
prefers local knowledge, which refers to ‘the knowledge(s) of rural and indigenous peoples not
only in developing countries but also in the developed world.’ To that extent, he says that
scientific knowledge is also a local knowledge, that shares its ‘localness’ like any other type of
knowledge (see also Millar and Curtis, 1999).

3.5.2 The nature of soil and water conservation by farmers

Scientists and farmers address the soil erosion problem in different ways, even though they
have the same goal. This difference has far-reaching consequences for finding the method to
achieve the solution. Scientists perceive soil erosion as a process of three steps: the detach-
ment of particles of soil by wind and water from the surface, the transportation of the parti-
cles and the deposition of these particles in another place (Sfeir-Younis and Dragun, 1993:22).
Richards and Brouwers point out that farmers see the movement of soil from place to place as
a result of deposition, and could see small rills. They observe the development of gullies
merely by water erosion. The question is whether they credit this process as one of the top pro-
duction problems and how they minimise erosion. However, like scientists, farmers also know
and are concerned about plant nutrients even though they cannot give them chemical names,
like scientists whose knowledge systems allow for that level of sophistication.
Unlike outsiders who often maintain a single objective, farmers are faced with multiple

56
.tesfaye.thesis 29-01-2003 15:12 Pagina 57

objectives in their livelihood. In addition, farmers’ indigenous SWC does not aim at merely
protecting the soil or improving the moisture level. They make compromises with their mul-
tiple objectives, resources, level of the erosion problem, urgency of the household needs, prof-
itability, etc. Therefore, ‘the best soil conservation practice from the farmers’ perspective is
not necessarily that which conserves the most soil’ (Kerr and Sanghi, 1993:260). In view of this,
farmers often favour SWC practices that give them a quick benefit, while minimising soil ero-
sion. For instance, supplementary fodder, income, and food that is commensurate to the
household objectives instead of the off-site benefits that are perceived by the state or its
agents, who base their calculation on a public interest point of view rather than on individu-
al households. Farmers’ efforts in soil and water conservation are responsive to economic fac-
tors. Indigenous SWC methods are often constructed with local labour exchange groups that
shift the direct immediate cost to the community. When the practices have to be undertaken
individually, they always tend to space the financial and physical burdens over time.
Generally, farmers’ responses to externally imposed SWC methods are highly shaped by their
indigenous practices that are embedded in their local institutions and culture (Kerr and
Sanghi, 1993; Yohannes, 1998,1999; Östberg and Reij, 1998; Belay, 1998).

3.6 Social learning

Goldstein (1981:236-240) provides some cogent accounts of social learning. He states that
social learning is a form of learning (what) occurring in a social context (where) for the purpose
of adaptation (why) of individuals and society (by whom). Woodhill and Röling (1998:47) see
social learning as the process of social change, cultural transformation and institutional
development necessary to achieve these changes. They further add that social learning must
address issues of social structure (op. cit., 53), that signify its social nature. This view is further
echoed in Woodhill’s recent definition of social learning. ‘Social learning is a process by
which society democratically adapts its core institutions to cope with social and ecological
change in ways that will optimize the collective well-being of current and future generations.’
(Woodhill, 2002:323). The social context provides a transactional character whereby symbols
and values enhance a conducive environment for learning. The quest for learning could be
purposive or emergent, in influencing human behaviour. It originates from the very nature of
a human being who is goal-directed and a solicitor for change when faced with unfavourable
environments.
Social learning helps us to adapt the environment around us in such a way that we prepare
ourselves to intentionally face the reality as it unfolds in front of us. Thus, through social
learning, people shape their conceptions of the past as they interpret it, the present as they
construct it, and the future as they envision it (Goldstein, 1981: 54). Even more important is
that social learning takes place in a collective environment (Röling and Jiggins, 2001). The
partners in collective learning change according to the issue of deliberation. These include
groups, organisations, communities, etc. (Maarleveld and Dangbégnon, 1999). For instance, in
soil and water conservation, farmers (individually and collectively), soil scientists, social
scientists, policy-makers and administrators need to be involved for better watershed man-
agement. Breaking institutional barriers that separate farmers and other stakeholders from
each other makes it possible for the long-term exchange of knowledge and information
between them (Millar and Curtis, 1999). The barriers include culture and attitudes, financial
resources, centralisation of research, lack of training in participatory approaches, and top-
down extension services. Learning among the stakeholders can be facilitated by a variety of

57
.tesfaye.thesis 29-01-2003 15:12 Pagina 58

facilitation tools and concepts developed since the 1980s, and pioneered by the ‘farmer–first’
school (Chambers, et al., 1989). These range from rapid rural appraisal to experiential learning,
to PRA, to PLAR, to RAAKS, to social learning, to mention a few. It is worthwhile to note that
these facilitation methods and concepts, though they appear with different names, share the
same epistemological and ontological backgrounds, which is the constructivist paradigm
(Guba and Lincon, 1989). King and Jiggins (2002) suggest that the most opportunistic learning
venue is at the interface between people and their environment. They warn that dealing with
only people or the environment would be ineffective.
Understanding contexts for social learning merits special attention in order to fully achieve
learning in the real-life situation, not only to a ‘person’ as learner, but more importantly to a
‘person as resource manager’. Learning to solve problems requires not only perception but
also resources from the environment. The household or the community cannot just learn
without the availability of the complementary resources for the learning to take place
because they learn for immediate use.
More importantly, social learning is not learning among a few individuals in the commu-
nity that is observed by outsiders. The learning process should spread sooner or later. In this
connection, Woodhill and Röling (1998:54) argue that ‘social learning should be thought of as
a society-wide process. It is not an exclusive or elite task for ‘scientists’, ‘experts’ or ‘intellec-
tuals.’ Their observation is a very useful precaution in view of the booming studies on inno-
vation and innovators.
The interrelationship between the individual and society facilitates social learning in a
community. Whether the quest for an alternative originates from an individual or group or
the entire society, practices that are consistent with the social system are likely to spread in
the community. When more people are involved in the practice, it is likely that it will be mod-
ified and developed to fit different members of the community. To realise learning opportu-
nities, one needs a conducive policy environment. This is particularly important in resource
management such as in soil and water conservation that is affected by a multitude of social,
economic and institutional factors.
Another issue is: Who perceives social learning? Outsiders, the community itself or both?
What outsiders call learning could emerge from a painful process of crisis management or
from a gradual adaptation process without a severe alteration to livelihood practices. Care
should be taken not to romanticise the life struggle of the poor community as ‘learning’ and
‘adaptation’. The interpretation or labelling of processes in the society deserves an appropri-
ate meaning, not always learning as perceived by the outsiders. Learning could also lead to
socially undesirable outcomes. That is, people could learn what is bad according to the exist-
ing social norms.
In this section, I have discussed social learning and how people construct knowledge col-
lectively. The following section deals with the individual decision-making model based on the
traditions of social psychology.

3.7 The attitude-behaviour model

The relationship between attitude and behaviour has been one of the crucial issues in social
science research. ‘Attitude is a disposition to respond favourably or unfavourably to an object,
person or institution or event.’ (Ajzen, 1988:4). He adds that the characteristic attribute of atti-
tude is its evaluation that must reflect a positive or negative evaluation of the attribute object.
Attitude is a non-overt attribute that can only be inferred from verbal or non-verbal responses.

58
.tesfaye.thesis 29-01-2003 15:12 Pagina 59

Ajzen identifies three response categories that help us to infer about attitude. These are cog-
nitive, affective and conative responses. These notions help us to better understand the cognitive
side of the human behaviour that underlies every decision, even though its impact and effects
are mediated by non-cognitive elements such as socio-economic status and external environ-
ments, as will be shown in the subsequent theoretical discussions and the empirical studies.
Cognitive responses are responses that reflect the perception of, and information about the
attitude object. Taking the example connected to this study, the information and knowledge
that farmers have on land management show their beliefs. For instance, the perceptions that
soil erosion reduces soil fertility and that steep land is prone to erosion are likely to result in
a positive attitude towards measures of erosion control. This indicates that more information
on the object of attitude is helpful to shape the attitude of people towards the object.
Affective responses are related to the evaluation of, and feelings towards, the attitude object.
Farmers who think that soil and water conservation increases crop production and who think
that soil and water conservation is the farmers’ responsibility hold a more favourable attitude
towards soil and water conservation than those who think that farmers should be paid for
such work or that it is the responsibility of some other agency.
Conative responses are near behavioural responses that show inclinations, intentions, com-
mitments and actions to the attitude object. These stem from the notions of what people say
they do, plan to do or would do under appropriate circumstances. Farmers who appreciate
farmers who have strong terraces in their farms and wish to strengthen their own as well are
more likely to take action in that line than those who emphasise the negative side effects of
terracing.
The causal link between attitude and behaviour was later conceptualised through what is
called the ‘theory of reasoned action’ (Ajzen and Fishbein, 1980), where intention determines
behaviour and is in turn affected by the attitude towards behavioural and subjective norms.
The model was developed on the basis of observation and pattern of volitional 13 behaviour
whereby an individual is able to accomplish what s/he intends to do. Through time, it was
realised that these behaviours face a limitation in their translation into action. This limita-
tion was termed as ‘perceived behavioural control’ (Ajzen, 1988). When an individual
perceives (s)he lacks control, he or she will not be able to perform the action because of a lack
of opportunities to perform the intended act. Understanding incomplete volitional control led
to Ajzen’s theory of planned behaviour that was developed from its precursor, the theory of
reasoned action (Ajzen, 1985, cited in Ajzen, 1988). The latter theory makes a special case of
the former when perceived behavioural control is not important to predict the behaviour.
Ajzen’s model predicts behaviour, with intention as an intervening variable. Attitudes
towards the behaviour, subjective norms and perceived behavioural control interact with each
other as antecedent variables to influence the intention to yield the expected behaviour. The
second possibility to predict behaviour is a direct link between perceived behavioural control
and behaviour (Ajzen, 1988: 132-136). In the latter case, strength of intention is not important.
I disagree with Ajzen’s second trajectory of behaviour, which emphasises perceived beha-
vioural control. This option of the model seems to undermine the importance of belief. I argue
that every behaviour has an underlying belief, whether the belief has or does not have a major
effect on its taking place. That means that belief alone is not sufficient to take action, as the
shift from the theory of reasoned action to the theory of planned behaviour made clear. By
the same token, the availability of opportunities, i.e., perceived behavioural control, without
a firm belief and desire to take action does not guarantee a continuation of a behaviour even

13 This includes actions such as reading or not reading a newspaper, going to church, losing weight, raising your arm, quit-
ting smoking, having an abortion, voting for one party or another, etc.

59
.tesfaye.thesis 29-01-2003 15:12 Pagina 60

if it takes place at one point or another. This has a direct implication as to where we apply the
model in a field such as agriculture and natural resource management or in a context such
as urbanised life in which the model has been tested widely. Barels (2002) states that
behaviour is a function of the individual and the situation (environment). In view of this, one
can hardly talk of a planned behaviour, because there are several intervening variables that
are not amenable for planning. In addition, I consider intention as a general state of mind
that results from the cognition and affect rather than as a separate state. The theory of
planned behaviour shares its epistemological shortcomings with intervention theory.
As indicated above, most behaviours cannot take place in the absence of the control belief,
which ensures the practicality of the behaviour, whereas the converse is not always true; that
is, fulfilment of the control belief when the behavioural belief is negative. Such a broader look
into the determinants of behaviour is helpful to better understand why people do what they
are doing, though it is not always achievable with certainty.
Leeuwis with Van Den Ban (forthcoming) introduced a model for understanding farmers’
responses, by adapting the works of Ajzen and Fishbein and others. In this model they empha-
sise evaluative frames of reference, perceived environmental effectiveness, social relations
and perceived social pressure and perceived self-efficacy from the domain of social psycholo-
gy, while showing the feedback side of the model from the practices of social actors (see also
Leeuwis, 2002). This model is cyclical rather than linear, unlike that of the theory of planned
behaviour.
According to Röling and Jiggins (2001:159-160), ‘the theory of reasoned action establishes
levers for changing the human cognitive agent, such as anticipated feedback, ‘desirability,’
and self-confidence. They further add that, ‘… the theory emphasizes the socially constructed
nature of the human cognitive agent’. In view of this, the structure and theoretical rationale
of the theory of planned behaviour is useful to study farmers’ soil and water conservation
practices, particularly with respect to the role of the underlying beliefs and the notion of a
perceived behavioural control (Beedell and Rehman, 2000). However, unlike attitude, that
could be either negative, neutral or positive, perceived behavioural control involves wider pos-
sibilities that are subject to many influences even for the same level of control from one
respondent to another. Therefore, it is not convenient to apply a scale for mathematical appli-
cation. On the other hand, applying the model in rural communities where illiteracy
dominates and where criteria for measurement of belief levels widely vary such as in agricul-
ture and natural resource management, is likely to lead to substantial bias. Therefore, this
study limited the application of the model to the conceptual link between behaviour, the
underlying beliefs, and environmental factors without using the mathematical relations
between the behaviour and composite explanatory variables. These are the attitude towards
behaviour, perceived social norm, perceived behaviour control and intention (intervening).
Considering soil and water conservation as a desired behaviour, farmers should first and
foremost know and perceive the erosion problem. Information on these subjects is likely to
shape their attitude towards erosion and its control measures. The cognitive base could be
obtained from outside or from the community or personal experience. The farming system
sets the norm for land-management practices. A favourable outcome from the cognitive base
and the normative belief are not sufficient to warrant implementation of a conservation prac-
tice in the absence of perceived behavioural control; that is, the resources. Put in another way,
availability of the necessary materials alone does not guarantee that farmers will implement
or maintain the conservation structure without the cognitive base. This study examines the
relative importance of these pillars in the assessment of determinants of soil and water con-
servation in selected regions of Ethiopia. More specifically, it examines to what extent

60
.tesfaye.thesis 29-01-2003 15:12 Pagina 61

knowledge and attitude towards erosion and soil and water conservation affect farmers’
behaviour regarding SWC practices.
The preceding sections of this chapter have dealt with some theoretical lines that are rele-
vant to the research area. In the next section, reviews of empirical cases are briefly presented.

3.8 Empirical studies on determinants of soil and water conservation

Determinants of soil and water conservation have been studied together with thousands of
agricultural interventions deployed by projects and mainly the public extension services.
Determinants of soil and water conservation can be analysed at four levels. These are: the
household, focusing on personal attributes and household characteristics; the farming sys-
tem; characteristics of technology; and external institutional factors. The farming system
includes the social, institutional and natural environments, whereas the external factors
mainly deal with macro-economic policies, institutional services and rules and regulations.
Even though they are separately treated for ease of presentation, all of these factors are high-
ly interwoven at different levels of analysis. The body of literature dealing with these factors
is too vast to be canvassed here. In view of this, a few empirical studies are reviewed to show
the salient issues determining farmers’ decisions regarding soil and water conservation prac-
tices.

3.8.1 Household variables

Since the ‘diffusion of innovation’ school (Rogers, 1983; 1995) opened up a host of studies in
the area of adoption and human behaviour, a number of socio-economic studies have been
carried out to identify factors affecting the adoption of SWC among farmers. These include:
age, gender, education, family size, farm size, labour availability, income, risk perception,
perception of erosion, perception of soil and water conservation technologies, off-farm
employment, etc.
Cramb, et al. (1999) have studied factors affecting soil and water conservation technologies
(terrace and hedgerows) in the upland areas of the Philippines. The study concluded that the
personal attributes such as age 14 and education were not important in explaining adoption.
Out of seven sites, age of the household was found to be a significant factor in only one loca-
tion where laborious work of stone wall construction was involved. Better education was
expected to influence farmers’ comprehension of the training materials. Like old age, heavy
tasks were not suitable for female-headed households. Adopters and non-adopters did not dif-
fer much with respect to their family size. The study by Semgalawe (1998) in the Tanzanian
Mountains did not show any relationship either, between farmer age and education on their
decisions about SWC measures. In the results of the study on ‘tassa and earthen mounds’ in
the Sahel by Baidu-Forson (1999), age did not show a significant effect on the adoption of these
practices. However, Bekele (1998) in his study of resource degradation and adoption of land-
conservation technologies in the Ethiopian highlands, found age and family size to be nega-
tively and significantly related to the farmers’ decision to retain conservation structures,
whereas education was not significant in this relationship, because most farmers were

14 Rogers (1995: 269) indicated that responses of adopters show inconsistent evidence about the relationship of age and
innovativeness; about half of the same 228 studies shows no relationship, a few show that earlier adopters are younger,
and some indicate that they are older.

61
.tesfaye.thesis 29-01-2003 15:12 Pagina 62

illiterate. In another study in northern Ethiopia, Ludi (2002) found that with increasing age,
the probability that SWC structures are maintained slightly decreases. Lapar and Pandey
(1999) found a positive correlation between farmers’ years of education and their adoption of
SWC measures.
Farmers’ perception of soil erosion and their subsequent conservation behaviour have
mixed results. In some studies, there was no substantial relationship between soil erosion per-
ception and farmers’ conservation behaviour, whereas in others, there were direct links. For
instance, the perception of erosion was found to be important to the adoption behaviour of
SWC in the Philippines (Cramb, et al., 1999) and at Andit Tid, Ethiopia (Bekele, 1998). Bekele
particularly showed that farmers’ decisions to retain conservation structures are positively
and significantly related to soil erosion perceptions, attitude towards new technology and
exposure to new practices. Such was not the case in Tanzania where farmers’ perception of
the soil erosion problems fails to explain household behaviour towards adoption of improved
SWC practices. And even more, the Tanzanian study found that non-perception of the soil ero-
sion problem does not always translate to households being unwilling and/or unable to use
improved soil conservation measures. Summing up such observations, Napier and Sommers
(1993:10) state that: ‘positive environmental attitudes towards the protection of soil and water
resources are a necessary but not sufficient condition to bringing about the adoption of con-
servation programmes at the farm level’.
The farmers’ decisions on SWC technologies are highly influenced by risk perceptions
(Ruthenberg, 1985; Baidu-Forson, 1999). The study by Baidu-Forson (1999) suggests that farm-
ers exhibit reluctance to adopt technologies that expose the farm enterprises to greater risk
and must be convinced that technical change will indeed bring about greater rewards than
existing practices.
Farm size is one of the crucial household variables that enter into the decision-making by
households with respect to soil and water conservation technologies (Kerr and Sanghi, 1993;
Sain and Barreto, 1996; Cramb, et al., 1999; Bekele, 1998; Semgalawe, 1998). In the Philippines
(Cramb, et al., 1999), farm size affected the adoption of soil and water conservation methods
in two of seven sites where the technology was introduced. The study found that a larger farm
size enabled adopters to increase the maize area to offset the area lost to hedgerows, thereby,
among others, maintaining total food production. Comparable to that of farmers in the
Philippines, farmers in Tanzania, tend to be willing and able to use improved soil and water
conservation techniques when their farm size is larger (Semgalawe, 1998). Bekele (1998) also
found that per capita availability of cultivated land, parcel area and productivity of the
technology positively and significantly affected farmers’ decisions to retain conservation
structures. Baidu-Forson (1999) takes this issue even further by linking his finding of a
negative correlation between the adoption of improved ‘tassa’ and earthen mounds and the
land-to-labour ratio with the hypothesis of adoption of intensification as population grows
relative to land resources.
Researchers who emphasise soil fertility management closely link livestock ownership with
the use of organic matter (Eyasu, 1997; Alemayehu, et al., 2001). Others try to find out the prob-
ability of ownership of livestock or specific animals (e.g., oxen) affecting farmers’ decision-
making regarding SWC technologies (Bekele, 1998; Ludi, 2002). The common observation is
that livestock serves as insurance for family food security, settlement of financial commit-
ments, provision of manure and traction power.
Labour is one of the few crucial inputs in soil and water conservation practices. Given a
relevant technology, studies have reported varying responses to labour in SWC decisions.
Cramb, et al. (1999) found that labour requirements for the technologies were an important

62
.tesfaye.thesis 29-01-2003 15:12 Pagina 63

consideration, but the household’s labour supply was not a major factor in itself. Rather, it
was related to cash flow concerns and the need to use spare labour off-farm in preference to
carrying out SWC. Therefore, access to cash was more important in giving them access to
labour and the use of other inputs to improve production, such as fertilisers. This was clearly
shown by higher cash income of adopters in all the sites compared to the non-adopters in their
study. Semgalawe (1998) in a similar study in Tanzania, found that households with cash crops
were more likely to adopt improved soil conservation measures than those who did not grow
at least one cash crop. On the other hand, off-farm income had a negative effect on the use of
improved SWC measures. Moreover, she also found that farmers who were members of a
labour-sharing group were more likely to adopt improved SWC measures than those who were
not. Interestingly, the result indicates that family labour does not influence the decision to
establish improved soil conservation structures, which implies the importance of the labour-
sharing group in SWC practices. Lapar and Pandey (1999) reported a similar result regarding
local labour exchange groups in the highland Philippines. In Thailand, labour demand was
more pronounced due to agronomic practices that need continuous attention from planting
to tending (Renaud, et al., 1998). Similarly, Pender and Kerr (1999) found that labour market
imperfections in India resulted in significantly more conservation investment in households
with more adult males, fewer adult females (an unexpected result), low caste households, and
on smaller farms.

3.8.2 Farming system

Physical factors
Biophysical factors and climate, mainly rainfall characteristics, play an important role in
farmers’ decisions regarding soil and water conservation practices (Sain and Barreto, 1996;
Semgalawe, 1998; Bekele, 1998, Cramb, et al., 1999; Lapar and Pandey, 1999). Farm attributes
such as distance from the homestead and its physical conditions such as its slope, soil type,
topsoil depth, stoniness, etc., were found to discriminate between adopters and non-adopters
of SWC in different parts of the world. These features may also lead to different decisions by
the same farmers with respect to different plots owned, which vary in these physical features,
let alone by different farmers. The study by Cramb, et al., (1999) showed that the adoption of
hedgerows in particular was more likely on fields, which were larger, steeper, had more erodi-
ble soils, and were located close to homestead areas. Bekele (1998) who studied 452 plots in the
highlands of Ethiopia, found that the higher the slope category of the plot, the higher the
probability that recognition of soil erosion would be above any fixed level.

Local institutions
Note that the ‘diffusion of innovation’ school of thought hardly addresses the importance of
local institutions as an important contributor to the adoption of suitable land-management
practices. Instead, subscribers of the school often consider local institutions to be a barrier to
diffusion of innovation, or at best overlook it as an irrelevant force. The role of local institu-
tions and the embedding culture became important in land management studies that looked
at ‘the soft-side of land’ (Röling, 1997). The importance of the soft side of land management
has been demonstrated in a number of studies; for instance, Gonzalez (2000), and Mazzucato
and Niemeijer (2000). Mazzucato and Niemeijer (2000) illuminate the role of local institutions
in adaptive land management. Their study underlines the role of the cultural economy that
operates on the basis of a mixture of principles emerging from both the local and cultural

63
.tesfaye.thesis 29-01-2003 15:12 Pagina 64

contexts and the external market economy. In Katheka, Kenya, the local self-help group
helped to promote bench terracing which was successfully reinforced in the early 1980s while
it had not been adopted under colonial rule (Laing and Ashby, 1993). This case was also able
to report on the exemplary collaboration between a local self-help group and the government
extension service in land improvement. Local self-help groups pursuing wider local develop-
ment issues or those organised for specific functions such as labour exchange develop their
own rules of the game, which are of vital importance when development interventions of any
sort are envisaged (Herweg, 1993; Tsehai, 1994; Hounkonnou, 2001).

3.8.3 Characteristics of technology

The ‘diffusion of innovation’ school has always paid attention to the characteristics of inno-
vation itself in addition to its traditional focus on the ‘innovators’ in trying to determine the
rate of adoption of an innovation (Rogers, 1995). According to Rogers, 49 to 87 percent of the
variance in rate of adoption is explained by five attributes of the innovation: relative advan-
tage, compatibility, complexity, trialability, and observability. This characterisation of tech-
nology is one of the best achievements of the school. To this, he adds other factors that affect
an innovation’s rate of adoption. These are: (1) the type of innovation decision (optional, col-
lective, authority-related); (2) the nature of communication channels; (3) the nature of the
social system; and (4) the extent of change agents’ promotion efforts. According to Rogers, five
perceived attributes of technology together with the four agent-channel components deter-
mine the rate of adoption. Generally speaking, Rogers’ model is complementary to the theory
of planned behaviour as discussed earlier, by providing clarity in terms of decision-making
processes. In addition, among the five attributes of technology, the relative advantage reflects
what other studies present as economic factors. Compatibility reflects adaptability of the tech-
nology/innovation to the cultural, institutional and technological settings of the adopting
society, while other attributes manifest the adopters’ personal characteristics with respect to
the innovation, including their socio-psychological characteristics.
Like any other technology, characteristics of technology also affect soil and water conser-
vation decisions. Farmers do not accept practices that merely reduce erosion rates from a
purely agronomic point of view unless they also address the soil-fertility problem. This often
brings a choice between the physical measures that are known to control erosion rates, and
the biological measures that have a quick response to soil fertility objectives, while also pre-
venting erosion and being easy to install or manage. However, this does not mean that farm-
ers do not refrain from adopting biological conservation technologies, as these too have their
own disadvantages depending on who has tried to adopt them (Hudson, 1993; Renaud, et al.,
1998). For instance, competition with annual crops for nutrients, water, and light; high labour
demand that often coincides with routine farm operations; serving as alternate hosts to dis-
ease and pests. From the technological characteristic point of view, farmers go into all these
details based on their day–to-day observation of each technology. When they are new to the
technology, their reservation may be due to complexity, observability, which also prohibits
farmers’ evaluation of the relative advantages and compatibility of the technology. In crop-
management techniques, farmers may continue to grow an erosive crop, with the knowledge
of its disadvantage when it fetches a high price. The same consideration goes when farmers
remove crop residues for animal feed or home consumption in different forms. They do this
not because they do not understand the contribution of crop residues to soil fertility, but
because they are constrained in the supply of alternative materials. Added to this farmers’ eco-

64
.tesfaye.thesis 29-01-2003 15:12 Pagina 65

nomics is the livestock management system. Where there is open grazing after harvesting,
farmers try to collect their residues for alternative uses (Sain and Barreto, 1996). These con-
siderations are important when one thinks of biological conservation measures and mini-
mum tillage practices where crop residue management is crucial. The relative advantage issue
also raises the need to incorporate SWC practices that give quick benefits to offset the side
effects of the technology, with low cost and low labour demands (Renaud, et al., 1998). This
issue is vividly presented in the study by Lapar and Pandey (1999): ‘engineering solutions such
as rock walls, check dams, and terraces have had very little success in generating a wider
impact due to their high costs. Other less costly technologies such as contour hedgerows, con-
tour ploughing, and cover management have also been adopted but sporadically.’ This obser-
vation underlines the danger of generalising farmers’ decisions regarding SWC in different
parts of the world, without taking into account the specific context. SWC technologies also
involve side effects that increase farmers’ risk of crop loss (Belay, 1992; Herweg, 1993). These
side effects include: waterlogging, rodents, weeds, land occupied by conservation structures,
obstacles to routine farm operations. These features highly reduce the adoption of SWC tech-
nologies among small farmers.

3.8.4 External institutional factors

External factors could enhance or inhibit farmers’ decisions on soil and water conservation.
These factors include: land tenure, access to credit, subsidy, extension services and infra-
structure (Napier and Sommers, 1993; Sain and Barreto, 1996; Cramb, et al., 1999; Bekele, 1998;
Renaud, et al., 1998; Pender and Kerr, 1998; Samgalawe, 1998). The impact of land tenure in the
Cramb, et al. study (1999) varied from tenure system to tenure system. While there was no sig-
nificant impact of the official classification whereby land was under different public cate-
gories, tenancy was found to be the main tenure issue that affected adoption of SWC practices.
In this case, the absentee landowners did not allow the tenants to adopt the SALT (Sloping
Agricultural Land Technology) recommendation. In another study in the Philippines which
included one of the villages that was covered in the study by Cramb et al., tenure security was
highly valued, without which farmers would resort to a ‘mining’ strategy based on the rapid
exhaustion of soil fertility (Lapar and Pandey, 1999). However, they indicated that a legal title
to land was not necessary for ensuring the security of land tenure. Bekele (1998) suggests a
possible association of farmers’ perceptions of the security of user rights to land with a high-
er level of use of conservation structures, but his results were not statistically significant.
Given that the state ownership of land prohibits land markets, he indicated the limitations to
undertaking conservation investments with a long payback period, including conservation
structures considered in his study. In a study carried out in three Indian villages, imperfect
land markets caused lower conservation investment on leased land (Pender and Kerr, 1998).
Sain and Barreto (1996) reported the absence of a relationship between land conservation and
land tenure.
Extension is generally expected to show a positive impact on the adoption of external tech-
nologies (Rogers, 1995). This view has been testified in a number of studies including those
involving SWC (Bekele, 1998; Semgalawe, 1998; Baidu-Forson, 1999). According to Bekele
(1998), access to information about technological options for soil conservation had a signifi-
cant effect on perceptions of the erosion problem and on retaining the conservation
structures, whereby he considers a positive role of extension on adoption. Similarly,
Semgalawe (1998) reported that farmers’ participation in promotional activities of SWC

65
.tesfaye.thesis 29-01-2003 15:12 Pagina 66

programmes increases the probability for the household to use improved SWC measures.
However, top-down extension approaches that heavily depend on incentives rather than on
educational processes hardly help to change land-management practices of farmers (Renaud,
et al., 1998).
The other variable that has been synonymous with soil and water conservation, particu-
larly in Africa, is the incentive that is mostly stimulated in the form of food-for-work pro-
grammes. The literature on the diffusion of innovation regards the provision of incentives as
an attempt to increase the degree of relative advantage of the new idea or practice, but also
as having other functions such as increasing observability and trialability which enables the
testing of compatibility (Rogers, 1995: 219).
Incentives can be provided in a variety of ways, including credit, subsidies, grants and
infrastructure. Reij (1998:1417-1418) sees incentives as a triggering mechanism for the adop-
tion of SWC technologies. This is shown in the case studies on which he reported from
Burkina Faso and Indonesia (Java). In Burkina Faso, he reported on a new initiative to provide
a donkey cart for the transportation of stones, manure, water, firewood, etc., which directly or
indirectly contributed to land management, instead of the expensive option such as a lorry
that would not be sustainable after the project terminated. His example from Java shows how
livestock credit to attract farmers to plant grass on terrace risers and terrace lips failed as this
was not attractive to resource-poor farmers who are hardly affected by soil erosion, because
their soils are deep and fertile. Eligibility for these provisions required the adoption of SWC
practices in the form of a cross-compliance incentive 15. The most crucial issue to be considered
before deploying incentives, however, is acceptability of the new practices in the face of the
evaluative frames of reference of farmers envisaged in the incentive. The reason for this con-
dition is that the farmers are expected to continue the new practice after the phasing out of
the incentive. That means that the incentive was deployed merely to overcome risk-aversion
by farmers and to increase their relative advantage. This phenomenon is termed as ‘dynamic
dis-equilibrium’ where the use of incentives in the form of a subsidy could be considered as
economically justified (Giger, 1998). Deploying incentives without winning farmers’ accep-
tance risks the destruction of previous investments and the absence of maintenance as
demonstrated by hundreds of projects around the world.
The study by Pender and Kerr (1996) reported that credit-market imperfections are affect-
ing investment in each of the villages included, though the evidence was not completely
conclusive. The most important issue with respect to access to credit is its link with land man-
agement; i.e., is credit used for hiring extra labour, purchasing fertilisers or other inputs?
Otherwise, mere access to credit does not guarantee proper land management as credit could
be used for consumption. In this connection, Ludi (2002) reported that increasing the amount
of credit brought on a slightly to medium significant negative effect on the willingness to
maintain SWC structures. She noted that such unusual results regarding credit occurred
because the conservation technology was labour intensive rather than capital intensive,
whereby cash constraint plays a minor role.

3.9 Analytical framework

As illustrated in the empirical studies, land-management as a complex process is not only the
result of the will or an act of land users. Its problems and achievements go way beyond the
household domain of operation, to include actors in the surrounding environment who also

15 For the detailed policy implications of cross-compliance for targeting economic incentives, see Bekele, 1998.

66
.tesfaye.thesis 29-01-2003 15:12 Pagina 67

affect their decisions. These actors operate at various levels and affect farmers’ land-manage-
ment decisions and thereby their behaviour. At the macro level, policies of the national
government are formulated. The global economy and political systems in turn affect these
policies. Macro-level policies affect policies of the regional governments at the meso level.
National and regional government policies become operational at the micro level. At this
level, household characteristics, institutional factors, technical factors and physical factors
affect farmers’ decisions on land management. The interrelationships among these factors
and how they affect soil and water conservation are shown in Figure 3.1. In the subsequent
paragraphs, the perspectives of the study with respect to these factors are outlined. Note that
the households are embedded in their biophysical and institutional environments. The latter
one encompasses a range of human agencies.

Figure 3.1: Analytical framework with the determinants of soil and water conservation

External context
Households
Knowledge and attitude on soil erosion
and SWC

+
Socio-economic status
Local and national
institutions Physical factors

Soil and water conservation

Characteristics of
technology

Household characteristics
Households in developing countries like Ethiopia are both production and consumption units
where both functions are closely linked (Ruthenberg, 1985). Therefore, the primary source of
inputs for production emanates from within the household and the local institutions and
family networks. Household goals are dependent on its resources that include both physical
and social capital.
With respect to soil and water conservation, the first issue is the state of soil erosion and
the level of soil fertility. When the household feels the impact of soil erosion in its farm, it is
expected to take the necessary measures. Its response to soil erosion depends primarily on the
level of its knowledge on soil erosion and the resulting attitude. If the household is aware of
the side effects of erosion on its livelihood, it is likely to take the known control measures
against soil erosion. After all, it is threatening its means of livelihood. However, soil and water
conservation does not fall within the domain of behaviours in which one can easily do what
one intends, as posited in the theory of reasoned action (Ajzen and Fishbein, 1980). Therefore,

67
.tesfaye.thesis 29-01-2003 15:12 Pagina 68

it requires other factors in addition to knowledge and attitude. These are: household resource
base (land, labour, capital), education, risk perception; institutional factors (extension
services, credit availability, labour organisation, markets, land tenure, etc.); physical factors
(topography, rainfall and soil factors); and the characteristics of the technology or practice per-
ceived to solve the problem.

Institutional factors
Farmers make decisions within a broader environment or context. One of the elements in the
environment consists of institutions. These can be seen at the local and national levels. The
national institutions in agriculture are often linked to research and extension services. They
are typical routes for external intervention in rural communities. Extension in particular pro-
vides farmers with information on soil erosion and methods to combat it. In collaboration
with other organisations, it could channel credits and other incentives to the farming com-
munity to improve its production through proper land management techniques. The type of
information provided by research and extension institutions affects farmers’ knowledge and
their attitude towards soil and water conservation. External organisations can also exert
pressure on local people through persuasive incentives such as food-for-work. In addition,
land-tenure patterns play their role in the farmers’ land-management decisions.
Farmers’ risk perception on the options for soil and water conservation depends on their
perception of the institutional environment, i.e., the degree to which it may support or
impede the practice. With respect to indigenous knowledge-dominated farming systems, the
local institutions, labour organisations and the moral economy are very important to their
decision-making processes and to the livelihood of the family (Mazzucato and Niemeijer, 2000;
Gonzalez, 2000). Among farming systems that are integrated within the modern institutions,
extension services, financial organisations, markets, input supply, and transportation compa-
nies are very essential for its decision-making and success. Nowadays, both institutions are
accessed by farmers as deemed necessary.

Characteristics of technology
A decision-making process could be enhanced or impeded partly by the intrinsic characteris-
tics of the technology, such as observability, complexity and divisibility/trialability and partly from
its extrinsic characteristics such as compatibility and relative advantage to the household domain
(Rogers, 1983, 1995).
When a technological option is easily observable, a farmer can easily form an idea and pro-
ceed to evaluate the option. Observability of the option helps the farmer to easily compare it
with previous experiences or the problem at hand, compatibility of the practice and its rela-
tive advantage (see below). When the options are less observable, decision-makers would not
have the chance to arrive at a decision to try it. When an option is perceived as complex, it is
likely that their risk perception increases, while their motivation declines, therefore nega-
tively affecting the cognition process and thereby their attitude. When this happens, a farmer
will end up sticking with the original experience and preliminary reflection without a for-
mation of positive ideas to try out the option. Similar to lower observability, complexity also
hinders farmers’ reflections on the compatibility and relative advantage of the option.
Arriving at a decision to take one action among many necessitates the divisibility of the
option (if applicable) to individuals in order to encourage trying out the practice. Note that
some technologies need small-scale trials before committing a large investment.
When the intrinsic characteristics of an option allow the farmer to proceed in the decision-
making process, the farmer would be in a position to evaluate compatibility and relative

68
.tesfaye.thesis 29-01-2003 15:12 Pagina 69

advantage of the option, which are partly in the domain of the environment and partly in the
household domain. When an option proves compatible, it would be tested for its relative
advantage; that is, not adequately assessed from other sources or previous experience.
Depending on the test, the practice would be included in the repertoire of practices for fur-
ther observation.
Compatibility compares the option with the life-world of the farmer such as farming prac-
tices, value and belief system, knowledge, skills, resources, etc. Relative advantage compares
technical performance (on-site and off-site effects), economic considerations (demand for
labour, energy, resource competition, financial return, outside resources), social aspects
(power relation, status, religion), institutional relationships (individual efforts, association,
external assistance) and political/legal issues.
Reflections on compatibility and relative advantage of the option are dependent on the
environment in which the decision-maker operates, which shows complexity of decision-mak-
ing in a real-life situation, and differentiates it from learning for the sake of learning.

Physical factors
Natural landscapes are one of the triggering factors for human interventions. Plain landscapes
and hilly landscapes do require different methods of farming. Different soil types also need
different management practices. Added to these factors, rainfall characteristics raise yet
another land-management issue. Without going into details of other biophysical characteris-
tics that affect farming, these three physical factors are sufficient for shaping farmers’ soil
and water conservation decision-making processes in conjunction with the ‘soft’ side of land,
after Röling (1997). Given a steep slope and erodible soils, a high rainfall period under low
vegetation cover is likely to play an important role in the perceptions and behaviours of farm-
ers on soil erosion and the role of SWC.

Practice and post-decisional evaluation

The analytical framework discussed above refers to pre-decisional processes. Another impor-
tant element of the decision-making process, which is often undermined, is the post-deci-
sional situation. What does the decision-maker do after a decision is made? This is an issue of
great importance when decisions are related to the livelihood of the people involved.
The decision-maker does not close his/her mind from all the processes gone through to
arrive at this decision. The original situation is always revisited by comparing it with the new
situation. Social learning plays its role by showing what others do in similar situations. What
have they gained or lost? Such questions run through the decision-maker’s mind with reper-
cussions to the ongoing practices depending on his or her attitude. When the decision-
maker’s evaluation is consistent with previous evaluative frames of reference, s/he will choose
to continue with the new practice or idea. Otherwise, it may be modified to suit his/her needs,
or rejected altogether, going back to the original practice. Similar action is taken for emerg-
ing issues that are related to this decision-making. This process is explained by the cognitive
dissonance theory coined by Leon Festinger (1957) (cited in O’Keefe, 1993).
Dissonance often consists of two clusters of elements, which on the one hand reflect a set
of favourable cognition to the decision made and on the other a set of unfavourable cognition.
The degree of dissonance experienced depends on the proportion of these two sets of cogni-
tion and on the importance attached to either sets by the decision-maker. Consequently, a
reduction of dissonance requires information to change the relative proportion of the two
knowledge bases and/or to reverse the importance attached to the cognitive elements involved
in the decision-making process. Unless dissonance reduction is carried out at the right time,

69
.tesfaye.thesis 29-01-2003 15:12 Pagina 70

there is a possibility that regret takes place as a post-decisional cognitive process that might
reverse the decision made. This signifies how practice influences the decision–making process
as indicated by the backward arrow from the practice to the rest of the elements in Figure 3.1.
The array of variables indicated above shows potential determinants of local level decisions.
For instance, local level institutions have to react to the directives of the formal institutions
on resource use, on obligations or rights of citizens, etc. As the clusters of factors are closely
interwoven at the local level, change in one cluster would sooner or later affect the other clus-
ters. The pace at which effects on other clusters may be observed depends on several factors
and tends to vary from community to community within and across farming systems. The cru-
cial element is the adaptive nature of the social system (Gold, 1980; Bartels, 2002). Adaptation
in this case assumes the establishment of a new social equilibrium, which is in the process of
a continuous change to form yet another equilibrium and so on, leading to a dynamic system.
Similar trends take place when the force of change emerges from within the local system and
apparently affects the ‘upper’ level systems.
The analytical framework is used to examine the landscape of each study area in a histori-
cal perspective. At the household level, it shows why one household differs from the other in
terms of its land-management decisions. However, there are broad overlaps on the choice of
land management among certain households with others who share common objectives and
constraints, whereas farming systems exhibit more pronounced variations with respect to
land-management practices. In addition, the analytical framework reflects the roles of human
beings who make decisions as situations are unfolding. This makes the issues of land man-
agement a social phenomenon – a socially constructed process and outcome - in the sense that
people adapt to their environment owing to their cognitive abilities, which help them to learn
from day-to-day environments that largely result from human actions. Learning changes
human behaviour because of new insights about the environment. Once again the backward
arrows in the model indicate this spiralling relationship between factors that affect farmers’
resulting behaviour.
In conclusion, this chapter discusses theoretical debates and empirical studies from which
the research questions in this study emerge, and which enabled me to suggest the analytical
framework (see section 3.9) for the case studies.
Research question 1 (‘What are farmers’ responses to the introduction of SWC measures?’)
links to the discussion on planned changed and development intervention (see section 3.3).
Together with the discussion on indigenous knowledge (see section 3.5), this allows me to lay
the basis for comparing technical interventions in chapter 6 and 7 (and their drawbacks) with
indigenous practices in Chapter 5 of this study.
The focus of research question 2 on farmers’ knowledge of and attitudes about soil erosion
and SWC relates to discussions on indigenous knowledge (see section 3.5) and to the link
between attitudes and behaviour (see section 3.7). Of course, rural peoples’ knowledge is not
limited to local knowledge only, particularly given current communication channels and
intensity. However, under the specific conditions of my research areas, local knowledge plays
an important role in land management as the case studies in Chapters 5, 6 and 7 will show.
My third research question deals with what farmers do to conserve soil and water on their
farms. This shows how the academic Boserupian versus Malthusian debate (section 3.2) is mir-
rored at field level in farmers’ current land management practices, showing them to be adher-
ing to the ‘Eco-dynamic’ school, the ‘land degradation’ school, or otherwise. An overview of
the case studies in this respect is presented in chapter 8.
Research question 4 emerges from the presentation of existing empirical studies (section
3.8), in which I summarise where current thinking had reached and to which my study con-

70
.tesfaye.thesis 29-01-2003 15:12 Pagina 71

tributes. My answers to question 4 on the determinants of SWC practices among small farm-
ers within and across socio-economic and agro-ecological environments are also informed by
theoretical insights on the link between attitudes and behaviour (section 3.7).
My analysis of the empirical material of the three case studies, within the context of the
theoretical perspectives presented in this chapter, allow me to answer my fifth question of
how constraints to promote soil and water conservation can be overcome.
I wish to stress the importance of a farming systems perspective and social learning.
Systems thinking allows us to see how one level of analysis is contextualised and embedded
in the larger systems, depending on the scale of analysis. More importantly, it shows the cou-
pling of the hard biophysical component and the soft social and institutional components
when we deal with land-management considerations, e.g., soil and water conservation.
Therefore, a farming systems perspective helps me to examine different issues related to soil
and water conservation within and across farming systems. The idea of social learning (see
section 3.6), as a process that facilitates the construction of knowledge in the society through
joint learning also cuts across all five research questions of the study. For example, it helps to
explain farmers’ reactions to the external interventions and how they construct and share
indigenous insights.
The next chapter discusses the core of the field methodology used to operationalise the
analytical model presented in section 3.9.

71
.tesfaye.thesis 29-01-2003 15:12 Pagina 72
.tesfaye.thesis 29-01-2003 15:12 Pagina 73

4 Research Methodology

4.1 Introduction

This chapter presents an overview of the field methodology and related issues. The overall
framework of the study is the case study research as presented by Yin (1984). It consists of qual-
itative research and quantitative survey research. The data collection process was divided into
two phases. Phase one entirely dealt with qualitative methods, while phase two focused on the
survey methods, though the qualitative data gathering continued until the end of the field-
work.
Section 4.2 presents an overall design of the study. The preparatory activities for the field-
work are presented in section 4.3. This is followed by an overview of the qualitative research
methodologies in section 4.4. The survey research methodologies are outlined in section 4.5.
Section 4.6 presents the experiences concerning compensation to the communities involved
in the field research.

4.2 Overall design

The study presented in this thesis made use of a golden opportunity to assess the impact of a
major soil and water conservation effort in Ethiopia upon its termination, the Soil
Conservation Research Project (SCRP). This project was implemented in six research sites that
introduced and evaluated the technical impact of soil and water conservation measures that
were largely established through food-for-work programmes to motivate local farmers to
implement the measures on their land. In order to make use of this opportunity I selected
three case study areas: Two from the SCRP and one based on an indigenous SWC system.

73
.tesfaye.thesis 29-01-2003 15:12 Pagina 74

1 Wolaita (Chapter 6): a relatively high potential area with deep soils, but high population
pressure and high rainfall, and hence high erosion, in southern Ethiopia where the SCRP
had carried out a particularly interesting experiment. It picked one micro-catchment (with
92 families) and introduced, through food-for-work, all the anti-erosion measures possible,
though focused in particular on establishing bunds on farmers’ fields. A neighbouring
micro-catchment was used as a control site (with 89 families). All possible erosion control
bunds in this latter area were forcibly removed. The sediment load and run-off from the two
micro-catchments were monitored through a specially constructed device installed at the
confluence of the streams in the two micro-catchments, with the main river (see Plate 6.2).
My study uses as case study areas (a) the experimental micro-catchment (treated), (b) the
control (untreated), (c) an adjacent catchment, and (d) a distant catchment site. Qualitative
work was done in all these areas, followed up by a quantitative sample survey.
2 South Wello (Chapter 7): a relatively poor, famine-prone and degraded area in the north of
the country where SCRP has carried out a major project in the Maybar catchment to intro-
duce a variety of soil conservation measures. My case study simply studied the whole catch-
ment without imposing any controls. The major purpose was to study the impact of the
experimental project and the reaction of the local people to it.
3 Konso (Chapter 5): an area characterised by a farming system that is nationally famous for
its indigenous soil and water conservation practices. Vast hillsides are covered by terraces
and provide scenes not unlike those in the Northern Philippines (e.g., Ifugaos: see Gonzalez,
2000), Indonesia or other Asian countries. I studied this area to get some idea of what it
takes for an indigenous farming system to pay serious attention to SWC. The case study in
Konso has more of an anthropological nature since I wanted to understand the habits,
labour deployment, institutions and spiritual dimensions of SWC in that area. In Konso, I
did not follow a watershed approach, as I did in the other two case studies. I selected vil-
lages based on an exploratory visit I made in 1998, dividing the Wereda into two areas,
namely an area characterised by a hand-tool farming system and one based on ox-plough-
ing.

These three case studies form the empirical base of my study. The choice means that I have
operated on different system levels. In the first place, I operate at the case study level, com-
paring the three different case studies. In the second site, in Wolaita, I compare the four dif-
ferent areas, the treatment area, the control, the adjacent and distant watersheds. In the sec-
ond place, I compare households (i.e., household heads) with respect to their attitudes,
knowledge on soil erosion and SWC on one hand and their practices regarding SWC on the
other. I have spent considerable effort to operate also at the field or plot level. The way farm-
ers systematically treat different plots is sufficiently different to warrant this effort.
The overall design used is therefore a case study approach, where case studies were select-
ed to provide me with as wide a range of insights into farmers’ SWC behaviours as possible in
order to answer the research questions (see Chapter 1). Within each case study, I have con-
ducted quite a major study devoting 1.5 years to qualitative work during intermittent visits to
the three areas, and following this up in Wolaita and South Wello, with sizeable quantitative
random samples with questions based on the results of the earlier qualitative work carried
out.
Below, I explain in greater detail the methods of data collection I used in the study. In each
chapter I will further provide details of the methods followed in each case.

74
.tesfaye.thesis 29-01-2003 15:12 Pagina 75

4.3 Preparatory activities: exploratory visits, meeting communities and

field staff
In order to gain better insight into the selected sites, the researcher visited all of the study
areas for a period of two to three weeks. The first visits targeted the entire zones to develop an
understanding of the broader farming system, and the institutional and economic environ-
ments. Contact was made with the bureaus of Agriculture in study regions, followed by visits
to zonal and Wereda offices. During these visits, extensive discussions were held with key indi-
viduals to learn about the general picture of the agrarian system, major problems, types of
interventions, areas of focus, etc.
Subsequently, the researcher became acquainted with the SCRP research stations and
identified the watersheds to be included in the study. In the process, the researcher was intro-
duced to the community and the grassroots workers.

4.4 Qualitative research

4.4.1 Group interviews and selection of individuals

Group interviews
Group interviews and focus group discussions were used depending on the nature of infor-
mation sought. Focus group discussions included; the elderly, participants and non-partici-
pants in SCRP research and school drop-outs who had become full-time farmers.
Group interviews were set up progressively through the first study period. Attempts were
made to develop a rapport with the community through short informal discussions that did
not focus on taking notes. These discussions deliberately took place in the usual public meet-
ing places in the village. Sometimes, people gathered while an individual interview was in
progress. Looking at the composition of the voluntary audience, topics of a more general
nature rather than focusing on the household level were introduced for discussion. In this
manner, efforts were made to show the purpose of the study to the community. The informal
group discussions were useful for identifying key informants and participants in group inter-
views. Names of some participants were also suggested by other participants and extension
agents. Transect walks at an early stage of the fieldwork and visits to individual farm plots
were good opportunities to meet community members and to become acquainted with the
diversities of the land-management practices in each of the farming systems.

Selection of individuals
Individual interviews during this phase were exploratory, but also explanatory in the sense that
people were directly providing their explanations for actions.
A thorough understanding of farmers’ decisions was achieved through farm-plot sampling
rather than per farmer, as is commonly done. The idea of farm-plot sampling emerged in the
field during the exploratory visits of the study areas. Some degree of variation was observed
among farm plots in the selected watersheds. Some farm plots have well-constructed, tall
stone or soil bunds while the adjacent ones have no, or poorly maintained, bunds. Some bunds
have vegetation cover, while others do not. Some farmers have removed bunds or shifted their
original location and so on. In addition, it was learned that one household treats his/her plots
differently for various reasons (distance from home, gradient, availability of construction

75
.tesfaye.thesis 29-01-2003 15:12 Pagina 76

materials). Looking at such variation under more or less similar biophysical conditions, the
researcher found it more informative to use the farm-plot as a sampling unit during the qual-
itative phase.
Individuals were identified based on variations observed in the farm. Field assistants iden-
tified the owners of the selected farm plots. Efforts were made to include as much diversity as
possible.

4.4.2 Methods of data collection

Discussions with managers and experts at various levels in both government and non-gov-
ernmental organisations were held with the help of key questions on the topical areas. The
methods of data collection discussed below refer only to group and individual cases identified
according to the procedure outlined above.

Narrative interviews
Narrative interviews were used mainly with groups. Interviews with the elderly began with
historical events and converged on the contemporary development problems. The researcher
began the interview with a brief outline of major topical areas. Details of the interview ques-
tions were formulated from responses of the group. Discussions were allowed to take place to
reach a consensus on dates and particular events of importance. Major historical landmarks
were provided to facilitate the narration and to help recall details. Such interviews were held
for two to three hours depending on the resourcefulness of the group. After the first narrative
interview, the researcher thoroughly examined the field notes and formulated further ques-
tions to fill the gaps during the second round of interviews that shortly followed the first
round. During the second interview, unique issues were explicitly clarified to ensure the cor-
rect interpretation of the message. In limited cases, more visits were required if comparable
responses were not found from a similar group elsewhere in a given case study area.
Formulating the questions during the first and second rounds of interviews on critical gaps
have enabled the researcher to gather adequate information on the topics of interviews. A his-
torical background of each farming system was achieved through this process.
Group interviews on contemporary issues were usually completed in one session.
Additional issues were treated during similar group interviews in each area. Group interviews
in a particular site were finalised when variations of responses on issues initially considered
and those that emerged during the session nearly converged or reached a point of saturation.

Semi-structured interviews
Semi-structured interviewing was the major instrument to gather information from individ-
uals selected through the farm-plot sampling technique. Some questions were prepared during
the preparatory phase of the study. The field guide prepared on the ‘researcher’s desk’ was a
general field guide. It was based on preliminary ideas and concepts conceived from literature
reviews and the personal experience of the researcher. The field guide was enriched through
discussions with experts in the field, exploratory visits and observation of the research sites.
The interview schedule was periodically improved in light of the insights obtained in the
course of the 18-month stay in the home country. It should be noted that visits to each
research area were carried out one after the other.

76
.tesfaye.thesis 29-01-2003 15:12 Pagina 77

Key informant interviews

Key informants were used at a later stage of the study on a few issues of which adequate infor-
mation had not been acquired from the other informants. Key informants were used for
generation counting 16 in Konso and community-based organisations in Wello.

Observation
The researcher periodically visited the specific research sites and the wider agrarian system in
which they are situated during the course of the year and a half of the first field research
period. This involved both dry and wet seasons in 1998, 1999 and 2000. During these periods,
he was able to learn several things by observation. Issues that emerged from observation were
used to guide interviews and discussions with people of each study area.
Transect walks and visits to individual plots created opportunities for observation. Closer
observations were made in selected fields where measurements on a few physical and biolog-
ical characteristics of the farm plots were carried out (see below).

Measurements
Biophysical data used in this study mainly come from secondary sources. However, seven
parameters that were found to be desirable for clarifying qualitative observations and respons-
es from the community were measured during the course of the first field period. These
parameters are: slope, altitude, bund height and width (where applicable), tree and shrub
counts, and plot size.

4.4.3 Methods of data analysis

The type of data analysis depends on the type of data itself. There are different, less structured
methods of data analysis for qualitative research. Grounded theory commonly uses coding, tech-
nical memos and diagrams to systematically scrutinise the data from time to time (Strauss
and Cobin, 1990). Yin (1984) also suggests data analysis techniques that are suitable for case
study research. These are: pattern-matching, explanation-building and time-series analysis.
Crabtree and Miller (1992:17-18) categorise data analysis strategies along a continuum. At one
end of the spectrum, they see some techniques of analysis as ‘objective’ in the sense that they
tend to isolate the researcher from the object of the research, towards the other end, tech-
niques exist that are subjective, experiential, context-dependent, existential, interpretative,
and generative.
In this study, data collected with qualitative methods were subjected to in-depth analysis
using codes, pattern-matching and explanation-building as deemed necessary (Yin, 1984).
Then different categories were created according to the objectives of the study. For example,
land tenure, SWC practices, soil fertility, labour organisation, etc., The categorised informa-
tion was used as a pool of information for writing up the different sections where applicable.
Note that these categories were used to guide the data collection process too.

16 The Konso people have ceremonies whereby the shift of social responsibility passes from one generation to another. Each
generation is marked by a stone, which lies half buried in the ground at the location specified for this purpose.

77
.tesfaye.thesis 29-01-2003 15:12 Pagina 78

4.4.4 Field assistants

At each site, two assistants who know the community and the agriculture in the area were
selected to provide assistance during each visit to the site. The individuals were either high
school students who quit their studies for several reasons, or employees of the SCRP/Ministry of
Agriculture. In Wolaita, a high school graduate who became a full-time farmer was used, where-
as in Konso, a senior employee of the Ministry of Agriculture, who graduated from Alemaya
University of Agriculture with a Bachelor of Science Degree in Plant Sciences, was used as a
field assistant. The Konso assistant was also assisted by a senior employee of the Department of
Agriculture who has insight into the local culture.
Assistants in Wolaita and Konso were also employed as interpreters of the local language,
even though people who speak the official language of the country were also often encoun-
tered, particularly in Wolaita. Language was not a problem in Wello as the researcher speaks
the same language.

4.5 Survey research

The survey research was conducted in two of the three cases selected for this study. These were
Wolaita and Wello. Its main purpose was to provide further insights into variables identified
during the qualitative phase. The variables used in the survey instruments were tested through
semi-structured interviewing in each location. The two sites were selected due to intensive
research and extension interventions having been carried out in soil and water conservation,
among other practices.

4.5.1 Secondary data

Data/information at the macro and meso levels provided mainly secondary data. These data
were gathered at various levels ranging from the federal to the smallest administrative unit of
the Peasant Association. These sources of information include: research and extension organi-
sations, authorities dealing with statistics, NGOs and Peasant Associations.

4.5.2 Design of the survey

Units of study
The units of observation were the selected watersheds in the case study areas, whereas the unit
of analysis was the household, its members, farm plots and livestock.

Sampling frame
The sampling frame for the survey was developed from a list of farmers who own farms in the
watersheds where the exploratory study had been carried out (see Chapter 6 and 7 for details).

Sampling technique
A two-stage cluster sampling method was used in Wolaita, while simple random sampling was
used in Wello. The variation in sampling technique is partly due to the experimental design of
the SCRP and partly the nature of the research investigation (See chapters 6 and 7 for details).

78
.tesfaye.thesis 29-01-2003 15:12 Pagina 79

In Wolaita, the objective was to assess possible variation among the watersheds included in
this study consisting of four sub-samples, whereas in Wello, only one sample was involved.

Sample size
The sample size was determined taking into account the degree of heterogeneity in the pop-
ulation with respect to soil and water conservation practices. Accordingly, 107 and 231 indi-
vidual interviews were carried out in Wello and Wolaita, respectively.

Survey instrument
A structured survey questionnaire was used to collect the data. Variables included in the ques-
tionnaire were identified during the exploratory phase.

Use of enumerator
Unlike the exploratory phase where the researcher alone conducted the interviews and obser-
vations, enumerators were used for this phase, following an intensive training course and crit-
ical assessment of their performance during peer interviews, researcher interviews (role play)
and finally, a pre-test.

Methods of data analysis

The study employed both qualitative and quantitative research methods. The Konso case study
emphasised qualitative methods and secondary data, while in Wolaita and Wello, both types
of data were applied. The statistical analyses used were: descriptive analysis, measures of asso-
ciation, the general linear model (univariate, post-hoc multiple comparison), and non-para-
metric statistics (chi-square test and Kruskal-Wallis analysis of variance).

4.6 Compensation of the community

Nowadays, some communities are expecting remuneration after an interview. This practice
was introduced by western researchers who conducted research under great time pressures,
which compelled them to do anything to get their interviews - virtually buying them.
However, researchers using local funds, or funds administered through local channels can
hardly do this.
The expectation of the community also varies from place to place. In this study, the Konso
and Wello areas were found to be more demanding than Wolaita. Konso is now a touristic site
and en route to the south where tourists have introduced cash payment for everything. The
Wello site has developed this expectation due to research fatigue from foreigners and exter-
nally funded local researchers.
In principle, the researcher has no objection to compensating the community members for
the their time and efforts contributed to the research process. However, as the ultimate owner
of the history and knowledge of the natural resources and agriculture in a given area will be
the community, the researcher preferred to contribute to the collective services rather than
paying individuals who participate in group or individual interviews. In his view, providing
some collective entertainment in accordance with the local social practices will help to devel-
op a greater rapport with the community than monetary handouts. In the end, a small grant
was transferred to the community-based organisations in Wello site and the provision of
social entertainment (local drinks) in Konso and Wolaita. In addition, a token amount was
paid to individual farmers at the end of the study period.

79
.tesfaye.thesis 29-01-2003 15:12 Pagina 80
.tesfaye.thesis 29-01-2003 15:12 Pagina 81

5 Society and Indigenous Soil and Water Conservation in

Konso, Southern Ethiopia

Abstract

Notwithstanding the introduction of modern practices over the last forty years at least, indigenous land
management is the major feature of Ethiopia’s agriculture which is said to date back over 2,000 years.
Land management in Ethiopia has evolved into various farming systems with different levels of intensifi-
cation. The indigenous agricultural system in Konso was studied with a focus on soil and water conser-
vation. The level of intensification in Konso shows a practice that is far ahead of the Boserupian thesis. It
is characterised by stone-based terraces and well integrated agroforestry practices. It has existed for at
least four hundred years. The strength of the system is the embodying culture and its institutions that fos-
ter this kind of agriculture. More specifically, its organisation of labour, indigenous population control
methods and system of governance that protect the farming system have allowed this system to thrive. The
uniqueness of the location makes its methods less applicable to other areas, yet it is an exemplary situa-
tion that can give a boost to conservation attitudes of other cultures. In spite of such a legacy, Konso’s ter-
race-based agriculture is showing signs of deterioration due to the alteration of its internal and external
driving forces: mainly, the abolishment of the population control methods, integration into wider markets
and institutions, and the shortage of rainfall. The study concluded that the continuation of livelihoods in
Konso now requires heavy external support to ease the pressure on land, by promoting irrigation and
developing valleys for agriculture.

81
.tesfaye.thesis 29-01-2003 15:12 Pagina 82

5.1 Introduction

The Konso people live in southwestern Ethiopia, in the former Gomu Gofa province. They are
well known for their stone terraces 17 that are believed to have existed for over four hundred
years. In spite of external interventions, the system has maintained its characteristics, albeit,
with its own pace of dynamism. In appreciation of their contribution to natural resources con-
servation, the United Nations extended an award to the Konso people. In addition, Ethiopian
Scientists (mainly, Anthropologist/Sociologists) and their associates are preparing to present
Konso as a candidate World Cultural Heritage centre (Dr. Yonas Beyene, 2001, p.c.).
Stone terracing provides a typical soil and water conservation structure, covering most of
Konso. There, land management is highly integrated and implemented within watershed
development, not easily replicable elsewhere. Hill treatment is initiated from the bottom of
the valley to the peak. These stone terraces have been built from internal motivation and insti-
tutions, and from the personal experience of the Konso people without any external influ-
ences or forced labour programmes. In a nutshell, this system has produced an architecture
that was developed in a particular landscape and institution. Explaining the uniqueness of
Konso’s terraces, Hallpike (1972:21) writes: ‘perhaps nowhere else in traditional Ethiopia has
the hand of man so impressed itself on the landscape as in Konso.’ With an exception of
Watson (1998) who brought a new insight about land tenure in Konso, previous studies have
not adequately addressed the interrelatedness of the social and technical dimensions of agri-
culture in Konso (Meseret, 1990; Ambron, 1984; Teferi, 1992; Kruger, et al.; Otto, 1994).
Konso’s agriculture has received wider popularity both in Ethiopia and to some extent
internationally. Unfortunately, there is little understanding on how to undertake measures
that help to restore its internal dynamics. The major reason for this misunderstanding is lack
of information on the genesis and processes of Konso’s agriculture.
This chapter presents Konso’s agriculture in its broader context. Its emphasis is on farmers’
land management behaviours and their institutions. The key questions are: How did Konso’s
land management emerge and survive up to now? What is the knowledge base for this land
management? What institutional supports are behind this knowledge and the practices?
What are mechanisms of social learning among the community members? Can we adapt
Konso’s experience elsewhere? What are the current situation and future prospects of Konso’s
agriculture?
A description of the study area is presented in section 5.2 followed by a methodology in 5.3.
The socio-economic situations and an overview of the production systems are discussed in 5.4
and 5.5, respectively. With these current backgrounds, the chapter goes on to the historical
roots of the Konso people and their institutions. Thus, origin and identity (5.6), basis of social
differentiation (5.7), land tenure (5.8) and local institutions (5.9) are discussed subsequently.
Following that, section 5.10 presents the different forms of labour organisation in Konso. The
place of beliefs, institutions and social learning in Konso’s agriculture are examined in 5.11.
Thereafter, the chapter discusses at length farmers’ knowledge, land management practices
and problems in sections 5.12 and 5.13. Finally, the conclusion is presented in section 5.14.

17 Note that terrace is used in this chapter to indicate the physical conservation measures that have already attained a
bench-level status as opposed to the soil/stone bunds in chapters 6 and 7 that were designed to increase in height slowly
(slow-forming bunds).

82
.tesfaye.thesis 29-01-2003 15:12 Pagina 83

5.2 The study area

5.2.1 Location and population

Konso is located in the Southern Nations, Nationalities and People’s Region (SNNPR), in south-
western Ethiopia (see map 1 in chapter 2). It is located south of Lake Chamo at about 85 km
south of Arbaminch Town (the capital of the former Gamu Gofa province). According to the cur-
rent administrative classification, it is a special Wereda accountable to the Regional
Administration in Awassa, about 350 kilometres away due northeast. It covers very ragged ter-
rain with little available flat land. The altitude of Konso varies from 500 masl in the lowlands
to 2,000 masl in the highlands.
The population of the Wereda is estimated at about 200,000. Its growth is said to be remark-
able during the last few decades. About three decades ago, the population of Konso was about
50,000 (Hallpike, 1972). Possible reasons for the population increase are: the abolishment of
the traditional marriage-age restriction system Fereyuma (see section 5.9.1) and improved
access to modern treatment, particularly after the commencement of the Norwegian
Evangelical Church in 1954. The total land area of Konso is estimated at 2,430 square kilome-
tres (Bureau of Agriculture, 2000). The population density of the Wereda is 82 persons per
square kilometre.

5.2.2 Climate

Rainfall: Rainfall in Konso has a bi-modal pattern (NMSA, 2000). It is higher during the Belg
season (March to June) than in Kermt. The latter occurs from September to November (Fig. 5.1).

Figure 5.1: Mean monthly rainfall in Konso (1971-2000)

140
120
Rainfall (in mm)

100
80
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12
Months

Source: Data Base, Ethiopian National Meteorological Service Agency (1971-2000).

Moisture is perhaps the most limiting factor of production in Konso. Farmers in all corners of
Konso mentioned that lack of moisture is a principal cause of poor production and crop fail-
ure. The average annual rainfall for 29 years is 551-mm. Decade-wide average rainfall are 464,
476 and 685 mm for the 1970s, 1980s, and 1990 (includes data for the year 2000), respectively.
The minimum total rainfall record in 29 years is 210.6 mm which occurred in 1984, the recent
great famine year in Ethiopia. Apart from total rainfall, the erratic nature and skewed distri-
bution of the rainfall caused crop failures for three consecutive years (FARM-Africa, 2001).

83
.tesfaye.thesis 29-01-2003 15:12 Pagina 84

Temperature
Konso has warm to hot temperatures that range between 12 to 33 degrees centigrade. The
highest average maximum temperature is 32.7 degrees, which occurs in the months of
February and March. The lowest minimum temperature is 12.2 degrees centigrade, occurring
in the months of June to August.

5.2.3 Soil

There are six major soil groups in the Konso Wereda. These are Eutric Regosols, Lithosols,
Chromic Vertisols, Eutric Nitosols, Chromic Luvisols and Eutric Fluvisols (Kelsa, 1996). Konso
is one of the areas in the Southern Nations, Nationalities and People’s Region, classified as
highly degraded (Tefera, 1996). As will be shown later, the conservation effort of the people is
unprecedented.

Farmers identified the following soil characteristics 18 based on their functions:

1 Bolbolta: brown soil from alluvial deposits with good depth.
2 Borbora: black vertic nature that sticks between the fingers when wet and cracks when dry.
It is difficult for farming tools to penetrate though the depth is adequate.
3 Kelkelita: reddish, slightly sticky, resembling Borbora, but cracks less. It has a good depth.
4 Ateta: grayish (ashy) with fine texture.
5 Tahita: a mixture of sand, rough texture.
6 Amata: soil with a mixture of stone.
7 Mokosha: white in colour and very shallow soil due to erosion.

5.2.4 Research and development interventions

Konso, like many marginal areas in Ethiopia, has been isolated from most development inter-
ventions. The most notable ones are the development programmes of the Norwegian
Evangelical Church that began in 1954. This religious institution has supported a number of
projects for the development of the Konso people on its own and in collaboration with the
Ministry of Agriculture. Even though several experiments mainly on cropping were initiated
under the auspices of the Ethiopian Evangelical Church Mekane Yesus, agriculture in Konso
still has a long way to go to achieve food self-sufficiency. Recently, other NGOs, such as FARM-
Africa initiated a capacity-building project in the area. By and large, Konso has so far attract-
ed more anthropological and archaeological researches than agricultural research.

5.3 Methodology of my study

An understanding of indigenous agriculture in Konso requires a historical journey back into

their history. Historical analysis helps us to understand the nature of the indigenous institu-
tions, how these institutions shaped their production systems and the general landscape.
However, this chapter does not attempt to provide the ethnographic accounts of the Konso

18 Note that farmers in different villages use different names for the same soil characteristics, which is common in a local
classification. For instance, Tahita’s description in Gamole fits that of Ajaita in Buso. Therefore, care must be given to the
description as one moves from one village to the other.

84
.tesfaye.thesis 29-01-2003 15:12 Pagina 85

people that is extensively dealt with by Hallpike (1972), using classical anthropological field
methods. What is examined in this chapter is the link between the institutions and agricul-
ture, and more specifically, soil and water conservation.
This study covers the entire Konso special Wereda. In line with the specific research ques-
tions mentioned earlier, the farming systems in Konso were first assessed to determine the
villages for study. Accordingly, the entire Konso was divided between hoe-farming and ox-
plough cultures (see 5.5). Following that, major villages were selected in these settings. In the
hoe-farming culture, Buso, Gamole, Gareti/Doketu, Kolme were selected. Turo area was selected for
the ox-plough culture. In the previously mentioned study (Hallpike, 1972), these villages were
also classified in their geographic locations, with Buso, Gamole and Gareti in the East, while
Kolme and Turo are in the West and North, respectively.
A local staff member of the Office of Agriculture in Konso with a B.Sc degree in plant
sciences assisted in facilitating the fieldwork and in the interpretation of my communications
with the local people.
Data collection was carried out with the help of a range of qualitative and quantitative
methods. These include: historical interviews, key informants’ interviews, individual and
group interviews, measurement observations, and secondary data.
The interviews began with a few key questions, elaborating on different issues. Emerging
issues were discussed until a point of consensus or saturation was reached in a given village.
Following each interview, the researcher reviewed the issues raised and prepared relevant
questions to be raised in that village. The key questions raised in one village would then be
raised in the next village as well, as meeting with one village is not necessarily consecutive.
When new issues emerged in one of the villages, another round of visits was carried out in the
previous villages to get their views on those issues. In this manner, ‘ground truthing’ was car-
ried out throughout the qualitative phase. Individual interviews were carried out with a non-
random sample of 24 farmers. Key socio-economic questions were administered to get an
overview. Measurements were carried out for terrace length and field plots.
Observation was an integral part of the entire fieldwork. For better opportunities, I trav-
elled over the testing hills of Konso under the blazing sun, encouraged by women and chil-
dren transporting manure to their fields. As noted in Chapter 4, a formal survey was not car-
ried out for this case. Instead, information from other surveys were used when appropriate.
The collected qualitative data were then entered into the computer, seeking patterns in the
written text. Thereafter, information on similar issues was found with the help of a word pro-
cessor by creating further patterns and categorisation of the text. This report is the result of
such a process.
The formal interviews with the 24 farmers (non-random) were summarised and were
included in relevant sections to support the qualitative information in addition to the sec-
ondary data sources.

5.4 Socio-economic situation

According to a recent survey (Eyob, 2000) 19, average family size was 5.7 persons. Sex composi-
tion was nearly equal. Based on the non-random sampling of 24 farmers, the family size
ranges from 3 to 12. Large family size often occurs among farmers who marry more than one
wife.

19 Note that this survey involved 2490 households in 11 PAs in the Konso special Wereda.

85
.tesfaye.thesis 29-01-2003 15:12 Pagina 86

The majority of the rural families is illiterate. School enrolment was attended by 15% of
school-aged children (Eyob, 2000). Children enter school at a late age and then quit for various
reasons.
Prior to 1954, when the Norwegian Evangelical Church started its Christian mission in
Konso, most of the Konso people were atheist. Even though the Ethiopian Orthodox church
started operating in the area since the late 19th century, its followers are few as compared to
that of the protestant church. The proportion of protestant Christians was 51% of the popula-
tion, followed by atheist (35%), Orthodox (11%) and 3% Muslim and others (Eyob, 2000).
Nowadays, there is a strong religious movement among the Christian churches due to a sort
of competition between different branches. This movement is penetrating the traditional
labour organisation and mutual assistance mechanisms.
Farming is the main source of livelihood for the majority of the Konso people. Additional
sources of income are obtained from weaving, bee-keeping, trading, sale of forage and forest
seeds, local brewing and meat retailing. In the past, weaving was restricted to craftsmen.
However, since the incorporation of Konso into the northern state in 1897, weaving is carried
out by cultivators as well, while craftsmen received access to land by purchasing it and
through a local network. Of the total of 2,490 farmers, 23% participated in daily labour and
trade (Eyob, 2000). Daily labour is mostly obtained in the rural areas in the form of perga (sec-
tion 5.10).
Average land size is 0.9 ha, for those who own one plot and 1.5 ha for those who own more
than one plot. Average cereal production is 0.5 and 0.3 t ha -1 for Belg (main rainy season) and
Mehir (short rainy season) seasons, respectively (Eyob, 2000). Household variability in terms of
socio-economic status goes back to the history of land acquisition, a system of inheritance that
favours the elder son at the expense of younger brothers, and marginalisation of women from
social, economic and institutional roles. Shortage of land in sufficient production often forces
people to migrate internally or out of agriculture, usually to the nearby towns or large com-
mercial farms. In a recent survey, Eyob (2000) reported a migration rate of 9%.
Konso is one of the food-insecure areas of the country. Growing food insecurity is partly
rooted in the changes in the balancing mechanisms of natural resource utilisation. These
changes are: increased population size, declining use of manure, expansion of bush farms
without manure and conservation structures, shortage of labour, and short or no fallowing.
Change of weather during the last few years has also contributed to the crop failures in Konso.
When crop failure occurs as in the case of 1999 and 2000, children and the elderly become
the victims of famine and subsequent death. In order to save their lives, the EECMY Konso,
Terapatic Feeding Centre admitted 602 children between October 1999 and December 2000 20.
Among these children, 4% died at the centre, 3% were referred to the hospitals while the
remaining were rehabilitated and subsequently discharged from the centre. With the excep-
tion of five sets of twins, all the children were from separate families, which shows the extent
of the severity of the problem in the entire community.

20 The data were obtained from the Ethiopian Evangelical Church Mekane Yesus Clinic, Konso.

86
.tesfaye.thesis 29-01-2003 15:12 Pagina 87

5.5 Overview of the production systems

5.5.1 Crop production

Konso’s agriculture is principally a sorghum-hoe-terrace complex (Westphal, 1975). This fea-

ture still dominates the majority of Konso today. However, there is also an ox-plough system
at Turo, Kolme 21, Gawada and Gomide where the land is relatively gentle in slope. This distinc-
tion is often overlooked in the description of Konso’s farming system.
Under favourable agro-ecological conditions, community food culture and rituals primari-
ly determine crops grown. Besides, exchange values of crops play no less important roles in
crop choices.

Plate 5.1: A typical terraced field in Konso

Konso farmers grow many types of crops. Crop diversity within small terraced fields is a
remarkable feature of the cropping system (see Plate 5.1 for a typical terrace structure in
Konso). The most common crops are cereals of which sorghum is the most important, followed
by maize. No less than 20 varieties of sorghum are grown in the area. Other cereals include
finger millet, teff and wheat barley. Cereals are often intercropped with pulses, among others,
for risk aversion, soil fertility and land saving. Pulses grown in the area are: haricot beans,
pigeon peas, lablab, peas, chickpeas and cowpeas. Owing to their drive to self-sufficiency, the
Konso people grow fibre crops such as cotton. The use of cotton became popular after their
clothing style shifted from skins to textiles some hundred years ago. The exchange of cotton
with the neighbouring tribes was also important. The growing diversity of their food habits
and knowledge of crop associations also led them to grow tuber and root crops such as yam,
cassava, sweet potato and taro. Oil crops are also grown on Konso land; common oil crops are
safflower and fenugreek. The warm climate of the area also favours production of citrus crops,
though on a small scale. Some farmers have a few lemon, orange and guava trees. In addition,
they also grow banana and papaya. Crop diversity, in a seemingly hostile environment also

21 The use of oxen for ploughing at Kolme is, unlike other places, in response to a labour shortage rather than suitability of
the landform for the ox-plough.

87
.tesfaye.thesis 29-01-2003 15:12 Pagina 88

favours the world’s celebrated stimulant crop, coffee. The ritual practices and traditions in the
society also call for the growing of chat and tobacco.

5.5.2 Farm tools

The major tools are: sipilota pona (metal tools for the dry period), liwa, payra, panga, and the ox
plough. The major features and use of these tools are explained below.
The indigenous farm tools are displayed in Plate 5.2.

Sipilota: a digging stick with a metal tip used to uproot sorghum stalks during dry season.
Because of the laborious nature of the work, this tool is mainly used by men.
Liwa: the same as sipilota, except smaller - it is used for weeding by both men and women.
Payra: a tool composed of wood, twin metal tips attached to short sticks. The sticks are
wrapped together with leather string and finally tied with a knot together with the main
wooden frame. This tool is used by both male and female farmers.
Panga: mainly used for cutting and slashing of shrubs and weeds. The panga is used by all
members of the community.
Ox-plough: this tool is widely known in Ethiopian agriculture and as such does not need fur-
ther description here. It was introduced after the incorporation of Konso over a century ago.

Plate 5.2: Farm tools used in indigenous Konso agriculture (from left to right; payra, liwa, panga
and sipilota)

5.5.3 Livestock production

Livestock production has been part and parcel of Konso’s mixed agriculture. The types of live-
stock commonly raised are cattle, goat, sheep, donkeys and poultry. Generally, the system
lacks draught animals for transportation. Livestock are highly appreciated for their economic
and ritual purposes. Unlike other parts of the country, the majority of Konso people raises live-

88
.tesfaye.thesis 29-01-2003 15:12 Pagina 89

stock for manure and savings while traction is important in only some parts of Konso. Milk and
butter are greatly used in the ritual services. Animal sacrifices that require a particular colour
and sex encourage a close integration between livestock production and crop production.
There is a system of animal sharing 22 whereby poor families can obtain access to livestock,
which also means access to manure. Of the total of 24 farmers interviewed, 38% were involved
in animal-shared rearing. Cattle ownership ranged from none, to four oxen, cows and sheep,
whereas the ownership of goats reached up to eight heads of animals per household. However,
animal disease coupled with the present extreme shortages of forage, and grazing land have
limited the production of livestock and thereby the availability of manure.

5.5.4 Forest production

Trees, like livestock, are an integral part of Konso agriculture for their ritualistic, economic,
social and ecological services. Multiple uses of trees have led to different patterns of tree pro-
duction and management. These are: natural vegetation, ritual forests, homestead forests, agro-
forestry and plantation forests.

Natural vegetation
This covers the remnants of trees in the high elevation areas that are not accessible to man, and
bushes in the hot lowlands. Some of the bush land is being cultivated for bush farms while
other uses are hunting, firewood and grazing.

Ritual forests
There are a few natural trees used for ritual services that are associated with the regional
Poqallas.

Homestead forests
Konso villages have a remarkable tree cover. I was particularly surprised by the coverage of trees
in the villages when I first visited Konso. The dominant trees grown are Morinaga stenopetala and
coffee trees. Their presence is not surprising after I learned the importance of these trees to the
daily diet of the community 23.

Agroforestry
The survival of Konso’s agriculture is due to its integrated nature, which might not be the case
if only stone terraces were used. This integration is partly realised with extensive use of agro-
forestry practices. The vertical and horizontal spaces are occupied by plant species that meet
family dietary needs, and that are used for cash crops, apiculture, animal feed, shade for other
crops and people, erosion barriers, soil fertility and as a farm boundary.

Plantation forest
Tree plantations began following the arrival of the Ethiopian Evangelical Church Makane Yesus

22 Animal sharing is common throughout Ethiopia. It takes place between two or more parties for mutual benefit. The
donor provides an animal to the recipient who will tend it. The recipient will use the by-products and power of the ani-
mal in his or her custody depending on the type of animal. For instance, if it is a heifer, the first offspring may go to the
recipient with a possibility to share the milk and butter. The second offspring would go to the donor and so on. The
arrangement varies for different animals involved. It may even be sold to share the price according to the contribution of
the parties.
23 The Morinaga leaf is mixed with cereals and cooked with water and served in a daily meal. Coffee leaves are boiled with
milk and used as both stimulant and food.

89
.tesfaye.thesis 29-01-2003 15:12 Pagina 90

(EECMY) in 1954. In the course of community development and rehabilitation programmes,

EECMY has introduced exotic tree species. Some of these trees such as Eucalyptus spp, Gravillea
robusta, Accia cyanophylla are now commonly seen around settlement areas and roadsides.
However, their coverage hardly matches the efforts that have been put into their promotion
in Konso.
Common trees species in Konso are: Juniperus procera, Euphorbia spp Terminalia browenii, Olea
africana, Ficus sori, Cordia africana, Sterculia africana, Accia abysinica. Among these, Juniperus pro-
cera has a high significance in Konso’s rituals (Olahita) and for its wooden statues (Waga).
Olahita is an important part of the ‘generation grading’ system while the Waga represents an
individual’s achievement, for instance hunting successes, wealth, and the like (see sections
5.9.1 and 5.8, respectively for more details).
Wildlife is an integral part of the forest ecosystem. In the past there were large herds of
large mammals such as lions, elephants, leopards, antelopes, wild pigs, gazelles, deer, wild
goats and hyena. Except for the latter two, most of these animals have vanished by now. As a
result, hunters must travel a great distance for the ritual hunting expedition. The disappear-
ance of wildlife is one of the indicators for the expansion of agriculture and increased popu-
lation density.

5.6 Origin and identity

The Konso people are one of several nationalities living in the SNNPR. They came to their pre-
sent territory several generations ago. Earlier settlers arrived at Konso by different routes and
often after interim settlement elsewhere. Their original settlement was based on families who
later grew into clans. Individuals outside blood relations were also integrated into clan net-
works for access to resources and protection. Over time, those clans created the Konso nation-
ality by maintaining their ancestral roots.
There are nine clans (kaffa) in Konso: 1) Keertitta, 2) Paasanta, 3) Tokmaleta, 4) Eelayta, 5)
Mahaleta, 6) Sawdatta, 7) Isalayta, 8) Arkamayta, and 9) Tikissayata (Hallpike, 1972; Otto, 1994).
Their settlement pattern is, however, not according to clan composition. These clans are exog-
amous, a practice that affects women’s identity and access to resources and active participa-
tion in politics.
The Konso people have common ancestral roots with many neighbouring nationalities that
are evident from their linguistic and cultural relationships. Earlier travellers and anthropol-
ogists documented these relationships (Cerulli, 1956; Hallpike, 1972). Hallpike (1972:3) in par-
ticular emphasised the linguistic similarities between the Oromo (Borena), Gidole and other eth-
nic groups in the area with the Konso. Elders know the origin of each clan leader passing
down information orally from generation to generation. The oral tradition serves as a promi-
nent means of information transfer in this society.
According to Cerulli (1956:51), there were Konso (and Burji) settlements in Kenya, in Moyale
and Marsabit. Konso people still live in Kenya with their distinct culture (Yonas Beyene, 2001
p.c.). Recently, I learned that there is a large group of Konso and Burji settlers in Marsabit
(Adeno Wario and Karen, p.c, 2001). Their settlement dates back to the colonial period in the
early twentieth century. The settlement took place after the British colonial government
agreed with the then Ethiopian government to allow the Konso and Burji people to start seden-
tary farming in Marsabit Mountain. The purpose was to supply food for the army and the civil
societies instead of importing from interior Kenya. The sources added that sedentary
agriculture was not common in that area at that time.

90
.tesfaye.thesis 29-01-2003 15:12 Pagina 91

Looking at their present settlement pattern, one may wonder as to why their ancestors
decided to settle here whereas they could have settled in a relatively fertile and flat terrain in
the neighbourhood. From repeated group and individual interviews, I learned that there was
little or no chance of settling elsewhere than the present land. There were territorial compe-
titions among different nationalities to claim as much fertile land as possible. In addition, the
fertile valleys were hot and infested with malaria. As a result, they resorted to their present
place because of its cooler weather, reliable protection from enemies and a reduced risk of
malaria.

5.7 Basis of social differentiation

Social differentiation in Konso is based on how early a person has settled, seniority (age) and
gender. These elements have direct implication for one’s access to resources and power in the
society (Hallpike, 1972; Watson, 1998).
The early settlers had easy access to land (see section 5.8). They were believed to have sub-
dued devils that resided in the jungle. As a result, they emerged not only as the landed social
groups, but also as spiritual leaders, who came to be named Poqalla. Poqalla-hood is the high-
est rank and honour among members of the society (Hallpike, 1972; Otto, 1994). Over time,
Poqallas emerged as clan leaders in the day-to-day socio-economic, political administration and
in religious matters. Their religious role and image within the society is the basis of their
power. They are believed to have been born with a special mark on their hands, speaking with
messengers of God (Waqaa), making rain; removing curses from crops, livestock and people.
Poqallas were central to the society’s spiritual and material wellbeing. They were not only the
gateway to peace, but also to prosperity as a source of land, for blessings for obtaining a good
rainfall, a good harvest, good livestock husbandry and many children. After the formation of
the Xela system of administration (section 5.9.1), which is a cross-clan system of administra-
tion, the Poqallas’ political leadership turned out to be an individual merit rather than an
outright position as during clan administration. However, their religious role and economic
position continued as before the formation of the Xela.
There are three regional Poqallas in Konso, namely, Quffa, Bamelle and Kalla, who are respon-
sible for Fasha-Kolme, Gareti-Doketu, and Gamole-Turo areas and their environs, respectively.
Regional Poqallas emerged with growing rituals and political influence among the lineage
Poqallas. These Poqallas are responsible for many villages irrespective of lineage. In addition to
regional Poqallas, there are clan Poqallas whose responsibilities are to bless their clan. Hallpike
(1972: 88) indicated that the number of lineage Poqallas in four villages, namely Idigle, Buso,
Degato and Gaho, had a total number of lineage Poqallas of 113 at the time of his data collection
(i.e., 1965 to 1967). The number of Poqallas who first occupied Buso, Gamole, Turo and Gareti vil-
lages was 12, 7, 11, and 18, respectively. From this figure, the ratio of Poqallas to the non-Poqallas
has been very small. The Poqalla, however, transcends his religious role as recently illustrated
by Watson (1998). Unlike previous anthropological studies (Hallpike, 1972; Ambron, 1984;
Otto, 1994) that posited Poqallas merely as priests, Watson clearly showed the economic roles
of Poqallas due to their access to land as earlier settlers.
In Konso, seniority is established at various levels. In the household, the first-born son has
the privilege of taking the largest share of his parent’s land. Unlike his younger brother(s), he
stays with his parents to take over the family compound, whereas others have to settle in
another place in the village. At the community level, seniority is established according to the
generation-grading system (section 5.9.1). According to this, the junior grade (under 20 years),

91
.tesfaye.thesis 29-01-2003 15:12 Pagina 92

has no right to marry, claim resources nor assume responsibility in the community. The turn-
ing point in manhood starts with the entrance to the warrior grade at which point they can
marry, assume social responsibility, and serve the community as a peacekeeping force. After
the warrior grade, there is the elderly grade with sub-grades. The elderly mainly bless the war-
rior grade and the community on various ritual occasions (Hallpike, 1972).
In terms of gender, males have the upper hand over their female counterparts. Land is
inherited by males rather than females. Females do not participate in most of religious cere-
monies, political leadership nor in conflict negotiations.
The socio-economic consequences of the basis of social differentiation discussed above in
the early days of Konso society is highly reflected in political power and the main resource,
land. The following section discusses land tenure in Konso in a historical perspective to show
its importance in land management.

5.8 Land tenure on a timeline

In view of its importance to explain the background of the strong conservation effort in
Konso, land tenure is presented in three historical epochs. These are: 1) Prior to the incorpo-
ration into the northern system, 2) From the incorporation until the Derg Regime (1897-1974),
and 3) From 1974 to date.

5.8.1 Prior to the incorporation

The establishment of land rights in Konso dates back to the time of influx of different clans
and occupation of the present land territory. Land acquisition of different clan leaders is one
of the important social decisions that shaped the landscape of Konso. Land acquisition used
to take place through a practice of ‘land burning’ (see also Watson, 1998). The first person who
came to a piece of land, lit a fire to burn the trees and bushes to turn it to residential and farm
land. The next person who came to the territory also lit a fire and let it go out by itself, which
subsequently marked his claim. The same practices took place for the following settlers. The
fact that the fire was allowed to go out by itself was associated with God’s (Waaqa’s) will to allo-
cate that much land to that family. Therefore, one’s luck determined the coverage of the fire
that in turn determined the land size. Some families opted for slash and burn 24, instead of
burning the land. Owing to the land-acquisition system, some people came to have a smaller
size of land than others. For instance, at Buso village, Etigele family had the largest land size,
whereas the Kayo family had the smallest land size, all because of the coverage of the fire. Of
course, not all Poqallas held land by fire. Most of the Poqallas in Buso held land by slash and
burn, which provided a relatively smaller piece of land as compared to obtaining land by burn-
ing it.
After the original settlers occupied the available land, a new pattern of ownership began to
emerge. The latecomers would obtain land from the original owners with the commitment to
associate, respect, give gifts, and provide labour service to them during peak farming seasons.
This way of access to land is called Piyolada, a term for ‘borrowed land’ from Poqallas in return
for labour and other services. Borrowed land is not limited between Poqallas and non-Poqallas.
The latter groups also exchange land, though there is a difference in obligation, for Poqallas
have multiple roles in the community. Land acquired in this manner is not subjected to sell

24 Land claimed by slash and burn is called Fax land to denote the means of access.

92
.tesfaye.thesis 29-01-2003 15:12 Pagina 93

or transfer to a third party. The original owner can claim it back whenever he so wishes
because of: poor management of the land, failure to provide labour, and any act of dishonour
to the owner.
The other form of access to land is through grants by Poqallas to the lineage members as
sole accumulators of land for their clans. The land obtained in this manner is called Kodeyata.
Such land can be sold or lent to, or inherited by children. These are rights exercised by the
receiver because of lineage. Kodeyata is also given to non-clan members because of special rela-
tionships with Poqallas.
Children obtain land from their fathers whereby most land goes to the oldest son. As in
many other societies, there is no land inheritance for women.
As private ownership of land was institutionalised in the society from the very beginning,
land sales are very common. Because of this long tradition of a land market, the number of
fields purchased by a man is inscribed on a traditional statue called Waga, in addition to his
hunting successes, victories over his enemies and number of his wives. A cluster of stones indi-
cates the number of fields one has purchased. Wagas are erected along public roads and near
the entrance of the residential areas, which further conveys the message to the generation
that follows (Metasebia, 1997).
The system of access to land in Konso ensures that everyone gets some land to start his liv-
ing. Expanding the amount of land however depends on one’s hard work, achievement and
motivation, reflected in the land to be inherited to his children, which is also indicated on the
Waga.

5.8.2 From incorporation until the Derg Regime (1897-1974)

Konso was incorporated in the northern system in 1897 by the Army of Milinik the second.
Following the incorporation strategies of the system, Konso’s local land lords, Poqallas were
designated as local chiefs, and given the official names of the northerners, Balabat, to assume
responsibility of the local administration, being accountable to the resident administrator
from the north. Accordingly, they maintained their original lands, but transferred their right
to claim the tribute and tax to the northerners. Because of their co-operation with the central
government, most of Konso’s Balabats maintained their traditional roles.
Apart from a few areas in the northwest that have relatively gentle inclines, there was hard-
ly any land that attracted northerners to establish their clean field-cropping system (Amborn,
1984). They were less interested in the terraced and agroforestry land of the majority of the
Konso people, which did not fit into their ideal image of a farm. They instead capitalised on
the juniper trees for construction in the new administrative centre they created adjacent to
the old Konso villages, now called Bequale. The rest of the trees were transported to other
administrative towns in the north and east (Gidole and Chencha) by human labour, over a dis-
tance of approximately 80-120 km. Apart from that, the incorporation did not interfere with
the religious and other local institutions. Of course, two Orthodox churches were established
in the main settlement areas of the northerners.

5.8.3 From 1974 to date

After overthrowing the monarchy in 1974, the military Regime promulgated a radical land
reform proclamation in 1975. The proclamation indicated that land belonged to the state,

93
.tesfaye.thesis 29-01-2003 15:12 Pagina 94

while it entitled each citizen with a maximum of 10 hectares of land for personal use. Like any
other rural people, tenants were granted a usufruct right to till the land for which they had
been paying a tribute to the landlords. In the same manner, Piyolada was claimed by the tillers
against their indigenous land-tenure system. Since then, the production relations between the
tenants and the landlords officially discontinued. This means that the Poqallas’ role as a
source of land for the junior lineage ceased with this change.
Following the land reforms, new formal institutions were established in rural Ethiopia,
where Konso was no exception. These were Peasant Associations, Service Co-operatives and
Producers’ Co-operatives. The revolutionary leaders and students who participated in the
‘Development through Co-operation National Work Campaign’ dismantled the indigenous
institutions and replaced them with the Peasant Associations.
In 1991, the military regime was in turn overthrown and replaced by a Tigrian-led, coali-
tion force called the Ethiopian People’s Revolutionary Democratic Front (EPRDF). The consti-
tution of the new government stipulates that the land remains the public property while
people have usufruct rights. As far as the fundamental land policy is concerned, there was no
change since 1974.
In spite of changing land policies during the last hundred years or so, farmers in Konso still
believe in private property of land. A common view on land is that ‘land spiritually belongs to
the Poqallas, as a property it belongs to the state, as for use, it is private’ (see also Watson, 1998).
In one of the group discussions, a farmer expanded this outlook on land as follows: ‘Land
belongs to the state. Below the state, it belongs to Kalla (one of the regional Poqallas). But, actu-
ally the land belongs to me. I got it from my father. When he generalised, he said, ‘what was
in my hand is mine.’
There is still a land market in Konso, though people do not declare it as it used to be.
However, it is believed that those who sell land are weak farmers, as purchasing land rather
than selling it is praised in Konso. The land category that is called ‘Fax’ land is readily sold
because this land was obtained by slash and burn. It is common to use land as collateral. As
long as the credit has not been paid, the creditor uses the land. When the land is Piyolada, sale
is still not practised, because of the original ownership to a Poqalla. But, this land too is mort-
gaged under the guise of the name Maresha, meaning ‘the plough’, indicating, that it is not
the land that is mortgaged, but the plough. In confirming the special treatment of Piyoladas,
one farmer gave his own situation as an example: ‘We received land from the Karpena family.
In the past, we used to give them one leg of meat and one side of ribs each year. Even though
we have stopped that gift-giving since the Derg, we cannot sell the land in spite of the state
ownership of land because we know that the land actually belongs to that family.’
This section has shown that indigenous land tenure played an important role in perpetu-
ating good land-management practices by ensuring the transfer of land to children, reflecting
the economic, social and political status of the people. Because of this, borrowed land
(Piyolada) has been maintained very well to continue the access to land. This is contrary to the
land-tenure theory that is based on the western land-ownership pattern wherein, non-private
lands are said to be mismanaged. The other important aspect of land tenure in Konso is that
the northerners showed little interest in their land’s topography and therefore their large
tracts were not threatened by claims of the new landlords in the 1890s. Due to the special
arrangement made by the northerners in leaving the land rights to the regional Poqallas, there
was little alteration in the land-tenure pattern during the land reforms of 1975. Note that in
other parts of the country, land that includes hills and forest which was transferred from the
previous landlords to the tillers, suffered marked mismanagement when falling under public
property. Therefore, the Konso case presents an example of a more or less stable land-tenure

94
.tesfaye.thesis 29-01-2003 15:12 Pagina 95

system over generations, which contributed to the continuation of indigenous land manage-
ment even after the 1975 land reforms.

5.9 Local institutions: indigenous governance, its structure and

foundations
5.9.1 The origin of indigenous governance

Ancestors of the Konso people who settled in the present territory had scattered settlements,
segregated by their lineage due to the land acquisition system. Because of attacks from neigh-
bouring ethnic groups, they came to settle in permanent villages. Each Konso village was
autonomously administered by its own elected administrative council. The election procedure
and constituencies of this council are discussed in section 5.9.2.
The system of governance was developed from the beginning of the Konso settlements until
being replaced by the formal Peasant Association in 1974. According to historical interviews,
the development of a common language began with the establishment of common rules 25.
Traditionally, gatherings were signalled by drums (Timbas) made from the cordia tree. In
order to show their unification, two sacred drums were constructed from this tree, and
named Ketennas for eastern villages and Keeha for western villages in all of Konso. In addition
to these, there are village Timbas that serve the peace and unity of each village. Coming
together was under the conviction of the spiritual drums - the binding forces of the Konso
people. These drums have played and are still playing very important roles in the life of the
Konso people. People still believe in the power of the drums. They symbolise the power of the
mighty Waaqa (God). Prayers for rain and protection against pest damage are made with the
drums. However, this does not mean that the society is always peaceful and harmonious.
There are conflicts of various natures, ranging from individual to clan, to between villages.
However, they have established a mechanism to reconcile arising conflicts in their own way.
Various bodies take action to ensure that peace and stability in the community follow the sym-
bolism of the spirit. The drum is like a switchboard in case of conflicts. Once the drumholder
(Apa timba) and the peacemaker (Nama dawra) arrive at the place of fighting, both parties must
refrain from further assault. This shows the importance of the social institutions in main-
taining peace and order in society. These agents were the counterparts of the police in the
modern peacekeeping machinery.
The rule of the elderly or the Xela or its delegation, is accepted in the community. Fines and
ritual cleansing take place accordingly. If someone violates the ruling, he or she will not be
permitted to borrow utensils or fuel from the neighbours. No-one will accompany a violator
in fetching water, fuelwood, etc. If the person does not comply with the ruling, he/she will be
expelled from the community, as the final stage.
Signifying how protection from an enemy was an important issue, villages were construct-
ed on hillsides that guarantee protection from the three sides of the village, with the possi-
bility that the fourth side can be defended. To this end, euphorbia species and other trees were
grown for use as a fire extinguisher and to ambush an enemy who may attack their villages.
This forest plot is called Dina. The growing of forest plots for defence and fire control is deep-
rooted in the Konso culture. According to a local proverb, ‘a village without a Dina is like a
woman without a dress.’

25 Language development among people who have come from different origins requires a detailed anthropological study.

95
.tesfaye.thesis 29-01-2003 15:12 Pagina 96

In order to defend themselves from enemies, the villages have been concentrated in a few set-
tlements areas. As a result, the size of villages often grow larger than the size of traditional
villages in Ethiopia, and perhaps in Africa and elsewhere. Even during the time of Hallpike’s
study, the average size of a village was 1,500 people (Hallpike, 1972:27). Actually, Hallpike uses
the term town, instead of village due to the large and dense population living within a walled
settlement. With the population increase, new walls have been constructed around all the vil-
lages. Nowadays, the original site is fully inhabited, with rather too close compounds, and
houses within compounds. As a result, some villages are encroaching into the traditional
forested plots (Dina). Walled villages are a typical scene in Gereti, Gamole and Fasha areas.
Turo settlements are relatively scattered and unwalled. The major reason is the lack of stone
in the area. A section of one Turo village that was walled was destroyed during the war
between the Italian army and an Ethiopian folk army called Fano. The latter targeted the
village because of their co-operation with the Italians. As the remnant stones were eroded
throughout the valley a long time ago, one can make the claim that there must not have been
a single walled town in the area.
A continuous system of administration in Konso began with the establishment of Xela, an
administrative assembly. This assembly is referred to as Keladule Dhegeya (literally, ‘based on
agreement’). It is elected from among people who have reached the Xela grade according to a
generation grading system in Konso. The generation grading system is a typical feature of the
indigenous administration. It defines an eligible age for leadership that is controlled by a
mechanism called Fereyuma, as discussed below. Konso’s generation grading system was initi-
ated to ensure that children remained two generations behind their fathers. The main pur-
pose of this restriction is to maintain the physical strength of men. Men are given important
roles in Konso’s political and spiritual world. They need to be strong in order to shoulder the
societal responsibilities such as engagement in war, and withstanding rigorous labour
demands in agriculture and hunting expeditions. The generation grading system practised in
the eastern villages of Konso has seven grades (see Table 5.1). For a detailed treatment of this
issue in all of Konso, see Hallpike (1972).

Table 5.1: Generation grading system in the eastern Villages, Konso, SNNPR, Ethiopia

Grades Fraida Xela Kat Orshita Qura Qulula Ukudeta

Approximate age < 20 20 - 36 38 - 54 56 - 72 74 - 90 92 - 108 100 - 126

Description Junior age Warrior Elderly Elderly Elderly Elderly Elderly

The Xela system began its operation by dividing the cultivators in Konso, excluding the crafts-
men, into two categories. The first group was named Kalkusa and the second, Herpa. Herpa and
Kalkusa are Xeltas 26 that show membership in the administrative assembly (Xela). The first term
of 18 years was given to Kalkusa. For this purpose, 16 families were identified for each Xelta.
One Xelta is responsible for the Xela for 18 years. The families that began the first cycle held
the symbol of the generation, Timba, for two additional years in addition to its own [term].
Therefore, the number of families involved in the circulation is 16 instead of 18. An example
from one of the eastern villages (Doketu) that shows the list of families responsible for the Xela
in each Xelta is presented in Table 5.2.

26 This refers to the particular generation that is holding the office term.

96
.tesfaye.thesis 29-01-2003 15:12 Pagina 97

The western and northern villages have their own version of Xelta cycles. For example,
Kolme, located in the west, practises three generation grades instead of two. These are: Kalkusa,
Melkusa and Herpa. The generation shifts took place every nine years instead of 18. Some 150
years ago, they shifted every twelve years. At Turo (north), which shares a regional Poqalla with
Gamole in the east, Xelta names are called Xelpihayta and Perpihata corresponding to Kalkusa
and Herpa, respectively. The generation shifts takes place every five years instead of 18 years
like in Gamole.

Table 5.2: Sequence of Timba circulation in Doketu between Kalkusa and Herpa Xeltas, Konso, SNNPR,
Ethiopia

Years Kalkusa Social Status Years Herpa Social Status

1-3 Pata Poqalla 1-3 Telliyya Poqalla

4 Kasimma Non-Poqalla 4 Kaanparo Poqalla
5 Pariyye Non-Poqalla 5 Palette Poqalla
6 Katto Poqalla 6 Kompa Poqalla
7 Koata Poqalla 7 Pittitta Poqalla
8 Porre Poqalla 8 Taanta Poqalla
9 Qura Poqalla 9 Kema Poqalla
10 Kajjo Poqalla 10 Mareeta Non-Poqalla
11 Sugta Non-Poqalla 11 Puulla Non-Poqalla
12 Lasha Poqalla 12 Ashuma Poqalla
13 Qaaja Non-Poqalla 13 Karshammo Non-Poqalla
14 Yimme Non-Poqalla 14 Ullupa Non-Poqalla
15 Robo Non-Poqalla 15 Dinke Dinke
16 Kerko Non-Poqalla 16 Milee Poqalla
17 Alexo Poqalla 17 Telliyya See 1-3
18 Jalule Poqalla 18 Milee See 16

The principle of generation grading was enforced by a system of Fereyuma, which forbade mar-
riage until about the age of 30. The spacing between children and fathers was thus main-
tained through Fereyuma. Women who conceived before their marriage age were forced to
abort, or as in the northern and western villages, they were expelled from the community.
Sexual relations with one’s senior opposite sex were a taboo, though older men sometimes
married women who are their juniors. The other reason for Fereyuma was to control the situ-
ation whereby children and parents produce children at the same time, something that was
believed to cause God to withhold rain.
In addition to Fereyuma, men’s contacts with their wives were constantly kept under con-
trol by having to sleep in the men’s sleeping quarters. The common belief was that men
should minimise their sexual activities in order to cope with the heavy fieldwork and for war.
Bearing a child before three years have elapsed since the last birth was also looked down upon
though there was no sanction against such happenings. All of these practices have been abol-
ished since 1974. The effect of Fereyuma on generation shifts is shown with the help of an
example (Box 5.1).
During each Xela there are two Karas, hunting expeditions undertaken by the youth. This
exercise is a crucial stage in manhood in the culture. For the Kalkusa group, Kara is done when

97
.tesfaye.thesis 29-01-2003 15:12 Pagina 98

Herpa’s Xela is at the end of its seventh year. The next Kara is carried out at the end of the 18th
year. The same holds true for the turn of Herpa’s Kara.

Box 5.1: Example of how Fereyuma works

When will a child from Kompa family (see table 5.2) be allowed to marry if he was born when
his father has already served in the Xela for seven years? Note that Kompa’s family (Herpa) takes
over the Xelta at the sixth year from the first family (Telliyya). Therefore, in his father’s genera-
tion he has to wait for 11 years in the junior generation (Fraida, see Table 5.1). Then, the other
Xelta (Kalkusa) takes over the Xela for another 18 years. After that, his generation takes over, but
they have to wait for five years when the restriction to marry (Fereyuma) is removed. This cere-
mony is called, Kenota. Finally, marriage is allowed after the sixth year. In total, Kompa has to
reach 35 years before he is allowed to marry and conceive a child. When Kompa’s son enters the
Xela, his father enters the Orshita grade. Accordingly, he remains two generations behind his
father. The time for his juniors is shorter, depending on which cycle of their father’s Xelta in
which they were born. In some villages, e.g., Gareti, both sex are treated in the same way. In
other villages, restrictions on women are rather relaxed as compared to that for men.

According to the generation grading, social organisation in Konso has a record of over 500
years (see also Metasebia, 1997). When generation shifts take place between Kalkusa and Herpa,
the incoming generation erected a symbolic stone at a designated Mora (a ‘sacred place’) in the
village, often the main Mora, but also the entrance to the village. This stone is called Deka
Diruma (literally, ‘stone of manhood’), for it is erected after the painstaking hunting expedi-
tion (Kara) which could last two months or more (in the old days), at times with fatal conse-
quences. During the generation shifting ceremony, new generation members also erected
Olaahitta, a bunch of Tid poles (Juniperus procera). This ceremony takes place around November-
December.
As Deka Diruma is strictly maintained, the number of stones directly shows the number of
generations in each village, at least since the system began its operation. I was told that prior
to stones, wood was used for a good number of years and erected as a symbol for returnees
from a hunting expedition. In the eastern villages, Doketu has 55 stones (28 for Kalkusa and 27
for Herpa). In this village, they practised hunting expeditions (Kara) twice a year after which
they erect the stone. This means that generation grading was practised for 504 years, (18 x 28),
on an 18-year cycle. In a neighbouring village, Buso where they erect Deka Diruma once every
18 years, there are 21 Deka Dirumas - 11 for Kalkusa and 10 for Herpa. In this case, we have
evidence of a 378-year old record of the village and institutions. The current Xeltas (Kalkusa) in
these villages had already served for 11 years by 2000 and will hand over in 2007. Considering
the case of Doketu village where the records have reached a maximum number of Deka
Dirumas, the system of generation grading in Konso began 504 years ago (i.e., in 1497). The pre-
viously published record by Hallpike (1972: 192) noted 21 stones in 1966, for eastern villages,
which by adjusting for the leap years allowed him to arrive at 361.5 years (i.e., since 16041/2).
As the number of years since the cycle starts vary according to whether the village in question
practises Kara or not or even one or two Karas, the basis of Hallpike’s findings on this needs
further assessment.

98
.tesfaye.thesis 29-01-2003 15:12 Pagina 99

5.9.2 Institutional structures

Traditionally, administration in the customary Konso villages did not have a supreme chief.
Power and authority were distributed horizontally and vertically among different social
actors. Each village had its own internal administration, which with some variation fulfilled
the necessary social functions for peace and unity.
A Xelta has a leader, who is still recognised to the present day. These leaders have different
names in different villages: Apa tula at Buso, Senkelita at Gareti and Gamole, Qolota at Turo and
Shoregota at Kolme. In the eastern villages, two individuals take on leadership as Senkelita
during the service period of 18 years. An individual who belongs to the senior group serves for
13 years while the one from a junior group serves for five years. Seniority is established accord-
ing to participation in the Kara, based on one’s age group. Regardless, the generation stays in
service for the entire 18 years according to its respective cycle. There are families designated
for this position. For instance, in villages where the Senkelita position is exercised, there are
four families, two for Kalkusa and two for Herpa. These families are said to be designated by
Waaqa (God). The same pattern of designation takes place in other villages too.
Xela leaders are assisted by Apa pila (‘holder of knife’) and Apa lisha (‘holder of whip’) at Buso
and Gareti. Assistants to the Xela leader at Turo and Kolme are Sara and Lishota, respectively.
Gamole do not have an Apa pila and Apa lisha. The outgoing Xela leadership nominates these
leaders. The criteria for nomination are: wisdom, physical fitness, personal conduct, bravery,
wealth, fairness, and a commitment to the community. Final approval is carried out at the
biggest Mora of each village after the annual assembly (for example at Morkito in Buso).
While Xela is a generation whose service terms varies from village to village, there are also
religious positions that are closely associated with it. These are Apa timba and Apa kuma. The
former one is the holder of the village sacred drum, whereas the latter holds the sacred stick
that is used during the Kora ceremony. Kora is the annual general assembly of each Xela
constituency (Kora Perra). This assembly takes place far away in a quiet valley bottom. During
this summit, major socio-economic, institutional and political issues that concerns the com-
munity are discussed, and consensus reached. New Xela leaders are chosen during this meet-
ing and announced when the meeting ends. According to the system, females are excluded
from these discussions. The end of the ceremony is marked by spraying of milk with the
sacred stick held by the Apa kumas. This last function is carried out when the assembly par-
ticipants enter the biggest village Mora.
Apa timba changes every year according to the sequences (see the example indicated in table
5.2). Note that there are many Poqallas who hold important leadership positions because of
their spiritual positions and wealth. Apa kumas are elected in the same way as the Apa pila and
Apa lisha. In the eastern villages, however, this election is from two families, one permanent-
ly represents Kalkusa and the other, Herpa.
Most of these roles are now nominal, apart from the religious/cultural aspects.
Traditionally, administrative decisions were made at the house of the Apa timba. He had mes-
sengers called Sara who announced the meeting. The importance of Apa timba is exemplified
by a root, while Xela is associated with a branch. This is because the leader came from the early
settlers of Konso and the founders of the system.
Porshodas and Hiyodas assisted the executive body. Porshodas were elected based on their per-
sonal qualities from designated families in each village. Hiyodas were selected from the com-
munity, based on their quality to lead and advise on a broad range of matters. These people
were said to have had political skills. Because of their frequent involvement in public matters,
their farms were poorly managed. Unlike Hiyodas, Porshodas were accountable to the Xela.

99
.tesfaye.thesis 29-01-2003 15:12 Pagina 100

Porshodas were like village administrators. They used to manage the traditional forest plots
(Dina); the follow-up of road and compound boundaries; the inspection of village gates for
their timely closing; the mobilisation of people; assistance to the sick, organisation of labour
for funerals, and for maintenance of roads, ponds and water points.
These institutions formally ceased operating actively since 1974 when the socialist
government overthrew the monarchy. In their place, the Peasant Associations were estab-
lished, while the traditional roles were relegated to the settlement of local disputes within the
villages, lineage or family. Even so, some cases are still referred to the traditional authorities
when the parties agree to be mediated by the local institutions instead of the judicial com-
mittee of the Peasant Association or the regular court.

5.9.3 Foundation of indigenous institutions

Indigenous institutions in Konso are spirit-led. The binding forces within and between villages
are the Timbas in which, according to them, certain powers are vested. They have served as an
instrument to unite people from different origins and for making peace. Spirit have governed
a wide range of the community’s life, including the following activities:

1. Land acquisition: As indicated earlier, early settlers accepted the coverage of fire as a land
demarcation. There was no second chance to claim more land with fire, the first chance
was the last chance.
2. Production processes: Production processes such as time of sowing and time of harvesting
are carried out after the annual spiritual ritual is practised by a designated elder on a spe-
cific field. Individuals who perform this kind of ritual must observe certain restrictions. For
instance, they do not sleep outside their home, attend a funeral before the burial ceremo-
ny, shave their hair, urinate outside their land area. These people also put curses on pests
and diseases.
3. Water development: Elders from a designated family bless water wells. This family carries
out rituals to ensure that the well gives adequate water to the community.
4. Forest protection. There are two types of protected forests. The first type belongs to region-
al Poqallas, called Mura poqalla, and the second, Mura dowra. The community, because of the
spirit associated with such forests, does not cut these forests down. In the past, their use
was possible only with the blessing of the Poqallas. Otherwise, the offender would face a
misfortune, sooner or later.
5. Human reproduction: Fertility of human beings, like fertility of the land, is expected to
take place if and only if there is peace and unity within the community. Therefore, the
entire system strives to ensure peace and unity at all times. Whenever, the peace is dis-
turbed because of human behaviour problems, a cleansing ritual takes place depending on
the nature of the conflict. One of the reasons to observe Fereyuma is to avoid child produc-
tion by parents and children at the same time, which according to their belief, God does
not permit.
6. War and hunting: One of the main tasks of the elderly is to bless the warrior group to
succeed in war and hunting.
7. Election of leadership: Leaders of the community are endorsed during the annual assembly
(Kora). The decision made during that meeting is final. The community believes in the deci-
sion of the Kora as a symbol of the spirit.
8. Human health: Human diseases were traditionally treated by spiritual doctors. They carried

100
.tesfaye.thesis 29-01-2003 15:12 Pagina 101

out all the prescriptions of the local healers because of their conviction that spirit is behind
poor health.
9. Land market: A market for land was common in old Konso. The final oath of sale of land
was carried out in the presence of local land sale facilitators who sanctioned the oath by
cutting a blade of grass from the land.

The major venues of religious exercises are Moras (sacred huts). Ancestors built these places.
They were originally located outside the residential areas, but, the village now encompasses
these Moras due to the expansion of the settlement. In addition to the sacred Moras there are
non-sacred Moras where men meet, play gebeta (a board game) and sleep at night. Male visi-
tors also sleep at these Moras. Generally, Moras are the place for men. During religious pro-
ceedings, women who are in their active reproductive cycle are not allowed to participate.
There are certain Moras that women cannot walk through for fear of becoming infertile.

5.10 Labour organisation

Access to farm labour plays a crucial role in complex and drought-prone environments like
Konso. The analysis of an informal interview of 24 households shows that ownership of assets
such as livestock and land size affects farmers’ access to labour. Out of the total, 21% have suf-
ficient labour for their farm. However, this does not mean that this category of farmers has a
comparable level of food availability. The rest of the respondents had annual labour costs that
range between Eth. Birr 20 to 600 27 for the cropping season of 1997/98, with an average labour
cost of Eth. Birr 110. Families who have a longer period of food availability have better access
to labour as compared to those who have a few months’ available food. Among the case study
villages, farmers in Turo tend to spend more on labour than in Buso and Gamole villages,
because of better access to land in Turo than the other areas. Even though there are genera-
tion-long labour organisations, food shortages are impeding the people’s mutual assistance
mechanisms as discussed below. In the past, clan members, neighbours and friends used to
bring their own food to assist families who could not afford food for the labour party.
There are many types of labour arrangements employed in Konso. The major features of
labour institutions reported in Gereti, Buso, Gamole, Kolme and Turo villages are outlined
below.

Alumella
This involves 20-30 members. Members may be charged depending on their labour demands,
but not at the same rate as non-members. The money collected from labour charges is used to
buy oxen for meat. This practice is common throughout Konso. Those who use the group have
to prepare food and Caqa, a local drink brewed from cereals. In every Alumella, there are 2-4
female group members. Their task is to organise food and drinks in the field. This is com-
monly observed in the field during peak working seasons. However, married women are not
involved in this kind of work. The same is true for Perga, discussed below.

Perga
The size of this labour/savings group is usually up to 20. Members do not pay for the service.
The group saves money obtained through working for others. Women charge their own mem-
bers grain and butter (seasoned) that is measured by a local measurement system. At the end

27 The exchange rate for 1 US dollar = 8.5 Ethiopian Birr in October 2002.

101
.tesfaye.thesis 29-01-2003 15:12 Pagina 102

of the season, they make a feast from the grain and enjoy themselves. Membership in Perga ceas-
es with marriage.

Keffa
This is a labour party organised from one’s clan. People resort to their clan members at times
when the head of the household is sick, dead or, in recent years, travelling for military service.

Xirasha
This is a married woman’s working group, which operates on the basis of a working norm. Each
task is divided among the members to allow ease of mobility and exchange of favours. The
women often help each other finish their tasks.

Ugenta
This labour party does not involve repayment of labour. Someone who wants to organise Ugenta
needs to prepare food and drinks for the party. In any case, there is an implicit commitment as
long as a person has participated in your Ugenta sometime in the past. The number of partici-
pants can be 5 to 20. Ugenta can also be organised from one’s family members. If someone is old
and has no labour and maintenance tools, he can ask his neighbour to support him. He can also
ask his friends for construction materials. The beneficiary prepares only food and drink. In case
he has no food, his relatives can help him with grain and drinks. This food arrangement is
called Xarana.

Fedeta
Like Ugenta, Fedeta does not involve payment of money for the service. It is organised by personal
request of the individual interested. Participants may be drawn from neighbours, clan mem-
bers and friends. The size depends on the amount of work involved.

Kenta
This is a kind of labour party organised by neighbours, in addition to relatives and friends par-
ticipating in the work. It takes place when a family head is imprisoned, or sick or dead. In this
case, everyone brings his own food.

Piyolada
The name literally implies ‘land borrowers’. People who have no or less land often go to their
Poqallas to get land. One of the commitments of the Piyolada is to provide labour services to the
poqallas no less than three times a year, in return for the use of the land. This is a type of labour
source that was the exclusive domain of the Poqallas. Participants in the labour party, togather
with others, are people who have borrowed land from Poqallas. The type of work could be
anything from sowing to weeding to terrace maintenance. This labour arrangement was offi-
cially abolished in 1974. But, owing to the spiritual links between the Poqalla and the Konso
people, they still provide labour services to Poqallas under the guise of spiritual significance and
clan relationships.

102
.tesfaye.thesis 29-01-2003 15:12 Pagina 103

5.11 Beliefs, institutions and social learning in Konso’s agriculture

5.11.1 The role of belief systems and institutions

Agriculture in Konso is founded on the strong link between the spiritual world (God - Waaqa),
land and people (see also, Watson, 1998). The association between these three entities is impor-
tant for peace, human fertility and production.
The overriding belief system in Konso is that Waaqa is a supernatural force above the earth
and mankind. Their belief system however does not fit into the same context as that of the
Christian or Muslim religions. The Konso people believe that Waaqa lives far away from earth,
though whatever they do is under His supervision. They see Waaqa’s manifestation in the peace
and order of the society, rather than in regular repetition of prayers in order to have a spiritu-
al link with Waaqa. Waaqa gives rain to the earth so that the earth can give food to the people
as they till it to sustain life. In the same way Waaqa blesses the womb of women to give off-
spring. In order to maintain this relationship, mankind is expected to respect the rules of
Waaqa which require peace and order in all spheres of community life. The elderly are the medi-
ators of these rules. They are the ones who give the blessing in the name of Waaqa for the main-
tenance of the rules. The wider codes of conduct of the people emanate from these rules. For
instance, if two people quarrel over a farm border, the Waaqa would withhold rain and the
whole community would suffer. If people kill each other or disobey the rules of Waaqa in other
ways, then women will become infertile, livestock will produce still- or deformed births, people
will perish due to disease, etc. In order to avoid these consequences, the society must quickly
act through its institutions. The joint administrative body that encompasses the warrior group
(Xela), elderly and the religious figures (Poqallas) takes action against the offender(s). The offend-
ing party may be identified by witnesses or by magico-spiritual procedures.
On top of these ties between the spirit world and the people, it is crystal clear to the Konso
people that they have to work the land to get the grain they eat. What they expect from Waaqa
is rain, which is the most important physical element for the Konso people. Furthermore, they
know that production will be at stake without adequate soil and water conservation measures.
This is a fundamental issue affecting the character of agriculture in Konso - the source of
unique labour discipline that is widely appreciated.
The dominant paradigms in the structure of the society are masculinity and seniority. Male
is the dominant sex whether it is among human agencies or among animals used for ritual sac-
rifices. During blessings, they emphasise the birth of sons, though girls are also mentioned.
Senior citizens, elderly, early settlers, older brothers (not sisters) are the manifestations of
seniority in the society. Thus, power, wealth, respect and their derivatives favour seniority.
Collective decision-making and collective works are the major features of the Konso people.
It is important to note that discussion and collective management is rooted in the Konso sys-
tem of governance. This is clear from the name of their institution, called Keladule Dhegeya
(‘based on agreement’). Therefore, every matter is decided upon in meetings (Sharita). The dis-
cussions carried out during the ruling of Xela and the annual assembly (Kora) have been very
important in this system (see section 5.9).
Collective action is clearly manifested through their organisation of labour (see section 5.10).
The labour institutions are the people’s response to the challenges of the landscape.
Undoubtedly, it is an intentional struggle for survival, which they maintain up to the present
date. The link between the people and natural resources is reflected in their local proverbs.
These proverbs praise their terraces, ponds, streams, trees, livestock, hard-working men, etc.

103
.tesfaye.thesis 29-01-2003 15:12 Pagina 104

5.11.2 Mechanisms for social learning

Institutions and land management practices established by the ancestors were passed down
to the present generation through social learning that is embodied in the system. In this
regard, institutions can be seen as a self-perpetuating system. This property shows the
strength of a system or an institution.
Konso values and belief systems have proved resilient to the many forces that took place
during the last century, from the incorporation by the north in 1897, to the interim period of
Italian occupation (1936-41), to the volatile policies of the Derg government (1974-1991) and the
market forces of the 1990s.
Social learning takes place through religious ceremonies and their venues such as Moras
and other public meeting places (Kora), blessing speeches made by elders, folk dances,
proverbs, songs sung during different farming operations, hunting expeditions (Kara), gender-
and age-based labour parties, wedding and mourning ceremonies, which convey the individ-
ual and societal features. Folk songs during the seasonal and daily routines of the communi-
ty, such as markets, water collection praised good farmers and good farming practices such as
strong terraces, big trees, permanent water points and productive farms. These are parts of
the culture that strengthen the current practices and their values are also passed down to the
future generations.
Besides collective social learning mechanisms, fathers share a great deal of issues with
their children, particularly sons. The head of the household is expected to inform his mem-
bers about public decisions. Undoubtedly, women are highly marginalised from social deci-
sion-making. Women’s marginalisation is attributed by the Konso’s to the widely-held belief
that Waaqa disliked women because of their deviation from His wants. The greatest marginal-
isation is reflected in the women who are in their active sexual maturity years. In addition,
some men also ascribed their marginalisation to the fact that women actually come from
another clan rather than their own, in which case they might leak out the secrets of the
society. Girls are prohibited to mix with the warrior group as they are believed to influence
the moral and physical strength of men who should cope with heavy community duties such
as policing, defence of their territory and hunting.

5.12 Farmers’ knowledge, land management practices and problems

5.12.1 Farmers’ knowledge and land management practices

Land management in Konso is a result of a continuous adaptation of the environment to meet

the needs of the community. This adaptation involves:
1. integrated soil and water conservation;
2. controlled livestock husbandry; and
3. an irrigation system based on flood harvesting.

These practices and associated problems are highlighted in the subsequent sections with the
purpose of showing the complexity of the farming system and the land management prac-
tices, people’s knowledge and skills that were developed over time.

104
.tesfaye.thesis 29-01-2003 15:12 Pagina 105

1. Integrated soil and water conservation

Soil and water conservation in Konso is marked by the combination of physical and biological
conservation measures. These include: stone terraces, tied-ridges, thrash lines, agroforestry,
intercropping, fallowing, manuring, Kraal shifting, burning of debris, minimum tillage and
commercial fertiliser. Each of these practices is briefly discussed in the subsequent pages.

Stone Terraces
As mentioned earlier, soil and water conservation measures have evolved over time in Konso.
Earlier settlements began in the valley. This was a relatively flat area where the use of flood-
ing was the dominant practice to get water and fertile soil. In addition, the use of thrash lines
and tree branches were reported to have begun in the valley. At the beginning, population size
was small, so there was no need to cultivate the hills. Later, people moved onto the highlands
for both their settlement and farming because of incessant attacks from neighbouring ethnic
groups. However, land management practices developed in the valleys were inefficient for hill
farming. As a result, the Konso farmers began building stone terraces to prevent the soil from
running down the slope.
Looking at the use of stone in both fields and residential areas, it would be interesting to
know from where the skills of terrace construction in Konso originate. That is, was it in the
fields or in the residential areas? Or did they develop simultaneously?
Two ideas came out of group discussions on this issue. One point of view holds that the ter-
race construction skills (Kawwatta daweeta) were obtained through the building of towns and
houses. This group mentioned the special design of Konso houses that have two major com-
partments in individual compounds. The upper layer (uita) where residential huts and the
kitchen are built, and the lower layer (arketa) where livestock are kept. These two layers are
always separated by a stone terrace. The lower layer should be strong enough to keep the ani-
mals in their compartments. It is seen as a sign of bad luck to the household should animals,
particularly cattle and sheep, enter into the human residential layer. Therefore, an animal
that manages to pass through this layer will be killed. In addition, the external wall around
the village is constructed from stone. The height of such walls ranges from 2-5 metres.
The other point of view holds that the skills of stone construction came from the farm to
residential areas. They believe that earlier settlers had the experience of constructing stone
barriers in the valley bottoms along with thrash lines and that that experience was used when
hillside settlement began. This group has emphasised the water-harvesting structures that
had stone reinforcements.
Whether or not the skill for stone terrace construction originates from one of these two
options or from both, Konso farms and villages demonstrate an interrupted chain of stone ter-
races and walls respectively.
The very idea of using thrash lines for moisture conservation and water-harvesting struc-
tures for use of flood water in the valleys was an important turning point for soil and water
conservation efforts. Over time, the effect of barriers on soil movement was learned after
observing layers of soil accumulating behind a fallen tree in the mountains. Stone lines
replaced tree branches and logs. Terraces have been constructed from the floor of the valley
to the top of the hill. They were built generation after generation. Father constructed part of
the field, the children took it further and so on. In this way, full coverage of the slopes was
achieved. This actually shows how the culture of soil and water conservation transferred from
generation to generation.

105
.tesfaye.thesis 29-01-2003 15:12 Pagina 106

Terracing in Konso begins from the lower part of the field and moves upward along the
slope. During construction, people dig a basement of 30-50 cm, with a width of about 25 cm.
The height of the terrace depends on the slope of the area. When the area is gently sloping,
the height of the terraces and the width between terraces decreases and increases respective-
ly. There are several terraces of over two meters while the terrace width in between mostly
ranges from 2.5 to 6 meters. The stone wall is built against the newly cut wall face. The soil
from the upper part is moved down behind the stone wall and is carefully piled. Space creat-
ed due to irregular stone faces are filled by soil. In this manner they attain a bench terrace
immediately. The height of the terrace is often maintained above the ground. This structure,
among other components, ensures infiltration of raindrops that fall on each plot.
Soil conservation experts start conservation of a watershed or a field from the top and go
down to the lowest point. They work on the premises of slow growing terraces that were to
attain bench structure over time. This is expected to be achieved through periodic mainte-
nance and soil movements between terraces. In view of this, the experts’ view was found to be
contrary to the farmers’ generation-long experiences.
Everyone in the hoe-farming system acquires the skill of terrace construction. Maintenance
of terraces is part of routine farming practice. They construct the lines on a contour without
any measuring or guiding instruments, depending on their visual judgement of the ground’s
features. When the ground is steeper, they maintain terraces with a short width. Sometimes,
they use a subsidiary terrace (Paqayta) to support the big one. Paqayta is built on land with
irregular patterns. This is done to maintain a continuous line of the main terrace on one
hand, and to prevent loss of soil and water due to difference on the slope.
After finishing the terrace, they add organic matter to compensate the covered topsoil. This
is achieved by mixing the topsoil (Kefeffa) with the big Payra. The fact that the piece of land
immediately under a terrace loses more soil in the process is well understood. That piece of
land is given special attention during subsequent cultivation. In addition, the farmers have
noted that backward movement of water within the terrace brings more soil and organic mat-
ter from the thrash lines on the edge of the terrace to the back of the terrace where most of
the topsoil had been lost during construction. Women play a great role in transporting
manure from the villages to the farms. During the construction process, they assist in col-
lecting and heaping the stones.
Based on the stone-erection ceremonies discussed in section 5.9.1, that shows a record peri-
od of more than five hundred years, it can be safely estimated that the Konso people started
using terraces for at least four hundred years. However, archaeological findings that are not
yet available would provide the best evidence.
From the foregoing discussion, it is clear that soil and water conservation efforts in Konso
are very labour intensive. The high demand of labour for construction and maintenance of
soil and water conservation structures have proven to be one of the major limitations else-
where. However, the Konso people have their labour organisations to meet their labour-inten-
sive agriculture. Their understanding of soil erosion stimulated them to develop a coping
mechanism. As a result, they established their ‘niche’ in a rather ragged and difficult terrain.
Farmers in Konso have an amazing appreciation for stone terraces. The most important
advantages of stone terracing are:
• protection of soil;
• water-retention in the field;
• serving as shade for coffee trees at a young stage;
• serving as a support to climbing crops such as lablab;
• making hill farming possible;

106
.tesfaye.thesis 29-01-2003 15:12 Pagina 107

• increasing soil fertility; and

• increasing production.

Of course, their appreciation is not without an understanding of the challenge. Almost all
farmers who participated in individual and group interviews confirmed the high labour
requirement to build and maintain stone terraces. Physical injuries arising from stone move-
ments, physical fatigue and the need for more food was among commonly mentioned
problems. The issue of food requirements is particularly important as the time for terrace
maintenance coincides with the time of food shortage in Konso. During this time, the pre-
ceding harvest comes only from ratooning, which fails during drier years.

Tied-ridges
Tied-ridges are one of the components of soil and water conservation in the agricultural sys-
tem. It is an age-old practice in the system that also originated in the valleys. Tied-ridges are
constructed in order to reduce the run-off from the upper terraces and to retain moisture.

Thrash lines
Thrash lines are used as a mulch and source of organic matter. The material used for thrash
lines are sorghum stalks that are uprooted during the dry season and maize stalk left in the
farm. Cereal stalks easily decompose under the hot climate and the effect of termites in some
fields. When the previous year’s production has failed or the crop stand was poor, there will
be a shortage of materials for mulching and thrash lines. This in turn affects the production
of the subsequent year.

Agroforestry
Agroforestry is a typical feature of the hoe-farming system in Konso. This practice is less con-
spicuous in the ox-plough farming system.
Most fields in the hoe-farming system are characterised by a multiple cropping system. The
dominant high tree species is Morinaga stenopetal. This is because of the importance of its leaf
in the people’s daily food. The next important tree is Terminalia browenii. It is grown for forage,
farm tools and building materials. Other tree species that are grown in association with crops
are Olea africana, Ficus sori, Cordia africana, Strerculia africana.
The next storey is filled by crops such as coffee, chat (Catha edulis), yam, cassava, pigeon pea
and cotton, as the case may be. Coffee, chat and cassava are planted under the terraces, where-
as yam is planted both below and above the terrace depending on the soil depth. When it is
planted above the terrace, farmers maintain an adequate distance from the terrace to avoid
damage of the terrace during harvesting. Those farmers who have shallow soil depth do not
plant yam at all, as it requires deep soil for its conspicuous rooting system. Pigeon pea is plant-
ed near the edge of the terrace, except when it is sown as a ley crop whereby the land is left
to fallow in the subsequent season(s). It is confined to the border in order to minimise the
shade effect as it stays in the field for 2-3 years. It is worthwhile to note that not all farmers
understand the beneficial effect of pigeon peas on soil fertility. Some farmers generally main-
tain low crop diversity due to a decline in depth and soil fertility. The pigeon pea is also
excluded. Cotton is thinly broadcasted in the whole field, unless the largest share of the plot
is allocated to it. Sorghum, millet and maize are the base crops for intercropping (see below).
Pulses and oil crops such as fenugreek and sesame cover the lowest ground. All in all, one can
count 10-15 crops in a field in addition to a high tree. Some studies have reported over sixty
different crops grown for different purposes (Mesert, 1990). Note that farmers often try to

107
.tesfaye.thesis 29-01-2003 15:12 Pagina 108

maintain the intensity of intercropping with the level of soil fertility in each plot.
Integration of high trees in the ox-plough farming system is declining over time. With the
introduction of the ox-plough about a century ago, farmers in Turo and the surrounding areas
have moved from the hillside farming to gentle slopes with the help of oxen, thus abandon-
ing terraces. Extensive bush clearing took place in the area with few scattered tree species left
in place. As terraces are less compatible with ox-ploughs, farmers have abandoned the terraces
that were built by their ancestors. Instead, they have focused on agronomic and soil manage-
ment practices rather than physical conservation practices that typify Konso. Unlike the hoe
system, cropping patterns in the ox-plough system involve less intensive intercropping. In the
latter system, single or two crop intercropping is common rather than multiple cropping.
Adaptation of the ox-plough to terraced plots is observed in Kolme, which traditionally
practises hoe farming. This adaptation was induced by a shortage in labour for land prepara-
tion, in order to be able to sow at the time when there is moisture.

Intercropping
The general pattern of crop association is a mixture of cereals with pulses. A cereal-cereal-
pulse arrangement may take place depending on the level of fertility of the land. In most
cases, 50-75% of the space is allocated to sorghum.
Farmers practise intercropping in order to ensure the availability of food from different
crops and to obtain feed (thinning products) on continuous supply. Some crops are drought-
tolerant (sorghum, cotton, pigeon pea), while others (e.g., haricot bean) are early-maturing if
there is adequate rainfall. The harvesting sequence is haricot bean followed by millet, maize
and sorghum, in that order. Pigeon pea and cotton last longer in the field. This arrangement
helps them to minimise risks from moisture stress. In addition, pigeon pea and cotton provide
a soil cover, particularly in the ley system wherein pigeon pea gives adequate coverage. The
effectiveness of this system, however, needs to be studied. Farmers have observed that some
crops can be vigorous if they are sole cropped (e.g., sorghum). On the other hand, they stated
that haricot bean cannot withstand direct sunlight. They understand that haricot bean gives
shade to the roots of sorghum and maize and thereby conserves moisture under its shade. In
addition, they have underlined the use of haricot bean leaf to improve soil fertility when it
decomposes. Apart from this, they do not understand the use of nodules in leguminous
plants, which they consider to be a disease 28. Different farmers gave different interpretations,
including leprosy, worm infection, physical damage and the like. Most farmers understand
the anchoring service of pulses’ roots. Unlike others, one farmer who had attended 8th grade
level education and who had attended a farmers’ training centre in the 1980s, reported his
knowledge of the nitrogen-fixation ability of legumes. He also indicated that his fellow farm-
ers do not understand this capability.
Intercropping also helps to minimise risks from disease and pests. When pests affect
sorghum, farmers may get production from haricot bean and pigeon pea and vice versa. It is
also seen as a method to minimise erosion. Few farmers noticed this phenomenon. Sorghum
remains in the field for ratooning after the other component crops are harvested, making use
of the short rainy season. Ratooning provides soil cover to the fields to minimise erosion.
Nowadays, farmers in ox-plough areas commonly practice sole cropping of haricot beans to
avoid the undesired shade from maize and sorghum as well as nutrient competition.
Moreover, in ox-plough system, many farmers, particularly those who have large farms, do not
practise ratooning, which could have given good soil coverage during the short rainy season.
They are indifferent because they can produce enough during the main rainy season. They

28 Farmers in Wolaita and Wello (Chapter 6 & 7) gave a similar explanation.

108
.tesfaye.thesis 29-01-2003 15:12 Pagina 109

think that they had better put their land under fallow instead of using it for ratooning. Apart
from that, the ox-plough system is dominated by cereals and pulses. There is a poor integra-
tion of root crops in this system.
Intercropping also helps farmers make better use of their limited land and scarcity of
labour. Farmers who have a small farm can grow the major crops they need for their home
consumption, though this hardly ensures family food security. Cultivation and weeding
practice also benefits different crops in the same field. Unlike sole cropping, weeding in
intercropping system requires careful operation, to avoid damaging the roots of crops with
different rooting arrangements.

Fallowing
Fallowing as a land management practice is carried out from short (half a year to two years)
to longer periods (three to five years). Generally, farmers allocate their plots to different land-
management practices according to the opportunities they have. For instance, almost all
farmers do not fallow their main and accessible farm. In this case, they devote most of the
available means of soil-fertility management practices to such plots. Due to the bigger field
size in the ox-plough system, farmers still practise fallowing of five more years.

Manuring
The major sources of organic matter are manure, household refuse and thrash lines. Pulses
and tree leaves and branches of harvested trees also provide organic matter.
Manure is obtained from animals raised at home and from the community garbage square,
located outside the village fence, adjacent to the traditional Dina. Unlike in northern and cen-
tral Ethiopia, animal dung is not used as fuel 29. Until about 30 years ago, farmers in Konso
used to apply dried human faeces to their field (see also Hallpike, 1972). This practice was
abandoned after they came to know the disease cycle, particularly, drinking-water contami-
nation. The issue of dignity was also involved to cease the practice. Use of community garbage
is not appreciated by some farmers because of risk of weed seeds, bad seeds and excess ash
that burns crops during dry spells.
It is normal to see a Konso woman crossing the hills every morning with some organic mat-
ter on her back wrapped with skin, a gourd of caqa and her payra or sibilota depending on the
season. In that, whether she is a single head of a household or a partner, Konso women con-
tribute a lot to the agriculture of the area, while also fulfilling the home management and
reproductive responsibilities.

Kraal shifting
Those farmers, who own cattle and labour, practise kraal shifting. This method is practised for
distant farms where the application of manure by human labour is difficult, in spite of the
availability of manure. Unfortunately, it is not suitable for families who do not have labour to
protect the animals from wildlife and theft.

Burning of debris
Burning of debris takes place in the months of January and February. During this period
Konso and Deresha Weredas seems to be setting off fireworks because of all the smoke blowing
into the sky. Farmers collect non-decomposed stalks of cereals, tree branches, residues of

29 I was ridiculed by a farmer for raising this question.

109
.tesfaye.thesis 29-01-2003 15:12 Pagina 110

pigeon pea, cotton branches and weeds and burn them by making the heaps scattered
throughout the field. This ash is then ploughed under. According to farmers, this practice
increases the soil depth and fertility. In the latter case they have used the analogy of butter 30.
They added that the soil would look like a woman who has polished her skin with butter,
when debris is burned and ploughed under. During the period of shortage, competition aris-
es between the use of crop residue for soil fertility or animal feed.

Mininum tillage
Tillage in Konso involves minimum tillage. The major tillage is carried out every second or
third year. The range depends on the fertility status and labour availability, and other
management practices such as manuring and fallowing. Due to the shortage of labour,
women-headed households often opt for manure application instead of tillage. Tillage is car-
ried out by the larger payra to turn the subsurface soil to the top for aeration and exposure to
sunlight. During the minimum tillage years, they simply uproot the sorghum stalks in the dry
period and then cultivate (without tilling) the land during the rainy period for sowing.

Commercial fertiliser
Nowadays, a few farmers have begun using commercial fertilisers in areas with better mois-
ture availability and access to irrigation from rivers.

Indicators of declining soil fertility

Farmers easily identify indicators of soil fertility. Commonly mentioned indicators are:
decline of crop productivity, difficult workability, emergence of noxious weeds, particularly
striga. For instance, the decline of crop productivity is further explained by stunted crop
growth, a yellowish leaf colour, and small head-setting. While they can identify these indica-
tors right away, other related soil fertility indicators come out only after facilitated discussion.
These are soil depth, change of soil colour and water-retention capacity.
Farmers related soil depth to workability. They said that less fertile soils are difficult to till,
unlike the fertile ones. Fertile soils are soft like humus. They have noted that water percolates
through these kinds of soils instead of carrying the topsoil away. Moreover, they have observed
the cohesion between the farm tools and the soils. Unlike the light soils, fertile soils have the
property to stick to the farm tools and are also ‘heavier’ when held in hand.
They associate change of soil colour to poor fertility. However, they could not explain how
the colour change is created. In the course of further discussion, however, they were able to
relate the change of colour at least to removal of topsoil by erosion. The different colours
reported for poor soils are: light, pale, grey and whitish (cf. 5.2.3). It was noted that a whitish
colour is an indication of an eroded soil that is called Mokosha. Fertile soils have red, brown or
dark colours.
Soil fertility management practices among farmers result in different trends in soil fertili-
ty. Most farmers reported a decline of soil fertility in different plots while a few farmers had
managed to at least maintain or increase fertility of a few plots.
Some farmers who managed to increase soil fertility in at most two plots used manure on
a continuous basis. The other means was fallowing for five years. Plots where fertility has been
maintained for the last 5-10 years were able to due to manuring, terracing and flood harvest-
ing.
Decline of soil fertility is a widespread phenomenon. The common problems identified by
farmers are summarised in table 5.3.

30 Surprisingly, I have heard a similar analogy from farmers in the Dire Dawa region, over 1,000 km away in eastern Ethiopia.

110
.tesfaye.thesis 29-01-2003 15:12 Pagina 111

2 Controlled livestock husbandry

The Konso people have developed a livestock husbandry system that is compatible with their
soil and water conservation practices. Aftermath grazing is one of the major obstacles to bio-
logical conservation and structural measures in most cereal ploughing farming systems in
Ethiopia. People in Konso strictly keep their animals away from the crop fields in order to pro-
tect their integrated soil and water conservation system. They achieve this by tethering meth-
ods and Fora, a ‘camping grazing’ system wherein the herd is taken away for a period of time
to distant areas. Fora helps to minimise pressure on nearby grazing areas and forage plants on
the farm. Without these arrangements, age-old terraces, thrash lines, and agroforestry prac-
tices could not have taken root.

Table 5.3: Summaries of problems of declining soil fertility, Konso, SNNPR, Ethiopia

Problems Reasons

1 Repeated sowing Shortage of land

2 Short or no fallowing Shortage of land
3 Less or lack of manure Shortage of livestock, distance of farm plots
4 Soil erosion Lack of terrace, or poor maintenance due to shortage of
labour and food
5 Highly erodable land Due to steep slope
6 Poor management of land Shortage of labour and distance for the available labour
7 Shortage of materials for thrash lines Poor crop production, use for animal feed

Obviously, Konso farmers did not practise controlled livestock husbandry at the early stage of
farming in the valleys where there were no perennial crops and physical structures sensitive
to animal trampling. Therefore, controlled livestock husbandry is a stage in the farming sys-
tem development rather a practice that has developed with state rules and regulations.

3 Irrigation system based on flood harvesting

There are a few seasonal and perennial rivers in Konso. These are: Segen, Woito, Gato, Delbena,
Gaba maga, Yanda fero and Regede. Delbena and Gaba maga are relatively accessible rivers. Other
rivers are located in the valley. Because of this, river irrigation has not been developed in
Konso in spite of the people’s vision and hard-working nature. However, they have developed
amazing flood-harvesting structures in the seasonal riverbanks.
The irrigation structure built at Teshmele (1,240 masl) is an engineering miracle, for bare-
handed workers. Massive flat stones are lined to divert the direction of the flood from its river
flow. Networks of canals were then constructed (see Plate 5.3). Passage bridges were con-
structed from local stones to distribute water from its main course to different fields. Excess
water is finally guided to join its natural way. The fields at Teshmele have deep soils as a result
of years of silt accumulation. The height of the structure ranges from 1-2.5 m in the fields and
3-4.5 m in the riverbanks.
Flood harvesting in Konso is a round-the-clock task. A drop of rain has great meaning in
that moisture-stressed land. Whenever there is rainfall, every able member of the family runs
to his field. They do this whether it is midday or midnight. At night they light their torches

111
.tesfaye.thesis 29-01-2003 15:12 Pagina 112

kept aside for this purpose and reach their farm. Everyone opens his canals by monitoring the
field capacity, to protect loss of soil. This practice is common in the Buso, Aba Roba and Nalia
Segen Peasant Associations.

Plate 5.3: Terraced fields irrigated by flood rivers in Konso

River course

An example of flood inlet

5.12.2 Replicability of soil and water conservation practices in Konso

Are soil and water conservation practices in Konso replicable to other parts of the country? I
am prompted to raise this question because of:1) a growing ‘conservation tourism’ organised
to Konso, mainly by NGOs, but also by Regional Governments; and 2) a frequent mention of
the Konso case by scholars with the implicit intention to adapt it elsewhere.
Answers to this question can be confirmatory or non-confirmatory depending on the set or
sets of practice we are talking about. Whatever the result may be, what can be adapted would
be the technology or the practice, not the embodying culture. It is the latter aspect that dis-
tinguishes land management of one society from the other. Whether adaptation would take
root or not depends on the existence or creation of internal motivation, and of a deep
commitment and understanding of the biophysical processes to manage land in a certain way.
For instance, if a given society happens to begin building terraces after visiting Konso, this
does not guarantee the same success. However, there is always room for sharing experiences
among farmers from cross-cultural backgrounds for there are several common experiences
among farmers in spite of their geographical locations. For that matter, many of the agro-
nomic practices employed by Konso farmers are common all over Ethiopia. However, the total
system of land management that developed in Konso, with both its ‘hard’ technical side (the
terraces), and its ‘soft’ side (the institutions, culture, and political system) cannot easily be
adopted elsewhere.

112
.tesfaye.thesis 29-01-2003 15:12 Pagina 113

5.13 Problems and prospects of agriculture in Konso

Agriculture in Konso has supported its people for several centuries. Its carrying capacity was
maintained mainly due to relevant land-management practices and was partly assisted by
slow growth of the population, both of which are the results of their institutions. Population
growth remained slow due to Fereyuma, and the social limitations on sexual activities even
within marriage.
Long ago, agricultural lands were put under long fallow periods. Vegetation cover was then
much denser than today. As a result, soil fertility was high. Due to the availability of animal
feed from various sources, competition for crop residue was minimal. For the same reason,
animal productivity was very high. Nowadays, the general perception is that the past was bet-
ter, particularly in relation to wealth and prosperity.
Crop failure occurred from 1998 to 2000, consecutively, whereby many lives were saved by
food aid. Crop failure occurred due to shifts of rainfall patterns, and not because of poor land
management. In Konso, soil and water conservation structures of most farms are still in place.
Variation in land management, undoubtedly, brings about differences in crop productivity,
but not indiscriminate crop failure. Therefore, it can safely be said that the declining trend of
agriculture in Konso is primarily because of a change in the rainfall regime. This is probably
due to gradual effects of changes in the microclimate and perhaps the influence of the glob-
al or at least the regional climate. Even though the collapse of the population control system,
Fereyuma, plays its part in the decline of soil fertility through its impact on the fallow system,
it cannot explain the total failure of crops in Konso.

Plate 5.4: The growing scars in the once- glorious Konso stone terrace network.

When crops fail, people sell their animals to buy food for the family. They cut the trees down
that would have had very little chance to rejuvenate. As a last resort, able family members
migrate in search of a job to support the family. Women and children embark on cutting fire-
wood from the bush lands for sale. As a result, conservation structures in the field are neglect-
ed to the extent that some families are seriously frustrated from maintaining them. In recent
years, the maintenance of terraces has become a difficult task because of the shortage of food
to organise a labour party. People become physically weak to cope with the arduous tasks of

113
.tesfaye.thesis 29-01-2003 15:12 Pagina 114

terrace maintenance under the blazing sun. Consequently, the at one time uniformly main-
tained and celebrated terraces of the Konso land are showing immense scars of damage (see
Plate 5.4). In the past, the majority of the people migrate internally within Konso or to the
rural areas of the neighbouring Weredas (Messeret, 1990). As these opportunities are now over-
saturated and also failing under pressure, more and more people are migrating to the urban
centres and the large farms nearby.
When productivity in the terraced fields declines, farmers expand the bush farms which
are often cultivated only once a year and without a stone terrace. The common conservation
practices in such fields are thrash lines. However, the stalks lined for these purposes are eaten
by animals grazing in the bush forest, as these areas are normally not used for crop produc-
tion.
Given the present scenario, what is the future of agriculture in Konso? How can people live
from such a fragile environment? The rainfall pattern is very much undermining an unprece-
dented culture of land management. In view of this, it is high time to look for alternative
means of livelihood. Unfortunately, alternatives are not readily available. In my view, a lasting
solution remains in the development of valleys. Such measures require the development of
infrastructures for irrigation, improved health facilities, improvement of roads, supply of
electricity for deep wells, building marketing infrastructures and creating mechanisms for
resolving conflicts about resource use within and between ethnic groups. These alternatives
are likely to ease the pressure on the cultivated hills and possibilities to restore the indigenous
land-management practices, such as long fallow and improved vegetation.

5.14 Conclusions

Konso’s system of agriculture has unparalleled complexities, at least in Ethiopia, that involves
a unique human impact on the landscape. It is a typical example of a multiple production sys-
tem where an adequate balance has been created in relation to the needs of each household.
The system encompasses crops, livestock and tree production with interrelated management
practices. The unique feature of Konso agriculture is a stone terrace-based integrated soil and
water conservation system. These terraces cover major parts of the agricultural land in Konso
and stretch over tens of thousands of kilometres. Farmers apply several methods of soil fertil-
ity management practices. Among these, the agroforestry pattern developed in that environ-
ment plays an important role, among others, in soil fertility management and food and feed
production.
The emergence of this unique agricultural system is rooted in the history of the Konso
people, mainly in their belief system and institutions that have an indigenous record of over
500 years. Their belief system is the centre of their institution that caters for all aspects of
societal functions, from the production system to internal stability to the protection from
interference of neighbours. The indigenous institutions adequately integrate the spiritual and
the emerging political actors in pursuing the societal goals. In this regard, the harmonious
collaboration of, among others, the Xela, Apa timbas, Apa kumas and Poqallas is very interesting.
These institutions proved resilient to the many external forces that interfered with the Konso
system over the course of the last century.
With respect to land management, the social organisation of labour and indigenous land
tenure played a great role. The scope of their labour organisation transcends a mere labour
exchange mechanism by creating labour markets, saving method, mutual trust and assis-
tance.

114
.tesfaye.thesis 29-01-2003 15:12 Pagina 115

Konso people have developed a culture of hard work and a knowledge-intensive system of
production. They have a knowledge and skill of tested agricultural practices that is passed
down from generation to generation. The new generation learns the agricultural practices
and the culture of hard work from the functions of the social structure and the family circles.
The heritage of Konso’s agriculture characterised by the stone terraces is one of indigenous
intensification far ahead of the Boserupina thesis (1965). There was no research or extension
support nor was there slave labour to construct it. All was the result of the ingenuity of the
Konso people. Thanks to the institutions born out of societal necessities, that were main-
tained and passed down to the present generation, the present shape of Konso’s agriculture
was ensured. Thus, Konso’s intricate system of soil and water conservation is based on (1) the
technical measures, with stone terraces, that are so visible when one visits the area, and (2)
the institutions that support it. These include the organisation of labour, the erstwhile
human fertility management system, the leadership and disciplines, as well as the spiritual
system that underpinned it all. In this respect the Konso system of land management is simi-
lar to that of the Ifugaos (Gonzalez, 2000). The necessity of this combination of the hard and
soft sides of land use (Röling, 1997) makes impossible to contemplate introducing only the
hard side without paying attention to developing the soft side.
However exemplary the Konso soil and water conservation system may be, it cannot be
copied wholesale elsewhere, as the embodied culture remains its own. Even then, farmers in
other parts of the country can learn a great deal from Konso’s agriculture to improve upon
what their culture and physical environments provide.
Unfortunately, indigenous agriculture in Konso has begun to show a sign of exhaustion
from its long wrestling with the many forces that have been interfering with it. The
exhaustion appeared because of unfavourable internal and external situations. Internally, the
population size increased over the last 30 years due to the disintegration of the indigenous
population control mechanisms (Fereyuma). Externally, wider integration into the regional
markets and institutions, peace and stability that allowed temporary migration 31, and short-
age of rains which caused crop failures over consecutive years. Therefore, chains of stone ter-
races now show immense damage due to lack of maintenance. Farmers are now unable to
carry out the periodic maintenance due to a shortage of food. This situation has forced farm-
ers to migrate to other places in search of alternative survival strategies. As a result, their
indigenous labour organisation, which was the backbone of the soil and water conservation
efforts, has failed to operate as usual.
The Konso agricultural system, particularly in its hoeing culture has moved very far from
one’s idea of farming. It rather consists of energy-draining exercises with which people are
forced to live. This is even all the more true as the age-old terraces are falling apart on many
farms. Therefore, maintaining livelihood in that environment requires heavy external
support to reduce the land pressure so that long fallowing can be practiced to ease soil fertil-
ity-restoration practices. This can best be done through land reclamation within Konso, by
developing an irrigation system from the limited accessible rivers and development of the
valleys. The latter requires immense infrastructure facilities, rural development packages and
collaboration with the neighbouring ethnic groups to avoid conflicts on resource use. In the
short run, however, food should be provided to families who are food-deficit in addition to
supporting food to organise the traditional labour parties credit to expand off-farm and non-
farm income sources.
The importance of the Konso case for this thesis is that it makes clear the importance and

31 Note that these social forces, while beneficial in creating alternative livelihoods, work negatively against the mainte-
nance of the indigenous agriculture.

115
.tesfaye.thesis 29-01-2003 15:12 Pagina 116

nature of the ‘soft’ side required for a working system of soil and water conservation. The fol-
lowing chapters will examine soil and water conservation in two areas whose external inter-
ventions about two decades ago introduced purely technical soil and water conservation mea-
sures. The elaborate labour organisation, human fertility management and other institutions
developed in Konso provide a backdrop against which to understand what happened.

116
.tesfaye.thesis 29-01-2003 15:12 Pagina 117

6 Farmers’ Response to Interventions in Soil and Water

Conservation in Wolaita, Ethiopia

Abstract

The Soil Conservation Research Project (SCRP) that was financed by the Swiss government succeeded the for-
mer Wolaita Agricultural Development Unit (WADU) in the early 1980s, to develop and disseminate a suit-
able soil and water conservation programme in the wet Woyna Dega agro-ecological zone in southern
Ethiopia. The project was stationed at Gunno, Wolaita. The SCRP project was actively operational from 1982
to 1995 and gradually phased out in 1998. This study was initiated to assess the impacts of the project on
farmers’ soil and water conservation (SWC) practices, to analyse farmers’ responses to the SCRP’s introduced
technology and the reasons behind their responses, to understand farmers’ knowledge and attitudes on ero-
sion and SWC practices, to assess farmers’ SWC practices and to find out suggestions for improvements. The
study used a combination of qualitative and quantitative research approaches. The study shows that farm-
ers in Wolaita have a fairly adequate knowledge and understanding of the erosion processes and their neg-
ative consequences. The soil bunds introduced in Wolaita were found to be effective in minimising runoff.
In spite of that and in spite of farmers’ appreciation of the problem, there ultimately resulted an increas-
ing non-acceptance of the soil bunds introduced by the SCRP through food-for-work. The reasons for the poor
acceptance of the soil bunds were: lack of farmers’ participation in the problem identification and subse-
quent research processes; use of food-for-work as an incentive without a profound and continuous educa-
tional process; negative side effects of the soil bunds that increased farmers’ risk of crop loss, which the
research-extension system failed to improve; and a termination of previously active labour co-operation
within the community for heavy farming operations such the construction of bunds. With the exception of
land size, livestock ownership and family size, farmers’ responses to the soil bunds were not affected by other
household and individual characteristics such as age and level of education. In addition, no strong evidence
was found on the negative impact of land tenure on the utilisation of soil bunds in Wolaita. Finally, the
study concluded by underlining the need to bridge the gap between physical and biological conservation
measures in Wolaita in view of the increasing incorporation of steep land under agriculture.

117
.tesfaye.thesis 29-01-2003 15:12 Pagina 118

6.1 Introduction

Wolaita is located in southern Ethiopia, and is one of the overpopulated areas in the country.
The production system of the area is that of mixed farming. Short V-shaped valleys dominate
the terrain. The higher edges of the valley are used for settlement while farming takes place
along the concave slopes down the valley. Farming also takes place on steep slopes in the
mountains and hills.
In the early 1970s, the Wolaita Agricultural Development Unit (WADU) was initiated to
promote modern agricultural technologies including a range of soil and water conservation
practices in the entire Wolaita zone. This unit was terminated in the early 1980s and
succeeded by the Soil Conservation Research Project (SCRP). The SCRP used Gunno as its exper-
imental site while the government’s extension system continued attempts to promote soil
conservation practices in the whole of Wolaita after the termination of WADU.
In spite of the continuous efforts that span more than three decades, the soil and water
conservation techniques introduced did not take root among the farmers. In the experimen-
tal catchment at Gunno, farmers did not maintain their soil bunds except during the first two
years after the construction of the bunds. In some cases, farmers removed soil bunds com-
pletely or selectively. Understandably, the dissemination of the soil bunds to wider areas in
Wolaita did not succeed in spite of increasing erosion threats due to the expansion of farm-
ing to steeper land.
The purpose of this chapter is to show how farmers reacted to the soil and water conserva-
tion intervention that began during the 1970s and to analyse their reasons in order to identi-
fy effective interventions for sustainable land management in the area. These issues lead to a
number of questions: What is the farmers’ understanding of soil erosion processes? What is
the farmers’ perception on soil and water conservation as a land management practice? What
do they expect from soil and water conservation? What do they do to conserve their soil from
erosion? What are their reasons for not carrying out periodic maintenance and construction
of new soil bunds?
Section 6.2 presents a general background to the study area. This is followed by the history
of interventions in soil and water conservation (section 6.3). The methodology followed in this
case study is discussed in section 6.4. Section 6.5 presents the results of the study, i.e., as an
integration of both qualitative and quantitative studies. Finally, the conclusions are given in
section 6.6.

6.2 The study area

6.2.1 Location of the study area

Wolaita was recently upgraded to a zone in the Federal Administrative System, which was ini-
tiated by the government in 1991. It is 400 km south of Addis Ababa and located between the
Omo river in the west, Lake Abaya in the south and the Blate river in the east (see map 1 in
chapter 2).
According to the 1994 census of Ethiopia (CSA, 1996), the Wolaita Zone has a total popula-
tion of 1,147,691. Of this 93% is rural, whereas the urban population accounts for only 7%. The
zone has a total land area of 3,928 km2. A population density of 550 persons per km2 was
recorded at the Soil Conservation Project site at Gunno (SCRP, 1996b). Population density of
all of Wolaita, based on the 1994 census, is 292 persons per square kilometre (CSA, 1996).

118
.tesfaye.thesis 29-01-2003 15:12 Pagina 119

6.2.2 Historical background

Socio-political history
Wolaita is composed of 110 clans that are divided into two ethnic groups: Mala with 12 clans
and Dogela with 98 clans (Wana, 1998). According to a historical interview with elders in
Gunno, the Wolaita people originate from a place called Kindo that is located in the Southern
Region.
Historical evidence shows that the present territory of Wolaita is a result of successive bat-
tles with neighbouring ethnic groups and nationalities (Wana, 1998). War has been one of the
dominant historical features of the Wolaita people much like that of the Northern Kingdom,
which subjugated the Wolaita in 1894. During those days, successive kings of Wolaita
mobilised tens of thousands of soldiers with traditional hand-held weapons and horses for
war expeditions or territorial defence. As such, war has played its part in shaping the politi-
cal history of the region, particularly in terms of creating the motive to increase population
size in order to maintain a strong army. To that end, efforts were made to settle outsiders
using land and power as incentives.
Before incorporation into the Northern Kingdom, Wolaita had its own kingdom that ruled
from the thirteenth century (1268) (Cerulli, 1956).
Christianity was introduced well before the final incorporation of Wolaita into the
Northern Kingdom in 1894. The introduction of Christianity before Wolaita’s incorporation
was partly due to the process of empire building of the Wolaita kingdom, which brought
people from diverse origins together.
Before its incorporation, Wolaita had been prominent in terms of agriculture and trade.
The latter was, however, based on bartering until a cash economy was introduced after the
incorporation (Tsehai, 1994).

Land tenure
Early settlers acquired land by slash and burn. During the period that it was the Kingdom of
Wolaita, the king administered the land. Newcomers including those from the north were
admitted into the kingdom to increase its power base. War heroes were given special treat-
ment: they were offered the best land irrespective of whether it had been occupied by others
or not. According to the elders, this led to tenure insecurity during the time of the kingdom.
Distribution of land to war heroes and nobility occurred particularly during the reigns of King
Amado and his son, King Gobe (Wana, 1998). Nevertheless, the land tenure system allowed
owners to inherit, give, lend and sell their land without any restriction. Purchase of land was
the best means of capital formation and carries a significant social prestige.
With the 1975 land reforms, rural people obtained access to usufruct land, which was
placed under the administration of the Peasant Associations. Even though the land reform
was received with enthusiasm on its early days, bottlenecks later began to emerge, such as cor-
ruption by the Peasant Associations, villagisation, removal of rights to land owned outside a
given Peasant Association and insecurity of tenure as land became scarce.
In spite of these changes in land tenure, the basic features of the farming system based on
perennial tree production continued in the whole of Wolaita to the present date.

119
.tesfaye.thesis 29-01-2003 15:12 Pagina 120

6.2.3 Local institutions

Apart from the administrative institutions of the earlier kingdom, local institutions existed
that were functional at the village level, and survive to the present day. These are communi-
ty-based organisations for funerals and weddings (Idiriya) and local thrift (savings) groups
(Shufa). These organisations are very important for mobilising resources and labour that serve
members of the community irrespective of their wealth status. However, people with more
resources enter into wider social relations that transcend the neighbourhood boundary,
including age group, kinship, etc., for refuge during bad times.
In addition, farmers who do not have enough land, but have enough labour, get access to
land through the well-established crop-sharing arrangements. Similarly, livestock-shared rear-
ing is practised. Local institutions also provide access to labour through different reciprocal
or non-reciprocal labour arrangements.

6.2.4 Description of the farming systems

Climate
In terms of moisture regimes, the study area is classified as moist sub-humid. According to the
traditional classification, it is in the Weyna Dega altitudinal belt (Weigel, 1986a).
Wolaita has a bi-modal rainfall pattern that extends from March to October. The first rainy
period (Belg) occurs in March to May, while the second rainy period (Kremt) covers July to
October, with its peak in July/August. The average annual rainfall over 43 years is 1,014 mm.
The mean annual rainfall for the decades of the 1970s, 1980s and 1990s was 1,015mm, 920 mm
and 1,290 mm, respectively (NMSA, 2000). According to the SCRP’s data (SCRP, 1996b), the
mean annual rainfall at Gunno station is 1,314 mm. As elsewhere in Ethiopia, the major
problem of rainfall in Wolaita is its distribution.
The mean annual temperature in Wolaita is 19.5 o C. The coldest and warmest months are
September and March respectively.

Soils
According to Weigel (1986a:7), who carried out a detailed soil survey (scale 1:5000), about two-
thirds of the survey area is covered by Eutric Nitosols associated with Humic Nitosols. These
are dark reddish-brown soils with an extremely deep profile that reach depths of up to 10
meters (ibid.: 13). This is visible from deep gullies and road construction works. While high
rainfall availability and high moisture storage capacity of these soils favour a good crop pro-
duction, low availability of phosphorous and nitrogen are severe limitations under the cur-
rent soil management practices (ibid.).

Land use, farming patterns and crops grown

Wolaita has an enset-based mixed farming system, where enset (Ensete vetricosum) is a co-staple
food together with cereals, roots and tuber crops (Westphal, 1975). Like everywhere in the
highlands of Ethiopia, livestock is an integral part of farming. Farmers rear cattle, sheep,
goats, chickens and equines.
Wolaita has a unique farming pattern that creates a beautiful landscape and vegetation
cover. The residential areas are often situated along the concave slope of short V-shaped val-
leys that dominate the landscape. A communal area used for grazing, public activities and
roads often divide a village. In lower altitude areas, the road along the concave slope and with-

120
.tesfaye.thesis 29-01-2003 15:12 Pagina 121

in the residential area is fairly suitable for motorised transportation. Farms generally start
from the back of the house and extend to the bottom of the valley, ending in a woodlot. Found
close to the house are a mixture of spices and a cabbage plot that covers a narrow strip of land.
Adjacent to this is a coffee plot. Next is a block of enset, followed by root crops such as yam or
taro. Maize, the major staple food crop among the cereals, is next to the root crops. The bound-
ary of maize marks the low end of the fertile part of the farm called Darkua, and is followed
by the Shoka zone. Ranges of crops are grown in this part of the farm. Below maize, sweet pota-
toes or potatoes are planted. Small cereals such as wheat, barley and teff are planted in the
lower part of the Shoka zone. Pulses are planted either next to the maize, or after the tuber
crops. Beyond the small cereals, grass is planted nowadays, along with some trees. Crop rota-
tion takes into account this general cropping pattern in the two farming zones. Finally, the
woodlot intercepts the river or the lowest point along the valley floor (see Plate 6.1).
While the above is the general pattern, farmers often make modifications depending on
their specific needs. These include: family food and cash needs, farm size and labour avail-
ability. Apart from the crops mentioned, some farmers grow banana and sugarcane in strips
or in a given spot. In the same manner, they grow fruit such as avocado, papaya, oranges,
guava and mango in different patterns. A few farmers also grow chat (Catha edulis) by inter-
cropping it with other crops. Vegetable production is rare in the area. However, a few farmers
produce rainfed potato, cabbage and onion.

Plate 6.1: A typical land use pattern in Wolaita

6.3 History of external interventions

Wolaita district is one of the few areas in Ethiopia where externally financed projects have
been deployed since the 1960s. The most notable ones were the Wolaita Agricultural
Development Unit (WADU) and its successor, the Soil Conservation Research Project (SCRP). In
addition, the Wolaita Rural Education Project (WREP) that was sponsored by a French aid
organisation and later by Agri-service Ethiopia, was a well-known external intervention before
the mushrooming of NGOs following the 1984 famine. While rural development was the con-

121
.tesfaye.thesis 29-01-2003 15:12 Pagina 122

cern of all external projects, WADU and SCRP had a major focus on soil and water conserva-
tion, though the latter one was limited to research only. The following two sections give an
overview of WADU and SCRP. However, the latter is given greater attention because of the tech-
nical data on runoff and soil loss it collected from 1981 until the early 1990s.

6.3.1 The Wolaita Agricultural Development Unit (1970-1982)

In line with the dominant rural development paradigms of the 1960s, the government of
Ethiopia initiated a project called the Wolaita Agricultural Development Unit (WADU) that
catered for the present Wolaita zone. The project was funded by the World Bank. It was oper-
ational from 1970 to 1982. The WADU was designed as an integrated rural development pro-
ject with multi-faceted project activities, which included the provision of fertilisers and
improved seeds, crop protection, soil conservation, livestock improvement, agricultural cred-
its, organisation of co-operatives, assistance to rural artisans, store and road construction, and
water development (WADU, 1981).
The project implementation began in the lowland part of the Wolaita to promote a settle-
ment programme whereby people from the northern and central highlands were settled. In
the highlands, the project focused on extension services combined with the aforementioned
components.
The WADU’s soil and water conservation practices included construction of bunds, gully
establishment, spring protection, water harvesting, and afforestation. To that end, it intro-
duced soil bunds in Gunno in 1974-75 because of severe soil erosion in the area. At the end of
the project in 1980/81, a total of 31,511 ha of land had been covered by soil bunds (WADU,
1981). Before WADU, soil bunds were not in the repertoire of soil conservation practices of
highland farmers in Wolaita. Cash was used to construct the bunds, whereas food-for-work
was used to maintain them. As farmers were short of money due to the poor marketing infra-
structure, attracting labour for bund construction was not difficult. Some farmers destroyed
the bunds shortly after their construction because of the inconvenience they presented to the
ox-plough. The state agents sent those farmers to prison to ensure that they refrained from
similar acts in the future. Farmers who told me about this situation estimated that about 20
farmers had been involved.
In order to lay the foundation for comprehensive data collection that would serve the
entire country, WADU initiated a hydrometric site at Gunno in 1977. Recording began in
January, 1981.
After WADU was phased out, all project activities were taken over by the Ministry of
Agriculture, with the exception of the research component which was taken over by the SCRP,
which began its operation in Ethiopia in 1981 (see below).

6.3.2 Soil Conservation Research Project (1982-1995)

Inception and termination

The Soil Conservation Research Project (SCRP) was initiated in 1981. This project was launched
upon the request of the Ethiopian government to guide the massive soil and water conserva-
tion programme supported by the World Food Programme (WFP) since the 1973 famine. Soil
and water conservation activities in Ethiopia comprised the largest food-for-work project in
Africa supported by the WFP. The Swiss and Ethiopian governments covered the financing of
the project.

122
.tesfaye.thesis 29-01-2003 15:12 Pagina 123

The SCRP established six research sites in Ethiopia. The headquarters of the project was
located in Addis Ababa, where a laboratory was based and part of the database management
was carried out. The major data processing took place in Switzerland (Karl Herweg, p.c. 2002).
The SCRP sites are located in different agro-ecological zones of the country in order to gen-
erate research results that would be relevant for soil and water conservation activities in the
country. There are marked differences among the zones in terms of human and livestock pop-
ulation density, access to markets and diversity of agriculture.
Each research site has operated in a watershed of approximately 10 km2 with a limited
catchment area for treatment with soil and water conservation practices (1-2 km2). The
research activities were more or less similar in the research units, as we shall see in chapter
7.
The SCRP fell under the auspices of the Ministry of Agriculture until the establishment of
the Ministry of Natural Resource Development and Environmental Protection in the early
1990s. The association of SCRP with the departments of these ministries was nominal rather
than functional in nature. As a result, the SCRP managed to create its own autonomous space
of operation without establishing modalities and mechanisms to work with ‘host’ institutions
and other related organisations with a mandate for research and/or extension in the country.
For its first ten years, the project did not have a mechanism for internal and external research
reviews and planning processes (Waters-Bayer, et al., 1998). The lack of clear institutional
arrangements contributed to a growing tension between the project and the host institution.
More than a decade and a half’s worth of technical research failed to save the project from the
scrutiny of the new government in 1995, leading to its closure.
In 1996, the SCRP research sites were decentralised and put under the management of the
Regional Governments in which they were located. Finally, the project was closed in June 1998
(Waters-Bayer, et al., 1998).

The Gunno research site

Gunno is located in the former North Omo zone (now Wolaita zone). It is situated on 37 0 39’
E and 6 0 51’ N, 15 km NE of Sodo town. Its elevation extends from 1,800-2,100 masl.
Unlike the other research sites, an exceptional design was employed at Gunno, where a
twin catchment was used to compare the effect of conservation measures, with one catch-
ment under treatment, while the adjacent ‘untreated’ catchment, receiving no bunds was
used as a control.
The treated catchment is called Zerwa, (68 ha) whereas the untreated one is Goppo (89 ha).
Treatment entailed the introduction of soil bunds (SCRP 1996a). According to a census taken
in 1983, there were 92 and 89 households in the treated and untreated catchments respec-
tively. The total populations in the respective catchments were, 413 and 408 people (Gunten,
1993).

Description of technical intervention

The SCRP commenced its work at Gunno with a visit of experts to the area who informed the
community about their intention to construct certain structures, which they said were useful
for soil and water conservation. Leaders of the Peasant Association communicated the plan of
the project to the community. Ten farmers apparently had to give up land because of the
establishment of the research site at Gunno, but did not receive any compensation for this
loss.
The SCRP forced farmers in Goppo, the untreated catchment, to destroy the bunds and
diversion ditches constructed by the WADU. Those who protested against that action were

123
.tesfaye.thesis 29-01-2003 15:12 Pagina 124

threatened and actually brought to court. The SCRP was assisted in this action by a local devel-
opment agent. During the group interview, farmers were able to mention the name of the
development agent involved, who still works for the ministry. The agent confirmed and regret-
ted this action of 20 years ago, when I interviewed him.
Farmers from both the treated and untreated catchments participated in the construction
of soil bunds 32 in the treated catchment. The incentive used by the SCRP was food-for-work.
The food items were cereals and edible oil. In addition, farmers in the catchment were given
a hand tool (akaffa). The labour was organised into groups of 20 people per working team.
Technicians were marking the spots for bund construction, which were based on technical
parameters. Maintenance was carried out by using food-for-work as the incentive. This main-
tenance was carried out during the first two years after construction of the soil bunds.
Thereafter, no food-for-work nor maintenance took place, apart from a few individuals who
maintained some bunds on their own initiative.
General awareness-raising conservation education was organised for the farmers from both
catchments by SCRP technicians. The WADU had carried out similar training in its farmers’
training component. Owing to the original plan of the hydrometric station, farmers from the
untreated catchment were not allowed to put the knowledge and skill of bund construction
into practice because that would jeopardise the interest of the scientific research. Quoting the
farmers from the untreated catchment - ‘we were told to learn for the future’ - a future that
they did not know at that time. In this connection, they explained the frustration which
resulted from their exclusion from the treatment by soil bunds while the SCRP technicians
were continuously collecting information from their plots every season in the same way they
did for the treated catchment. During my discussions, local people did not hide their stereo-
typing of educated people, whom they think are highly engaged in ‘writing, but doing noth-
ing for the rural people’.

The conservation education given to farmers included:

• erosion monitoring by looking at indications in the field (mainly rills);

• contour ploughing;
• planting grass on soil bunds;
• use of agroforestry;
• using branches and banana stems for gully stabilisation; and
• maintenance of soil bunds (moving the soil deposited alongside the bund onto the bund to
raise its height).

These educational activities were, however, presented on a piecemeal basis rather than incor-
porating them into the farming system through an informed or organised deliberative par-
ticipatory learning process. Besides this, farmers already knew some of these practices such
as contour ploughing, multiple cropping, ‘agroforestry’, and the use of banana stems for gully
stabilisation.

Runoff and soil-loss results

The SCRP collected a wide range of data to help understand the land management dynamics
in the agro-ecological zone. Runoff and soil loss were measured at the plot level, whereas river

32 Bunds occupy a considerable amount of land that vary according to slope. On a 8% slope, 2-5% of land; on 8-16% slope, 8-
12% land; on a 16-30% slope, 15-20% land and on a higher than 30% slope, 25% of land (FAO/MoA, 1986, cited in Belay,
1992:52).

124
.tesfaye.thesis 29-01-2003 15:12 Pagina 125

discharge and sediment yield were measured at the catchment level. The test plots and catch-
ment-level measurements distinguished between the treated and untreated catchment. The
plot-level measurements showed on-site effects, whereas the catchment-level measurements
showed off-site effects (see Plate 6.2).
Runoff and soil loss from individual plots were assessed according to different land use, soil
type, slope length and gradient using plots of 30 m2 (2x15m) and 3 m2 (1x3m) (see Appendix 1
for details). Farmers who owned the land where the test plots were installed co-operated with
field staff in removing and replacing the structure during the seasonal farm operations.
Farmers themselves selected the crops grown in the test plots. Yield samples of each crop on
test plots and other plots were taken from farmers’ fields in both catchments. Yield data from
non-test plots were taken from fixed and non-fixed plots.

Plate 6.2: Hydrometeric station at Gunno, Wolaita

Outlet from the untreated

catchment

Outlet from the treated

catchment

Four test plots were treated from 1982 to 1992 at Gunno. Test plot 1 was continuously laboured
to expose the soil in order to determine soil erodibility. Test plots 2 and 4 were cultivated plots
with different crops. Test plot 3 was maintained under a grass cover. The former three plots
were on a slope of 16%, while the last plot was on a relatively steep slope (42%). Two micro-
plots (MP 5 and 6) with a 16% slope were staged next to test plot 2. Test plot 2 and micro-plot
5 were monitored from 1983 to 1991, whereas micro-plot 6 was continuously hacked and
therefore represented bare surface conditions. Data from micro-plots were collected from
1983-1991 (SCRP, 1996b; SCRP, 2000a).
According to the SCRP (1996b), soil loss at an on-site level is highest in August. Erosion at
the site level is also considered as high, with a very low runoff (Table 6.1). High refers to soil
loss of 50-100 t ha-1 whereas runoff was considered to be low, at 200-300 mm per year.

125
.tesfaye.thesis 29-01-2003 15:12 Pagina 126

Table 6.1: Summary of experimental results at Gunno, Wolaita, Ethiopia

Particulars On-site effects (plots) Off-site effects (catchment)

Plot 2 Plot 3 Plot 4 Plot 5 Plot 6 Treated Untreated

Runoff (mm/y) 155 15 171 321 347 149 246

Erosion rate (t ha-1 y-1) 80 .02 48 49 110 0.1 0.7

Source: SCRP Database 1982-1993 (SCRP, 1996b).

In the case of the off-site results, sediment yields of both the untreated and treated catch-
ments were considered very low (<1 t ha-1). The river discharges from the untreated and treat-
ed catchments were moderate and low, respectively (SCRP, 1996b). A river discharge is consid-
ered to be moderate when it yields up to 200-300 mm, and low when this figures between 100-
200 mm.
In addition to experimenting with plant covers, test plots were also used to evaluate the
effect of soil loss and runoff under different conservation techniques. These techniques were
grass strips, graded Fanya Juu 33, graded bund, level Fanya Juu, and level bund. In this case,
runoff and soil loss under these techniques were compared with a control plot from 1987 to
1991 whilst under different crops.
Soil and water conservation tests at Gunno were found to be satisfactory in terms of ero-
sion reduction. The reduction of runoff was, however, associated with a potential for water-
logging above the bunds. Production under different soil and water conservation practices
tested at the station showed a slight increase. The impact on the catchment hydrology was,
however, found to be minimal (SCRP, 1996b).
When I started my study to assess the impact of SCRP at Gunno in 1998, the project had
ended officially with very little follow-up by the regional government with the exception of
the climatic data and hydrometric readings.

6.4 Methodology of my study

6.4.1 Unit of analysis

I used the design of the SCRP project to identify the watersheds to match the objectives of my
study. Thus, farmers were selected from four locations (watersheds). These were farmers in the
(1) treated and (2) untreated catchments, and in (3) adjacent and (4) distant catchments. The
latter two catchments were defined by the researcher, while the former two were used in the
SCRP’s experimental design. Farmers from the treated and untreated catchments were select-
ed to measure the direct positive and negative project impacts, whereas the adjacent and dis-
tant farmers were considered to assess the spatial impacts of the project. ‘Adjacent’ refers to
farmers who are farming just next to the two experimental watersheds. The distant location
(at 8-15 km away from the SCRP sites) falls in the Waja river catchment that covers most of the
southwestern side of Damota Mountain with an altitude of 2,908 masl. Because of its steep
slopes, it was believed to serve as a critical test for the conservation measures promoted in

33 Fanya Juu is a Swahili term, which indicates throwing the soil up the slope from below to create a contour bund.

126
.tesfaye.thesis 29-01-2003 15:12 Pagina 127

Wolaita. Thus, the main reason to include the distant location is the fact that it is erosion-
prone, something that is expected to increase the acceptance of soil bunds.
The SCRP sites and the adjacent catchments generally range from 1,800 to 2,100 m. The dis-
tant location in the Waja river catchment approximately ranges from 1941 m at the bottom to
2,908 m at the peak of Damota Mountain, situated north of Sodo town.
The watersheds covered by the study are located in three Weredas (i.e., districts). These are:
Kindo Koisha, Sodo Zuria and Boloso Sore (Table 6.2). Farmers selected for the formal interview
were drawn from four Peasant Associations: Dogoshakisho, Damot Waja, Doge Mashedo and
Gunno.
Most of the farmers who participated in the survey were men (90%). Women were targeted
during the qualitative phase of the study. Households within respective watersheds were inter-
viewed and observed. Watershed and community-level observations were drawn from group
interviews within and across watersheds. At the household level, the study dealt with both the
total farm as well as plots to gain a better understanding of farmer decision-making and con-
straints. Therefore, survey results are reported by locations (watersheds) and/or plots as appro-
priate.

6.4.2 Sampling

A survey was conducted after one and a half years of qualitative fieldwork. A sampling frame
for the survey was prepared of farmers who live and farm in the four locations mentioned ear-
lier. A farmers’ list was prepared by enumerators from the locality with the assistance of exec-
utive committee members of the respective Peasant Associations and development agents who
are responsible for the area. Preparation of these lists took 3-10 days, depending on the size of
the watersheds. The total number of households listed in the sampling frame in each watershed
were as follows: 103 treated, 120 untreated, 195 adjacent and 680 distant.
Household heads were selected using two-stage cluster sampling. At the first stage, the four
watersheds were purposively selected. At the second stage households within these watersheds
were randomly selected. Finally, a total of 231 farmers 34 were selected for an interview (Table
6.2). These farmers represent a fraction of farmers in Wolaita. However, the areas selected for
the sample are representative in terms of farming practices and are typical of soil and water
conservation efforts carried out in Wolaita.

6.4.3 Methods of data collection

Both primary and secondary data sources were used. The primary data were collected using
both qualitative and survey research methods. The qualitative research featured group and
individual interviews, focus group interviews and observations. A semi-structured interviewing
schedule was used in all the interviews. The qualitative aspect of the research featured narra-
tives around the research issues on soil and water conservation using key questions to spark the
discussion. All other relevant issues were built up on the previous questions until the point of
saturation on the issues was reached. The entire qualitative data set was collected by the
researcher using interpreters. A structured survey instrument was used to collect the survey
data. Enumerators who speak the local language were used at this stage of the research. They
were given a one-week intensive training course including two days of pre-testing.

34 The sample gives a precision of ±7%, with a 95% confidence level based on a 55% coefficient of variation for land size
from the same sample.

127
.tesfaye.thesis 29-01-2003 15:12 Pagina 128

Table 6.2: Number of sampled farmers per Wereda, village and location in relation to Soil Conservation
Research Project site

Location in Relation to Soil Conservation Research Centre

Name of Weredas Name of Treated Untreated Adjacent to Distant from Total

Villages the centre the Centre

Kindo Koisha Shega 25 24 49

Feteta 40 1 7 48
Maheldoge 6 6
Centra 3 3
Huletegna 15 5 20
Shakisho
Mashedo 6 6
Subtotal 65 46 21 132

Bolo Sore Offa 7 7

Subtotal 7 7

Sodo Zuria Ketena 1 21 21

Ketena 5 23 23
Ketena 3 18 18
Ketena 7 13 13
Ketena 4 17 17
Subtotal 92 92

Total 65 46 28 92 231

6.4.4 Methods of data analysis

The qualitative data were partially analysed during the process of data collection to immedi-
ately be able to identify gaps to be filled through subsequent data collection. The final analy-
sis was carried out through pattern-matching to suit the research objectives. The patterns
were obtained from the recorded field notes. This process involves the selection of a key word
that fits the main research questions and the elaboration of the information gathered on that
theme. Elaboration was facilitated by the in-built grounding process of the data collection
phase, which attempts to answer all relevant questions until the point of saturation.
The quantitative data were first summarised on a data summary sheet prepared for this
study and subsequently entered into a computer. Both activities were carried out by assistants
under the supervision of the researcher. Before the final analysis, the researcher carried out
data screening and cleaning whereby he discovered a serious data entry mistake. As a result,
he checked all 231 cases involving 900 primary variables. Finally, the data were analysed with
appropriate statistical procedures based on the level of measurement of the variables
involved, i.e., the nominal, ordinal and interval levels. The Statistical Package for Social
Sciences (SPSS version 10.0.7) was used. The statistical analyses include: descriptive statistics;

128
.tesfaye.thesis 29-01-2003 15:12 Pagina 129

frequencies, averages, cross-tabulation; association; general linear model; univariate, post-hoc

multiple comparison and non-parametric statistics (chi-square test and Kruskal-Wallis analy-
sis of variance).

6.5 Results and discussion

6.5.1 Socio-economic situation of farmers in Wolaita

Household characteristics
Households in the study area consist of both nuclear and extended families. Therefore, fami-
ly members included parents, grandchildren, cousins, nephews and nieces. The majority of
the households is, however, of the nuclear type. Even so, there is multiple dependence on the
same farm plot, a situation that can be termed ‘farmholds’ rather than households.
Table 6.3 presents key household characteristics related to farmers’ land-management deci-
sions. The subsequent discussion in this section is partly supported by the information in this
table. The table presents the status of the entire sample (231 farmers) and the four locations
with respect to these variables. Test of difference among the four locations is given by chi-
square or F-ratio, as appropriate.
Of the surveyed farmers, 88% were married. Of the 198 married farmers, 87% were married
to one wife. The remaining 12% and 1% of the farmers were married with two and three wives
respectively, which indicate that polygamy is not very common, unlike in the past. The fami-
ly size ranges from 1 - 16 persons, with an average size of 6.5 persons. Family size does not
show a significant difference between locations (Table 6.3).
Due to the small farm size per capita (average 0.62 ha), it is common to find two genera-
tions of subsistence farmers in the same household. An age range of 60 years is an indication
of this situation. The minimum age of the surveyed farmers is 20, whereas the mean age is 43
years. Age too does not show a significant difference between locations.
Economically dependent age groups (0-14) and elderly (65 and above) vary from nil to 12 per-
sons per family, with a mode of four persons. Ninety-six percent of the cases has seven depen-
dents. The economically active age group ranges from 1 to 10 persons with a mean of 3.2. The
mode of the economically active age group is two persons, mostly a wife and a husband. The
sample does not show a significant difference between the locations with respect to the eco-
nomically active age groups. Amare (1988) found economically active age groups of 2.6 and 2.5
for treated and untreated catchments respectively. The dependency ratio at the household
level varies from 0.2 for 2% of the farmers, to 8 for 1.4% of the farmers. The modal dependen-
cy ratio is 1, covering 21% of the households in the survey. In 48% of the households, the
dependent age group exceeds the economically active age group, whereas 31% of the house-
holds had a more economically active age group as compared to dependent age groups. With
respect to a resource base, cropland distribution of the families with the modal dependent
population (4 persons per family) ranges from 0.13 ha to 1 ha.
Even though the study area is only 200-250 km from major urban centres, migration is not
common. Out of 231 cases, 30% and 16% of the surveyed farmers has reported temporary and
permanent migration, respectively. Most of the temporary migrants were male, whereas
females exceed males among the permanent migrants. The number of households that obtain
remittance income from migrants was negligible. There is no significant difference between
the locations with respect to migration.

129
.tesfaye.thesis 29-01-2003 15:12 Pagina 130

Table 6.3: Household characteristics and access to resources in Wolaita, Ethiopia

Variables Total Locations

Sample
Treated Untreated Adjacent Distant Tests (χ2/F) Significance

Family size (mean) 6.5 7 6 7 7 0.376 ns

Age of household head 43.4 42 44 41 45 0.889 ns
Dependent family members 3.4 3.3 3.4 3.7 3.5 0.244 ns
Economically active labour 3.2 3.4 2.8 3.0 3.2 1.130 ns
Migration
Temporary (%) 30 20 15 8 27 0.202 ns
Permanent (%) 16 16 4 2 14 7.471 ns
Education of Household Head 5.001 ns
Illiterate (%) 147 29 18 11 42
Basic Education (%) 18 22 28 22 28
Grade 1-8 (%) 53 30 21 13 36
Grade 9-12 (%) 12 17 25 8 50
Crop land (mean in ha) 0.45 0.49 0.51 0.44 0.40 2.686 *
Grass land (mean ha) 0.09 0.13 0.09 0.08 0.07 10.107 ***
Forested land (mean ha) 0.09 0.13 0.10 0.09 0.07 9.021 ***
Total land (mean in ha) 35 0.62 0.71 0.69 0.62 0.51 5.359 ***
Livestock owned (number) 36 3.2 3.5 3.0 3.8 3.0 6.864 ns
Use of fertilisers 1998 46.1 52.6 47.8 44.6 37.9 0.928 ns
(mean kg)
Use of fertilisers 1999 44.3 47.6 45.9 46.1 38.8 0.556 ns
(mean kg)
Visits by extension agents 11.282 *
(%) Yes 27 38 26 41 17
(%) No 73 62 74 59 83
Access to credit 10.776 *
(%) Yes 72 60 88 81 71
(%) No 28 40 12 19 29

Note: ns, * and *** shows: not significant, significant at p ≤ 0.05, and at p ≤ 0.001 for χ2 or F test. Sample size: treated 65,
untreated 46, adjacent 28 and distant 92.

Most farmers (64%) are illiterate (n=230). Some of the farmers who participated in the literacy
programme in the 1980s (8%) are now unable to read or write. The educational attainment of
28% of the farmers ranges from grade 2 to grade 12. In terms of grade distribution, 18% fall in
the range of grades 2 to 6, 5% had achieved grades 7 and 8, and 5% were between grades 9 and
12. The overall chi-square test for education does not show a significant difference between
the locations. Only 5% of the total farmers interviewed in this area reported ownership of
radio, which indicates poor access to cheap and relevant source of information.
The majority (99.6%) of the respondents in Wolaita are Christians of different churches,

35 The total land figure does not include the communal land to which all farmers’ households have access, such as village
and distant grazing areas.
36 Excluding chickens.

130
.tesfaye.thesis 29-01-2003 15:12 Pagina 131

composed of Orthodox (62%), Protestants (37%) and Catholics (0.9%). Before the introduction
of Christianity a century and a half ago, the Wolaita people were pagans. Their object of belief
was large trees (e.g., ficus). Some people believed in witchcraft (Wukabi). During military rule,
paganism and witchcraft were abolished, and thereafter people began to neglect tending the
big indigenous trees and shifted to eucalyptus, which was introduced in the last quarter of the
19th century. Ancestral cemeteries were covered by indigenous tree species. After people con-
verted to the Orthodox Christian religion, the ancestral cemeteries were no longer used.
Nowadays, only Protestants bury their dead in the private burial places. Followers of the
Orthodox religion celebrate four days of every month, apart from Sundays, and one week of
Meskel holiday in October. Protestants on the other hand celebrate Good Friday, Christmas and
on Sundays. Farmers do not report any difficulties because of religious holidays. There are also
no days where farm work cannot take place due to superstitious beliefs.
Rural people in Wolaita mainly depend on agriculture for their livelihood. Farming is the
only means of living for 84% of the 231 farmers surveyed. The major off-farm employment in
the area is trading. Only very few farmers earn a meagre income from pottery, wood carving,
basketry, or wage labour. School dropouts and returnees from military service go back to their
parents’ land. Individuals with no access to land can hardly find other means of livelihood
due to lack of know-how to create self-employment, lack of finances and poor rural infra-
structure.

Land ownership and land fragmentation

Due to high population pressure, landholding per capita is small. The average landholding of
the surveyed farmers is 0.62 ha, with a mode of 0.5 ha (Table 6.3). Farmers who own land in
excess of 1 ha number 11%. Total land size, grassland and forested land show a highly signifi-
cant difference across location (p ≤ 0.001), while cropland was significant at p ≤ 0.05. The coef-
ficient of variation of the total and croplands were 55% and 57% respectively. Those farmers
who have a greater land size can practise more mixed farming than those who have little or
no land for hay production or grazing.
Inheritance from parents was the main source of land for most of the surveyed farmers. A
small proportion of farmers got small plots from the Peasant Association both during and
after the Derg Regime.
Land ownership in terms of land size has been fairly stable over the last 25 years.
Approximately 53% of the farmers did not see any change in their holdings, while 37% have
reported a decrease in farm size during this period, and 6% managed to increase their hold-
ing through purchase and inheritance. The remaining 4% started their own farm in the last
five years.
Important reasons for the decline in landholding size are soil erosion (38%), sharing the
land with children and brothers (30%), sale/lease (12%) and redistribution by the state (8%). A
few farmers (12%) faced a combination of situations that led to a decrease in their holdings.
For instance, three farmers reported that their land was taken by the SCRP without compen-
sation. In addition, these farmers added that they shared their land with their children and
brothers, sold/leased and lost some portion due to soil erosion.
Farmers in the area are very well aware of the government’s claim on land. Of the total
farmers, 33% of them replied that the land belongs to them, while the rest stated that the land
belongs to the government. Even then, only 4% of them see this claim as a problem to their
land-management practices. This issue has also been pursued through small group and key
informant interviews, which confirmed the same trend.

131
.tesfaye.thesis 29-01-2003 15:12 Pagina 132

In Wolaita, as in many societies in Ethiopia, land is inherited by the male child. In the next
ten years alone, 54% of 231 farmers are expecting to share land with their children. In spite of
extremely small farm sizes, one to five male children are in line to inherit land from their par-
ents in the next ten years.
Due to increased land shortages, farmers shift grass or wood plots to crop production.
Small landholdings are one of the reasons for the destruction of soil bunds in the area.

Slope gradient
Farmers’ land management practices and production problems are better understood if anal-
ysis is carried out at both the plot and farm level and at the watershed level for community-
level decisions. Farmers in Wolaita divide their farms into small plots for operational reasons
even though their landholdings mostly fall in one location. Because of this, farm distance is
not a very important factor except in terms of transportation of manure. Following farmers’
practices, farm plots were numbered starting from the plots next the house to the last one
located before the grass or wooded plots. This study found a maximum of twelve such plots,
with an average of plots being 5.9. Only the first six plots will be presented here. The other
plots were omitted because of limited response. Note that the percentages of each plot do not
add up to 100 because of multiple responses. This may also hold true for some additional per-
centages provided in the subsequent discussions, when the interest is on the specific response
other than the sample distribution.
The slope of the plots greatly influences the rate of erosion by water (cf. Universal Soil Loss
Equation). The slope is one of the observable factors in the erosion process that prompts farm-
ers to monitor erosion. In view of this, farmers were asked to identify their plots according to
four descriptive gradient classes: namely, steep, moderately steep, moderately flat and flat. In
addition, the farmers were asked if plots were under the ridge, a situation, which was inher-
ently affected by erosion (see Table 6.4).
Apart from the first three plots, the percentage of respondents with moderately flat to flat
slopes was found to be greater than that of moderately steep to steep slopes in all other plots.
The percentages of moderately steep to steep were greater than those who rated their plots as
moderately flat to flat. The difference in percentage of farmers reporting moderately steep to
steep slopes increased when moving towards the plots located down the valley, the Shoka field.
Most of the farmers with steep farm plots are located in the distant location (Waja catchment),
whereas most of those who consider their farm as flat are located in and around the SCRP
catchments.

Use of external inputs

Nowadays, commonly used external inputs in the farming system are improved seeds and fer-
tilisers. The major sources of seeds in the study area are the local market, from the farms
themselves and from government organisations, in order of importance. Supply of improved
maize seed from government sources is fairly adequate as compared to other small-grain cere-
als.

132
.tesfaye.thesis 29-01-2003 15:12 Pagina 133

Table 6.4: Slope gradient of farm plots, Wolaita, Ethiopia

Gradient Classes Plot 1 Plot 2 Plot 3 Plot 4 Plot 5 Plot 6

n=230% n=227% n=215% n=201% n=172% n=126%

Steep 20 20 24 28 30 34
Moderately steep 15 17 18 25 31 33
Moderately flat 24 27 23 19 19 16
Flat 40 36 32 24 18 14
Land under steep ridges - - 3 3 2 2

Farmers in the study area use both urea and DAP 37 fertiliser. Most farmers (68%) have used fer-
tilisers on their farm at one time or another. In last two years (1998 and 1999), 62% of the farm-
ers had used fertilisers in the cropping seasons, whereas the remaining 38% did not use any
fertiliser.
Looking at the rate of usage of fertilisers, the percentage of farmers using DAP was greater
than that of using urea. The percentages of farmers using urea in 1998 and 1999 was 38% and
41% respectively, whereas, the users of DAP over the same period numbered 55% and 60%
respectively. Mean fertiliser use for 1998 (46.1 kg) was slightly greater than that for 1999 (44.3
kg). An analysis of the amount of fertiliser used in 1998 and 1999 shows that mean fertiliser
usage by farmers from the treated catchment was the highest in both years, whereas corre-
sponding figures for the distant farmers were the least over the same period. There was no sig-
nificant difference between locations in terms of fertiliser usage during those two years (Table
6.3).

Labour organisation
Major sources of labour in the study area are family members, labour parties, and relatives
and friends. Labour is mostly organised according to its importance for: cultivation or hoeing
(88%), harvesting (77%), house construction (64%), and sowing (62%). Land preparation and
maintenance and construction of bunds were mentioned by only a small percentage of
respondents (8% each).
The organisation of a labour party is rooted in the tradition of the Wolaita people (Data,
1998). Labour parties are organised for house repairing, wood splitting, grass cutting and rou-
tine farm operations. Different labour groups are organised according to labour demand of
the work and household resource status. These are categorised as: Dagua, (small 10-20, large
120-130 people), Yurpia (10-20 people) and Zayia (5-10 people). The latter two forms of labour
groups are mostly reciprocal unlike the first one. In addition to these three forms of labour
organisation, the community has another labour support system called Wossa. This does not
involve reciprocal labour because those who qualify for Wossa are often handicapped, the
elderly or widows.
Apart from actual labour participation, community members help each other with a con-
tribution of grain for the preparation of food and drinks, collecting firewood, fetching water,
and the preparation and serving of food and drinks. Women fulfil these roles, while men par-
ticipate in the field operations.

37 Di-Ammonium Phosphate.

133
.tesfaye.thesis 29-01-2003 15:12 Pagina 134

Nowadays, the organisation of a large labour party is uncommon, partly because of the
small land size and the extensive use of the ox-plough that has eased part of the labour pro-
cess.
A new source of labour other than the traditional labour organisation is found in the con-
gregation of church members. Whenever a church member faces a labour shortage, members
set a date and provide labour for him/her.

Trends in crop production

As indicated earlier, Wolaita is suitable to a range of crops. In spite of its ambient environment
for agriculture, many farmers are not able to produce sufficient food for their families. The
average yields of barley, haricot bean, maize and teff at Gunno research station from 1986 to
1993 were: 0.6, 1.2, 2.5 and 0.8 t ha -1 (SCRP, 1996a). Looking at the trend of crop production over
the past decade, the majority of surveyed farmers (82%) experienced a decrease in total crop
production. An increase in crop production was reported by only 12% of the farmers, while
the remaining (6%) did not see any change in their production.
The increase of production was achieved by using commercial fertiliser with local or
improved seeds, the use of manure and household refuse, share-cropping, leasing more land,
and shifting grass/woodland to cropland. Some farmers use a combination of strategies to
achieve their objectives.

Table 6.5: Farmers’ prioritisation of problems in crop production

Crop Production Problems Ranks

1 Decline in soil fertility 2

2 Pests 8
3 Disease 6
4 Drought* 3
5 Shortage of rainfall* 1
6 Soil erosion 4
7 Shortage of land 5
8 Lack of fertiliser 7
9 Shortage of manpower 9

* According to farmers, drought refers to a prolonged absence of rainfall, whereas shortage of rain refers to its erratic nature
whereby crops are affected due to little rainfall or to its interruption after its onset.

Production gaps are often filled by market supply. Most families (93%) purchased food in 1999
and 2000. In 2000, 95% of the surveyed farmers purchased food grains. The average period over
which food grains are purchased was six months. The few months before the next harvest are
critical to families with many members. Food shortage in Wolaita is very common even out-
side the years of widespread famine in the country. For instance, Amare (1988) stated that 78%
and 76% of farmers from treated and untreated catchments respectively reported a food short-
age during the year. The periodic vulnerability of Wolaita to food shortages is widely docu-
mented and aggravated by erratic rainfall, among other factors (Dagnew, 1995). The farmers
identified different problems related to crop production (Table 6.5). The problems that were
identified are closely interrelated. Priority problems lean towards biophysical constraints. The
importance of soil erosion was reflected in the relation to a decline in soil fertility, discussed

134
.tesfaye.thesis 29-01-2003 15:12 Pagina 135

in later sections. Socio-economic problems such as a shortage of manpower occur when some
farm operations in the Belg season and that of the Kremt season overlap. Due to indigenous
institutions that give access to labour and oxen through social arrangements and rents, farm-
ers did not emphasise a shortage of these resources. Their focus leaned towards factors that
are outside the control of their institutions.

Livestock production system

Livestock rearing is a typical feature of the farming system in Wolaita. There was no signifi-
cant difference between locations with respect to the ownership of livestock. During the sur-
vey period, 93% of the 231 households in Wolaita had some kind of livestock. Cattle are the
dominant livestock species integrated with the cropping systems of Wolaita. Farmers who
owned one and two heads of oxen are 34% and 5% respectively. Draft animals such as donkeys
and horses are rare compared to cattle. In terms of location, farmers in the distant location
who live in the mountainous area own more draft animals. Sheep are more common than
goats.
Shared livestock rearing 38 is very common in the area. Fifty-seven percent of the farmers
have been involved in shared livestock rearing during the last two years. Shared livestock, like
any other partnership, requires common interest, mutual benefits, social ties and trust.
Shared animals are cows, heifers, sheep and goats. Rearing of these animals gives the recipi-
ent farmers the opportunity to get offspring, milk, butter and manure. Oxen and bulls are
shared for fattening purposes in addition to traditional services like manure and draft power.
Livestock feed comes from individual grass plots, communal grazing, crop residues, thin-
ning products and weeds. In the residential areas and on individual grass plots, animals are
mostly tethered. Free grazing is applied in bigger communal grazing lands. Due to strict ani-
mal-control systems and live fences around farming plots, the livestock do not interfere with
trees and grass on soil bunds.
Animal production has declined in the past ten years; 84% of the farmers have experienced
a declining number in their stock, while only 8% managed to increase their stock. These are
young farmers who established their families in the past ten years. Few farmers (8%) main-
tained their stock through breeding or purchasing.
Disease was the major problem that reduced livestock in the area. It was reported by most
of the farmers (72%). This situation is said to be improving. The second most important bot-
tleneck to livestock production is a shortage of feed (69%). This is aggravated by the shortage
of land, droughts and erratic rainfall that affect the production of biomass from the farms
and the grazing areas. Other problems that account for a decline in livestock include a low
price per head of animals, physical injury to the animals and a lack of cash.
Farmers sell animals for various reasons: purchase of food, house construction, funerals,
improved stock composition, wedding ceremony, shortage of feed, disease outbreak, purchase
of cloth, profit, loan repayment, health expense and other miscellaneous social reasons (e.g.,
shared rearing, court expense). Among these reasons, ‘purchase of food’ and ‘house construc-
tion’ were the most important ones. Cattle, specially oxen, were the most frequently sold ani-
mals in the area.

Access to information and credit

Nowadays, the extension service of the regional state is the main source of information in
agriculture. The SCRP has served the community in the experimental watershed with soil and
water conservation practices during its active period. Unlike farmers in the lower altitude,

38 See footnote 22 in chapter 5.

135
.tesfaye.thesis 29-01-2003 15:12 Pagina 136

farmers in the upper catchment of the Damota Mountain that is designated as a ‘distant loca-
tion’ in this study did not get adequate extension services due to the landscape and accessi-
bility. The New Extension Intervention modeled after Sasakawa Global 2000 provides farmers
mainly with seeds and fertilisers, but also with improved animal breeds and soil and water
conservation methods. The major extension method for the participants of the new extension
propgramme is the individual method. Among farmers in this study, 27% was contacted by
the extension agent during the last two years. Of the total 231 farmers, 68% obtained credit
from different sources (Table 6.3). The major source of credit is the regional state, which pro-
vides agricultural credit to the rural community followed by micro-finance. Consumption
credits are obtained from relatives and friends.

6.5.2 Farmers’ knowledge, perceptions and practices in soil and water conservation

Farmers’ knowledge and perceptions on soil erosion and soil and water conservation
Our understanding of farmers’ knowledge and their perception of factors that influence their
land-management practices is of paramount importance for promoting sustainable land man-
agement. It is also interesting to know if and when farmers practise what they know and per-
ceive. As discussed in chapter 3, the relationships between knowledge, action and attitude are not
straightforward. The following sections of this Chapter deal with how farmers’ land-manage-
ment behaviour is shaped as a result of an interplay between socio-cultural, socio-economic,
institutional and technical factors.
During the qualitative phase of this study, efforts were made to understand farmers’ knowl-
edge and perception on soil erosion, soil fertility, incentives for soil and water conservation
and their view about the future generation. This was achieved with the help of a long list of
statements on these issues that were used in individual interviews. Following the interviews,
farm plots were visited with farmers who were involved in the interviews. Their opinions on
knowledge and perception were compared with their behaviour in land management. This
exercise provided a useful insight into why farmers do what they are doing and how they perceive
what they are doing. More specifically, we were interested in why they showed reluctance to
adopt/maintain the soil bunds that were introduced by two successive projects since the
1970s. In addition, there is a need to know how many farmers know, can and do carry out
some land-management practices. However, it was not known how many farmer were practis-
ing land-degrading farming in a population of a given watershed or otherwise. It was possible
that conservation-oriented farmers are few as compared to those who follow land-degrading
practices and vice versa. In view of this, selected statements were included in the survey
instrument to assess farmers’ responses within the population of the study area.
The statements were measured at three levels: (1) not clear, (2) disagree and (3) agree. This
short set of alternatives was considered in the interest of clarity to the farmers based on the
experience from the pre-test during the qualitative phase.
Table 6.6 shows results of farmers who agreed with the statements on knowledge. Cases
with ‘disagree’ range from 0 to 2.6 percent, while cases with ‘not clear’ range from 0.4 to 2.6.
Thus, responses with a small percentage were not included in the table 6.6.
The responses across the four locations for the most part do not show differences, except
for three statements that showed significant difference at p ≤ 0.001 (item 2 and 6) and at p ≤
0.01 (item 4). Responses of farmers from the treated catchment differ significantly from the
untreated and distant locations with respect to farmers’ knowledge on the consequences of
overgrazing and continuous cultivation of land to erosion. The responses of farmers from the

136
.tesfaye.thesis 29-01-2003 15:12 Pagina 137

treated catchment also significantly differ from all the three locations with respect to the con-
sequences of high population to land degradation. Interestingly, a few farmers from the treat-
ed catchment area who were exposed to soil and water conservation experimentation for near-
ly three decades, with intensive coverage during the last 15 years, show some reservation on
the knowledge statements treated in this study. The reason for this reservation could be cog-
nitive dissonance whereby farmers changed their actions likely due to a conflict with their
objectives (Chapter 3). The large majority responded ‘disagreed’ or ‘not clear’ while they
agreed with most of the statements. Apart from that there were no differences with respect
to the other knowledge statements, nor was there any difference among the other three loca-
tions (i.e., 2, 3 and 4).

Table 6.6: Farmers’ responses of agreement to statements on knowledge of soil erosion and soil fertility
processes, Wolaita, Ethiopia

Locations
Statements (1) (2) (3) (4) n χ2 Between -
Treated Untreated Adjacent Distant (p-value) Location
(%) (%) (%) (%) Differences

1. Steep lands are prone to 99 100 100 100 229 2.538

erosion (0.448)
2. Overgrazing exposes land to 91 100 100 100 224 15.568 1-2*
erosion (0.001)*** 1-4**
3. Land with poor vegetation 97 100 100 100 228 5.099
cover is easily eroded (0.165)
4. Continuous cultivation 91 100 100 99 223 11.845 1-2*
reduces soil fertility (0.008)** 1-4*
5. Soil erosion reduces soil 99 100 100 100 229 2.538
fertility (0.468)
6. High population pressure can 82 100 100 100 218 32.976 1-2**
increase land degradation (0.000)*** 1-3*
7. Farm bunds have to be main- 97 100 100 100 228 5.099 1-4***
tained regularly (0.165)

Note: *, ** and *** show significance at p ≤ 0.05, 0.01 and 0.001 respectively, non-significant otherwise. Between-location dif-
ferences were analysed with the Kruskal-Wallis one-way analysis of variance. Sample size: treated 65, untreated 46, adjacent
28 and distant 92.

It is needless to state that farmers do understand the severity of erosion on steep lands from
their life experiences. They understand the relationship between overgrazing and erosion as
animal hooves expose soil by removing grass and other vegetation. Widely cited examples
were footpaths and animal-resting places, which showed signs of vegetation removed by ani-
mal hooves.
With regard to vegetation cover, most farmers understand the anchorage function of plants
without understanding their association to reducing the erosive force of raindrops. The effect
of roots to enhance infiltration is hardly understood among farmers. What is clearly observed
by farmers is that their land is less eroded when there is good plant coverage. The farmers’
choice of crops involves trade-offs between the benefits from the crops in question and the fer-
tility status of the plot location. In view of this, they classify crops according to the demand

137
.tesfaye.thesis 29-01-2003 15:12 Pagina 138

of their nutrient requirements from the soil, which they call ‘energy’. Accordingly, crops such
as maize and legumes enrich the fertility while a crop such as teff reduces the fertility.
However, how the soil enrichment process takes place is not very clear except through leaf
decomposition. Because of this, I asked several farmers about the function of nodules in the
leguminous crops. The common answer was that it is a disease, a kind of deformation.
Contrary to a widespread belief that soil erosion is not noticed by farmers (Hurni and
Kebede, 1992; Kassaye, 1997), they could easily identify soil erosion on their farms, and its
impact on soil fertility. However, as indicated in Table 6.5, their focus is not on soil erosion per
se, but on soil fertility. They mention the symptoms of soil erosion only when specifically
asked. Their common indicators are small and large rills that are seen in the field after the
rainy period. Owing to their day-to-day experience with these phenomena, they all understand
that erosion affects their productivity through reducing the ‘energy’ of the soil. However, they
do not mention sheet erosion. One way whereby they indirectly mention sheet erosion is that
‘runoff’ takes away the ‘cream’ from the soil. They have observed that the ‘cream’ of the soil
is restored when erosion is controlled. In another discussion they associated ‘energy’ and
‘cream’ to the quality of a good, fertile soil. Continuous cultivation is one of the reasons cited
for the decline in soil fertility. They generally believe that the land should rest for a period of
time after some years of cultivation (for more information on indigenous soil knowledge in
Wolaita see Eyasu, 1997; Data, 1998).
The relationship between population pressure and land degradation is not that clear to
farmers. Their observation on this issue is linked with land inheritance. Therefore, the depen-
dence of a large population on the ecological services of the land needs further contextuali-
sation when discussing with farmers. For instance, the queuing at the water points, the space
used for residential areas, etc., are marked indications that they can easily understand and
relate to a population increase. Dwelling on this issue, farmers indicated that about a centu-
ry and half ago, there were no villages as such. All the families built their houses on their own
land. As the land owned by a family was very extensive, each family had been residing no less
than 2-4 km away from each other (see also Data, 1998:9). Farmers in Wolaita now understand
the effect that more people limit the farmlands. Only approximately 10-30 meters now sepa-
rate the houses of cousins and brothers. One farmer put his problem as follows:
‘The farm work is so demanding; however, productivity is going down from year to year.
Besides this, there are more people to feed than those who work on the farm. I have nine fam-
ily members, but I work alone. When there is erratic rainfall, I also sometimes lose part of the
little that I managed to grow in a season.’
Many farmers who participated in group interviews have shared the above view. For
instance, one farmer added:
‘What my fellow brother said is true. The land needs bribing, fertiliser. In the past when we
started using fertiliser with the WADU (early 1970s), we used to buy 50kg for only 12 birr. Now,
the same quantity of fertiliser is 150 birr. What we harvest is too small to feed our family three
times a day and to also buy shoes and clothes for our children who go to school.’
For obvious reasons, they tend to emphasise the issue of soil fertility. Farmers in all loca-
tions are aware of the importance of soil bunds in their area where soil erosion is the enemy
to the soil fertility they have built up with great effort. They also know the importance of soil
bunds and its management from several years of group and mass communication in the area.
In spite of that, bunds 39 are not yet part of their repertoire of land management practices.
In order to learn how farmers’ knowledge would influence their perception, statements of

39 Other studies (Data, 1998; Alemayehu, et al., 2001) report that farmers in the mountain areas practise stone bund con-
struction on their farms, but these too are limited to few bunds in selected sites rather than a typical case for Wolaita.

138
.tesfaye.thesis 29-01-2003 15:12 Pagina 139

perception related to the knowledge statements in Table 6.6 were presented to the farmers
(Table 6.7). The level of measurement was similar to that of the knowledge statements used in
Table 6.6.
Farmers’ responses were strong in terms of both the ‘agree’ and ‘disagree’ statements. The
response for ‘not clear’ was too small to be included in this table, ranging from 0.4 to 2.6 per-
cent. Among the six statements presented in Table 6.7, items 3 and 4 were highly significant
(p ≤ 0.001) while item 6 was significant (p ≤ 0.01). As in the case of knowledge statements, farm-
ers from the treated catchment significantly differ from the other three locations, whereas
there is no difference between the latter three locations, with respect these statements (see
between location difference, Table 6.7). There is no difference across locations with respect to
the role of soil and water conservation for crop production (item 1), the need to maintain soil
fertility for the future generations (item 2) and farmers’ responsibility in soil and water con-
servation (item 5).

Table 6.7: Farmers’ perceptions of soil and water conservation and crop production, Wolaita, Ethiopia
Locations
Statements Responses
(1) (2) (3) (4) n χ2 Between -
Treated Untreated Adjacent Distant (p-value) Location
(%) (%) (%) (%) Differences

1. Soil and water con- Agree 97 100 100 100 228 5.099
servation increases (0.165)
crop production
2. We have to main- Agree 92 96 93 96 217 1.004
tain soil fertility for (0.800)
the future genera-
tion
3. We have to feed the Disagree 74 98 100 100 212 42.025 1-2***
present generation (0.000)*** 1-3**
instead of thinking 1-4***
for the future
4. A high production Disagree 79 100 96 99 214 29.804 1-2***
now is preferable (0.000)*** 1-3*
than getting the 1-4***
average for the next
five years
5. Soil and Water Agree 88 98 93 97 216 6.703
Conservation is (0.082)
farmers’ responsi-
bility
6. Farmers should be Agree 77 96 96 92 205 13.500 1-2**
paid for soil and (0.004)** 1-3*
water conservation 1-4**
in their farms

Note: *, ** and *** show significance at p ≤ 0.05, 0.01 and 0.001 respectively, non-significant otherwise. Between-location dif-
ferences were analysed with the Kruskal-Wallis one-way analysis of variance. Sample size: treated 65, untreated 46, adjacent
28 and distant 92.

139
.tesfaye.thesis 29-01-2003 15:12 Pagina 140

As underscored in their responses to the knowledge statements in Table 6.6, farmers value
the importance of soil and water conservation to crop production. In view of declining land
size from generation to generation, farmers were asked about their views on the future gen-
erations. The concern for future generations is strong and is directed to each family. As farm-
ing is the major means of livelihood that they know and which is accessible to them, they see
the maintenance of soil fertility as a saving for their children. In view of this, they strongly
agree with the importance of soil and water conservation for crop production, which they see
as the farmers’ responsibility. Contrary to this opinion, close to 90% of the farmers expects
payment for any SWC measures they practise on their own farms. Farmers from the treated
catchment widely differ on the payment issue from farmers from the other three locations.

Shifts in land use and emergence of soil erosion

Farmers do not sow seeds on land that is threatened by erosion without some means of min-
imising the erosion. In the first place, erodible sites are not selected for production on a tillage
basis. The agrarian history of Wolaita also shows this general observation.
Due to a judicious allocation of land resources for crop and livestock production systems by
farmers, soil erosion was not the major bottleneck of crop production in highland Wolaita
until the early 1970s. According to farmers, it came to occur widely after the expansion of
cereal production from the flat to sloping land, which used to be operated by hand tools (gesso
and ayile). The expansion was enhanced by the shift from the gesso to the ox-plough for land
preparation. Note that the ox-plough has been found suitable for flat terrain while gesso was
used for steeper land near the homestead areas. Cultivation was and still is carried out by ayile.
In the past, Wolaita farmers used to grow coffee, enset, taro and some maize only on the gar-
den plot, called Darkua. The outer fields, Shoka, were used for grazing only. Darkua is the only
cropland known in the system that is managed with organic fertiliser. In the early 1970s, they
expanded their farming horizons to the Shoka land, using ox-ploughing and fertiliser intro-
duced by the WADU. According to the farmers, the WADU encouraged them to utilise fertilis-
er in their Shoka plots. For its part, the WADU promoted crop diversification to supplement
enset and root crops (see also Alemayehu et al., 2001). Credit for oxen was one of the WADU’s
programme activities. Prior to the 1970s, farmers were in a position to maintain a balance
between the Darkua and Shoka lands according to the food requirements of the household,
livestock size, grass requirements for house construction, etc. Grass from the Shoka land has
been supplementing the Darkua by supporting a large herd per family which produces
manure for fertilising the Darkua plot (see also Eyasu, 2000). After shifting large portions of
Shoka land to the cultivation of cereals, livestock depended on a small portion of grassland
located mostly towards the lower end of the farm. In addition, the scarcity of grass for con-
struction increased farmers’ expenses. A shortage of land led to the diversion of using crop
residues for soil fertility to animal feed. With the shrinking of woodlots down at the bottom
of the valley, crop residues, including enest stems, came to be used for firewood, which fur-
ther contributed to the decline of soil fertility. Farmers underlined that erosion is aggravated
when the fertility of soils declines.
As a result of the growing population, farmers in the middle altitudes were forced to con-
vert part of their home yards to crop or tree production. Farmers in the Domota mountain
area (the distant location) have widely converted the forest to farmland. This shift of land use
was carried out without sound physical conservation methods on very steep slopes cultivated
on all sides of the mountain, ranging from 1,941m to 2,908m at its peak, in the watershed cov-
ered by this study. These farmers do not have access to feasible technical means nor do they
receive security from the lands brought under cultivation.

140
.tesfaye.thesis 29-01-2003 15:12 Pagina 141

The state and causes of soil erosion

Farmers were asked whether there were erosion problems in any of their plots on the farm. Of
the total 231 farmers interviewed, 68% perceive erosion problems in some of their plots. In
terms of their location, the breakdown was: treated (28%) 40, untreated (20%), adjacent (10%)
and distant location (42%). With respect to the causes of erosion, some farmers have a clear
idea of why they have erosion problems, whereas others only have general ideas. Of the total,
56% of the farmers identified steepness (also see Table 6.8); 44%, water erosion; 25%, poor man-
agement, whereas 33% of them simply said it is because of natural problems. One farmer said
‘I do not know’ while another person attributed the cause of erosion to removal of the soil
bunds. Soil erosion occurs as a result of a combination of factors as mentioned by 68% of the
farmers (steepness, water erosion, and poor management).
In addition, farmers were asked why soil erosion occurred on their farms on a plot per plot
basis (Table 6.8). ‘Lack of bunds’ is generally mentioned in all plots except for P1. Even then,
floods that enter from the residential area affect this plot.
Erosion due to ‘lack of bund’ occurs in plot 1 on farms in the distant location, contrasting
with P1 in the other three locations. The reason has more to do with the steeper land features
in the distant location than with other factors associated with the proximity to the station.
Erosion in the other plots is randomly distributed among farmers from all four locations.
‘Lack of bunds’ does not show a significant difference between plots except for in P4 where
the between-location response was significant, at p ≤ 0.05. Only a small proportion of the
farmers at the distant location mentioned lack of bund in P1. Farmers from the other three
locations did not mention lack of bund as a problem to soil erosion in P1. This is probably due
to the direct negative experience they had in the treated catchment.
Among reasons indicated in Table 6.8, ‘steep lands without bund’ 41 and ‘lack of diversion
ditches’ were frequently mentioned problems with a substantial percentage across locations
and plots. The response of ‘steep land without bund’ varies across location and plots with p ≤
0.001 for P1 and P2, while P3 is significant at p ≤ 0.01. Responses to this item for P4, P5 and P6
do not show variation across locations. Farmers from the distant location emphasise steep
lands and absence of bunds more often than those from the other three locations. This dif-
ference can be explained by the fact that farming is practised on predominantly steep agri-
cultural land in the distant location.
‘Damaged bunds’ occur in the treated catchment and the distant location where soil bunds
were installed through food-for-work. The response on this issue was significant for P3 (p ≤
0.01), P4 (p ≤ 0.001), P5 (p ≤ 0.05) and P6 (p ≤ 0.001). Farmers from the treated catchment asso-
ciate soil erosion with damaged bunds more strongly than in the other three locations.
The diversion ditch is a very important physical conservation measure that is widely prac-
tised in the area. Farmers felt that the higher occurrence of erosion in some plots compared
to others was due to an absence of diversion ditches. However, their response with respect to
the upper two plots (P1 and P2) shows a significant difference across locations (p ≤ 0.05). This
is caused, among others, by the exposure of some plots in the upper location to floods from
outside the farm boundary. Unlike other locations, farmers in the treated catchment did not
mention a lack of diversion ditches as a reason for soil erosion, which is particularly apparent
in the first four plots. A variation in the degree of flood hazards from farm to farm was also
reflected in the group discussions. The difference between farmers from the treated location
and other locations could be because of less intensive flooding from the pathways and better

40 These farmers constitute 68% of those from the treated catchment (65 in number).
41 ‘Steep land without bund’ was deliberately included in the survey instrument in order to capture a more precise response
than that of ‘lack of bund’, which may occur on less steep lands as well.

141
.tesfaye.thesis 29-01-2003 15:12 Pagina 142

fertility in the upper plots, as soil fertility has increased in more plots in the treated catch-
ment over the last ten years.
Some farm plots are inherently prone to runoff because of their location. ‘The land is under
a steep ridge’ is one of the features that farmers mention to show erosion prone plots. These
kinds of plots are distributed across the location, though it does not show a significant dif-
ference between locations.

Table 6.8: Farmers’ reasons for soil erosion, Wolaita, Ethiopia

Reasons Location P1 P2 P3 P4 P5 P6

Lack of bunds Treated (%) - 2 6 9 2 5

Untreated (%) - 11 15 9 7 2
Adjacent (%) - 11 11 14 7 7
Distant (%) 5 11 7 1 2 -
χ2 7.721 5.358 3.643 8.399 3.555 5.819
(p value) (0.052) (0.147) (0.303) (0.038)* (0.314) (0.121)

Steep land without Treated (%) 2 3 11 15 9 5

bund Untreated (%) - 13 17 17 4 -
Adjacent (%) 11 14 18 11 7 4
Distant (%) 36 45 34 21 8 3
χ2 46.443 42.575 13.268 1.737 0.999 2.038
(p value) (0.000)*** (0.000)*** (0.004)** (0.629) (0.801) (0.565)

Damaged bunds Treated (%) 5 3 14 17 11 12

Untreated (%) - - - - - -
Adjacent (%) - - - - - -
Distant (%) - 2 3 1 1 1
χ2 7.762 2.101 14.539 25.361 9.636 17.222
(p value) (0.051) (0.552) (0.002)** (0.000)*** (0.022)*a (0.001)***

Lack of diversion Treated (%) - 3 3 11 8 6

ditch Untreated (%) 8 13 15 13 4 4
Adjacent (%) 7 7 14 7 7 7
Distant (%) 14 19 17 12 8 2
χ2 10.133 9.421 7.127 0.682 0.618 2.090
(p value) (0.017)* (0.024)* (0.068) (0.877) (0.892) (0.554)

The land is under Treated (%) 2 3 6 12 11 8

steep ridges Untreated (%) - 2 7 4 7 4
Adjacent (%) - 7 7 7 11 11
Distant (%) 2 8 7 9 3 1
χ2 1.553 2.749 0.033 2.257 4.133 6.175
(p value) (0.670) (0.432 (0.998) (0.521) (0.247) (0.103)

Note: *, ** and *** show significance at p ≤ 0.05, 0.01 and 0.001 respectively, non-significant otherwise (χ2 test). The letter ‘a’
shows adjusted chi-square and p-values. Sample size: treated 65, untreated 46, adjacent 28 and distant 92.

142
.tesfaye.thesis 29-01-2003 15:12 Pagina 143

Another factor that was identified by farmers as a cause of soil erosion is a shift in the rain-
fall pattern. Wolaita farmers prepare land for the Belg season starting from October. Rainfall for
the Belg season used to start in January, but the onset of the rains has shifted to March or April
over the last thirty years. Farmers stated that the soil is easily eroded if exposed to sunlight for
too long after ploughing. However, the farmers’ complaint about the rainfall pattern could not
be ascertained based on only 43 years of rainfall data from the National Meteorological Service
Agency (see section 6.2). If what farmers claim is true, it warrants a corresponding shift in land
preparation to overcome excessive exposure of the soil. However, this is the issue to be deter-
mined between agronomists and farmers through action research supported by climatic data
over a longer period.
During the qualitative work, I held interviews with farmers in their fields. I asked them how
and why their responses to the statements on knowledge of soil erosion and soil fertility, and
their perceptions of soil and water conservation and production (Table 6.6 and 6.7) were relat-
ed to their field conditions. During those interviews, farmers showed me what they did, while
confessing the shortcomings, which they said they would remedy in the future. Through non-
participant observations I learned that some farmers did not carry out their plans to control
some rills which they had agreed to during our discussion in the previous seasons. This method
also gave me the opportunity to learn that farmers do different things in different plots, yet for
different reasons. The point is that they always have reasons for their actions, be it shortage of
labour, lack of motivation, or lack of other resources. Generally, I have witnessed the depth of
farmers’ knowledge about their fields. They know the history of each and every pocket of their
land. Owing to this, they know which part of the plot suffers from erosion and why, where a
lack of bunds has caused most problems and where drainage ditches should have been
required. Erosion at the plot level may be explained by one or more reasons as shown in Table
6.8. That means that in some cases most of the major problems mentioned above may occur in
the same plot, while in others, two or three of the problems occur in one plot, while none in
others. This observation generally calls for plot-level measures where farmers themselves play
a leading role, taking into account the farm and the watershed features.

Trends of soil fertility

Farmers were asked to identify trends of soil fertility for each plot in their fields from 1990-2000
(Figure 6.1). With the exception of the upper Darkua plots where there is a substantial percent-
age of plots reporting increased infertility, in all other plots, the percentage of farmers with
increased soil fertility declines as one moves from the residential to the outer fields (Shoka).
Eyasu (1997) has carried out a nutrient input and output balance of fields and farms in the
study area. According to his findings, Nitrogen balances of most of the highland farmers in the
Darkua and taro fields were either slightly positive or at equilibrium. At the same time, Shoka
fields showed a pronounced negative N balances. The same study shows a Phosphorous balance
that was close to equilibrium. In terms of socio-economic groups, rich 42 farmers had accumu-
lated P in enset, taro and even in Shoka fields, whereas the poor farmers experienced a slight
deficiency of P in the taro and Shoka fields.
Sheleme (2001) attributes the variations in soil characteristics from soil profile pits to dif-
ferences in slope and soil management. His findings show that soil profiles of a convex crust
and depression have deep dark surface layers with a relatively higher organic matter content
than those at mid and top slopes. He also added that soil bunds have helped in the accumula-
tion of soil at the lower border of the bunds, resulting in a deep ‘A’ horizon.

42 Wealth indicators used in this study were ownership of oxen, herd size, and size of landholding.

143
.tesfaye.thesis 29-01-2003 15:12 Pagina 144

Figure 6.1: Trends of soil fertility from 1990-2000 in Wolaita area

70
60
Percentage of farmers

50
40
Increase
30
20 Decrease
10
0 Same
1 2 3 4 5 6
Plots (1-6)

The decline of soil fertility takes place for different reasons between households (Table 6.9).
Farmers identified ‘shortages of manure’ and ‘soil erosion’ as the most important reasons in
all six plots. The responses on soil erosion were significant for P1 (p ≤ 0.05), P2 (p ≤ 0.001) and
P3 (p ≤ 0.01), whereas the responses do not vary significantly in the lower plots. Farmers in the
treated location faced soil erosion because of removal of soil bunds from their plots. Farmers’
observations are consistent with experiments carried out on erosion rates in different plots of
the area which show the importance of erosion in the decline of soil fertility (SCRP, 1996b;
Eyasu, 1997). Erosion in plot 1 (under perennial crops) takes place because of runoff from the
residential areas, such as pathways rather than by runoff built into the plot, which is insignif-
icant, because of the permanent plant cover and high organic matter. Lack of opportunities to
fallow land coupled with unabated erosion has aggravated the decline of soil fertility in Shoka
plots. The high price of fertiliser has limited the use of commercial fertiliser as an alternative
source. This problem is felt in all locations. Due to this problem farmers complain about
‘tired’ soil that is loose on the plough shear.
According to farmers’ practices, an increase in soil fertility can be achieved by expansion of
the fertile farm horizon (Darkua) towards the infertile zone (Shoka) and/or by targeting specif-
ic plots. Over the last ten years the increase of soil fertility in terms of average number of plots
was highest for farmers in the treated catchment, followed by untreated, distant, and adja-
cent. The average number of plots with decreased soil fertility, over the same period, is high-
est in adjacent, untreated, distant and treated plots, in descending order. Note that the treat-
ed catchment shows the highest and the lowest averages for plots with an increase and
decrease of soil fertility, respectively. This favourable situation among farmers from the treat-
ed catchment could be because of the soil bunds which could serve as an entry point to pro-
mote a participatory learning process on soil fertility management in the area.
In view of the soaring fertiliser price, the lack of rural financial markets, the decline of
grass land, and the decline of productivity on the Shoka fields; the sustainability of whole sys-
tem depends on the opportunities that help farmers to expand the horizons of the Darkua
zone into that of the Shoka zone (Eyasu; 1997, 2000). For instance, providing livestock credit for
poor families to own livestock, among others, to improve their manure supply.

144
.tesfaye.thesis 29-01-2003 15:12 Pagina 145

Table 6.9: Farmers’ reasons for decline of soil fertility, Wolaita, Ethiopia

Reasons Location P1 P2 P3 P4 P5 P6

Repeated Treated (%) 9 11 14 11 11 8

Cultivation Untreated (%) 2 7 13 13 11 9
Adjacent (%) 4 - 14 14 11 4
Distant (%) 4 13 21 17 11 3
χ2 3.286 4.868 1.986 1.439 0.001 2.530
(p value) 0.350 0.182 0.575 0.696 1.00 0.470

Lack of Manure Treated (%) 20 17 25 20 19 11

Untreated (%) 20 22 33 26 20 15
Adjacent (%) 29 32 39 36 29 21
Distant (%) 21 29 39 24 15 7
χ2 1.056 4.239 4.026 2.678 2.562 5.703
(p value) 0.788 0.237 0.259 0.444 0.464 0.127

Soil Erosion Treated (%) 6 9 14 14 15 8

Untreated (%) 11 22 35 28 20 13
Adjacent (%) 7 18 25 29 18 11
Distant (%) 22 37 38 22 12 5
χ2 9.641 17.037 11.867 4.311 1.590 2.618
(p value) 0.022* 0.001*** 0.008** 0.230 0.662 0.454

Lack of Fertiliser Treated (%) 6 6 11 14 8 2

Untreated (%) 4 15 26 22 17 11
Adjacent (%) 7 7 18 21 21 14
Distant (%) 8 20 24 22 14 5
χ2 0.569 7.021 5.504 1.832 3.902 7.049
(p value) 0.903 0.071 0.138 0.608 0.272 0.070

Note: *, ** and *** show significance at p ≤ 0.05, 0.01 and 0.001 respectively, non-significant otherwise (χ2 test). Sample size:
treated 65, untreated 46, adjacent 28 and distant 92.

Farmers’ expectations of and commitment to soil and water conservation

In the previous sections, it has been shown that farmers highly appreciate soil and water con-
servation in the production processes. In this section, the expectations and commitments are
assessed.
The farmers’ major expectations from soil and water conservation is increased crop pro-
duction. They stated that soil bunds began to reduce erosion from the second year. After the
second year, the difference between crop stands above and below the bund becomes visible.
Farmers’ observations of agronomic characteristics of crops on bunded fields include: germi-
nation length (early, late), growth rate within a week, leaf colour (greenish or yellowish), stem
vigour, and production level. The crop stand above the bunds comes out better in all these
agronomic characteristics as compared to the ones below the bund.
Regarding changes in production, farmers indicated that fields with bunds provide better
yields that can be observed four years after the construction of the bunds. Increases in crop

145
.tesfaye.thesis 29-01-2003 15:12 Pagina 146

production by up to 50% were reported by farmers. They have indicated that use of fertiliser
declines when there is a bund in a plot. An estimated reduction in the amount of fertiliser
used was up to 50%. Even though it is difficult to be sure whether the increase in production
was due to efficient fertiliser utilisation or because of reduction in soil erosion, some farmers
argued that using the same amount of fertiliser on plots resulted in plots without a bund
being less productive than plots with bunds. Some of them stressed their effort to maintain
soil fertility using other means, including bund construction, because of increasing fertiliser
prices from year to year. Amare (1988:81) reported that maize production on plots with soil
bunds at Gunno was more than double that of plots without bunds. However, the effect of soil
bunds on productivity needs much more careful analysis because of many intervening biotic
and abiotic factors that affect land productivity.
In order to get an overview of the farmers’ view, I asked them to identify their expectations
from the soil and water conservation activities. They identified different combinations of
expectations. Of the total 203 farmers, 40% expect better production and feed, reduced soil
erosion and increased soil fertility. Another 27% focused their attention on increased produc-
tion, soil fertility and reduced erosion. The remaining 33% of these cases opted for miscella-
neous alternatives, ranging from a single purpose such as production, feed, soil fertility or
reduced erosion to combination of these. Most farmers expect about 100 kg (total production)
for their efforts in soil and water conservation from their entire fields. But there were few
farmers who expect as much as 500 kg. The average expected additional production due to soil
and water conservation was 144 kg.
In connection to this issue, I asked farmers about the number of days they would like to
work on soil and water conservation measures such as soil bunds and diversion ditches. On
average, they are willing to work for 21 days per year, though a few farmers would like to work
almost the whole year through. The period that is most suitable for soil and water conserva-
tion structures is December and January when major farm operations are at their minimum.
The 25th, 50th and 75th percentile for number of days to participate in soil and water con-
servation activities are 9, 15 and 25 days, respectively.

Farmers’ practices in soil and water conservation

In the previous sections of this chapter, farmers’ understanding of the biophysical processes
on soil erosion, soil and water conservation and expectations from conservation measures
were discussed. In this section, farmers’ soil and water conservation practices are presented in
order to assess the influence of their knowledge and perception on their land management
practices.
Wolaita farmers do practise both biological and physical conservation measures, albeit
with different emphasis. Table 6.10 presents profiles of some of these practices that are briefly
discussed below.
Biological conservation measures are widely used in soil and water conservation practices.
These include: manure, household refuse, mulching, live-fences, leaf litter and to a limited
extent row-planting of multi-purpose trees.
The most common and accessible methods are manure and household refuse application
that are the insurance of soil fertility in Wolaita. These inputs are applied to plots where crops
that highly contribute to food security (enset, taro, yam and maize) are grown. As shown in
Table 6.10, most farmers apply organic fertiliser in P1 and P2 that show significant differences
(p ≤ 0.001) across locations. The weight of manure application declines as one moves away
from the homestead areas, mainly because of the crops grown on those plots. In all six plots,
the percentages of farmers using this method in the treated catchment are greater than farm-

146
.tesfaye.thesis 29-01-2003 15:12 Pagina 147

ers from the adjacent and distant location, by twofold or more. A similar trend was observed
in the other plots, some of which are significantly different from each other (p ≤ 0.05).
Farmers in the untreated catchment follow those in the treated catchment in terms of
manure utilisation.

Table 6.10: Soil and water conservation practices, Wolaita, Ethiopia

Practices Location P1 P2 P3 P4 P5 P6

Manure Treated (%) 43 39 22 17 9 5

Untreated (%) 35 26 7 4 - -
Adjacent (%) 18 7 - - - -
Distant (%) 20 17 10 5 1 -
χ2 12.667 14.200 11.664 11.380 7.329 3.468
(p value) (0.001)*** (0.001)*** (0.039)* (0.016)* (0.062) a (0.325) a

Row-planted food Treated (%) 3 3 2 2 2 -

trees Untreated (%) - - - - - -
Adjacent (%) - - - 4 - -
Distant (%) 2 3 - - - -
χ2 0.509 0.546 0.972 3.019 0.972 -
(p value) (0.917) (0.909) (0.808) (0.389) (0.808)

Flood water Treated (%) 6 3 2 2 2 3

Untreated (%) 9 9 2 2 - -
Adjacent (%) 4 4 - - 4 -
Distant (%) 10 10 8 3 - -
χ2 1.497 3.371 5.851 0.212 3.019 1.982
(p value) (0.491) (0.162) (0.050)* (0.976) (0.389) (0.576)

Diversion ditches Treated (%) 2 2 5 8 2 -

Untreated (%) 2 4 9 9 7 7
Adjacent (%) 4 14 14 11 4 4
Distant (%) 34 34 27 14 2 1
χ2 43.930 35.421 17.033 1.919 2.638 5.111
(p value) (0.000)*** (0.000)*** (0.000)*** (0.175) (0.451) (0.164)

Soil bunds Treated (%) 11 12 12 14 6 6

Untreated (%) - - - - - -
Adjacent (%) - - - - 4 4
Distant (%) 1 1 - - - -
χ2 9.424 17.222 14.076 23.916 5.492 5.492
(p value) (0.024)* a (0.001)*** (0.003)** a (0.000)*** (0.139) a (0.139) a

Soil bunds with Treated (%) - 5 15 23 28 19

grass Untreated (%) - 4 4 4 2 9
Adjacent (%) - - - - 4 1
Distant (%) 3 4 3 1 3 1
χ2 1.206 1.300 12.339 30.018 34.703 22.570
(p value) (0.752) a (0.729) (0.006)** (0.000)*** (0.000)*** (0.000)***

Note: *, ** and *** show significance at p ≤ 0.05, 0.01 and 0.001 respectively, non-significant otherwise (χ2 test). The letter ‘a’
shows adjusted chi-square and p-values. Sample size: treated 65, untreated 46, adjacent 28 and distant 92.

147
.tesfaye.thesis 29-01-2003 15:12 Pagina 148

The coffee plot (P1), which is next to the house, also receives animal urine channelled to
the plot directly from the barn through the house wall. Composting, as prescribed by techni-
cal personnel, however, is not practised in the area. Very few farmers have had training on
composting and even those who have, have not applied the training. According to the farm-
ers, composting is tedious as it involves digging of holes and transferring the compost mate-
rials between the holes.
Application of manure and household refuse is mainly the task of women. Children of both
genders assist women in this task. Men identify specific locations where organic matters
should be applied. This is because of their close association with every inch of the field that
helps them to know the fertility status of the whole farm.
Even though manure is the major means of soil fertility management, there is marked vari-
ation among farmers in the rate of its use. Of the total 231 farmers, 17% use manure in all of
their plots, while those using it in part of their plot number 81%. The remaining 3% of the
cases do not use manure at all.
When we look into livestock ownership patterns of these manure-user groups, 14% of ‘non-
users’, 7% of ‘use on part of the plots’ and 13% of ‘use on all of the plots’ do not own livestock,
not taking chickens into account. Most of those who use manure in part of their plots and
those who use it in all of their plots own one to five heads of livestock, while a few of them
own up to 11 heads of livestock. But, this issue needs further investigation.
During informal interviews, farmers said that access to manure in Wolaita is very difficult,
as it is a very essential input in crop production. From these discussions it was learned that
there was no chance of borrowing manure from relatives or neighbours. As one farmer put it,
‘giving manure to someone else is like giving your door to your parents-in-law and being eaten
by a hyena.’ On this issue, a woman interviewed by Eyasu (1998:15) replied to a question with
the question, ‘would you give someone your life?’. As can be seen from the result of the sur-
vey, there are possibilities for access to manure without owning livestock at home.
The other factor that affects manure utilisation patterns is land size. There is no regular
pattern between use of manure and the size of cropland for those farmers who own a below-
average land size (0.62 ha). All farmers who have above-average cropland can cover only part
of their plots. The general observation is that farmers with a small land size could increase
the fertility of their land by using organic matter (Eyasu, 1997, 2000). Even though this obser-
vation is plausible, the other point is whether the increase in soil fertility would ensure their
food security, something that is highly affected by the family size, among other factors.
Apart from shortage of manure, farmers also mention distance of plots and shortage of
manpower to apply manure. This was mentioned even if the end of the plot is 150 to 200
meters away from their houses.
Live fences and high trees are very common in Wolaita. Live fences of mixed trees and
shrubs are good sources of leaf litter that are regularly collected by both men and women,
including children for incorporation into the soil. Leaf litter is an important source of organ-
ic fertiliser reported by 76% of 231 farmers. Green manuring is very rare, reported only by 4%
of the farmers.
In addition, farmers plant rows of banana, mango, cassava, local hobs and citrus fruits for
both soil conservation and the generation of income and food. However, this practice is at an
early stage. It began only among a few farmers in an effort to replace soil bunds or to plant
trees for soil conservation purposes at an erodable site. Some farmers mix food crops with
suitable agroforestry plants such as Cordia african, Croton macrostchyus and Erythrina brucei.
However, multiple storeys showing the vertical association of trees and crops is rare, with the
exception of a few Cordia trees in coffee plots. Erythrina is particularly common in the live

148
.tesfaye.thesis 29-01-2003 15:12 Pagina 149

fences, unlike the other two trees mentioned above. Therefore, the green landscape of Wolaita
is because of enset, coffee and live fences rather than agroforestry vegetation.
Plots with perennial crops, particularly enset are allowed to receive floodwater to increase
moisture availability in order to ensure its vigour during the dry period. However, such farm-
ers ensure that excess water is diverted away to the waterways to avoid run-off. Use of flood-
water seems common among farmers from the distant location who farm in steep areas with
a 70% slope, and followed by those from the untreated catchment.
The use of floodwater depends on the construction of diversion ditches, which drains
excess water to the waterways. To that extent, diversion ditches are very common physical con-
servation measures, particularly among farmers from the distant locations. These farmers
construct diversion ditches diagonally across their farms. The use of diversion ditches shows
highly significant differences across locations for the first three plots (at p ≤ 0.001).
Soil bunds with and without grass cover are markedly visible in the treated catchment as
compared to other locations in this study, obviously because of a continuous intervention in
this catchment since the 1970s. Of the total 231 farmers in the sample, 82 (36%) reported use
of soil bunds in their farms. Among these, 59% of them come from the treated catchment,
whereas those from untreated, adjacent and distant location were 16%, 6% and 20% respec-
tively. The presence of soil bunds shows a significant difference between locations and plots
(see Table 6.10). Farmers in the adjacent and untreated catchments have constructed soil
bunds on their own initiative. Unfortunately, these cases are too few to warrant a self-initiat-
ed critical mass for soil and water conservation in the area.
A few farmers have maintained a traditional bund that was gradually shaped out of a huge
piece of land. Such bunds have existed since their forefathers. Some farmers have dug these
bunds and incorporated them into their croplands.
The other widely practised conservation measure in the area is contour ploughing. While
farmers plough up and down during the primary land preparations, the final plough is done
following the contour, to minimise runoff.
Apart from the above specific conservation practices, farmers tend to apply a range of inten-
sive land-management practices such as repeated ploughing, cultivation to minimise weed,
burning of debris before planting and crop rotation to enhance soil fertility.
In the earlier part of this section, farmers’ responses on selected knowledge and perception
statements were presented (Tables 6.6 and 6.7). Later on, their current practices in soil and
water conservation were discussed, which includes that of soil bund construction and main-
tenance. When we compare the farmers’ knowledge and perception response with their
response on use of soil bunds, the deviation in percentage was 30, 72, 79 and 81 for treated,
untreated, adjacent and distant farmers, respectively. The reasons for the gap between knowl-
edge and perception response and practice are discussed in the next section.

6.5.3 Soil and Water Conservation at Gunno and surrounding areas revisited

The Soil Conservation Project began testing soil and water conservation measures at Gunno
with the purpose of identifying suitable conservation measures for the whole wet Woyna Dega
agro-ecological zone. Alongside this, attempts were made to popularise soil and water conser-
vation through the national extension system. To that effect, soil bunds were constructed by
some Peasant Associations through food-for-work incentives.
As noted earlier (section 6.3), the SCRP was phased out in 1998 with a gradual transfer of
the centre management beginning in 1996. In view of this, the present study is interested in

149
.tesfaye.thesis 29-01-2003 15:12 Pagina 150

assessing the changes in the land-management practices in line with the great effort made in
this part of the country.
In the survey, it was discovered that 54% of farmers in the treated catchment who had soil
bunds have removed one to four bunds. The number of bunds removed was 1, 2 and 4 (Table
6.11). Most farmers removed the upper bunds and the ones in the middle. Among these farm-
ers, 15% have totally removed the soil bunds, while the remaining 85% have one to six bunds
left in the their fields, though most farmers (77%) have less than four bunds.
The trend in the total sample also shows a similar situation. Out of 82 farmers in the total
sample who had some bunds on their farms, 52% preserved all soil bunds, while the remain-
ing 48% removed 17 to 100 percent of their soil bunds. Most of those who completely removed
their bunds were located in the treated and distant locations, each accounting for 39% of the
total. Farmers from untreated areas and adjacent to the centre both account for 11% each.
For both retained (A) and removed (B) soil bunds, farmers from the treated catchment sig-
nificantly differ from the other three locations (at p ≤ 0.001).

Table 6.11: Decisions on soil bunds, Wolaita, Ethiopia

Locations % of those who retained their bunds (A) % of those who removed their bunds (B)

No. of bunds removed

1 2 4
% % %
Treated 60 19 18 3
Untreated 94 4 2 -
Adjacent 89 4 7 -
Distant 93 3 4 -

A B

F-value 22.418 11.802

Significance 0.000 0.000

LSD (A,B) Treated Untreated Adjacent Distant

Treated *** *** ***
Untreated *** ns ns
Adjacent *** ns ns
Distant *** ns ns

Note: ns, *** show non-significant and significant at p ≤ 0.001 for the F-test. Tukey’s procedure was used to test the significance
of differences between locations.

Very few farmers started removing the bunds two or three years after they were constructed.
Widespread removal, however, took place after 1990, mainly because of shrinking field activ-
ities by the SCRP. Farmers remove the bunds progressively. Some farmers are confident about
their measures, while some of them regret removing the bunds after they observe rills, spe-
cially under heavy rainfall. Those who are confident about their action claim that they have
not seen any flood passing at the point because of the protection from the upper bund or the
perennial crops in the garden. Some farmers removed bunds simply because others who fall

150
.tesfaye.thesis 29-01-2003 15:12 Pagina 151

on the same contour or above them have removed them. When the bunds from above are
removed, farmers who are at the lower altitude are affected by vigorous runoff.
More than half (51%) of those who have preserved their bunds come from the treated catch-
ment. These farmers account for 46% of their stratum. However, the proportions of those who
have preserved their bunds is higher for farmers from the untreated and distant locations.
This difference is partly because of a deliberate construction among these farmers; i.e., the
design is made according to their own judgement for space, there is a felt need to do some-
thing about the erosion problem, and above all there was no obligation.
Farmers have been advised to maintain their bunds periodically by raising their height by
transferring the deposition on the basin onto the embankment as specified by Hurni (1986,
cited in Gunten, 1993). In this connection, they have mentioned that the bunds would form a
bench terrace in about 8-10 years. After two decades of constructing levelled soil bunds in the
Zerwa catchment (treated), no bund has attained the stature of a bench terrace. This is main-
ly because of the lack of periodic maintenance of the soil bunds. Of the total 231 cases, only
10% reported some maintenance from time to time. Among these, 73% are from the treated
catchment, followed by 14% from the untreated catchment, whereas there were none from the
distant location.
The percentage of bunds removed was correlated with selected variables that were used in
previous research studies in the Ethiopian setting (Amare, 1988; Bekele, 1998). These were:
total number of livestock owned, area of cropland, economically active labour force, age of the
household head, and family size (see Appendix 2). Cropland area, labour force and age were
negatively correlated, while livestock ownership and family size were positively correlated
though weakly correlated with the percentage of bunds removed. Unlike age and labour, crop-
land area was significantly correlated with percent of bunds removed (p ≤ 0.05). This result
denotes that the larger the cropland, the lower the percent of bunds removed and vice versa.
To that extent it is consistent with farmers’ views regarding competition with land. It is worth-
while to note, however, that the relationships in the sample show wider possibilities that are
not fully reflected by these statistical relationships. For instance, the cropland size of those
who preserved their bunds ranges from the smallest land size in the sample (0.13 ha) to 1.38
ha. Cropland size of those farms where the bunds are totally removed falls below the average
land size for the sample (0.62 ha), with the exception of two farmers who had land sizes of 0.81
and 0.88 hectares. With respect to livestock ownership and family, the tendency is that those
with more livestock and a higher family size tend to remove the bunds more often than oth-
ers, partly for reasons of land competition.
Even though age of household heads was negatively correlated with percentage of soil
bunds removed, some farmers of an average age of 40 and 25 years had removed 17% and 80%
respectively. On the other hand, the average age of farmers who preserved 100% of their soil
bunds ranges from 44.4-47.5 years.
The contribution of livestock to the farming systems regards the provision of traction
power, manure and food security. The use of cash from the sale of livestock in crop production
may affect the purchase of fertiliser and the organisation of labour rather than the decision
on soil bunds directly.
Education, which is often cited as one of the explanatory variables in farmers’ land-man-
agement decisions also shows an irregular pattern with regard to farmers’ decisions on soil
bund retention or removal. The general farming practice did however show some difference
between illiterates and literates.

151
.tesfaye.thesis 29-01-2003 15:12 Pagina 152

Constraints to construction and maintenance of soil bunds

Poor acceptance of soil bunds among farmers in Wolaita has been noticed both during the
WADU and SCRP periods (Hailemariam, p.c; Belay, 1992b; Eyasu, 1997; Tilahun, 2001). This sec-
tion presents farmers’ reasons for not accepting soil bunds on their farms and why they fail
to maintain them after their having been installed through food-for-work.
In spite of a high level of knowledge and positive attitude about soil erosion and soil and
water conservation practices, not many farmers constructed and/or maintained the soil bunds
on their farms. Care for the bunds and a range of complementary conservation practices were
only taken in the first few years. According to a remark by a farmer, pointing to the SCRP staff,
‘we left it, when they left it’.
As shown in table 6.8, the lack of bunds was a highly noted reason for the prevalence of ero-
sion and the decline of soil fertility on their farms. In view of this, I asked farmers why they
do not construct and maintain bunds if they so appreciate its use. In answer, farmers gave a
range of reasons, most of which are applicable to the whole farm whereas some are more rel-
evant at the plot level. In view of this, the two sets of reasons are presented in the consecutive
tables (Table 6.12a and b).

Table 6.12a: Farmers’ reasons for not constructing and maintaining soil bunds, Wolaita, Ethiopia

Reasons Locations

Treated Untreated Adjacent Distant χ2

p
(% of farmers) (% of farmers) (% of farmers) (% of farmers) ( -value)

Soil bunds harbour mole-rats 9 17 4 2 11.342

(0.010)**
Soil bunds harbour weeds 2 7 4 - 5.466
(0.141) a
Bunds hinder ox-ploughing 11 30 - 14 14.717
(0.002)**
Soil bunds waste land 9 22 11 20 4.791
(0.188)
Shortage of labour 8 33 39 41 22.354
(0.000)***
Shortage of tools 3 - 4 1 1.956
(0.582) a
I do not know 2 17 18 22 13.090
(0.004)**

Note: *, ** and *** show significance at p ≤ 0.05, 0.01 and 0.001 respectively, non-significant otherwise (χ2 test). The letter ‘a’
shows adjusted chi-square and p-values. Sample size: treated 65, untreated 46, adjacent 28 and distant 92.

In order to learn farmers’ general opinions about the soil bunds they were specifically asked
to identify if soil bunds have disadvantages. Of the total 231 cases, 80% identified five major
problems. These include; high labour demand for construction, wastage of land, harbouring
of mole-rats (Heterocephalus glaber) and weeds, and maintenance requirements. Farmers’
responses show different weights for these five problems. Accordingly, 22% of the farmers
mentioned all five problems, whereas 10% mentioned the first three problems as stated above.

152
.tesfaye.thesis 29-01-2003 15:12 Pagina 153

Another 8% mentioned harbouring of weeds and mole-rats. The other half of those who
observed some disadvantage in soil bunds, mentioned miscellaneous combinations of prob-
lems, each with small percentages ranging from 0.4-5%. The remaining 20% believes that soil
bunds do not have any disadvantage. Some of these farmers underlined that the disadvan-
tages of soil bunds should be tolerated in view of the protection they give to their land. They
added that farmers need someone to ‘teach’ them (cases 94, 194 & 195). Most of the farmers
who favour soil bunds come from the distant location. However, 22% of those who responded
that bunds do not have a disadvantage have destroyed all of their soil bunds. The view on the
disadvantages of soil bunds is shared by both farmers in the treated catchment who had direct
experience and also farmers in the distant location, most of whom did not have any soil
bunds. To illustrate, the latter group accounts for 43% of those who mentioned all five prob-
lems listed above, followed by farmers in the untreated catchment (28%).
When comparing the responses of treated and untreated (Table 6.12a), farmers from the
untreated catchment seem to fear the negative effects of soil bunds more than others, includ-
ing those from the treated catchment who experienced it. Among these responses, mole-rats
and ‘hindrance to oxen ploughing’ were significant at p ≤ 0.01.
Farmers’ complaints about rodents and weeds were documented a few years after the soil
bunds were promoted by the SCRP (Amare, 1988; Belay, 1992b; Herweg, 1993). However, their
occurrence seems common among plots distant from homestead areas. Farmers reported that
mole-rats expand in the field and damage crop roots including those of coffee. Traditionally,
farmers destroy mole-rats by digging out their holes and killing them. Maintaining the bunds
means allowing the rodents to breed further, threatening their crops. To that extent, the
introduction of soil bunds has interfered with this traditional mole-rat control method. In
addition, overflow enters into the bunds through holes that are created by the mole-rats and
makes the bunds susceptible to damage. This is often aggravated by expansion of the cropping
area close to the base of the bund. Some farmers cut the width of the bund to expand the
crop’s area and to bring down the fertile soil of the upper field to the shallow portion of the
lower field. At an early stage, the project provided farmers with chemicals to kill the mole-
rats, but this was ceased long ago. No one has ever tried to reconstruct another bund after
removing the original one because of mole-rat or weed infestation. No action was taken by the
project or any other body regarding the weeds, particularly running grasses. The best method
known to farmers is uprooting the weed with hand tools. In the process, they destabilise the
bunds.
Socio-technical factors such as incompatibility with the existing farm equipment (i.e., ox
plough) and competition of bund space with cropland were mentioned by a substantial num-
ber of farmers. Those in the untreated and distant locations seem to emphasise these prob-
lems. Land preparation, particularly for teff needs a repeated ploughing. Because of this, they
plough up and down several times to ensure that the soil is well mixed before they finally
plough along the contour. In this case, the soil bunds hinder the up and down ploughing prac-
tices and prompt farmers to remove them.
A shortage of manpower was mentioned as a reason by 30% of the 231 farmers. Most of
those who reported this problem come from the distant location, followed by adjacent,
untreated and treated locations. The responses were significant across locations at p ≤ 0.001
(Table 6.12a). As discussed in section 6.5.1 (Table 6.3), the number of economically active per-
sons per family is higher for the treated catchment followed by the distant, adjacent and
untreated catchments. However, there was no statistically significant difference between the
locations with respect to an economically active labour force per household.
A few farmers mentioned the lack of tools and know-how for construction of bunds. The

153
.tesfaye.thesis 29-01-2003 15:12 Pagina 154

majority of farmers who mentioned these problems come from the distant villages.
Lack of know-how (‘I do not know’) to construct soil bunds was mentioned both during the
group interview and in the survey. The response was significant at p ≤ 0.01. Even though young
farmers from the SCRP catchment did not have a chance to participate, older farmers partici-
pated in the food-for-work bund construction programme organised by the SCRP to treat one
of the catchments. Even then, the need for technical support is often mentioned when farm-
ers are asked why they do not construct bunds in their fields.
Lack of government aid in the form of food and tools was repeatedly mentioned in the
series of group interviews. After I learned the general opinion of the community on this issue,
I carried out a short dialogue on this point with farmers (see Box 6.1).
The other common practice of soil and water conservation in the area is building diversion
ditches. However, there is no common effort to control flooding at the catchment level.
Everyone tries to divert the flood from his/her land as much as s/he can, though without
much success. Frequently mentioned problems in this respect is the lack of co-operation
among community members (see also Alemayehu, et al., 2001).

Box 6.1: Farmers’ rationale about bund construction and maintenance

Researcher: Why do you have to wait for food aid when the bund will benefit you and only you?

Farmers: When we constructed the bund in the first place, it was only with a view of getting the
food and tools. We did not appreciate the bund very much at that time and only recognised
its advantages after a long time.

Researcher: I really do not understand why you have decided to leave your land without a bund
or maintenance for 17 years, while waiting for government aid that is unlikely to come.

Farmers: In the past, those who were in the station were following it up like owners. So we were
expecting food from them. After they left, we became careless too. If we get a person who will
help us with food and tools, we will be ready again.

Researcher: Why do you not use the local labour groups to construct the bunds like you do for
other farm operations?

Farmers: We lack food, which means energy, so we cannot feed the large labour party required
to construct the bund. That is why the government gave us food in the past, and that was how
we could construct the bund in one season. The government did it before, and they can do it
again. The government can do any thing on this earth.

Researcher: How long do you have to wait if there is no food-for-work for this purpose?

Farmers: We know that we are losing, but there would not be any improvement on our side
either. Finally, we need government support too, though we can also accept the government’s
ideas.

In view of this, I raised the issue in group discussions of why labour parties are not used for
diversion ditches and soil bund construction. The tradition of using labour parties for such
tedious jobs was discontinued a long time ago. In addition, labour exchange on these tasks is

154
.tesfaye.thesis 29-01-2003 15:12 Pagina 155

not possible because of the variation in the degree to which the problem prevails among farm-
ers; i.e., some farmers have erosion problems whereas others do not. Use of community labour
for such tasks labelled as tedious came to cease after widespread use of oxen instead of gesso
(early 1970s). It was also mentioned that the erratic nature of rain also pushed farmers to
focus on the ox-plough. The use of gesso for land preparation on steep slopes used to be a
motive for organising communal labour parties.
The way farmers perceive the gradient of their farm and thereby the degree of its suscepti-
bility to erosion affects their decision to construct the soil bunds. A large percentage of farm-
ers responded that they did not construct soil bunds because they perceive their land as being
flat (Table 6.12b). This response came from farmers in all four locations. However, there is a
higher percentage of farmers from the untreated catchment than from the other locations, at
least for the first four plots. The differences among the first five plots across locations were
significant (at p ≤ 0.001) for P1 and P5, (at p ≤ 0.01) for P2 and 4 and (p ≤ 0.05) for P3. The way
farmers perceive the nature of their farm in relation to the erosion processes is an important
issue to promote relevant erosion control measures. Perhaps the land that farmers think to be
less susceptible to erosion hazards may from the technical point of view prove to be a more
susceptible plot. One can also encounter farmers holding different views on exposure to ero-
sion in seemingly similar land features.

Table 6.12b: Farmers’ reasons for not constructing and maintaining soil bunds, Wolaita, Ethiopia

Reasons Location P1 P2 P3 P4 P5 P6

My plot is flat Treated (%) 26 22 22 12 2 2

Untreated (%) 57 50 44 33 22 7
Adjacent (%) 54 43 43 32 29 7
Distant (%) 24 26 24 13 4 2
χ2 20.906 13.185 10.098 12.666 26.311 3.906
(p value) (0.000) *** (0.004) ** (0.018) * (0.005) ** (0.000) *** (0.272) a

Steep slope Treated (%) - - - - - -

Untreated (%) - - - - - -
Adjacent (%) - - - - - -
Distant (%) 4 5 5 2 - -
χ2 2.079 3.123 3.123 0.581 - -
(p value) (0.556) a (0.373) a (0.373) a (0.901) a

Note: *, ** and *** show significance at p ≤ 0.05, 0.01 and 0.001 respectively, non-significant otherwise (χ2 test). The letter ‘a’
shows adjusted chi-square and p-values.

A high gradient was identified as one of the constraints to the construction of bunds in the
distant location (Table 6.12b) where forested land on Damota Mountain was converted to
farmland. In spite of that, there was no adequate conservation support in place to help those
who carry out risky farming. See Plate 6.3 for changing forest landscape.

155
.tesfaye.thesis 29-01-2003 15:12 Pagina 156

Plate 6.3: The view of Waja catchment in the Damota Mountain, Wolaita

6.6 Conclusions

Farmers in Wolaita have a fairly adequate knowledge and understanding of soil erosion, its
association with soil fertility (processes) and land productivity. They practise a number of tra-
ditional soil and water conservation (SWC) techniques that heavily depend on organic matter
cycling and other biological measures. Farmers clearly indicate that manure is their insur-
ance for soil fertility. Thus, access to livestock affects soil and water conservation since
manure is the most important factor that influences soil fertility. Farmers state that there is
an ongoing decline in soil fertility with a subsequent decline in productivity. They claim that
this is due to: (1) an increase in erosion; (2) a decline in rainfall; (3) a decrease in landholding
size that restrains fallowing; and (4) a shortage of manure due to a decrease in grazing land
and low biomass in crop production that provides feed.
Soil bunds that were introduced to conserve both soil and water in an experimental catch-
ment in the early 1980s, have shown clear physical water and nutrient conservation effects.
Farmers confirm that using soil bunds reduces fertiliser consumption and ensures prolonged
soil fertility. The few farmers who reported plots with increased soil fertility largely came
from the treated catchment.
In spite of this perception farmers are reluctant to utilise soil bunds in their land-manage-
ment practices. This reluctance already became clear from the very beginning of the SWC
intervention in 1980s. Since then, there has been poor acceptance of soil bunds. Farmers who
constructed bunds 15 years ago (through the food-for-work programme) do not maintain their
bunds anymore. Even worse, more than half of the farmers in the experimental watershed
have removed some bunds in the course of the years. A similar trend was observed in other
catchments where soil bunds were introduced through food-for-work. As regards soil bund
management, the proportion of farmers who preserved their bunds was highest among
farmers from the untreated catchments, followed by those from locations distant to the exper-
imental watershed as opposed to those from the treated catchment and farmers adjacent to
the research centre.
There was limited association between farmers’ land management behaviour and charac-

156
.tesfaye.thesis 29-01-2003 15:12 Pagina 157

teristics of households. It was observed that removal or maintenance of soil bunds takes place
irrespective of age, education or livestock size. Cropland area is the only socio-economic vari-
able that is fairly associated with bund management among common variables tested as
explanations for land-management decisions. In addition, land tenure that is often linked
with farmers’ decisions on soil and water conservation practices hardly affected Wolaita’s
farmers’ behaviour in soil bund management. However, the lack of secure land tenure plays
its role in open access land in Damota Mountain. Even then, these farmers practise diversion
ditches.
In view of the above the question arises why farmers do not use and even oppose the use of
soil bunds. There are four explanations: (1) lack of participation; (2) the wrong use of the food-
for-work incentive; (3) lack of attention to side effects; and (4) changing attitudes to labour co-
operation.

Lack of participation
The soil bund was introduced with the researchers’ understanding of erosion, whereas the
farmers were mainly interested in food-for-work. The project staff that guided the introduc-
tion took the food-for-work as a panacea to enforce a breakthrough in conservation behaviour
in the area but failed to involve farmers in the research and development processes. The
project phased out without active farmers’ participation, whereas their farms and the entire
watershed became a field laboratory for scientists. What is more, in the interest of these exper-
iments, farmers in the ‘untreated’ catchment had to destroy the previous bunds under legal
pressure.

The wrong use of food-for-work incentive

The incentive of ‘food-for-work’ was not properly integrated into educational processes. At the
end of the project it appeared that, in the absence of food-for-work, farmers were not willing
to carry out either periodic maintenance, nor construction of new bunds. It is well possible
that this food-for-work tragedy still has an impact on farmers’ contacts with outsiders. In this
respect, it is possible that farmers’ exaggerate the seriousness of erosion and the decline of
fertility and productivity, and also exaggerate about their confidence in the benefits of the
use of SWC technology in the hope that a new project will bring incentives like food-for-work
again. The nature of this study in Wolaita does not permit the independent verification of the
mentioned seriousness of erosion and fertility and productivity decline.

Lack of attention to side effects

Farmers mention a number of side effects of the soil bunds. These are: (1) a waste of land; (2)
harbouring of mole-rats; and (3) harbouring of weeds. In the eyes of the farmers, the soil bund
has increased the risk of crop losses. The research-extension system did not make any attempt
to show the value of the technology by giving farmers more information on the pros and cons
of farming with and without soil bunds that might change farmers’ perception of the side
effects. Their major preoccupation was the collection of massive experimental data. Farmers
were not able to find viable solutions on their own. So, apparently, the observed benefits from
soil bunds failed to outweigh these side effects. Owing to the negative perception of the tech-
nology that continues to date, farmers’ motivation towards adopting and maintaining it
increasingly declined. This not only holds true in the experimental watershed but also in adja-
cent catchments and even in the steeper farming areas at distant locations. Note that farmers
emphasise the side effects because of the shortage of land that compounds their responses in
addition to their heavy use of organic fertiliser in deep soils.

157
.tesfaye.thesis 29-01-2003 15:12 Pagina 158

Changing attitude to labour co-operation

Termination of labour co-operation for heavy land-management practices such as the con-
struction of soil bunds affected farmers’ perception on labour requirements. This termination
may be considered as an evolutionary process after the expansion of the ox-plough. Lack of
labour co-operation coupled with a lack of motivation led to less or a non-allocation of labour
for soil bunds. However, the shortage of labour that is often given as an explicit reason for not
maintaining soil bunds is not adequate in view of the relatively high proportion of economi-
cally active labour force per household, small-sized farm land, poor off-farm employment
opportunities and minimum migration.
This study has shown that farmers in Wolaita are knowledgeable of soil erosion, its impact
on crop productivity and the conservation potential of soil bunds. However, the observed ben-
efits from soil bunds failed to outweigh the side effects so that there is no adoption of soil
bunds in the absence of a food-for-work programme. Lack of participation and timely recog-
nition of and combating side effects worsened or even may have created this situation.
What needs to be done in the first place is a verification of the mentioned seriousness of
erosion, fertility and productivity decline. Essential is that this is done within a participatory
learning process. Heavy dependence of the farming system on organic sources of soil fertility
management may mask the importance of physical soil and water conservation options such
as soil bunds. After agreement on the seriousness is achieved, solutions for the side effects
should be developed using farmers’ experimentation. Finally, the aim of soil and water con-
servation should be sought in an overall approach to increase productivity. A shortage of grass
for animal feed (and hence manure), the lack of access to credit other than inputs and a high
fertiliser price will not make the strategy for encouraging SWC practices easy.
However, it is the lack of information on how much farmers lose or gain under their tradi-
tional and alternative practices that is the major challenge to be faced if sustainable land
management is to be promoted in Wolaita.

158
.tesfaye.thesis 29-01-2003 15:12 Pagina 159

7 Farmers’ Response to External Intervention in Land

Management, Maybar, South Wello, Ethiopia

Abstract

Intervention by the former Soil Conservation Research Project (SCRP) in Maybar, South Wello was studied
using both qualitative and quantitative methods. The intervention focused on soil and water conservation
(SWC) practices, which later included irrigated agriculture. The SWC practices consisted of physical con-
servation, area closure and biological conservation. Farmers in the area have a fairly adequate knowledge
about their resource base and land-management practices. Their traditional farming system is charac-
terised by a long fallow-based agriculture in flat and gently sloping areas whereas staggered structures
are used in the steep uplands. Farmers first reacted negatively to the stone bunds introduced by the pro-
ject. As time passed and soil fertility continued to deteriorate, they began to modify the original design,
maintaining bunds that resulted from an innovative bund-management practice, ‘moving bunds’. Their
responses to area closure, irrigation and biological conservation failed because of social, technical and
institutional reasons. The study concluded that an improvement of the agriculture in the area requires
diversification and intensification on one hand and in the long run minimisation of land pressure, on the
other.

7.1 Introduction

Maybar is located in South Wello, about 14 kilometres southeast of Dessie town. The first rural
intervention in the area was the ‘development through co-operation’ campaign known as
Zemcha, that was carried out by university and high school students in the early 1970s. A tar-
geted intervention took place after the largest famine ever in 1973, commonly known as ‘the
famine of Wello’. The intervention was carried out in the form of food-for-work. The purpose

159
.tesfaye.thesis 29-01-2003 15:12 Pagina 160

was to reverse land degradation, and to achieve food self-sufficiency. The major contribution
of` the food-for-work programme was to support the construction of soil and water conserva-
tion structures on arable land and hillsides, including afforestation or reforestation initia-
tives. This intervention was funded by the World Food Programme and implemented by the
Ministry of Agriculture.
The external intervention at Maybar was intensified with the establishment of an experi-
mental site by the Soil Conservation Research Project (SCRP) in 1981 that promoted soil and
water conservation based on a focused research study. The major interventions carried out by
the SCRP were: (1) physical conservation; (2) area closure (prohibition of both animal and
human interference in the hills for a period of 3-5 years); (3) irrigated agriculture; and (4) bio-
logical conservation measures.
The overall aim of these interventions was to achieve a sustainable land management sys-
tem in the Maybar area (Weigel, 1986b:48). The interventions were implemented as typical
watershed management practices. To that end, soil and water conservation activities began in
1983 and reached their peak in 1987. The SCRP carried out all of its conservation work in col-
laboration with the MoA. However, by the end of the 1990s, most of the intensive conservation
measures had been destroyed, and the region hardly differed from the other areas in the
north where such intensive programmes had not taken place.
The purpose of this study is to understand how farmers make decisions about soil and
water conservation (SWC). The practices introduced by the SCRP at Maybar are used as the
starting point for analysing how farmers responded to interventions and to draw lessons for
future interventions in similar situations, and to identify an entry point for supporting the
farmers in the Maybar area further. These inquiries require a deeper understanding of the fol-
lowing issues: What is the farmers’ understanding of soil erosion processes? What are the
farmers’ perceptions on soil and water conservation as a land management practice? What do
they expect from soil and water conservation? What do they do to conserve their soil from ero-
sion? What are their responses to and reasoning of each intervention?
Section 7.2 provides information on the location and population, historical background,
land tenure, local institutions and the physical setting of the area. Background on the Soil
Conservation Research Unit at Maybar is presented in section 7.3 where a description of the
technical interventions is discussed. Section 7.4 discusses the methodology of this case study.
Sections 7.5 and 7.6 present results and discussion, and the conclusion, respectively.

7.2 The study area

7.2.1 Location and population

Administratively, Maybar is found in Dessie Zuria Wereda of the South Wello zone (see map 1
in chapter 2). It is located about 14 kilometres southwest of Dessie town, which is the zonal
capital. The area is named after a crater-like lake called Maybar. The lake covers an approxi-
mate area of 1.5 km2 (Ketema, 1980:28; cited in Mulugeta, 1988).
The nearest rural town is Tossa Felena, which is located about five kilometres from the
SCRP station. Government institutions such as an upper primary school, health post and the
nearest market are located in this town. There is a primary school near Lake Maybar that ser-
vices the communities in that vicinity.

160
.tesfaye.thesis 29-01-2003 15:12 Pagina 161

The total population of the South Wello zone was 2.12 million in 1994, whereas the popula-
tion of Dessie Zuria Wereda was 201,433. It is a densely populated Wereda, which is almost twice
as dense as most of the Weredas in South Wello. In terms of its distribution, 99% of the popu-
lation lives in rural areas (CSA, 1996).

7.2.2 Historical background

Socio-political history
South Wello is part of the northern political system of the Amhara nationality. The former
Wello province that was divided into north and south Wello since 1989 represents a multiple
human geography and a succession of different nationalities in the course of three millennia.
As a result, Oromos, Falashas and Agaws in addition to the Amharas have inhabited Wello. Each
of these ethnic groups has its own representation and geographic territory that is recognised
in the recent Federal Government of Ethiopia. A muslim population that belongs to the
Amhara nationality inhabits the specific study area, Maybar.
The former Wello province is one of the northern regions where periodic famine and local
power struggle has taken place (McCann, 1987). While famine is not confined to the Wello
province alone, the 1973/74 and 1984/85 famines have substantially affected the population of
the region, which attracted the international donor community more than anywhere else in
the country.
Wello was the main route for repeated expeditions to the north in the 19th and 20th cen-
turies and was one of the most war-prone areas of the country. The peasant army, which did
not have its own supplies for food and shelter often depended heavily on the rural people with
serious repercussions for the properties of the peasantry along the expedition routes (Merid,
1986). As Maybar is located along the old route that joins Shoa, one of the centres of the
Amhara’s hegemony, and Wello, the people of Maybar must have fallen victim to such pres-
sures by the peasant army.

Land tenure
The succession of different nationalities in different parts of Wello has influenced land
tenure. For instance, in the areas where Oromo nationality dominated, communal ownership
of the grazing land and water points used to take place irrespective of kinship. In the areas
where the Amhara institutions dominate, such as in the present study area, variants of land-
tenure systems originated in the north have been institutionalised, where Rist 43 was the main
form of land-tenure system (see Chapter 2). These include: Gebbar Meret 44, Galla Meret 45, GinDibel
Meret 46 and Hudad Meret 47 (Gebre-wold, 1962). After the 1975 land reforms, land belongs to the
state. This same principle was adopted after the change of government in 1991 (Chapter 2).
The Amhara National Regional State has recently issued a proclamation to determine the
administration and use of rural land, in proclamation No.46/2000 (ANRS, 2000). Without

43 Rist is the right to claim a share of land based on kinship to ancestors held in common with other Rist holders (Cohen
and Weintraub, 1975).
44 This is a form of land tenure whereby the farmer (the gebbar) pays asrat (a tenth of his produce) to the government after
paying the principal land tax.
45 This is land given to soldiers for their service to the government or the local authorities. In principle, these soldiers were
not expected to evict the gebbars as long as the gebbars are willing to pay the tenth of their produce, provide labour and
other contributions to the soldier who became his lord (melkegna).
46 Owners of this category of land were obliged to go to war in lieu of paying tax. They also provide transportation of ammu-
nition by pack animals.
47 This category of land mainly caters for the kitchens of the royal camp. These lands are fertile to ensure the supply of food
items to the royal camp to feed the soldiers and the nobility.

161
.tesfaye.thesis 29-01-2003 15:12 Pagina 162

going into the details of the proclamation, few issues can be said in relation to land tenure.
The proclamation serves a good starting point to focus on the rural land problem and it
touches upon a number of operational issues. On the other hand, it bears directly on and
potentially contradicts rights to and duties demanded from those targeted by the proclama-
tion. Suffice it to mention a few articles along these lines: Article 5, sub-article 1 states that,
‘any person with an age of 18 and above and who lives in the rural part of the region has the
right to freely take possession of the land he/she uses for farming.’ The provider of such land
is an organ designated to execute the land administration. As indicated in article 6, sub-arti-
cle 1, ‘… land is the property of the government and the people. Thus, shall not be subject to
sale or exchange.’ In this scenario one may wonder where does extra land come from to be pro-
vided for every new generation? Would this not worsen the existing unintended malpractice
in farming? This point would be even more pronounced in the face of a substantial number
of land claimants at the moment who are eligible for land distribution. Among several useful
developments in the proclamation, article 6, sub-article 3 ensures the security of holding and
user rights, and article 8 addresses issuance and possessing of ownership book. Article 10
deals with one of the bottlenecks of rural land in Ethiopia planted since 1975. It states that,
‘so long as the act of giving land for free to farmers is maintained, redistribution of land shall
not be effective unless the land distribution does not otherwise affect the productive capaci-
ty, requested by the community supported by the study and decided by law.’
Besides common judicial issues enforced by such official documents, the proclamation
highly seeks ‘environmental stewardship’ from rural people, which is hardly achieved
through rules and regulations. Certainly, practice will teach us all as time unfolds.

7.2.3 Local institutions

An important local institution in the area is Kire. This is a village-based association of which
all family heads are members, to get help in organising funerals as well as for other social
events such as weddings. The role of Kire transcends social support by including arbitration
between husband and wife and all members of the community. Its role was much more cru-
cial before the establishment of the Peasant Associations (PA) after the 1975 land reforms.
Kire was the highest local institution to arbitrate local disputes mostly excluding the loss of
life. If someone files a complaint to the Kire leaders, designated elders scrutinise the case,
which is settled by charging the offender. The fine depends on the level of the offence and the
co-operation of the offender with elders who preside over the matter. If someone denies a case
and there is no witness, the case is finalised with the local ritual system called Bele. The latter
is the symbol of the local shrine in Maybar. Loss of property including farm products, live ani-
mals or other valuable property is settled in the same way. Because of their conviction to the
spirit world, community members confess their wrongdoing and pay the penalty. If someone
fails to pay the penalty he/she would be expelled from the community. The person will how-
ever not be accepted in another community either, because of the common understanding
among the rural communities. Bele is still practised on private property. However, the com-
munity does not practise this to settle the frequent thefts from the community forests.
In addition to Kire, the community also has a local savings group known as equib. This is a
main method of saving from the sale of butter and other small farm products.

162
.tesfaye.thesis 29-01-2003 15:12 Pagina 163

7.2.4 Physical setting of the area

Topography
Maybar falls in the Awash drainage system at an altitude range of 2,500 to 2,800 masl. Two-
thirds of the 9 km2 of the catchment around the SCRP site is characterised by steep to very
steep slopes (25-55% slope), which is a general topographic feature of the eastern escarpment
of Wello. The remaining part of the catchment consists of moderately steep to flat hill bot-
toms (13-25% slope) with a varied proportion with other landforms, including the Lake Maybar
catchment 48, which is the focus of this study (Weigel, 1986c).

Climate
The average annual rainfall at Dessie station, 14 km away from the study area, for the period
of 1962-1999 was 1,150 mm (NMSA, 2000). The mean annual rainfall at the Maybar station was
1,211mm with an erosivity value of 420J/mh (SCRP, 1996c). The highest precipitation occurs in
July and August. Rainfall in the study area follows a bi-modal pattern. The short rainy period
(Belg) occurs in April-May and precedes the main wet season (Kermt), which takes place from
July to September. There is no overall trend in the annual series at Dessie from 1962-1997
(Conway, 2000). The mean annual rainfall on a decade basis shows that the 1990s were the
wettest decade (1,248mm) followed by the 1960s (1,232mm), the 1980s (1,025mm) and the
1970s (1,106mm). The years 1993-1998 were particularly wet.
The major problem of rainfall at Maybar is its distribution rather than the total amount.
The Belg rain is particularly erratic in terms of its onset time and distribution.
The mean annual minimum and maximum air temperatures are 11.5 0C and 21.5 0C respec-
tively (SCRP, 1996c). There is little variation in the temperature of the region based on the cli-
matic conditions of Dessie from 1961-1990 (Conway, 2000). Temperature records at the Maybar
station for the period of 1981 to 1993 also show no annual variation, unlike the monthly tem-
perature that shows some variation from month to month (SCRP, 1996c). The warmest months
are from April to June, while the coolest months are from November to February. The hottest
period occurs between March and May.

Soils
The soils of the Maybar area are affected by relief land use and vegetation (Weigel, 1986c).
More than 50% of the soils of the Maybar station are shallow Lithic Haplic Phaeozems followed
by Lithosols (25%), very shallow Haplic Phaeozems (15%) and some moderately deep to deep
Haplic Phaeozes (10%).

Land use, farming patterns and crops grown

Agriculture in Maybar is rainfed, with a mixed farming system dominated by cereals.
According to Westphal’s classification system (1975), the area is in the grain plough complex
farming system (see Chapter 2). Typical land use of the area shows crop production on flat and
gently sloping uplands and grazing on some parts of the uplands and in swampy flat areas.
There is a traditional practice of hay production during the dry period, mainly for dairy ani-
mals and oxen. Natural vegetation is found on the uplands and in the gorges. The dominant
tree species in the uplands and in the gorges are Juniperus porocera, and Eucalyptus. There are
a few stands of Olia africana and Podocarpus gracileor. Trees are not permitted to grow in and
around the main crop areas. They are regarded as a safe harbour for birds and rodents that
attack crops. So, an ideal field is considered to be a clean plot of land without trees. In addi-

48 In the rest of this chapter, the name Maybar is used instead of Lake Maybar catchment.

163
.tesfaye.thesis 29-01-2003 15:12 Pagina 164

tion, fencing is not established between individual farms. Live fences are used along the path
to protect the field from stray animals.
Farmlands are divided into the homestead plot (Guaro) and outer fields (Ersha). While Guaro
is often adjacent to the house, Ersha is located and often scattered in the neighbourhood or
distant from the village. This pattern of land ownership was inherited from the land tenure
system rooted in the region that provides land to ‘descents’ 49 (Cohen and Weintraub, 1975).
Ownership of land at different elevations helps farmers to grow crops for different purposes
according to the suitability of the land to different crops over the various seasons.
Villages are either located at the upper limits of potentially cultivable land or on the divid-
ing ridges, overlooking the entire area, including the lake.
The main crops grown in the area are maize, teff, barley, emmer wheat, peas, lentils, horse
bean, and haricot bean. Among these crops, lentil, peas and emmer wheat can be grown on
steep slopes during the Belg to supplement production from the main cropland. Only very
short seasonal fallow periods are observed.
Socio-economic variables related to land management practices are discussed in section
7.5.1.

7.3 The Soil Conservation Research Unit at Maybar

7.3.1 Inception and termination

The Maybar research station has been functional since 1981. It is one of six such research sta-
tions that were established by SCRP in Ethiopia co-financed by the Swiss and Ethiopian gov-
ernments. The site was identified by one of the senior SCRP management team members. The
land on which the station premises were built was obtained from a farmer with compensa-
tion. The project has attracted the farmers by buying construction materials from them and
providing wage labour. The project is remembered by many for its food-for-work component,
which saved many lives during the 1984 drought and the employment opportunity for some
during the early days. The Maybar station was handed over to the Amhara Regional National
State after the mid 1990s.

7.3.2 The Maybar research site

The Maybar research station is located at 39 0 39’E and 10 0 59’N and 2,500 masl. The research
station covers the Kori catchment that covers an area of 116 ha. It is one of the small catch-
ments of Lake Maybar, which drains to the Awash drainage system, via the Borkena River. The
project designated the areas outside the Kori catchment as ‘outside’ the catchment.
According to Weigel (1986c), in 1982 there were 35 households within the Kori catchment
and 112 households outside the catchment. The population density in the catchment of 1.16
km2 was 138 inhabitants per square kilometre. The average family size was 4.2 persons. In
1988, the population density at the Kori catchment was 220 persons/km2 and 300 TLU/ km2 for
livestock (Mulugeta, 1988).

49 These are members of a lineage who are entitled to a land unit, through either parents, from elders controlling the allo-
cation.

164
.tesfaye.thesis 29-01-2003 15:12 Pagina 165

7.3.3 Description of technical interventions

The intervention at Maybar, unlike many other places presents a unique integration of land
management practices intensively supported by research and extension activities. In this sec-
tion, four major interventions are discussed. These are physical conservation, area closure,
irrigated agriculture and biological conservation. How farmers responded to these interven-
tions is revisited in section 7.5.3 after assessing farmers’ knowledge and perception of the bio-
physical environment and their understanding of the erosion processes.

Physical conservation
The SCRP, in collaboration with MoA, began treating the Maybar catchment with stone con-
tour bunds in 1983. Later on, farmers modified this bund by mixing it with soil. At the begin-
ning of the 1987, all the crop areas were treated with bunds. The work was fully supported by
food-for-work. At the beginning, farmers were not interested in the food-for-work programme
itself, but participation increased with a shortage of food in the area. The project technicians,
however, were making efforts to convince the farmers about conservation. The Development
Agents of the MoA who work with technicians of the project, reinforced the construction of
the stone bunds in the area. The construction of the stone bunds resulted in a lower width of
the field between the bunds than was the farmers’ practice. While the extent to which stone
bunds were constructed in the entire watershed is very considerable, in the small Kori catch-
ment alone, 54 km of stone-faced contour bunds were constructed between March and July
1983 (Hurni, 1984; cited in Kassaye, 1997:59).
Farmers’ education on soil and water conservation emphasised the value of the permanent
bunds. These bunds should be maintained every year in order for the height to increase until
it becomes a bench terrace. The width at the top of the stone bund was technically determined
to be 50 cm.

Area closure
Prior to the 1975 land reforms, the hills were held by individuals and were used for upland
crop production and forests. In those days, the land size per family was much bigger. Hillsides
were not exposed to erosion. For instance, there were only 11 households in Abo Ager village,
which has now grown to about 60. Due to private ownership of land, no one was cut trees from
another person’s woodlot. In addition, the local institutions and the rules were in full force.
Following the land reforms, landless people began to get access to land. New families were
in need of wood for construction. The source for all wood needs were the old private wood-
lands, which were said to be landlords’ wood. As a result, the large indigenous trees were
destroyed. In response, the military government introduced ‘community forests’ planted
through mass mobilisation.
In addition, the SCRP introduced area closures to achieve soil and water conservation at the
watershed level. As a result, the steep lands that were not suitable for agriculture were sealed
off. The closures remained in force for 3-5 years until 80% of the natural grass cover had
regrown (Hurni, 1986; cited in Kassaye, 1997). In the process, an ecological plan for Maybar was
prepared to achieve a sustainable land use pattern (Weigel, 1986b). Conservation of the upper
catchment began in 1983 according to the SCRP’s original plan. In a short while the work
extended to the entire Maybar catchment and was completed in 1987 (see below). All farmers
who were covered by the enlarged conservation plan used to belong to one PA (Maybar), which
was later merged with two other adjacent PAs.

165
.tesfaye.thesis 29-01-2003 15:12 Pagina 166

The drought of 1984 brought a new situation in the area. Many families have lost their valu-
able assets. As a result, the central government decided to resettle the drought-stricken people
into sparsely populated areas of the south, west and southwest of the country. A number of
families were sent to the resettlement, while 174 households remained at the three villages of
Maybar namely, Abo Ager, Aba Lucho Ager and Jerjero. Though the resettlement was not envis-
aged when the project started, the SCRP took this opportunity to implement the agro-ecologi-
cal plan it had developed (Weigel, 1986b).
In order to implement the plan, land redistribution was carried out which involved 174
households from three villages. Individuals who did not have land in the irrigable area
received land through the redistribution. The latter was decided in the presence of the
government representative from Dessie Zuria Wereda and the Peasant Association. The land dis-
tribution involved three types of lands, viz. irrigated plot, uplands and grass field. Had it not
been for the resettlement, the land redistribution would not have taken place in the project
area alone.
The SCRP has trained farmers in area closure and zero-grazing techniques, afforestation,
and introduced irrigation (see below). Farmers were allowed to cut and carry the grasses rather
than grazing the animals in the sealed off areas. Six individuals were recruited from the PA to
guard the enclosed area. These individuals were paid in food grain for their services.

Irrigated Agriculture
Irrigated agriculture at Maybar was initiated by the SCRP because of the 1984 drought. When
it began, it was believed that it would lead to more intensive agriculture that would help min-
imise the pressure on the land, while improving the livelihood of the people. The source of
water for irrigation was Lake Maybar. The irrigable area was 27 ha. The irrigation system com-
bines a motorised pump and surface irrigation using gravity. Therefore, the project imported
a 60 horsepower pump and a capacity of 50 ls -1. The pump was donated by a Swiss relief organ-
isation. After the pump arrived at the site, the project discussed the matter with the Wereda
administration. The latter used the provision of the irrigation pump as an incentive to carry
out community work in the same catchment, which was consistent with the project’s inten-
tions. The kinds of work carried out were digging an irrigation canal, hillside terraces, rural
road construction, and reforestation to protect siltation of the irrigation canals.
Before the project introduced irrigated agriculture, there was no practice of using the water
from Lake Maybar for irrigation nor was there a means because the lake is at the bottom of the
major cultivated area (see cover photo). However, very few farmers were growing onions using
small streams from the marshy areas scattered in the crop area. This situation has been report-
ed to expand slightly after the experience of the 1985 drought. What is surprising is that farm-
ers on either side of Maybar catchment use small-scale irrigation. Farmers from Albore area
located in the eastern part of the lake irrigate from the outlet of the water. Additional irriga-
tion channels were under construction, along very steep hills in another village.
In 1985, all farmers in the catchment have grown vegetables such as onions, potatoes, and
carrots using irrigation. In addition, they have used the water to irrigate small cereals. The
seeds for the vegetables were provided by the project and the MoA. Use of irrigation in the area
has created unprecedented enthusiasm among the community. Potatoes became abundant.
People shared the production of vegetables, specially potatoes according to the tradition
whereby sharing and gifts are common among neighbours, friends and kinship. With the
onset of the following season’s rain, everyone returned to their rainfed practices, thus ending
this enthusiasm. As a result, the heavy-duty irrigation pump has been standing unused for over
15 years.

166
.tesfaye.thesis 29-01-2003 15:12 Pagina 167

The Maybar PA has been paying the monthly salary of the guards who watch over the pump
since the third year of the project. I shall come back later to the reasons for the termination
of the irrigation project.

Biological Conservation
The SCRP’s biological conservation programme at first focused on area closure, reforestation
and grass cover on the stone bunds. In the late 1980s, experiments were carried out at the
Maybar station. The experiment focused on alternative planting methods of maize, teff and
horsebeans. The main purpose of the experiment was to test the effects of biological conser-
vation techniques on erosion control and yield improvement. Among the alternative planting
methods experimented with Vetch (Vicia dasycarpa) inter-row planting with maize, without
ridges, showed a very large increase in supplementary fodder, and substantial reduction in
runoff and soil loss compared to the farmers’ practices. In addition, observations were made
on different forage plants such as alfalfa, phalaris, rhodes grass, clover, pigeon pea and tree
lucern (Kassaye, 1997).

7.3.4 Runoff and soil loss results

Soil erosion experiments at Maybar were carried out in the Kori catchment. The terms runoff
and soil loss were used to denote the overland flow and soil particles leaving the test plots
respectively. Corresponding parameters for the catchment level were river discharge and
sediment yield. The main features of the test plots and erosion results are given in Table 7.1.

Table 7.1: Average results of erosion experiments (1982-93), Maybar, South Wello, Ethiopia

Particulars On-site effects (plots) 50 Off-site effects

Plot 1 Plot 2 Plot 3 Plot 4 Plot 5 Plot 6 (Catchment)

Slope (%) 16 64 43 37 64 64 various

Land-use type crop grass grass crop grass grass various
Runoff (mm/y) 191 55 90 103 111 211 323
Erosion (t ha -1y -1) 36 16 21 32 4 48 13

Source: SCRP Data Base (SCRP, 1996c).

Soil erosion at the Kori catchment was highest between the months of March and August, cor-
responding with the period of highest erosivity (SCRP, 1996c). The mean runoff on the culti-
vated test plots was 103-191 mm/y while the resulting mean soil loss was 32-36 t ha -1y -1 (Table
7.1). Runoff at the plot level was rated as low to very low 51. The soil loss was moderate. Maize
and horse bean were the dominant crops for plot 1 and plot 4, respectively.
Except for a variation in size, plots 2, 5 and 6 were located at the same location. Higher
runoff and soil loss from plot 6 as compared to 5 were considered due to some errors (SCRP
1996c). It was also noted that ‘the presence of limited transport conditions at Maybar lead to
similar material loss within a length of 15 m’.

50 Note that the area of the test plots (1-4) is 30 m2 (2x15) and the area of the micro plots (5 and 6) is 3 m2 (1x3).
51 These comparisons are based on the ranking among the SCRP station (SCRP, 1996c). For runoff: very low is < 200; very
high is > 500. For soil loss: very low is < 10; very high is > 100. For river discharge: very low is < 100; very high is > 500. For
Sediment yield: very low is < 1; very high is > 20.

167
.tesfaye.thesis 29-01-2003 15:12 Pagina 168

Soil and water conservation experiments involving both graded and levelled structures
were carried out on an area of 180 m2 (6x30). The methods were grass-strip, graded and level
Fanya Juu, and graded and leveled bund. These five methods were compared with local prac-
tices as a control. The result of the experiment from 1986 to 1989 showed that soil loss from
runoff was considerably reduced by all measures. Production of all the treatment plots were,
however, lower than the local practice (control). Slight improvements, but still lower than the
local practices, were observed under dry periods due to moisture conservation.
Based on this experiment, sustainable land management practices in the area were recom-
mended. These consisted of: steep slope cover by forest and grass, cultivation of gentle slopes
with adaptable soil and water conservation that fits the moisture regime of the area; i.e.,
excess water drainage during heavy rainfall and conserve the moisture during shortage of
rain.
The river discharge was high in the months of August and September, whereas the high sed-
iment yields occurred in the months of March and April due to less crop cover. The mean
annual river discharge was 323 mm/y while the mean annual sediment yield was 13 t ha -1y -1
(Table 7.1). Both the river discharges and the sediment yield were considered as high (see foot-
note 51). ‘The annual rate of sediment yield seems to have reached a lower level as of 1986,
with the exception of two years (1988 and 1992) during which river discharge was high’ (SCRP,
2000b:56). These changes were attributed to the shift of the closed off area in the catchment
to grazing and farming. The pattern of the annual river discharge between 1982 to 1992
requires further explanations by analysing the catchment dynamics 52.

7.4 Methodology of my study

7.4.1 Unit of analysis

The area of study is the lake Maybar catchment that includes the SCRP research site (Kori
catchment). The unit of observation was the catchment itself, whereas the units of analysis
were the households, household members, farm plots and livestock.

7.4.2 Sampling

A list of all farmers who own land in the Maybar catchment was prepared by the SCRP tech-
nicians who worked with the community for over 15 years. This work took four days. The farm-
ers included in the study are from three Peasant Associations, viz. Maybar ena Defati (83%),
Abawid Ager (14%) and Wilde Ager (3%). The total number of farmers in the Maybar catchment
was 182. In order to capture the distribution of the land management practices and the asso-
ciated reasons, 120 farmers were planned for the interview. Due to logistics and rainfall, a
total of 107 farmers were interviewed. The selection of farmers was carried out using system-
atic random sampling. Of the total, 100 (94%) were male, whereas 7 (6%) were female.

52 For instance, the river discharge shows 406 mm in 1983 and suddenly drops down to 176 mm in the following year, and
rises to 303 mm and 439 mm in 1985 and 1986 respectively. Interestingly, the 1987 figure drops to 145. More or less simi-
lar trends appear in the remaining years, though there is incomplete data. From this, one may say that the general pat-
tern is two years of a relatively high river discharge followed by a relatively small discharge. Why?

168
.tesfaye.thesis 29-01-2003 15:12 Pagina 169

7.4.3 Methods of data collection

The data collected aims to assess the distribution of farmers’ responses in the Lake Maybar
catchment 53.
The data for this case study were collected from both primary and secondary sources. The
primary data sources include informal individual interviews, group interviews, observations,
and measurements of bunds and plots and a questionnaire survey. The secondary data were
collected from published and unpublished sources.
The survey was carried out after one and a half years of qualitative data collection on the
specific study area and zonal level. During the qualitative phase, in-depth interviews com-
bined with plot measurements were conducted with a group of 30 farmers. These farmers
were selected on the basis of plot observation rather than the usual practice of selecting farm-
ers from a residence list. This was carried out by repeated transect walks in the crop areas and
across the villages to find out variations in the observable land management practices. The
variations found were: presence or absence of bunds, height of the bunds, removal of bunds,
movement of bunds and maintenance of bunds. Interviews were held with farmers owning
the specific plots. Views from individual interviews were discussed in groups. Thereafter, the
information from the qualitative phase was thoroughly reviewed to finalise the survey ques-
tionnaire.

7.4.4 Methods of data analysis

The analysis of the qualitative data was enhanced by an exploratory approach that guided the
fieldwork to progressively uncover farmers’ strategies and reasons. The relevant qualitative
information was integrated with the quantitative data for better understanding of the issues
covered in the study.
The survey data were entered into SPSS (version 10.0.7). The data were thoroughly checked
by the researcher before the analysis by directly comparing all 107 cases with the original
questionnaire. This procedure took a considerable time. Due to the nature of the research
questions and the design of the sample (one sample), the analysis of the survey data is limit-
ed to descriptive statistics and measures of association.

7.5 Results and discussion

7.5.1 Socio-economic situation of farmers in Wello

Household characteristics
The majority of the farmers interviewed had married (94%), except for one farmer. Among the
remainder, five were widowers and one was separated. The average family size was 5.5 persons
(Table 7.2). About 20% of the families have the average size, whereas about 30% fall below the
average size. The remaining 50% exceed the average, with a maximum of 13 persons for few
families.
The average age of the sampled farmers was 53, with a minimum and maximum age of 18
and 78 respectively. Relatively high average age is due to shortage of land for young farmers

53 Note that this study treated the entire Lake Maybar’s watershed rather than the SCRP’s Kori catchment which many stud-
ies also refer to as Maybar catchment.

169
.tesfaye.thesis 29-01-2003 15:12 Pagina 170

who have to depend on their parents for land. Therefore, they did not qualify for an interview,
as they do not own an independent farm.
Table 7.2 shows major household characteristics that are related to farmers’ land-manage-
ment decisions. The discussion in the remaining part of this section is partly based on this
table. Note that percentages used in some parts of the discussion may not add to 100% when
the interest is on the specific response rather than the distribution of the sample.

Table 7.2: Household characteristics, and access to resources, Maybar, South Wello, Ethiopia

Characteristics Values Characteristics Values

Family size (mean) 5.5 Cropland (mean in ha) 0.78

Age of household head (mean) 53 Grassland (mean in ha) 0.20
Dependent family members (mean) 2.4 Forested land (mean in ha) 0.15
Economically active members (mean) 3.1 Total land (mean in ha) 54 1.01
Migration Livestock owned (mean) 55 8.0
Temporary (%) 35 Use of fertiliser 1998 (mean kg) 5.2
Permanent (%) 15 Use of fertiliser 1999 (mean kg) 1.4
Education of household head Visits by extension agents
Illiterate (%) 42 Yes (%) 45
Basic education (%) 56 36 No (%) 55
Grade 1-8 (%) 17 Access to credit
Grade 9-12 (%) 5 Yes (%) 27
No (%) 73

Economically dependent age groups 57 (children 0-14 and elderly above 64) range between zero
to three, with an average of 2.4 persons. Of the total, 11% of the families did not have depen-
dents. In about 58% of the families, the number of dependents ranges between one and three.
In about 20% of the families, this number increased to 4 persons. Those with more than four
dependent family members were 12 %. Most members of the dependent age group are chil-
dren below 15 years of age. The number of economically active persons ranges from one, in
female-headed and single male households, to 8 in male-headed households. Of the total, 89%
of the farmers have one to four economically active family members, of which one-half are the
parents. The average of the economically active persons is 3.1. The dependency ratio varies
from nil to three persons per family. About half (54%) have less than one dependent, while
about one-third (29%) have more than one. The remaining 17% have just one dependent.
The livelihood of farmers in Maybar depends on farming. With the exception of one farmer,
who is sometimes involved in petty trading, all other farmers in the sample depend on
farming only for their livelihood. There are hardly any off-farm and non-farm employment
opportunities in Maybar (see also Ludi, 2002). Out-migration is 35% and 15% for temporary and
permanent migrants, respectively. Male temporary migrants total 87%, while all of the per-
manent migrants are male. Remittance income is obtained by only a few individuals (6%).

54 The total land figure does not include the communal land to which all farmers’ households have access.
55 This figure excludes chickens.
56 In Ethiopia, basic education refers to the literacy programme whereas grades 1-8 refer to formal elementary education.
57 Note that in Ethiopia, children as young as seven years of age and the elderly above 64 years of age actively participate in
agricultural activities.

170
.tesfaye.thesis 29-01-2003 15:12 Pagina 171

Most of the farmers in Maybar area are illiterate (42%), followed by those who have followed
basic education and religious school (36%). The remaining 22% have had formal education
with a maximum of grade 9. Most of these individuals quit school a long time ago and have
been farming for many years now. Formal education of those with lower grades is undermined
by lack of access to reading materials and to mass media. The number of farmers who own a
radio is 15% of the total (N=107).
The Muslim religious group, which constitutes of 98% of the sampled farmers dominates
the community of Maybar. The remaining 2% are Catholics. The regular religious holidays are
Wednesdays and Fridays. Farm operations that are not carried out during these days are
ploughing, weeding, cultivation and harvesting of the main crops. However, it is possible to
go to the market.

Land ownership and land fragmentation

The average total land size is 1 ha. The average for cropland is 0.78 ha. The grassland and
forested land cover 0.19 ha and 0.15 ha, respectively (Table 7.2). The total land ranges between
0.25 ha to 2.5 ha per owner. Thirty percent of the sampled farmers own land below the average
size, while 32% own just the average land size. Those who own land in excess of the average
number 38%. Ownership of cropland ranges between 0.13 ha to 1.5ha. Of the total, 10% and
65% of the farmers do not own grassland and forested land, respectively. This is partly because
of a shift of the grass plots to crops that was reported by 22% of the farmers.
The land parcels owned by farmers are scattered in the watershed covering both the flat
and uplands. The majority of the farmers (90%) own 1-4 plots, while the remaining own up to
seven plots in addition to the grass pastures and woodlots when applicable.
Most of the farmers obtained the land from their parents and the land redistribution of the
1970s and 1980s. The other source of land includes periodic redistribution by the Peasant
Associations, purchase and/or lease from other farmers. There was marked instability of land
ownership in the Maybar area since the land reforms of 1975. During this period, 65% of the
farmers experienced a reduction of their land ownership, 8% experienced an increase while
the remaining 27% remained the same. The main reason for the decline of the land size was
redistribution by the state (90%). The other reasons include sharing with children and broth-
ers, and land degradation. Sharing of land with children is becoming difficult for parents due
to the low productivity on the existing land. In spite of that, 51% of the families expect to
share land with their children (males) over the next ten years. The number of children who
expect such an opportunity ranges from one to four per family. In 84% of the families, there
are one or two children for such land distribution. These families are confident to be able to
share from their present land holding within the next ten years, which is indicative of the
farmers’ planning horizon. Very few cases of conflict between parents and children (male)
were reported because of a delayed marriage that is indirectly linked to land inheritance.
Farmers basically know that the land belongs to the state (84%) and they have the right to
use it in return for land tax. They also know that the state has the power to take away the land
for various reasons (e.g., redistribution, school, etc.,). These situations, however, did not
change farmers’ traditional beliefs of what they call ‘my father’s land’. Therefore, there was
no strong indication of land insecurity that deters them from undertaking the common land
management practices, even though some of their actions are not fully free from risk-aversion
strategies (see also Ludi, 2002). However, there was a marked difference between the arable
and non-arable lands, including trees (see section 7.5.3 for further discussion on this issue).
During the survey, farmers were asked whether the present ownership of land affects their
decision on their land–management practices. Of the total, 97% replied that they did not face

171
.tesfaye.thesis 29-01-2003 15:12 Pagina 172

any problem. The group and individual interviews also came up with similar results with the
exception of clarification on their points of view and the implicit uncertainty about the fate
of their children.

Use of external inputs

External inputs were not common in Maybar until the 1990s when the government launched
a package approach. The major external inputs that reached the community through the pro-
gramme were cereal seeds and fertilisers.
In spite of the government package, farmers still depend on their traditional seed sources;
namely, their own seed savings and the local market. The local market is an important seed
source as farmers regularly exchange or sell seed grains in the weekly open market at Tossa
Felana. Farmers who used seed from government sources are very few due to their low level of
participation in the package programme. Those who did obtain seed through the package pro-
gramme did so only once.
More than half of the farmers at Maybar do not use fertilisers at all, against 44% who have
used them at least once. Among these, 25% have used only Urea, 28% only DAP, while 47% used
both Urea and DAP fertilisers. Farmers’ use of fertilisers was assessed for the years 1998 and
1999 crop seasons. Of the total farmers, 32% have used fertiliser in the years 1998 and 1999.
Quantities of fertilisers used by each farmer vary from 2 to 50 kg. Those who used more than
10 kg per season do not exceed 25% of the sampled farmers. The mean of Urea fertiliser used
in the year 1998 and 1999 was 11.2 kg and 8.6 kg respectively. The mean of DAP fertiliser used
in 1998 was 10.5 kg, whereas this figured dropped to 6.3 kg in the following year. The drop is
mainly because of the rising fertiliser price from year to year.

Labour organisation and farm tools

Farmers depend on family labour for their farm operations and social purposes. When the
farm work has to be done in a relatively short period of time or when a lot of work is to be
done, farmers organise a labour party. The other source of labour is assistance from relatives
and friends. A labour party is organised primarily for harvesting (81%), including threshing.
The next important farm operation that requires a labour party is cultivation (hoeing) (58%).
The other farm operations done by a labour party are seeding (39%), weeding (24%), land
preparation (19%) and bund maintenance (18%). Farmers do not prefer a labour party for main-
tenance of bunds because of the skills and patience required for bund maintenance (see sec-
tion 7.5.3). Outside the farm operations, house repair is done by a labour party (50%).
In Maybar area, labour parties are classified according to the number of people involved
and the duration of the work. Debo is the common form of labour party that involves 10-20
people. Debo is organised for all of the farm operations listed above. At the end of the day there
is a feast at the house of the organiser of the party. Wenfel is organised for digging the upland
with hand tools, for cutting grass and for splitting wood for house construction. Women
organise themselves into Wenfel to transport harvested materials and straw to the homestead
areas. The number of people is determined by the volume of work. In terms of duration of
work, Ware is work for half a day, usually, for weeding. Yemewaya is work that would take the
whole day.
A typical farm tool in the grain-plough farming system is the plough that is used for both
land preparation and hoeing. Hand tools with metal tips are used for digging of steep lands
that are not convenient for the ox-plough. However, such practices are very rare, particularly
nowadays.

172
.tesfaye.thesis 29-01-2003 15:12 Pagina 173

Trends in crop production

The average yields of maize, horse bean, barley and teff from 1986 to 1993 were 2.3, 1.9, 1.6
and 1.2 t ha-1 (SCRP, 1996c). According to farmers in Maybar, the trend of crop production has
been on the decline for many years now. In view of this I asked farmers how they judged the
trend of their production during the last ten years. Of the total sampled farmers, 97% report-
ed decline, 3% the same situation. There was no case reporting increased production (see also
Yohannes, 1999; Ludi, 2002).
Food from their own production covers family needs for 5-7 months for many farmers (see
also Zealbowesen, 1998:37). The remaining months are covered by purchasing from the local
market or by food aid in years where crop failure is recognised by the government. The cash
for the purchase of food grains is often obtained by selling livestock, wood or borrowing from
relatives and friends. In years 1999 and 2000, 88% of the sampled farmers have purchased
food. The average number of months during which grain was purchased was 5. Grain sold in
the local market originates from other regions, at times as far away as the western part of the
country through the marketing channels. However, small farmers also sell very limited
amounts to meet other family needs or to exchange for other grains or pulses.
The reasons for the decline in crop production are indicated in Table 7.3. Note that drought
and shortage of rain in Table 7.3 show specific farmers’ observation on how weather affects
their production system. In this context, drought refers to a prolonged absence of rainfall,
whereas shortage of rain refers to its erratic nature whereby crops are affected due to little
rainfall or its interruption after the onset.
Farmers’ ranking shows the importance of climatic factors and soil fertility (see also Ludi,
2002:244). It is interesting how farmers perceive soil erosion, which is least ranked. This is
mainly because of the perception of researchers who often included ‘soil erosion’ in the lists
of response, whereas farmers perceive the problem in terms of soil fertility, rather than soil
erosion. Farmers do not talk about soil erosion unless it occurs in the form of rills, gully ero-
sion and landslides (for farmers’ perception on erosion processes, see 7.5.2). For this reason,
researchers often blame farmers. ‘… In effect, the farmers’ perception of soil erosion is still
low. Let alone perceive erosion as a life threatening problem, they do not even rate it among
such problems as pests, weeds, untimely rain, and hail.’ (Kassaye, 1997: 58-59)

Table 7.3: Rankings of farmers’ crop production problems, Maybar, Ethiopia

Problems Rank

1. Decline of soil fertility 3

2. Pests 5
3. Diseases 6
4. Drought 1
5. Shortage of rain 2
6. Soil erosion 7
7. Shortage of land 4

Farmers take collective measures to minimise their risk of pest and bird attack. Farmers in the
adjacent plots discuss which crop to grow the following season. These discussions are held at
social meeting places, such as the praying time, markets, river points, during ploughing, etc.,
The other critical decision in line with this is the sowing time. This decision is often made by
knowledgeable farmers whom the rest follow, at least in some practices. These farmers have

173
.tesfaye.thesis 29-01-2003 15:12 Pagina 174

their own basis for judgement to make such decisions. Wondering about this situation I fol-
lowed up what farmers would do in the Belg of 2001 at Maybar (Box 7.1).

Box 7.1: An account of indigenous knowledge about soil moisture for sowing

All Belg fields were prepared thoroughly. What was expected was a reliable rainfall to start sow-
ing. There were few showers in the previous days, but no one started sowing in the entire water-
shed. One morning in the same week, I suddenly saw a farmer sowing, assisted by another fel-
low farmer. I joined them immediately and asked the owner of the field why he started to sow
while others did not start yet. He replied that ‘they will start as soon as they see me.’ Explaining
my point, I asked him what prompted him to start the sowing. He just collected the soil from
the ploughed ground and pressed it between his palm and threw it away about 3 meters. Then
he picked up a clod and told me that that unbroken piece was a sign for the presence of enough
moisture to start sowing.

Livestock production system

With the exception of one farmer, all farmers in the sample own livestock. The animals raised
in Maybar are cattle, sheep, goats, donkeys, horses, mules and chickens. Among ruminants,
cattle are the most abundant species, followed by sheep. Of the total, 66% own one cow and
15% own two. Nearly half of the farmers (47%) own one ox and 36% own two. Farmers who do
not own an ox amounted to 17%. The other important animal in the farming system is the
donkey, owned by 41% of the farmers. Ownership of donkeys is crucial for the farmers’ liveli-
hood strategy. Farmers who own a donkey can rent it to someone for a sum of money that
varies with the distance and the type of load or carrying firewood to the market to sell. Mules
serve the same purpose. Adding all species of animals owned, excluding chickens gives an
average of 8 heads of animals per household. When sheep and goats are excluded the figure
falls to 4.8.
Farmers start livestock production through family gifts, purchase, or shared animal rais-
ing. In shared raising, one partner contributes the stock while the other manages it in return
for use of the service and sharing of the offspring. What applies in each case depends on the
initial agreement between the two parties. Among 107 farmers, 37% participated in shared
animal raising between 1999 and 2000. Animals commonly shared in or out are cows, heifers,
sheep, horses and occasionally other animals.
Animal production in the Maybar area is based on open grazing on communal grazing
land, hillsides, marshy areas and aftermath grazing. Hay is used as a supplementary feed in
dry periods and at night. Farmers obtain hay from two sources. These are their own grass plots
and communal grassland. The latter is protected during the wet season for hay production
whereas it is used for open grazing during the dry period. The management of the grasslands
is carried out by the community through village-level committees.
Animal production is deteriorating on the one hand because of the decline in grazing land
and low biomass production, and on the other because of disease (Table 7.4).

174
.tesfaye.thesis 29-01-2003 15:12 Pagina 175

Table 7.4: Farmers’ prioritisation of livestock problems, Maybar, Ethiopia

Livestock problems Rank

Disease 3
Shortage of forage 1
Drought 2

Livestock marketing plays a great role in the dynamics of livestock ownership and manage-
ment in Maybar. The sale of livestock is due to a range of reasons that vary from household to
household, but with considerable commonality. For instance, the purchase of food for the
family, the purchase of family cloth, a shortage of feed, and for house construction are the
most common reasons behind selling different species of animals, though house construction
often requires selling one’s oxen. Other reasons such as the replacement of a mule or donkey,
or young animal for fattening were part of the explanations for sale of animals, to mention
few.

Access to information and credit

The main information source for agricultural development in the community is the extension
service of the Regional State. Soil and water conservation practices were assisted by the SCRP
during the early stage of its intervention. Since the launching of the new Extension
Intervention that was modelled after the Sasakawa Global 2000, in 1994/95, the focus of the
extension system was on the distribution of seed and fertilisers on credit. The extension ser-
vice’s portfolio includes livestock and natural resource conservation.
Extension agents contact farmers on an individual basis, though farmers who have suitable
plots in adjacent locations are teamed up for demonstration. Almost half of the farmers have
been visited one or more times for implementation of the current extension programme.
Access to credit, however was very limited as only 27% of the farmers obtained credit.

7.5.2 Farmers’ knowledge, perceptions and practices of soil and water conservation

Most of this section is devoted to the major issues related to soil erosion and soil and water
conservation. Discussion on area closure, irrigated agriculture and biological conservation are
treated under farmers’ practices at the end of this section.

Farmers’ knowledge and perceptions on soil erosion and soil and water conservation
Farmers’ knowledge and perception of their environment provide essential information for
understanding their land management practices. As illustrated in Chapter 3, there is no clear-
cut association between knowledge, action, attitude and perception. The land management of
farmers in Maybar varies according to the biophysical characteristics of their plots, such as
soils, slope and land use and also their socio-economic situation such as access to labour, land
size, ownership of oxen. One can see a good farming practice on a given plot, but the reverse
on an adjacent plot or elsewhere in the catchment. More interestingly, such variations are
observed on plots of the same household.
During the qualitative study phase, I carried out 30 informal interviews involving several
farm plots that show different farming practices. Following that, I designed a formal survey
involving 107 farmers in order to know the proportion of farmers’ knowledge and perception

175
.tesfaye.thesis 29-01-2003 15:12 Pagina 176

on the biophysical environment and human interaction in the Maybar watershed. This was to
better understand their land management behaviour as discussed in the subsequent sections.
To this effect, I used similar knowledge and perception statements used in Wolaita (Chapter
6). The statements were measured on a scale of three categories: (1) not clear, (2) disagree, and
(3) agree. This scale was thoroughly tested during the qualitative phase. The distribution of
farmers’ responses is given in Table 7.5.

Table 7.5: Farmers’ responses to statements on knowledge of soil erosion and soil fertility, Maybar, South
Wello, Ethiopia

Statements Not clear (1) Disagree (2) Agree (3)

1. Steep lands are prone to erosion. - - 100

2. Overgrazing exposes land to erosion. 8 3 89
3. Land with poor vegetation cover is easily eroded. 8 1 91
4. Continuous cultivation reduces soil fertility. 5 1 94
5. Soil erosion reduces soil fertility. - - 100
6. High population pressure can increase land degradation. - - 100
7. Farm bunds have to be maintained regularly. 1 - 99

Owing to the steep nature of the landscape, farmers in Maybar fully appreciate the erosion
problem on steep lands (Table 7.5), particularly after the expansion of hillside agriculture.
However, their emphasis to soil fertility over soil erosion seems to understate their knowledge
on erosion (cf. Table 7.3), which is not the case. Long ago, hillsides were used alternatively for
grazing and upland cereal production rather than for crop production on a continuous basis
(see below). Occurrence of soil erosion is depicted by small rills in the farm, and accumulation
of silt on the grass or roadside after the rain. The effect of overgrazing is also well known to
them, even though few farmers cannot see the direct link between overgrazing and erosion.
Farmers are also able to distinguish between the livestock in terms of erosivity of their hooves.
They consider cattle with their sharp and even hooves as erosive as compared to equines that
have rounded hooves with a blunt hoof edge.
The role of the vegetation cover in protecting soil from erosion is common knowledge
coined in the local proverb: Yemeret libisu zaf, Yesemay libisu damena, which literally means, ‘tree
is the cloth for the earth, while the cloud is the cloth for the sky’. However, their observation
is limited to the cover per se, without a full understanding of the role of the root structure in
water infiltration and protective ability of the crop cover from splash erosion 58. This is also
true for field crops. Crops are however classified in terms of their effect on the soil nutrients.
Some crops are considered as nutrient depleting (teff, wheat, oats) and some as nutrient
enriching (maize, legumes) crops. In this case too, they do not understand how removing the
crop residue would affect the soil nutrient levels and how legumes enrich soil fertility apart
from their leaves. In spite of that, farmers consider this observation in their crop rotation
practices.
The very fact that farmers began fallowing before the intervention by modern agricultural
science resulted from their direct observation of a decline in crop yield on land that has been

58 Belay (1998) argues that lack of knowledge among farmers on the utility of crops in erosion control may negatively affect
the possibility of controlling erosion through agronomic methods.

176
.tesfaye.thesis 29-01-2003 15:12 Pagina 177

cultivated continuously. They expressed this fact with many analogies. For example, in com-
paring the power of an ox, which has not ploughed for certain seasons with the one that
ploughs continuously. They cultivate every season due to a shortage of land, which they
explained as being connected to population pressure. Traditionally, there were two types of
land, Belg and Mehir land. Belg land was cultivated in October and November with residual
moisture, called Eritib Ersha (wet land). This land is sown in March-April for Belg crops and har-
vested in June/July. Thereafter it is fallowed until the next Belg. The Mehir lands were cultivat-
ed in the dry period and sown at the onset of the main rainy season, Kiremt. These lands too
were sown only once a year. On top of that, cropping on uplands is carried out with a tradi-
tional soil conservation measure called Dib. This was efficient, allowing for frequent cropping.
The produce they used to obtain from this pattern of farming was sufficient for the family.
Some plots were left for a longer fallow period of up to ten or more years, depending on the
food availability. During the last 30 years, this situation has changed. The type of land owned
changed because of land tenure system changes introduced since the 1975 land reforms. Since
then, access to flat- and uplands gradually changed. These changes, coupled with a decline in
crop productivity, increased family size and decreased land size due to inheritance-limited
farmers’ adaptive land management practices (Röling and Jiggins, 1998) and led to continuous
cultivation. Under this situation, farmers are forced to convert the grasslands to croplands. Of
the total of 107 farmers interviewed, 22% have shifted an average 0.2 ha of land from grass- to
cropland. A number of farmers specifically mentioned that buying grass was better than buy-
ing grain. They hoped that their animals would get something from the crop residues too.
These are some of the farmers’ reasons that are hidden to outsiders. Some farmers even sug-
gested support from the government to minimise the number of people who depend on land.
They need jobs for their children who cannot make a decent living on the remaining land.
Their philosophy about this is that ‘those who have education go to town, and those who have
strength go to Jimma,’ their common distant migration site located in western Ethiopia, over
700 km away.
In the early 1980s, continuous bunds were introduced. Unlike the traditional stone bunds
that are constructed on very steep plots and garden plots, the introduced bunds need to be
maintained to meet the engineering design recommended by the soil conservation experts.
Farmers of Maybar, fully accepting the new bund, have even introduced an innovative modi-
fication of the bund that suits their farm realities (see section 7.5.3). As a result, bund
construction is now accepted as one of the soil fertility maintenance strategies of the farming
system in Maybar.
After discussions with farmers about their knowledge on soil erosion and soil fertility man-
agement discussed above, their perception was also assessed using the same levels of scale.
These were: (1) not clear, (2) disagree, and (3) agree. The findings of this measurement are
given in Table 7.6.
Farmers of Maybar have learnt the importance of bunds from their long-term observation
of those constructed through food-for-work on lands traditionally cultivated without a bund.
They now speak from their own field experience that plots with strong bunds have better pro-
ductivity than others. This is clear from a complete agreement with statements on this issue
(Table 7.6).
For farmers, the family is responsible for the future of their children and grand-children.
This is clear from the institution they have established (inheritance) that has been functional
since time immemorial. Land was one way of capital accumulation before the land reforms of
1975. Parents buy land as a gift to their children. It is not only that they fallow land for future
generations, but fallowed land is also seen as a saving for the future. In addition, Dib is seen

177
.tesfaye.thesis 29-01-2003 15:12 Pagina 178

Table 7.6: Farmers’ perception of soil and water conservation and crop production, Maybar, South Wello,
Ethiopia

Statements Not clear (1) Disagree (2) Agree (3)

1. Soil and water conservation increases crop production - - 100

2. We have to maintain soil fertility for the future genera- 8 92
tion.
3. We have to feed the present generation instead of think- 3 79 18
ing for the future.
4. A high production now is preferable than obtaining the 8 74 18
average for the next five years.
5. Soil and water conservation is farmers’ responsibility. 1 2 97
6. Farmers should be paid for soil and water conservation 9 51 40
in their farms.

as a storage place of soil for future generations. Therefore, they do not plough the Dib. Their
perception reflected in Table 7.6 is consistent with their practice as rooted in their land man-
agement. By the same token, the majority of farmers opposed the idea of focusing on the pre-
sent generation as compared to the future and a higher production now rather than the aver-
age for the next five years (item 3 and 4, Table 7.6). It is not farmers’ intention nor a practice
to reach a short-term goal of family needs, but it is an inevitable choice they are confronted
with due to unfavourable circumstances (see section, 7.5.3).
Owing to their observations over nearly two decades, farmers in Maybar now highly agree
with their responsibility for soil and water conservation. However, they are still not fully freed
from their ties with the food-for-work and related payments. Note that 40% of them were in
favour of paying farmers for soil and water conservation practices on their farms. During the
discussion, some of them pointed out that they needed assistance from the government
through construction tools and technical assistance rather than food. Food would only be
desired only when they face crop failure.

The state and cause of soil erosion

Soil erosion is a common phenomenon in the agricultural system of the Maybar area. The
problem of erosion was perceived by 76% of the 107 farmers interviewed. The qualitative inter-
view with 30 farmers before the survey also confirmed the same. However, the nature and
occurrence of erosion vary from plot to plot and also within a plot as vividly explained to me
by farmers. In order to build on the discussion on knowledge and perception, I discussed the
causes of erosion complained about by the farmers. These causes were then presented in the
survey instrument to assess farmers’ response in the watershed. Among the most important
causes for soil erosion, 20% of the farmers mentioned water erosion; 10%, poor management;
64%, steep slopes; and 36%, unspecified ‘natural causes.’ In view of the nature of the landscape
in the area, the choice ‘steep slope’ seems logical. However, quite a substantial proportion
(36%) of them did not specify which natural causes.
The subsequent question was why their land was exposed to one or more of these causes of
soil erosion. Therefore, farmers were asked to identify why erosion occurred in their plot. The
number of plots covered in the survey was 399. These plots are scattered in different locations
in the watershed, and identified by different names by the farming community. However, the
number of plots in some locations were very few, mostly 1-6. Therefore, plot locations with at

178
.tesfaye.thesis 29-01-2003 15:12 Pagina 179

least 10 plots were selected for the analysis of farmers’ reasoning. This selection led to 229
plots (57%) out of 399 being included, involving all 107 farmers. The purpose of developing
this selection procedure was to see similarity and difference of farmers’ response on location-
specific problems and conservation practices (Table 7.7 to 7.9) 59. Some descriptions of these
locations are given below.

Guaro (L1): These are the plots located adjacent to the house. As residential areas are mostly
situated in the upper and middle locations of the catchment, more than 50% of these plots
are located in moderately steep, to steep slopes.

Yedo (L2): This location is found between Abo Ager and Addis Amba villages stretching from the
foothills down close to the Office of the research station. Farmers classified this location as
moderately flat to moderately steep.

Liygenda (L3): This is the area located above Abo Ager village, which is part of the encroach-
ment into woodland. Sixty percent of the plots in this location are in moderately steep, to
steep slopes.

Meda (L4): As the Amharic name implies, this is the name given to the farm area adjacent to the
Lake Maybar, hence ‘flat’.

Aygebir (L5): This farm area is located adjacent to the office of SCRP research station due south.
Due to the effect of the bunds, farmers consider most plots within this location as flat.
Actually, the slope drops gradually and intercepts the L4 (Meda)

Golbo (L6): This farm area is nearer to Jerjero village in the north, below the road to Tossa
Felana. Immediately above the road is the state forestry zone where ‘fertile floods’ originate,
harvested by farmers in the upper part of the area.

Denu (L7): This farm location is close to Lake Maybar, on the southern side.

Atarimesk (8): This farm area is located north of the SCRP’s station office. It is close to the peak
of the landscape looking down upon Lake Maybar at around 2,700 masl.

The major reason that farmland was exposed to erosion was lack of diversion ditches that
affected plots from all farm locations (Table 7.7). ‘Lack of bund’ was reported for less than a
quarter of the farm locations with a small percentage of plots, with a much higher proportion
in Aygebir (L5). Some farmers from this location have removed their bunds with some replace-
ments in the form of ‘moving bunds.’ (see section 7.5.3). Damaged bunds were also reported
for Denu (L7). According to the farmers’ views, the necessary measures were taken in the
major farm areas apart from constructing more diversion ditches. The need for diversion
ditches was emphasised because of deforestation on the hillside that aggravates the runoff
from the upper slopes.

59 Note that the logic of analysing farmers’ land management problems and practices from the point of view of the plot
location was also followed in the Wolaita case (Chapter 6).

179
.tesfaye.thesis 29-01-2003 15:12 Pagina 180

Table 7.7: Farmers’ reasons for soil erosion Maybar, South Wello, Ethiopia

Reasons Responses in percent 60

L1 L2 L3 L4 L5 L6 L7 L8
n=102 n=22 n=11 n=37 n=14 n=11 n=18 n=14

Lack of bund 4 - - - 21 - 6 -
Steep land without bund 9 - - - 14 9 - -
Damaged bunds 1 - - - 7 - 17 -
Lack of diversion ditches 26 14 46 8 29 9 17 21

Note: L1= Guaro, L2=Yedo, L3=Liygenda, L4=Meda, L5=Aygebir, L6=Golbo, L7=Denu and L8=Atarimestk

Trends of soil fertility

Soil fertility is a central issue in soil and water conservation. To that extent, my discussion
addressed the trends in soil fertility of different plots among farmers who participated in the
qualitative phase. Farmers showed me interesting variability of plots in terms of their soil fer-
tility and therefore, difference in crop stands. They often mentioned three descriptive cate-
gories for soil fertility. These were ‘increased’, ‘decreased,’ and ‘the same’. Following this
observation, I included these categories to assess farmers’ views on the trends in soil fertility
for the period 1990-2000. On this basis, the fertility of 399 plots covered in the survey show
32%, 33% and 35% for increased, decreased and the same, respectively. From the farm location
point of view, the Guaro plots, which are close to the house are more fertile. The proportion
of plots in the Guaro location had increased soil fertility by more than six times than in other
locations (Figure 7.1). The proportion of plots with increased soil fertility was low at Golbo (L6),
Liygenda (L3), Denu (L7) and Aygebir (L5) locations. Here, fertility declined and at best main-
tained the same levels.

Figure 7.1: Trends of soil fertility in selected areas of the Maybar catchment, South Wello, Ethiopia
Soil fertility: Increase Decrease Same
70

60

50
Percent of plots

40

30

20

10

0
Guaro Yedo Liygenda Meda Aybebir Golbo Denu Atarimesk

Plot location number

60 Percentages are calculated based on the number of plots in each of the eight locations (L1 to L8).

180
.tesfaye.thesis 29-01-2003 15:12 Pagina 181

Table 7.8: Farmers’ reasons for decline of soil fertility, Maybar, South Wello, Ethiopia

Reasons Responses in percent 61

L1 L2 L3 L4 L5 L6 L7 L8
n=102 n=22 n=11 n=37 n=14 n=11 n=18 n=14

Lack of bund 4 - - - 21 - 6 -
Steep land without bund 9 - - - 14 9 - -
Damaged bunds 1 - - - 7 - 17 -
Lack of diversion ditches 26 14 46 8 29 9 17 21

Note: L1= Guaro, L2=Yedo, L3=Liygenda, L4=Meda, L5=Aygebir, L6=Golbo, L7=Denu and L8=Atarimestk.

Decline of soil fertility occurs for various reasons as clearly stated by farmers. Based on the
farmers’ reasons for the decline of soil fertility in the area, a similar question was addressed
in the survey instrument. Results of the survey are summarised in Table 7.8 and Figure 7.2.

Figure 7.2: Farmers’ reasons for decline of soil fertility, Maybar, South Wello, Ethiopia
Repeated cultivation Lack of manure Soil erosion Lack of fertiliser Lack of manpower Water-logging
60
55
50
45
Percent of plots

40
35
30
25
20
15
10
5
0
L1 L2 L3 L4 L5 L6 L7 L8
Guaro Yedo Liygenda Meda Aybebir Golbo Denu Atarimesk
Plot locations

As indicated in the survey results and by the group discussion with the farmers, the soil fer-
tility problem of farmers in Maybar is a vicious circle of land degradation. Farmers are forced
to cultivate their limited holding without fallowing the land due to the need to produce fam-
ily food. Cultivation is carried out with removal of almost all crop residues, without either
adequate additions of manure or other sources of organic fertiliser. In the worst case, some
plots do not have bunds or even if they have, they may not be high enough to protect the soil
from being removed. In some cases, between-bund width is so wide that erosion builds up
from the plot itself, due to removed bunds to ease land scarcity. As will be shown in section
6.5.3, these problems emerge from farmers’ coping strategies to offset the shortage of land.
Plots at L3 (Liygenda) have often lost their fertility due to soil erosion by water. Commercial
fertiliser is not available nor is there adequate knowledge among the extension agents as to
which fertiliser would be effective where. Farmers who were forced 62 to participate in the

61 See footnote 60.

62 During the fieldwork of this study, extension agents were simply marking a plot for the next season’s extension demon-
stration. They usually select the land in collaboration with the PA leaders. Farmers are then expected to co-operate.

181
.tesfaye.thesis 29-01-2003 15:12 Pagina 182

extension package between 1998-2000 complained that the extension agents wanted them to
use fertilisers on relatively better plots, which induces crop logging on some plots and aggra-
vates weed problems on others. Shortage of labour was not indicated in the group discussion,
as reflected by least-response in the survey. A few plots in some locations face water-logging
when there is excess rain due to the lack of a proper drainage system.

Farmers’ expectations from soil and water conservation practices and their commitment
Physical conservation by a continuous bund 63 in gentle and flat areas introduced a new prac-
tice into the farming system (see also Yeraswork, 2000). Farmers’ expectations from this tech-
nology are: better production (98%), better forage (34%), reduced soil erosion (70%), and
increased soil fertility (56%). Even though most of these objectives are interrelated, it was
interesting to learn about distinct objectives for their actions.
Farmers explained that increased production was achieved through control of erosion with
the help of bunds. The speed at which a given plot improves its crop production depends on
the bund management. Under favourable situations, farm plots began to show results in no
less than two seasons. Their estimate of increase in crop production from fields under soil and
water conservation practices was in the order of 50-70%.
Most farmers expect between 100-500 kg of total production from their soil conservation
efforts, more specifically bund construction and maintenance. The average expected addi-
tional cereal production from bund management was 220 kg. The average number of days
they are willing to allocate labour to bund construction and maintenance was 32 days per
year. Few farmers reported their willingness to work for more than half a year. Farmers pre-
ferred using the slack period (December-February) for bund construction, modification and
maintenance.

Farmers’ practices
This section gives an overview of farmers’ practices. The purpose of this discussion is to link
the farmers’ knowledge and perception in land management and to give some background to
their reaction to the introduced practices presented in section 7.5.3.
The physical conservation structures known in the area are Kab, Weber, Dib and stone mulch.
Kab is built from stones on the steep areas and in the homestead plots, which are often in
steep areas. Weber is constructed in the farm with a width of 20-30 meters depending on the
slope gradient. Farmers leave a strip of land at the desired location uncultivated, which is rein-
forced with debris and stones during periodic farming practices. This structure holds soils
that are moved down in the process of ploughing and water movement. Grass is allowed to
grow on it for better stabilisation. Apart from minimising soil movements on the farm, this
also delineates the land of one owner from the other. On the hillsides, the structure called Dib
is created from the irregular shape of the land to protect the soil from water erosion. This
structure has an irregular shape and length. It is created by shaping the outcropped land dur-
ing farming operations. Therefore, it is not continuous nor does it follow a contour. Unlike
contour bunds, it is convenient to turn oxen at points without much difficulty. A stone mulch
is created by scattering different sized stones in the field in an irregular pattern to protect the
soil from being washed away. In order to minimise the space taken by these stones, farmers
anchor them with their sharper end. Very big stones are removed aside during oxen plough-
ing and returned after sowing. As with crop covers, farmers appreciate the protection of
stones in relation to soil movement only, without understanding its effect on the rain splash,

63 Indigenous physical conservation (Kab, Weber and Dib) are typically staggered structures, even though Weber could be fairly
continuous depending on the landform.

182
.tesfaye.thesis 29-01-2003 15:12 Pagina 183

which is very important in fields without continuous bunds. Contour ploughing, diversion
and drainage ditches are used in combination with the above-mentioned structures 64.
A continuous bund was originally introduced with grass cover on relatively wider bund
width. In the process of adaptations of the practice (see below), some farmers have narrowed
the bund width to the extent that no grass can grow on it. However, bunds with grass cover
are still widely seen in the catchment, though there is variation between plot locations (Table
7.9) 65. Out of the total 399 plots surveyed, 51% have bunds with or without grass cover. In terms
of households surveyed, 36% of 107 farmers do not have a bund in any of their plots. Farmers
who do not have bunds in their farm use combinations of other soil and water conservation
practices discussed above according to their suitability to their household resources.

Table 7.9: Soil and water conservation practices, Maybar, South Wello, Ethiopia

Practices Responses as percentage

L1 L2 L3 L4 L5 L6 L7 L8
n=102 n=22 n=11 n=37 n=14 n=11 n=18 n=14

Diversion ditches 36 36 36 30 43 27 39 36
Stone and soil bunds 19 32 9 16 14 18 6 7
without grass
Stone and soil bunds 59 73 64 54 43 64 83 86
with grass
Earthen bund (Dib/Weber) 21 - 9 14 7 - 22 -
Leaf litter 8 5 9 8 - - 28 14
Fertiliser 14 18 - 19 14 18 11 -

Note: L1= Guaro, L2=Yedo, L3=Liygenda, L4=Meda, L5=Aygebir, L6=Golbo, L7=Denu and L8=Atarimestk.

Table 7.10: Farmers’ comparison of Dib/Weber and the introduced stone bund, Maybar, South Wello,
Ethiopia

Particulars Dibi/Weber Bund

The amount of grass produced High Low

Capacity to hold soil Low High because it is continuous
Wastage of land Less (because of more grass) More
Continuity Not continuous Continuous
Ease of modification Not easy Easy
Harbouring rats/moles Less More because of the stone
Strength High Low
Labour requirement for construc- Low High
tion and maintenance

64 For different features and performance of the indigenous conservation methods, see Belay, 1998 and Yohannes, 1999.
65 Note that this table presents farmers’ responses on the 229 plots, which are within the eight selected farm locations.

183
.tesfaye.thesis 29-01-2003 15:12 Pagina 184

As can be seen from Table 7.9, bunds with or without grass are the dominant physical con-
servation practice in the study area. As a result, I asked farmers to compare Dib/Weber and
bunds to better understand their rationale for shifting to stone-based bunds in spite of their
resistance in the early days of its introduction (Table 7.10). As can be seen in the table, farm-
ers opted for the bunds in favour of its capacity to hold more soil, despite its disadvantages as
compared to the traditional methods. This shows how much the soil matters to the farmers.
Most of the manure is used in the Guaro plots that are located around the residential areas.
Use of manure in distant plots is the exception rather than the rule. Even then, cattle dung is
separated and dried and used for cooking. It is also used for plastering the house. Droppings
from other animals and household refuse are also used for improving soil fertility. Contrary
to common knowledge, which says that farmers shifted to burning dung because of scarcity
of wood, the use of cattle dung for fuel was a common practice in the area since time
immemorial (McCann, 1995). Cattle dung is preferred to wood for heat and more importantly
to keep the fire going through the night to be restocked the next morning. Presently, howev-
er, cattle dung is one of the few options that farmers have for light and source of energy.
Manure is applied by women and children onto the Guaro plot. Men use donkeys or mules
to transport it to distant plots. Of the total 107 farmers, 94% used manure in part of their
plots, 2% in most of their plots and 4% did not use manure in recent years. The majority of
those who use it in part of their farm do so in the Guaro plot. Their reasons for not using
manure include lack of manure, distance of the plots, lack of a donkey or mule and shortage
of manpower. Therefore, availability of manure alone does not guarantee its use. In view of
this, farm size is not related to the extent of manure use. Most plots are located in fairly close
proximity to the residences. Farmers’ estimates show five minutes to about an hour for most
plots.
Residue burning and green manure were not reported, while leaf litter was reported for
very few plots. The source of leaf litter is limited to plots that are adjacent to the footpath, res-
idence or community forest. In the past, fallowing was one of the soil conservation strategies
that is nowadays reduced to inter-seasonal fallow.
Use of fertilisers is another new land management practice in the area. However, fertiliser
use is far from taking root in farming systems in the area (see section 7.5.1).
Farmers do not practise area closure. Hills are sources of woods, farm tools, games, grazing,
hay, honey, herbal medicine, scenery, spiritual place, etc.,
Generally, farmers apply a combination of soil and water conservation (SWC) practices to
maintain soil fertility. However, most of the surveyed plots (70%) were under one of the com-
mon conservation methods, of which 60% consists of stone bunds. About 20% of the plots was
under two types of SWC methods, whereas the last 10% consists of four to eight different con-
servation methods. This is an indication of farmers’ efforts to integrate different SWC meth-
ods into their farming systems when they find it necessary and they are able.
A comparison of farmers’ practices with their responses to the knowledge and perception
statements in Tables 7.5 and 7.6 shows that farmers’ practices are consistent with their knowl-
edge and perception. Out of 107 farmers, 90% of them practise SWC as expected. Further anal-
ysis of bund management with size of the cropland among those who practise SWC does not
show a significant variation among land-ownership classes.

184
.tesfaye.thesis 29-01-2003 15:12 Pagina 185

7.5.3 External interventions at Maybar: achievements and constraints

Physical conservation
Farmers did not accept the stone bunds during the early period of their introduction. This was
because the speed with which the construction of the stone bunds took place did not give
farmers the opportunity to internalise the benefits of the stone bunds. The original design was
a bund constructed from stones that was poor in retaining water. These bunds served as a per-
manent shelter for rats that had used the bushes before. The rats even increased their habitat
range as a result of the continuous stone bunds that stretch deep into the farm away from the
bushes and live fences. The other immediate problem was the spread of weeds across the stone
bunds. In addition, unusually narrow inter-bund intervals did not allow farmers to ox-plough
conveniently. A higher number of bunds per plot also means a larger amount of wasted land.
As the work was carried out hastily by and with farmers who were not convinced, a substan-
tial number of bunds were destroyed in the following seasons, littering the farm areas with
stones.
Due to these immediate problems, farmers’ complaints mounted in the subsequent years,
though in tacit protest because of the authoritarian 66 nature of the Peasant Association (PA).
Firstly, farmers began to modify the stone bunds by mixing the stone with layers of soil as
filler. Some farmers removed alternative bunds to ease the ploughing and to destroy the rats,
making it possible to see erosion in the farm plots at different spots. Having seen the damage
caused to their farms, they began to replace the removed bunds, though sometimes in a dif-
ferent location from before. This marked the second and major modification of the use of
stone bunds in the area, the ‘moving bunds’ (Yohannes, 1999). This began when farmers start-
ed distributing the soil that was accumulated under the bunds to the lower part of the plot,
which often lacks soil. Distribution of soil locally called Yebidir Afer (meaning ‘the borrowed
soil’) is the main reason for moving the bunds. The number of farmers who reported this prac-
tice was only 19% of 107 farmers, which is not substantial. However, the significance of this
practice is not in the number of people who practise it now; rather, it is in the confidence-
building and facilitation for the acceptance of the continuous bunds in the entire watershed
that became dominant, as shown in Table 7.9. For farmers’ bund management practices see
plates 7.1, 7.2 and 7.3.
Moving bunds began as a ‘blessing in disguise’. The pioneering farmers adapted the prac-
tice from three farmers, who had been given some land belonging to another farmer who was
sent to the resettlement. Prompted by the condition of their access to the land, these farmers
ploughed all of the land including the Dib that is not traditionally ploughed. They achieved a
very good production from the land, particularly the Dib. After observing this practice, a few
farmers in Abo Ager began to distribute the soil that was accumulated behind the bund and
to reconstruct the bund. They decided to make use of the deposited soil because of the land
insecurity that prevails in the region (Yeraswork, 2000: 214-215). Destruction of traditional
structures such as Kab and Weber to use the accumulated soil is a known practice in the farm-
ing system (Belay, 1998:11; Yohannes, 1999). Through this process, the practice of ‘harvesting
soil’ was widely copied by other farmers and became farmers’ knowledge in the area. This was
a major farmer innovation in bund management. This work was accomplished by moving the
bund 2-6 meters below the original bund. It became popular after the change of government
in 1991 when control of the bunds by government agents ceased.
Farmers follow certain criteria to move bunds. Some of these criteria are outlined below.

66 Peasant Associations are mandated to implement the government’s action plans at the local level. In view of this, those
who fail to co-operate do not get access to services channelled through the PA and will be subject to imprisonment.

185
.tesfaye.thesis 29-01-2003 15:12 Pagina 186

Age of the bund

Bund age is determined by the amount of soil accumulated above a bund. On average, a bund
accumulates adequate soil from 2-5 years depending on the location. Their usual indicator is
crop stand at different locations in the bund (above, below and middle). This emerges from
their day-to-day observation of their farm plots.

Conditions of the upper bund

In order to remove bunds, farmers observe conditions of the upper bund, mainly the border
bund (Weber) with other farmers. If the upper bund is strong enough to protect the runoff
from their farm, they move the next bund or more.

Flood incidence
They look at the historical occurrence of flood in that farm location before a decision to move
a bund.

Slope variation
They take into account the slope variation within their farms and the entire watershed to
judge the severity of flood if it occurs.

Rat/mole or weed control

A farmers’ decision to move a bund is also affected by the presence of rats/moles or weeds in
the bunds. The timing for moving the bunds is therefore affected by the extent of these prob-
lems. Plots close to the village and bushes are particularly susceptible to rats.

Uneven soil distribution

When the soil depth is affected because of the bund position, farmers move bunds selectively
to balance the soil distribution in their farm. In others words, when some bunds hold a
greater proportion of soil as compared to others, those bunds are moved to distribute soil to
other plots through water movement and ploughing.

Strength of the bund

When the original bund is not strong enough to hold an adequate amount of runoff and soil,
farmers move that bund and construct it in a new location.

On the whole, the moving bund practice increased the utilisation of stone bunds in the area. This
had been discouraged by the SWC experts and the government agents who considered it as a
rejection of the technology rather than as an adaptation. Due to their own adaptation of the
bund, farmers in Maybar have now internalised it within their farming system (see also
Yohannes, 1999). Most of them confessed that they used to resist bund construction because
of its disadvantages. Because of that, they were exaggerating the observed problems such as
rats and mole to destroy the bunds. Over time they have reduced the negative side of the tech-
nology by strengthening its positive sides. They now know that their land would turn to stone
very quickly without those bunds. Stating this situation, one farmer said that bund construc-
tion is now Yewideta Gideta 67 (necessary). As a result, bunds surpassed the traditional conser-
vation techniques such as Dib and Weber with their coverage of the cropping areas.

67 The farmer meant that bund construction is inevitable under the present farming practice which they have maintained
in order to protect the soil from erosion.

186
.tesfaye.thesis 29-01-2003 15:12 Pagina 187

Plate 7.1: Stone bund moved for utilisation of the accumulated soils (note that the plot is located near the
residential area)

Previous bund

Removed stones

Plate 7.2: A moved bund in the lower slope (note that the upper bund was left intact)

Apart from the bund-moving practice, farmers also remove some bunds when they find it
desirable. Accordingly, 14% of the 107 farmers removed 33-50% and in one case 100% of the
original bunds in some plots. Most of the farmers removed one bund per plot, while very few
of them removed two to four bunds per plot. Farmers remove bunds for various reasons. These
include: ease of ploughing (48%); distribution of the soil 14 (%); destruction of rats (14%); and
24% for a combination of the above reasons. One farmer removed the middle bund in one of
his plots, having an area of 0.14 ha. His justifications were that the upper bund was strong
enough with a height of 90 cm. The border bund (Weber) had a height of 1.4 meters covering
the length of 45 meters. Such situations are widespread in the Maybar catchment.

187
.tesfaye.thesis 29-01-2003 15:12 Pagina 188

Farmers’ decisions are often related to their socio-economic backgrounds, which are likely to
influence their decision. In this study too, notwithstanding the farmers’ rationale as
explained above, a few variables were selected to assess how they might relate with the farm-
ers’ decision to remove the bunds. Thus, the percentage of bunds removed was correlated with
the age of the household, livestock owned, labour, family size and cropland area (see Appendix
2). Among these variables, age, labour and cropland area had a negative correlation with the
percent of bunds removed, though this was only significant for age (p ≤ 0.05). The number of
livestock owned and family size on the other hand positively, but not significantly correlated
with the percent of bunds removed, indicating land pressure. Unlike age, the negative ten-
dency of cropland area and the percent of bunds removed are consistent with farmers’ prac-
tices. Even then there was no strong correlation.

Plate 7.3: Fields with selective moving of stone bunds

In addition to either removal or movement of the bunds, farmers also adjust the width of the
bund to allow for the increase of the crop area even if it is a few centimetres over a length of
the bund. The width of the bunds was reduced from 50-70 cm in 1984/85 to 30 cm in 1999.
However, there are some bunds whose width is close to 1 meter. These kinds of bunds are those
which are reinforced on the Weber.
Maintenance of the bund was reported by 94% of the 107 farmers surveyed. Those who
reported no maintenance did so because of a plan to remove the bund, the strong condition
of the bund and a plan to replace it with Dib and diversion ditches. As there is no compulsion
to maintain the stone bunds by the government, farmers allocate time and labour to maintain
the bunds in their farm. Because of the poor quality of the work, farmers do not appreciate
the campaigns that were organised by the Extension Agent and the Peasant Association to
maintain the bunds in the area. They would prefer community co-operation for the mainte-
nance of hillside bunds. In their farms, they prefer to do it alone or in co-operation with a few
farmers who know each other well.
With the exception of four plots that belong to different households, other plots in the
moderately steep, to steep area did not become bench terraced. However, this does not mean
that all the bunds are movable or that doing so would be desirable due to slope and location.

188
.tesfaye.thesis 29-01-2003 15:12 Pagina 189

Some bunds are reinforced on a massive Weber, which are not easily movable.
Farmers’ strategies are consistent with a recent cost-benefit analysis, which compared the
introduced with the adapted SWC designs at Maybar, Anjeni and Andit-Tid, in the Amhara
region (Ludi, 2002). NPV (Net Present Value) was used in this study to compare these two meth-
ods of SWC under different erosion conditions (sheet and rill erosion and sheet erosion only),
different slope gradients, labour costs (with and without subsidy), with and without fertiliser
use. This was carried out in a special case where soil depth has already decreased to a critical
level of 10 cm, where crop production has to be given up. The CBA (Cost-Benefit Analysis) of
this study shows that the introduced SWC is profitable only if slopes are gentle (≤ 8%), or
steeper (≥ 31%), with additional fertilisers and in some cases only if labour costs for both con-
struction and maintenance of the bunds are subsidised. Ludi explains these unfavourable
conditions of the introduced SWC by the fact that an annual soil loss rate at Maybar is low
whereby small yield reductions are tolerated as compared to the higher proportion of land
occupied by conservation structures and high labour costs.
The preceding findings quantify farmers’ economic criteria that, among others, take into
account the land occupied by the conservation structure. These are entered into the Decision
Tree in terms of crop production from that piece of land. However, this is only one of the
several criteria which farmers consider in their decisions about soil and water conservation.
These include: technical conditions of the plot (see criteria for bund moving), plot location,
land size and labour availability. Thus, farmers do not necessarily decide according to costs or
benefits alone.

Area closure
With the area closure, the entire uplands that are not suitable for agriculture due to slope and
prior degradation were treated with forest and hillside bunds. Following that, a marked reduc-
tion in runoff and gully stabilisation began to be seen in the entire catchment of Lake Maybar,
which was also confirmed by farmers during group discussions.
At the time of the change of government in 1991, a massive destruction of conservation
efforts took place. The conservation was implemented through mass mobilisation. The com-
munity primarily cut down trees, but also removed conservation structures directly and indi-
rectly by allowing their animals to trample them.
The destruction did not spare even the nursery at Tossa Felena. The forest in the closed area
was the target of everyone from within the area and distant villages. The guards assigned by
the MoA were not in a position to protect the mass destruction. The scene became a typical
case of a ‘tragedy of commons’ (Hardin, 1968), more specifically the ‘tragedy of open access’
(after Warren, 1991).
After the change of the military government in 1991, the army of the former government
was retrenched and the ex-soldiers came back to their home areas. In addition, people who
were sent to resettlements also came back to their place of origin. This influx led to a claim
on land, particularly among those returnees from the settlements whose land had been dis-
tributed to others in order to implement the agro-ecological plan. In order to solve this
problem, the Peasant Association leaders firstly asked farmers to share land voluntarily with
their compatriots. In return, those who shared land with returnees were given a plot of land
in the hillside for pasture. As this measure was not sufficient to accommodate all returnees,
the Peasant Association began to re-allocate the land in the formerly closed areas. This mea-
sure ruined the community forest area. Widespread grazing in the closed area began in 1991
and continued until the fieldwork of this study was completed at the end of 2000. The PA lead-
ers distributed land for returnees without responding to the requests of many youth, who are

189
.tesfaye.thesis 29-01-2003 15:12 Pagina 190

now also eligible for a farm plot. There are young farmers with families who still depend on
their parents’ farm for subsistence as well.
Opening up of the closed areas to grazing and agriculture aggravated the erosion in the
catchment, which resulted in a widening of the gradually narrowing gorge. Large stones and
heavy siltation were visible along the embankment. At times of heavy rainfall, the adjacent
fields become littered with stones and silt.
In early 1991, the operation of the station was interrupted for a few months, as the signs of
hostility against the project became more overt. The Research Assistants returned to the sta-
tion in June 1991. They resumed the work with support from the officials of the new
government. Farmers around the catchment area invited the Research Assistants for a meet-
ing to discuss their interest in grazing their livestock on the hillsides to ease their livestock
feed problem. During the discussion, the Research Assistants advised the farmers to use their
best judgement in view of the past conservation efforts and what they had learned together.
In spite of that, the farmers continued to graze in the closed area until 1997. This ruined the
hillside bunds and the vegetation by both animal and human interference. In fear of a repeat-
ed plea by the Research Assistants to the management of the PA to protect the hillside as in
previous years, some farmers went directly to the Wereda Administration to get permission to
graze animals in the closed areas. Following that, the Wereda Administration sent an expert to
demarcate the community and the state forests. According to the ruling, the former one was
to be managed by the community and the latter to be managed by the MoA. However, farmers
do not distinguish between the two. This demarcation legitimised grazing on the hillside that
the project had been protecting up until 1991. Farmers were happy with the decision of the
expert, as they did not accept the idea of the area closure from the very beginning. After the
closed area was reopened, grazing was prohibited during the wet season on the hillsides and
other grasslands in order to grow hay. The grass was divided among farmers. Management of
the wet season grass was carried out by a village (Got) committee. This committee has three
members, one from each village.
In 1997, the government issued a decree on the use of community forests. According to this
decree, the Peasant Association can levy a service charge to cover running costs for the Office
Administration. To the resentment of the farmers, the PA management began to charge them
for use of the grass from the areas demarcated as a community forest through the farmers’
own efforts.
Farmers have been planting and still plant trees, mainly eucalyptus, on their land. This
practice is of course more because of their confidence in the local institution, Kire, rather than
being a tenure issue only. As shown above, the open-access state forest and nominally closed
access community forest are suffering from over-exploitation by individuals. This is a clear
example of where the land tenure system of the country negatively affects natural resource
management. Insecurity of holdings on arable land is by far less than on the non-arable land.
This is aggravated by poor participation of the community. Therefore, the community does
not apply the local control mechanisms to protect the ‘community’ and state forests.
According to a Kire leader and other informants, Bele is not applied on the community and
state forest because everyone would be affected by the curse (cf. Section, 7.2.3).
The main reason for the illegal utilisation of trees in both community and state forests is
the lack of a defined utilisation mechanism for the rural people besides planting (Alemneh,
1990). The element of the tragedy of the commons is clearly reflected in the case of the forest.
Many farmers use tree species, which they either do not own or of which they have few stands.
At times, one could see fresh heaps of wood behind houses covered with grass or straw. Then
I asked farmers with whom I had developed a rapport where the source of such wood was.

190
.tesfaye.thesis 29-01-2003 15:12 Pagina 191

Laughing at my observation, two farmers confided in me that the trees came from the nearby
forests. When I asked why, one of them said: ‘if I do not cut it, others will’. This indicates a clear
absence of rules and regulations for access to the benefits of trees planted, which jeopardised
the tree production in the area and also hold true for the entire country.

Irrigated agriculture
The irrigation facility was used only for one cropping season, in 1985. In 1993 the project hand-
ed over the pump to the three villages in the Maybar PA that were involved in the land distri-
bution. After the handing over, some members of the Abo Ager village suggested to use the
pump for a flourmill. On the other hand, the Tossa Felana town requested to use it for power
generation. These proposals were not realised on the grounds of technical difficulties.
Thereafter it was suggested to sell the machine. The latter option was not implemented pri-
marily for lack of an official document, which was not traced nor known to have existed when
the machine was brought to the area. In the end, the machine has continued to stand unused
for many years. This was in spite of the repeated requests from the projects and visitors from
Addis Ababa and elsewhere, advising the community to use it for irrigation.
In 1997, the three PAs were merged into one. These were Maybar, Defatit and Albore. With
knowledge of the condition of the pump, the new villages entered into the claim on the pump’s
proceeds, if and when it is possible. And yet, the people who have a direct prospect for its use
did not show any interest at all. It is worthwhile to indicate that farmers from Albore, which is
located in the eastern part of Lake Maybar irrigate their farms from outlets of the lake. What is
even more interesting is that these people are regularly maintaining about 3 km of irrigation
channel that was constructed along the hillside. And the water distribution is managed by a
local water-use committee.
The common explanation given by farmers for not using the pump after the 1985 season was
that the water from the pump burns the field crops. According to the information from the
research assistants, that is impossible. In addition, inability to cover the running cost and logis-
tics was also mentioned by few, in spite of an overt lack of motivation to restart the pump by
different parties. In view of this, I continued searching for more subtle explanations that shed
light on the farmers’ reactions to irrigation. To this end, I held intensive interviews with the
research assistants, youths and elders in the community. Finally, I held group discussions in all
three villages focusing on this point. After this triangulation, the following issues were con-
cluded (see also Ludi, 2002:281-283).

1 Noisiness for the Maybar Shrine

The pump is placed opposite the Maybar Shrine (Shehochu). The noise it creates during operation
was believed to have been ‘disturbing’ the spirits. The legend about the lake says that it was
given to one of the earlier religious leaders to use for cleaning during praying, which was
explained by some elderly to show the link of the lake with the spirit. There was a difference of
opinion between the elderly and the youths on this issue. The religious leaders clean demons
from sick people by directly dipping people in the lake on the opposite side of the pump site.

2 Fear of eviction
There was a rumour that the state could lease the land to an investor if irrigation was devel-
oped in the area and a producers’ co-operative established, which had a bad reputation during
the time of the socialist government. Fear of eviction is likely in view of the insecure land
tenure in the region, and the direct experience of so many families who were sent to resettle-
ments.

191
.tesfaye.thesis 29-01-2003 15:12 Pagina 192

3 Obstruction by neighbouring villages

The villages below the lake, which view the lake as a table above them feared that the water
outlet would be lessened, particularly during the dry period if it is used with a motorised
pump by farmers in the lake catchment. Because of this they are said to have discouraged the
use of the pump by spreading different kinds of rumours (cf. no.2 above). In addition,
prospects of the sale of the machine raised a mixed opinion in the community.

4 Reclaiming for land

Land redistribution, which took place when the agro-ecological plan was implemented, affect-
ed some families more than others. There were farmers who did not have a piece of land in
the irrigated site (70 farmers) but obtained land after this land redistribution. Some families
who had to exchange their upland in return for land in the irrigated site did not do so after
the control was relaxed. This has left some resentment among those who associate the trans-
fer of their land with the irrigation initiatives. Therefore, they resist the continuation of the
irrigation with the prospect of getting their land back.
In spite of these reasons, the project did not make a concerted effort to resolve this dead-
lock against continuing the irrigation practices for over 15 years, a situation that could have
brought a different scenario to the watershed. In the absence of an alternative to their grow-
ing problem of food shortage, farmers returned to the uplands in pursuit of their short-term
goals for their families and livestock, which has meant the end of the agro-ecological plan.
With these findings, I decided to contact the concerned bodies to take the necessary mea-
sures to utilise the lake for irrigation. The process and the outcome of this effort is given in
box 7.2.

Box 7.2: Office back to office on irrigation in Maybar, South Wello, Ethiopia

At the end of my fieldwork in early 2001, I contacted the Wereda Administration and Agriculture
Office, the Zonal Office of Agriculture, and the Research Assistant of the former SCRP station
to inform them of my finding about the irrigation in Maybar. Subsequently, I organised a meet-
ing, which they all appreciated. During the meeting we discussed what should be done to re-
start the irrigation programme and who should be involved and so on. Following that, I con-
tacted the SIDA representative in Dessie and an NGO in Kombolcha in the presence of the expert
from the Zonal Office of Agriculture and the research assistant to discuss the possibility of
financing the irrigation activities. Both offices expressed their willingness to assist the initia-
tive in matters of their capacity. Following that, a team of five persons including the SIDA rep-
resentative (expatriate) went to Maybar in March 2001 to view the situation and discuss it with
the farmers. After that, the team decided to prepare a budget for the pump renovation to be
financed from the SIDA fund. Unfortunately, no action was taken until May 2002 when I
inquired into the latest development in this matter. The main reason was that the individuals
who agreed to prepare the budget failed to do so. Coming back to square one, I finally left the
matter by appealing to the Zonal Administrative Head, whom I believe will link the matter with
a country-wide new development initiative, which claims to decentralise the development
management at the Wereda level.

In view of this, I wonder when that lake will be utilised for irrigation again, that at least those
families with access to irrigation will be able to improve their food security without soil min-
ing or other land degrading farming practices.

192
.tesfaye.thesis 29-01-2003 15:12 Pagina 193

As can be seen from the foregoing discussion, the shortcoming is not only on the project side.
The local authorities also lack vision and more importantly the good will and initiative to
solve local problems. One means to break the deadlock on the irrigation issues is to openly dis-
cuss with the villages concerned and to offer the upland parcels to those who have shared
land from their irrigated area with others. The religious leaders should also be approached for
their blessing to the irrigation initiatives as part of their role.

Biological conservation
The research on biological conservation at Maybar ended up with a pessimistic conclusion.
Even though some biological conservation methods prove to be complementary to the physi-
cal conservation measures, the recommendations were not feasible under the existing farm-
ing system and socio-economic conditions of the farmers in the area (Kassaye, 1997).
Farmers who hosted the biological conservation experiments expressed their appreciation
for the forage plants they had tried. But there was a lack of seed to continue with the pro-
duction of Vetch even though the farmers adapted it to their own method of planting. The
other forage plants were partly less adaptable to the area, and the animals that graze in the
croplands after harvesting, damaged the vegetation that showed some vigour. The other
problem was that moles and rats were attracted to the introduced grasses. Therefore, farmers
abandoned the grass from the fields and continued with the local grass species. Grass cover
on the stone bunds was removed with the shrinking bund width. The pigeon peas were dam-
aged by oxen during ploughing. In the end, the entire biological conservation effort made on
cultivated land was lost and became history. Forage plants planted in the closed areas failed
partly because of adaptation to the area and partly because of the opening of the area to
animal grazing. The major causes for these losses are lack of a controlled livestock production
system, lack of suitable forage species, lack of alternative sources of forage and lack of farm-
ers’ participation.

7.6 Conclusions

In the early 1980s, interrelated land-management practices were introduced in Maybar catch-
ment by the SCRP. Apart from the bund, other interventions, namely area closure, irrigation
and biological conservation were a complete failure. The bund survived because of farmers’
adaptation of it. It became a widespread practice after the rigid control of the bunds was
removed after 1991.
Farmers in Maybar have a very thorough understanding of their natural resources, their
limitations and potential to agriculture. They have knowledge of soil erosion and its impact
on soil fertility and thereby crop production. Their agricultural practices were dependent on
an alternative fallow system between the flat lands and the uplands. This practice was under-
mined by population pressure and tenure insecurity. The upland indigenous physical soil
conservation practices (Dib/Weber) became less effective due to continuous cultivation. Use of
manure outside the adjacent plot (Guaro) is an exception rather than a rule. In addition, use
of other organic matter is very rare in the area. This is mainly due to the perception of the
community, which labels fields without trees and shrubs as a good farm. According to farm-
ers, soil fertility is declining due to repeated cultivation, lack of manure and soil erosion. On
the other hand, their responses on the overall trend of soil fertility between 1990 and 2000
show no change at the watershed level, while variations between plots remains.
The introduced bund was modified by farmers and according to their resources, indige-

193
.tesfaye.thesis 29-01-2003 15:12 Pagina 194

nous knowledge, in response to land pressure, technical, economic and social criteria. The
modifications involve mixing stone with soil, bund movement, reduction of bund width and
removal of alternative bunds. Over time, it has become a routine farming practice that suits
the increasing hillside agriculture. Farmers now regard bund construction as a necessity.
Farmers’ bund management practices do not show a major difference because of socio-eco-
nomic variables such as age, education, livestock ownership, labour and family size. However,
the intensity of modification varies with the size of cropland.
Interventions on area closure, irrigation, and biological conservation failed because of
social, economic, technical and institutional reasons.

Social
From social points of view the project did not take into account the local institutions such as
the spiritual centre (Shehochu), which is located at the edge of Lake Maybar. Its influence in the
community and the relationship with the lake was very important for the changes in the area.
In addition, past experience of the farmers in irrigation should have been assessed, at least,
after the famine of 1984 when irrigation was first introduced.

Economic
The land issue is central to the failure of area closure and to some extent to irrigation and bio-
logical conservation. Some people got access to flat land while others lost such rights because
of the resettlement and land redistribution carried out specifically in Maybar. This situation
would continue to work negatively on all the interventions related to land. Farmers did not
accept area closure because they did not have an alternative for their livestock. As irrigation
was associated with land redistribution, those who ‘gave’ land to others hoped that stopping
the irrigation would bring their land back. There is also the issue of equity for the pump
among the neighbouring villages. Growing forage crops on the cultivated land means compe-
tition with crop production that is hardly accepted by farmers.

Technical
The project did not think of biological conservation on cultivated land until the late 1980s nor
did it develop technical options that fit into the existing farm tools and farming practices.
With respect to irrigation, farmers did not have genuine participation in the plans because of
the emergency situation in which irrigation was introduced. As a result, the subtle reasons
behind their resistance that limited the use of irrigation later were not revealed nor assessed
by the project. This could have been avoided if a participatory approach had been employed
in the initiative.

Institutional
land security was not reliable both at the local and regional level. The resource was left with
open access that exposed it to bad utilisation. As the ‘community forest’ did not involve the
community, the community did not follow its rules to protect what they see as government
forest. This development led to an uncontrolled destruction of trees. The decree that was
issued by the government in 1997 about the utilisation of community forests has even aggra-
vated the problem.

What should be done?

In order to achieve sustainable land management and livelihoods, a two-pronged approach
could be followed. These are intensification and diversification on one hand and a minimisa-

194
.tesfaye.thesis 29-01-2003 15:12 Pagina 195

tion of land pressure through non-farm employment, on the other. In the short run, rehabil-
itation and promotion of the irrigation facilities is feasible as many families have irrigable
land. In addition, utilisation of manure in distant plots is a feasible action for intensification.
Alongside irrigation, other rural development components such as credit, road improvement,
family planning and education should be carried out. In the long run, most of the young gen-
eration and some of the new families should go out of direct farming and carry out non-farm-
ing activities. In addition, means to involve some rural community into the formal and non-
formal employment sectors should be sought, for example through skills training.

195
.tesfaye.thesis 29-01-2003 15:12 Pagina 196
.tesfaye.thesis 29-01-2003 15:12 Pagina 197

8 Discussion

8.1 Introduction

This chapter attempts to bring together the major findings of this study with the purpose of
understanding soil and water conservation behaviours of farmers in Ethiopia. The study used
a design that deliberately included case studies of two areas in which a major soil and water
conservation project, the SCRP, had intervened by introducing erosion-control measures using
food-for-work as an incentive. Major issues arose from studying the impact of the SCRP that
help us to understand how farmers manage their land. The study also looked at Konso, per-
haps the most famous example of indigenous soil and water conservation practices in
Ethiopia. Comparing these different case studies have given some insights into a wide range
of farmer behaviours and reactions to intervention.
This study has operated at different levels. It compares case studies (watersheds), house-
holds, and fields or plots. It is hoped that my study allows for drawing some vital conclusions
to influence future SWC policies and strategies in Ethiopia. Soil erosion is, of course, nothing
new to Ethiopia. The empires of the Egyptian pharaohs were, after all, based on the annual
load of silt deposited by the Nile, silt that for the most part came straight from the Ethiopian
highlands. In modern times, erosion has, however, become vastly accelerated by human activ-
ity. The Abay River, widely known as the Blue Nile, could now more appropriately be called the
‘brown’ Nile because of the millions of tonnes of fertile soil it carried away every year. Land
deterioration in Ethiopia is out of control and perhaps one of the major threats to the oppor-
tunities of the country to become part of the developed world instead of a permanent basket
case. Hence this study attempts to better understand how farmers’ practices and erosion are
related. Land use by farmers is not very amenable to regulation, fiscal policy or other inter-
vention. It is of extreme importance to understand the reasons that farmers have for using the
land in certain ways. This study has looked at these reasons. This discussion chapter revisits

197
.tesfaye.thesis 29-01-2003 15:12 Pagina 198

the research questions formulated in Chapter 1:

1. What are farmers’ responses to the introduction of SWC measures?
2. What are farmers’ knowledge and attitudes on soil erosion and soil and water conserva-
tion?
3. What do farmers do to conserve soil and water on their farms?
4. What are determinants of soil and water conservation practices among small farmers with-
in and across socio-economic and agro-ecological environments?
5. How can constraints to promote soil and water conservation be overcome?

Section 8.2 highlights the problematic situation of Ethiopia and some of the responses the
country takes to redress it. Section 8.3 then presents the main conclusions with respect to
farmers’ reactions to the interventions. Section 8.4 draws together the conclusions that can be
gathered from the study of Konso, the indigenous SWC practices. Farmers’ knowledge and atti-
tudes with respect to soil erosion and SWC are presented in section 8.5. This is followed by a
discussion on farmers’ practices in SWC (section 8.6). Determinants of soil and water conser-
vation among small farmers are presented in section 8.7. The discussion in section 8.8 deals
with possible suggestions to solve constraints to promote soil and water conservation in
Ethiopia, which includes specific policy implications. The final section of the chapter presents
issues for future research.

8.2 Ethiopia: a country prone to erosion that has needed to take measures
to redress it
Ethiopia is a mountainous country with a substantial proportion of its land (45%) in the high-
land zone at more than 1,500 masl. Due to this physical feature, the country is highly prone
to erosion. A very high population growth, limited expansion of the opportunities for off-farm
employment, severe droughts and extreme poverty combine to form a worrying scenario of
continued soil degradation.
The country has not been food self-sufficient since the 1973/74 famine. It has faced similar
famines in 1984/85 and 1993/94 (Markos, 1997:89). As I write this thesis (fall, 2002), millions of
Ethiopians again are starving as a result of the recent famine that has hit many African coun-
tries. The food problem is associated with land degradation, population pressure, a low level
of technology, ill-conceived policies and poor institutional services. The state has taken up the
‘land degradation triggered by the population pressure’ thesis with a vengeance. This view is
referred to as ‘a generalised Malthusian narrative’ (Keeley and Scoones, 2000; Hoben, 1996;
Dessalegn, 1998). The state’s position has been reinforced by large-scale studies carried out in
the country such as the EHRS and location-specific erosion and runoff studies by the SCRP
(Chapter 1).
With consensus on the causes of the food deficit and land degradation in the country, the
state has commissioned large-scale soil and water conservation activities that focus their
attention on the food deficit regions of the northern, central and eastern parts of the coun-
try. These initiatives, implemented by the MoA, have been supported by the WFP programme,
several NGOs and other multilateral organisations. As a result, about 15% of the country that
needed to be under appropriate conservation measures was covered in the 1980s. The other
key partner in the conservation initiative was the Soil Conservation Research Project (SCRP)
that became a technical arm of the conservation movements in the country. In addition to
assisting the national conservation programme, the SCRP has been carrying out an intensive

198
.tesfaye.thesis 29-01-2003 15:12 Pagina 199

SWC action research project in small catchments to generate information and appropriate
technologies. Based on the initial diagnosis of the problem that was believed to be a technical
problem, the SCRP and its associates embarked on technical solutions. During the nearly two
decades of its operation, the SCRP carried out several activities in Ethiopia. The major ones
are:
• establishing research stations in different agro-ecological zones;
• producing an extension handbook on soil and water conservation for extension agents
(first edition in 1986, then reprinted in 1995);
• supporting nation-wide food-for-work implemented by the WFP and MoA with technical
matters;
• generating data and producing several scientific publications, including a textbook in the
official language (Amharic) and a video film in English;
• collecting long-term data on soil loss and runoff in different agro-ecological zones (these
data form a resource base that makes Ethiopia unique among developing countries);
• testing and adapting soil and water conservation measures that are technically effective;
and
• training about ten high level professionals in the area of soil and water conservation and
assisting several national and expatriate graduate students.

Unfortunately, these interventions have not taken root in the farming systems they targeted,
nor did they change the extent of land degradation in the country (Hoben, 1996; Yeraswork,
2000, Alemneh, 1990; Stahl, 1990). Farmers’ negative reactions to the interventions became
more pronounced after the downfall of the military government that had been enforcing the
conservation measures through an administrative and authoritarian machinery that reached
down to the grassroots level.
This study was initiated to study the farmers’ responses to the SWC interventions in greater
detail and to compare the intervention scenes with indigenous SWC practices developed over
four hundred years by farmers themselves. Hence I start off this discussion chapter with try-
ing to answer the first research question.

8.3 Research question 1: what are farmers’ responses to the introduction of

SWC measures?
The SCRP’s impact was studied in two areas: Wolaita and South Wello. The Wolaita people
started agriculture in the lowlands and later on moved to the highlands. Their production
system was garden-based and featured perennial and root crops in addition to cereals and
pulses. The hand tools were soil friendly. The rich and deep soils of Wolaita did not necessi-
tate agroforestry in the same way as Konso, where the shallow soils pose a higher level of crop
production risks. The garden plots were productive for the low population size and the
integrated crop-livestock production system produced sufficient manure for sustainable
production. In view of this, the Wolaita people did not widely promote physical structures for
soil-fertility management in the highlands. Their labour organisation focused on soil and crop
management practices and social activities rather than bund construction.
Nowadays, things have changed considerably with the move to the higher grounds and
expansion into the land traditionally used for grass production. Labour organisation is
reduced to seasonal farm operations due to the use of ox-ploughs and the small farm sizes. As
in Konso, livestock in Wolaita now needs to be controlled because of the limited free grazing

199
.tesfaye.thesis 29-01-2003 15:12 Pagina 200

areas and because of the perennial, root and spice crops. Farm fields are fenced and trees are
an integral part of the farming system. Live fences, and farm trees provide multiple uses to
farmers, including production, conservation and soil fertility. However, many of the attitudes
and practices of old persist. Deep soils allow the Wolaita people to ignore the danger signals.
Wello developed an ox-plough, cereal-farming system that favours plain and gently sloping
areas, which suit the main farm tool, the plough. In view of this, physical SWC structures were
limited to uplands and garden plots, adjacent to the residential areas. As shown in Chapter 7,
these farming practices are effective only under a long fallow system. Sole cropping became
the dominant feature in Wello due to the influence of the plough that is used for land prepa-
ration and cultivation. In this system, trees are not appreciated in and around the farming
areas. Trees belong to the bushes or hills, not to the farms. In view of this, it is not easy to pro-
mote biological conservation measures in the Wello system, unlike in Konso and Wolaita.
Livestock production in Wello is still extensive in terms of its management. Because the Wello
farmers had nothing left needing protection after the crops were harvested, the age-old prac-
tice of after-math grazing was permitted and continues to date, making the initiatives of bio-
logical conservation measures impossible to realise (7.5.3). Nevertheless, the Wello people are
still very much aware of the fragility of their environment. They live in a famine-prone area
and need to spend a great deal of attention on soil fertility. In other words, Wolaita and Wello
had very different farming systems in which very similar SWC practices were introduced from
the outside. Yet the two areas reacted similarly in some cases, though very differently in
others, to the intervention by the SCRP. In this conclusion chapter we highlight some of the
main results.

8.3.1 Soil erosion vs soil fertility

Researchers and extension personnel tend to emphasise soil erosion for which physical soil
movement is a central issue. In this case, the solution is a physical barrier that maintains soil
in its original place. For farmers, soil movement is not that apparent unless rills and gullies
are created. Their concern with soil is its fertility that they measure using different indicators.
They see its colour (dark or light/white), they compare the trend of crop stand and yield, and
they see its physical characteristics such as weight on the plough shear and plough depth,
stoniness and the like. Based on this observation, they refer to soil fertility rather than to soil
erosion. Therefore, they tend to give soil erosion a lower priority than soil fertility when rank-
ing crop-production problems. In Wolaita, soil erosion was ranked fourth while soil fertility
was ranked second. The corresponding figures for Wello were seventh and third. This view
does not mean that farmers are not concerned with erosion as some scientists want us to
believe: ‘The personal perception of local land users is generally inadequate where erosion is
concerned. Sheet and rill erosion on steep lands are generally perceived as processes, but in
most cases not as hazards threatening agricultural production.’ (Kebede and Hurni, 1992: 4).
To this they add that: ‘…one should not forget that local wisdom coexists with ignorance’.
Implicitly, this means that knowledge transfer must come from outsiders. In my view, the
shortcut to a common understanding is to approach soil erosion through soil fertility, fol-
lowing the reasoning of the farmers, who are, after all, the ultimate decision-makers on land
use. In their view, soil fertility takes precedence over erosion. The entrance point to their
minds and hearts is through soil fertility.

200
.tesfaye.thesis 29-01-2003 15:12 Pagina 201

8.3.2 Side effects of SWC practices

What farmers do or not do is not without reasons. In fact, their reasons are rational from their
own points of view. When they are encountered with a new practice, they always assess its
many sides. One of these aspects is the relative advantage. They enquire as to whether the new
practice solves their problem at cost and how great this cost is to their livelihood.
Farmers in Wolaita and Wello have thoroughly assessed the side effects of the contour
bunds introduced on their farms. They compared the stated soil erosion problem and the los-
ses caused after the installation of the contour bunds. These include: land taken up by the
bund (an amount that reaches up to 25% depending on the slope); weeds and rats breeding in
the bunds; and inconvenience to the ox-plough. To this, they add the annual maintenance
required, that claims a separate labour organisation to their own practices that are mostly car-
ried out together with the annual farm operations.
Outsiders, in their single-minded concentration on soil erosion as a single target variable,
tend to overlook these multiple impacts on the livelihoods of the farmers.

8.3.3 Differential reaction to bunds

At the early stage of the intervention, farmers’ responses to soil and stone bunds in both
Wolaita and Wello were the same. As farmers were not convinced about the SWC practices at
the time of their introduction, they emphasised only the side effects and the temporary food
provision by the project. Implementation of the practices was based on problem assessment
by outsiders, without worrying about farmers’ knowledge and attitudes. The food-for-work
programme used as an incentive allowed for a total neglect of the educational process and
gave space to a top-down approach.
As time went by, farmers’ reactions to the introduced bunds began to diverge in the two
locations. In Wolaita, 54% of the farmers in the treated catchment had removed one to four
bunds by the time of this study (end of 2000). In addition, 15% of them removed all of the
bunds installed under the food-for-work programme. In the entire study area, 48% of the farm-
ers removed 17-100% of the bunds on their land. The proportion of farmers who preserved
their bunds is higher for untreated and distant locations, than for the treated and adjacent
catchments. Of the total farmers, only 10% reported maintenance of their bunds. Removal of
bunds after a number of years shows a ‘dis-adoption’ of the technology that is basically dif-
ferent from a decision not to adopt from the beginning. Such dis-adoption merits special
attention (Moser and Barrett, 2002).
Growth of the negative attitude to soil bunds in Wolaita is due to two main reasons. Firstly,
the research and extension systems were not and still are not proactive in providing farmers
with the necessary information to compare the advantages and disadvantages of fields with
and without soil bunds. Such information would have been worthwhile because farmers
cultivate steep land under heavy rainfall (1,300 mm), which partly occurs during times of low
vegetation cover, making farming prone to erosion. Secondly, farmers themselves did not take
initiatives because of the relatively high soil depth (Weigel, 1986a), heavy reliance on organic
fertiliser and the use of commercial fertiliser on outer fields (Shoka) that conceal the negative
effects of erosion on crop production. Therefore, given the information available to the farm-
ers, their behaviour is rational. Soil erosion or declines in soil fertility did not reach a thresh-
old level to motivate farmers to accept soil bunds.
In Wello, farmers reduced their negative attitudes towards the stone bunds some time after

201
.tesfaye.thesis 29-01-2003 15:12 Pagina 202

their introduction. They began to actively experiment to find out ways to minimise the side
effects they observed in the technology, such as waste of land, harbouring of rats and weeds,
and providing an obstacle to ox-ploughing. In the process, farmers were able to appreciate the
benefits of the bunds that were becoming indispensable as their agricultural zone stretched
into the steep lands and fallow periods declined, leading to a ‘vertical’ expansion which is a
third stage of evolution 68. Their experimentation resulted in the ‘moving bunds’ strategy,
which changed farmers’ perceptions of the introduced bund. The ‘moving bund’ was the
farmers’ own innovation. This bund-management practice involves the destruction of an old
bund and constructing a new one in a selected site on the same plot. The soil ‘harvested’ from
the old bund is distributed over the entire plot (Section 7.5.3). Movement of bunds was facili-
tated because of the mixture of stone and soil whereby they could easily move the stones to a
new location. Note that this would have been much more difficult if only soil bunds had been
involved, as is the case in Wolaita. This practice also shows farmers’ interest on soil fertility
rather than in keeping the soil in its original position according to the scientists’ design.
Hence, farmers in Wello are now voluntary adopters of bunds. However, the types of conser-
vation methods applied by each household are based on the resources available and on plot
characteristics (section 7.5.3). These features should be taken into account in intervention pro-
cesses. At the time of the study, 86% of the farmers preserved their stone bunds with a 94%
maintenance rate.

8.3.4 Conclusion

The SCRP’s 17-year intervention failed to meet its main objective, ‘to develop and promote eco-
logically sound, economically viable and socially acceptable conservation measures’ (Tesfaye,
2002) 69. Besides falling short of its objectives, the process of its implementation and its end
results confirm that the history of intervention world over, as ascertained by many anthropo-
logical studies of development discourse, has a ‘techno-reductionist’ character (Ferguson,
1990; Escobar, 1995; Grillo and Stirrat, 1997). Thus, the project phased out with a narrow tech-
nical achievement that did not lay an adequate foundation for sustainable land management,
as intended (see also Waters-Bayer et al., 1998). The major weaknesses were:
• a uniform, narrow technical intervention that does not take into account socio-economic
variations;
• virtual lack of use of the standard socio-economic data it collected to modify its technical-
ly oriented action over the years that displays a typical gap within monitoring and evalua-
tion efforts;
• soil and water conservation technologies that proved inferior to local practices in terms of
crop production; i.e., they lacked short-term benefit to farmers;
• a single-minded focus on collection of soil erosion and runoff data, not taking into account
other factors;
• in Wolaita, treatment of the community as an experimental plot, dividing it into ‘treated’
and ‘untreated’ sections (see section 6.3.2 for details);
• failure to integrate indigenous land-management practices from the beginning;
• lack of a participatory approach and heavy reliance on food-for-work;

68 Alemneh (1990:28) divides the evolution of agriculture in northern Ethiopia into two. The first stage is ‘outward’ expan-
sion in which new or fallow land is brought under cultivation. The second stage is ‘inward’ expansion, which involves
cultivating the same land more intensively without fallowing.
69 I have also personally discussed with one of the former managers of the SCRP who in retrospect agreed with my argu-
ments on the shortcomings that I point out in this chapter.

202
.tesfaye.thesis 29-01-2003 15:12 Pagina 203

• failure to establish working linkages with the national research and extension system. Even
though it was housed in the MoA, the project created its own space run by an independent
management team, which later backfired, to the destruction of its very existence;
• failure to create an adequate human capacity in the country that would help continuation
of the work that has now lost a national level momentum; and
• lack of sensitivity to the local institutions and changing policy environments.

8.4 Lessons from Konso: the role of the ‘soft’ side of land

Soil and water conservation has long been regarded as a mere technical issue whereby con-
servation engineers were requested to lay down contour bunds, check dams, waterways, etc.
This focus neglects social, cultural, economic and institutional aspects that are increasingly
being considered to be just as crucial as the technical options (Röling, 1997; Gonzalez, 2000,
Mazzucato and Niemeijer, 2000). These studies have shown that technical (hard) options are
viable only when they are embedded in suitable institutions and other soft aspects of land use.
The Konso case brings out the importance of these aspects with a vengeance.
The Konso people started terrace-based farming probably about four centuries ago. Their
terraces cover hundreds of square kilometres, their design and height being the most amaz-
ing part of the artefacts. They built the terraces because of the environment, which is charac-
terised by low rainfall, and by rugged and steep terrain that is highly susceptible to erosion.
They were forced to live in this territory because of political and biophysical factors (malaria
and hot weather). The dominant farming system in Konso is a hoe-complex where only hand
tools are used for all forms of farm operations. Emergence of the ox-plough in part of Konso
is a recent phenomenon.
SWC in Konso is of an integrated nature, based on the stone terraces built during the last
400 years. The terraces are supported by tied-ridges, thrash lines and agroforestry practices,
while soil fertility is augmented by fallowing, manuring and minimum tillage, among others.
These ‘hard’ practices are supported by an elaborate set of institutions invented by the Konso
society over the centuries to allow for the intensive system of land use.
In the first place, Konso society has developed mutual labour organisations that are instru-
mental for creating and ensuring the continuity of the terraces. What is even more amazing
is the treatment of watersheds to capture floods for irrigation that is hardly thinkable in most
arable farming systems of the country. Labour is not only a mere factor of production, it is also
a result of social organisation.
In the second place, the long fallowing system and balanced use of the bush and farm areas
was ensured through a traditional population-control method achieved through the practice
of Fereyuma, which prohibited marriage before the age of 30-35. Culture and institutions
allowed and still allow for different modes of access to land that support land-use practices
that are consistent with the environment.
Owing to the integrated SWC, Konso people moved to controlled grazing and fencing prac-
tices far earlier than the Wello did, and perhaps many other societies, in order to protect the
terraces and the perennial crops from being damaged by animals. These practices allowed a
widespread agroforestry system developed by farmers. These practices are supported by local
rules and regulations rather than a central planning and control system.
The watershed-management practices and care for trees highly benefited from the spiritu-
ality of the Konso people who regarded certain forest bodies as sacred. Their utilisation was
possible only with the blessing of the religious figures.

203
.tesfaye.thesis 29-01-2003 15:12 Pagina 204

All in all, SWC in Konso is not only a ‘hard’ set of technical practices, but it also necessarily
comprises the binding and facilitating ‘soft’ elements that ensure the successful and contin-
ued performance of the hard practices. All these soft elements are an essential feature of
Konso’s agriculture. It is worthwhile to note that the recent deterioration of Konso’s terrace
agriculture is partly due to the collapse of some of the soft elements. For instance, the
abolishment of the Fereyuma since 1974 led to a rapid increase in population that affects the
fallowing system, land use (bush vs farm) and farm size. Also, a food shortage due to the lack
of rain and poor soil fertility disintegrated the complex system of labour organisation.
The Konso case is comparable with the Ifugaos of the Philippines (Gonzalez, 2000) and the
Wakara of Ukara Island in Africa’s Lake Victoria (Netting, 1993) who independently construct-
ed a chain of stone terraces. Although these systems are also under threat because the insti-
tutions required to sustain them cannot be maintained in modern society, they, as does
Konso, teach us that sustainable land use cannot be created without attention to the soft side
of land-use practices.

8.5 Research question 2: what are farmers’ knowledge and attitudes with
respect to soil erosion and SWC
Farmers in Konso, Wolaita and Wello have an adequate knowledge of soil erosion and SWC.
Their attitudes towards the latter are positive with the exception of the Wolaita farmers who
still maintain negative attitudes towards the side effects of the soil bunds introduced by the
Soil Conservation Research Project.
In Wolaita, farmers from the four catchment areas used in that case study were exposed to
a number of questions during this study. Regarding statements intended to measure knowl-
edge of soil erosion and soil fertility, responses of farmers in the treated catchment differed
significantly on certain points from the other three locations (untreated, adjacent and dis-
tant) (Table 6.6). Their responses to the statements on perception also follow the same pattern
(Table 6.7). However, overall, the knowledge and awareness of issues related to erosion and
SWC was very high in Wolaita, approaching a 100% level. Scores on knowledge and perception
statements of Wello farmers showed similarly high results (90-100%).
From this outcome, it can be concluded that farmers in the two study areas were not
different in terms of knowledge and perception of soil erosion and soil and water conserva-
tion. The case of Konso farmers also confirms the same observation. In other words, a lack of
knowledge is not determining farmers’ soil and water conservation practices. They seem very
well informed.

8.6 Research question 3: what do farmers do to conserve soil and water on

their farms?
As shown in detail in Chapters 5, 6 and 7, SWC practices in the three study areas show some
distinct features. The SWC methods observed in the three case study areas include adaptation,
dis-adoption and continuation of practices.
Konso was and still remains an integrated system based on stone terraces. Wolaita by and
large continues with its indigenous practice that is dominated by a biological conservation
system based on organic matter. Use of soil bunds is very limited. Wello has shifted from crop
and soil management in flat and gently sloping land supported by long fallow periods to con-

204
.tesfaye.thesis 29-01-2003 15:12 Pagina 205

tour bunds that include stones (Table 8.1). Indigenous physical conservation was and still is
used on steep land. Contour stone bunds cover at least 60% of the land of the sampled farm-
ers. It was introduced in the early 1980s and further adapted by farmers.
At the plot level, farmers use a combination of SWC practices according to their opportu-
nities. Diversities of SWC methods at the plot level are highest in Konso, followed by Wello and
Wolaita, in that order.
Use of commercial fertilisers is limited. The survey at Wolaita and Wello shows that respec-
tively 68% and 44% of the farmers have used fertiliser at least once in their farms. Wolaita
farmers use more fertiliser than Wello farmers because the former group was exposed to this
input since the early 1970s. Fertiliser utilisation in Konso is the lowest among the three areas.
Low fertiliser use is due to the high price and poor rainfall conditions. When fertiliser con-
sumption among farmers is lower due to higher fertiliser prices, their acceptance of land-com-
peting conservation technology will be lower as well (Bekele, 1998).

Table 8.1: Major soil and water conservation practices in Konso, Wolaita and Wello, Ethiopia

Study areas Practices

Konso Stone terraces

Tied-ridges
Thrash lines
Agroforestry
Minimum tillage
Manure (extensive)
Diversion ditches

Wolaita Manure (extensive)

Soil bunds (limited)
Diversion ditches
Live fences

Wello Diversion ditches

Stone bunds
Traditional Earthen bunds (Dib/Weber)
Manure (limited)

8.6.1 The relationship between farmers’ knowledge, attitude and practice in SWC

Farmers’ knowledge on and attitudes to SWC have a basic, but limited role in determining
their practices. Semgalawe (1998:127) in her study on household adoption behaviour regard-
ing soil conservation in Tanzania found out that ‘the perception of the soil erosion problem
is not a necessary condition for using effective soil and water conservation measures as antic-
ipated. Neither does the perception of the soil erosion problem determine the level of con-
servation investments among adopters.’ However, Bekele (1998) in a similar study at Andit Tid,
Ethiopia, reported a positive and significant relationship between farmers’ decisions to retain
conservation structures and their perception of soil erosion.

205
.tesfaye.thesis 29-01-2003 15:12 Pagina 206

Decisions on a given practice requires suitable biophysical and socio-economic environ-

ments, in addition to adequate knowledge and favourable attitudes. When these factors are
not in congruence, farmers’ practices show divergence from their knowledge and attitudes.
This situation came to pass in the Wolaita and Wello cases. In Wolaita, farmers who agreed to
the knowledge and perception statements by almost 100% showed variation between the four
locations in the use of soil bunds, ranging from 30 to 81% (section 6.5.3). In Wello, thanks to
the farmers’ learning process, the deviation was only 10% (section 7.5.3).
Moreover, one season’s use of irrigation in Wello did not guarantee the continuation of the
practice let alone a mere knowledge, showing the complexity of change processes. It is worth-
while to mention that in Wolaita, 22% of farmers who responded that ‘soil bunds do not have
disadvantages’ removed 100% of the bunds (6.5.3). This observation hints that one cannot
establish a clear casual link between knowledge and attitudes on the one hand and practices
on the other. This means that practices are affected by factors other than knowledge and per-
ception, which brings us to the next research question.

8.7 Research question 4: what are the determinants of soil and water
conservation practices among small farmers within and across
socio-economic and agro-ecological environments?

8.7.1 Land tenure and SWC

There is no strong evidence for the negative effects of land tenure 70 on investment in soil and
water conservation practices where this study was carried out, particularly on arable land (cf.
Yohannes, 1999). However, the negative effects on SWC are more pronounced on the non-
arable (communal) land such as hillsides, roadsides and grazing land. Grazing land is shrink-
ing in size while the incentive mechanism for livestock ownership to increase one’s stock
remains the same. As discussed in connection to the risk perception (below), farmers are reluc-
tant to construct permanent conservation structures on land for which they have no sure title.
The situation for arable land is different. Farmers in Wolaita continued with their traditional
farming practices such as the planting of perennial trees, application of organic matter on
both Darkua and Shoka fields under the land tenure system that has existed since 1975. In
Wello too, farmers designed a mechanism to minimise their loss on soil conservation invest-
ment, but did not cease to maintain existing bunds after they realised their effectiveness in
minimising soil erosion. However, frequent land redistribution programmes that were carried
out in the Amhara region created enough doubt among farmers on the security of their
tenure. As indicated earlier, the situation of Konso is different because of a more stable land-
tenure system. Scoones and Toulmin (1999:62), on the basis of a comparative study from 12
countries in sub-Saharan Africa, state that ‘… providing farmers with security over access to
land, are rarely mentioned within the documents on which this comparison of case study sites
has been based.’ Note that one of these case studies is Wolaita.
Much of the literature on land tenure in Ethiopia argues on the basis of an economic
assumption without taking into account differences in land use in different parts of the coun-
try. Many experts claim that farmers do not invest in soil conservation without secure land
tenure (Merid, 1986; Amare, 1977; Dessaglegn, 1994; Zemenfes, 1995; Yigremew, 1999). The his-
tory of land tenure before 1975 and after differs in north and the south (Chapter 2). Therefore,

70 Note that land tenure in Ethiopia is characterised by state ownership whereby farmers have usufruct rights.

206
.tesfaye.thesis 29-01-2003 15:12 Pagina 207

it is not appropriate to generalise on the impact of land tenure in the entire country. Besides
this, it should be noted that land tenure is only one of several variables that affect farmers’
land-management decisions (Lapar and Pandey, 1999).

8.7.2 Incentives for SWC

The most common incentive for rural people to adopt the external soil and water conserva-
tion practices was food-for-work. However, this programme has been heavily criticised not only
in Ethiopia, but also elsewhere in the world (Giger, 1998; Reij, 1998).
In this study, it was found that the shortcoming was not in the scheme per se, but in the
general patterns of its deployment. The criticism waged against food-for-work here empha-
sised the farmers’ reactions to the introduced SWC methods that included afforestation and
area closures without scrutinising the underlying reasons for farmers’ negative reactions. The
farmers’ negative reactions to the SWC structures in the early days were in response to side
effects, land competition, change of land use, land insecurity, forced measures and faulty
designs (section 6.5.3 and 7.5.3). The community was persuaded to participate through a food-
for-work programme without creating the sense of ownership and learning. Instead, the food-
for-work campaign conveyed the message that both the land and the bunds belong to the
government. Owing to this, farmers commonly stated: ‘They gave us food to construct the
bunds, let them give us food for maintenance too.’ This dependency syndrome was implanted
because the implementation was carried out on the basis of technical diagnosis of experts,
which was not understood by the farmers. The history of their land-management practices
and indigenous knowledge were not taken into account.
In spite of all this, I would say that food-for-work has been a blessing in disguise in Ethiopia.
Had it not been for food-for-work, the proclaimed soil erosion rates in the study areas and per-
haps in the whole country, could have been much higher and the erosion damage more
pronounced. Therefore, I would suggest continuing to support local people to undertake land-
management practices that ensure land maintenance for future generations. This is, however,
not to suggest a reinforcement of past mistakes, but rather to redress it by linking conserva-
tion with poverty-reduction strategies within a learning process. I would argue that mere
withdrawal without helping farmers to be self-sufficient is not only unwise, but also expen-
sive in terms of resources and social stability of the country, at least in the long run. The state
is expected to pursue this policy on behalf of the present and future generations (de Graaff,
1996). In a nutshell, much has changed over the last three decades to think of withdrawing
and leaving everything to the market forces. The most important issue to keep in mind,
however, is that the practice will only be supported if it is accepted by farmers (Giger, 1998),
which can be guaranteed by the incorporation of a farmer learning process into conservation
incentives.

8.7.3 The role of local institutions in SWC

Local institutions are the cornerstones of local development whereby they initiate and ensure
local dynamics (Hounkonnou, 2001). The way people organise their day-to-day life affects the
environment within which they live. This is produced and reproduced in the mode of their
local economy that is affected by a mixture of the social network and the formal exchange
mechanisms of the market forces. Mazzucato and Niemeijer call this a ‘cultural economy’.

207
.tesfaye.thesis 29-01-2003 15:12 Pagina 208

This is based on the notion that people apply a mixture of allocative principles emerging from
both the local institutions and the external market economy. These principles result from the
cultural and historical contexts in which institutions develop (Mazzucato and Niemeijer,
2000).
This study has identified some cultural elements that fit into the cultural economy that
affect land-management practices. These are: access to land, crop and livestock sharing,
indigenous institutions, and mutual assistance and labour exchange. These elements are out-
line below.

Access to land
Land has played a crucial role in the institutions of the communities of the three case study
areas in different ways (Chapter, 2, 5, 6 & 7). Land could be given away as a gift or borrowed in
return for other favours or services to the landowners. In the past, these transactions created
social networks that had far-reaching consequences on the way farmers managed land.
Availability of land to others helps to maintain the traditional practices of land management
such as fallowing, which minimised land mining. In addition, it minimises expansion of farm-
ing into unclaimed bush or hillsides, which could have aggravated the land-degradation
process at much lower population growth rates. Nowadays, access to land is limited to leasing
arrangements that were allowed since 1991. As land becomes scarce, local institutions adjust
to this change through crop and livestock sharing arrangements, as discussed below.

Crop and livestock sharing

When either of the major factors of production (land, labour and capital) is in short supply,
the society creates ways to adjust to these situations. These adjustments are often overlooked
in the rural intervention programmes that often clash with local resource-sharing mecha-
nisms. Sharecropping has become one of the means to augment food production when either
land or labour/oxen is in short supply. Shared livestock rearing also takes place when land or
capital is in short supply. These sharing arrangements are based on mutual trust and are
backed by local rules in case a dispute arises.

Indigenous institutions and mutual assistance

Rural communities still depend on long established mutual assistance organisations. This
assistance includes: support during bad and good times, financial grants and borrowing.
These organisations are common in all case study areas in spite of differences in nationality,
geographic location, language and religion. This assistance essentially provided the rural
people with an insurance system. To that extent, these forms of assistance are the pillars of
the rural economies. Without such mutual assistance, people could have drained their assets
to cover their expenses. This in turn jeopardises land management due to increased migration
which removes labour, a lack of traction power, a shortage of manure and the selling of land
piece by piece. The role of indigenous institutions transcends day-to-day mutual assistance.
They include local governance, which encompasses all functions of society. These functions
were shown by the Xela in Konso, the Idiriya in Wolaita and the Kire in Wello (Chapters 5, 6 and
7). In each case, there are mechanisms to reinforce natural resource management and conflict
resolution. The power of these institutions has greatly diminished since the establishment of
Peasant Associations after the 1975 land reforms. And yet, PAs are very ineffective in ensuring
natural resource management, as illustrated in Chapter 7.

208
.tesfaye.thesis 29-01-2003 15:12 Pagina 209

Labour Exchange
Labour exchange plays a crucial role in land management. Besides getting the job done,
households use labour as a means of transaction to obtain other productive resources (see
Semgalawe, 1998; Lapar and Pandey, 1999). In all three cases, there were indigenous labour
organisations with or without reciprocal labour exchange, though they are stronger in Konso
(Chapters, 5, 6 & 7). The most important issue with respect to labour exchange is the stage and
type of land-management practice in the farming system. More specifically, what are the
kinds of jobs carried out with available labour? Unlike Konso, where terrace maintenance and
construction is in the portfolio of farm operations for which labour is organised, Wolaita and
Wello farmers do not organise labour for bund construction and maintenance. In Wolaita, it
is desired to bring bund construction into routine farm operations, whereas Wello farmers
preferred not to use external labour to undertake such a precise job, but do it by themselves
or in a small group of friends.
Labour exchange in Konso has remained stable until recently, when it weakened under the
threat of food shortages. Wolaita and Wello have been influenced by the substitution of the
ox-plough. As a result, heavy work that involves hand tools has almost been abandoned in
Wolaita. Being in the heart of the system where the ox-plough developed, Wello farmers did
not enter into widespread heavy work on steep lands using hand tools.

8.7.4 Household’s socio-economic characteristics and SWC

A range of household variables is commonly tested to explain farmers’ land-management

behaviour (Amare, 1988; Bekele, 1998; Zealbowessen, 1998; Ludi, 2002). These include: age of
household head, family size, cropland area, livestock owned, labour, access to food, education
and risk perception.
One of the critical farmer behaviours with respect to the introduced bunds was the dis-
adoption or removal of the bunds after many years. The relationship between socio-economic
variables and bund removal was tested using data from Wolaita and Wello. Accordingly, crop-
land, labour and age were negatively correlated with the percent of bunds removed, whereas
livestock ownership and family size were positively correlated. Of these variables, only age and
cropland area were significantly related to the removal of bunds (at p ≤ 0.05). The empirical
observation shows that the statistical relationships between percent of bunds removed and
cropland, family and livestock size are consistent with the actual farmers’ practices. There are
some farmers who removed some bunds because the land became too small for their large
family size, or because of large numbers of livestock needing a greater production of crop
residues instead of putting energy into maintaining more bunds that occupy land. The fol-
lowing paragraphs elaborate on the relevant variables.

Cropland area
As indicated by a fairly strong negative correlation between cropland size and percentage of
bunds removed, land availability seems to be the most important factor that guides farmers’
decision on their choice of SWC methods (see Cramb, et al., 1999; Semgalawe, 1998; Bekele,
1998; Baidu-Forson, 1999). The average total landholding size in Konso, Wolaita and Wello was
1.2, 0.6 and 1 ha per household, respectively. Shortage of off-farm and non-farm employment
is aggravating this problem, causing many families to expect to share land with their chil-
dren. In Wolaita, families who expect to share land with their sons added to 54% whereas the
corresponding figure for Wello was 51%.

209
.tesfaye.thesis 29-01-2003 15:12 Pagina 210

Age
Even though age is negatively and significantly correlated to percentage of bunds removed, an
overview of descriptive statistics and empirical observation do not show a clear pattern of
variation in the SWC practices considered in this study. As shown in section 6.5.3, age of farm-
ers who removed bunds and preserved bunds falls over a wide range. In addition, what is con-
sidered in the analysis is household age only. Moreover, it is contrary to the contribution of
day-to-day learning of farmers, which improves with age. Of course, there are technologies
that are less observable and complex for people under less exposure and who are old (Van den
Ban and Hawkins, 1996:101), from which SWC methods are no exception. For instance, farm-
ers cannot see the biophysical and chemical processes that take place in soil erosion and SWC.
Their constructs do not go to that level, but develop more or less comparable knowledge from
the most observable parts of the phenomena that are very close to their day-to-day experience
(see Cramb, et al., 1999; Semgalawe, 1998; Baidu-Forson, 1999).

Labour
The negative correlation of labour availability with percentage of bunds removed does not
reflect farmers’ practices. What farmers do or do not do is dependent on motivational factors,
among others, rather than a mere presence of labour in the household. No amount of labour
would make people do what does not motivate them. On the other hand, the scarcities of off-
farm employment, relatively low migration rates, and small farm size per family do not sug-
gest the importance of labour in farmers’ responses in this regard (see Cramb, et al., 1999;
Semgalawe, 1998).

Livestock ownership
In this study, livestock ownership affects family income, which has a direct relationship with
family food security (see below). With respect to soil and water conservation, livestock pro-
vides dung for soil fertility; that is, if use of manure is widespread in the farming system
(Eyasu, 1997; Alemayehu, et al., 2001). The income from the sale of animal by-products and live
animals by and large goes toward the home and other expenses (sections 6.5.2 and 7.5.2) rather
than to farm investment. Konso is an exception to these observations. Farmers in Konso invest
in labour organisation, which also includes physical soil and water conservation methods
rather than the soil and crop management practices alone, as in the cases of Wolaita and
Wello. The issue is: do farmers invest in soil and water conservation? If so, how does livestock
ownership contribute to that decision? These and similar questions should be answered based
on actual on-the-ground situations rather than merely through inferences from relationships
from the statistical results.

Access to food
Nowadays, many families are not able to meet their food needs from their own production.
Due to this, they depend on external sources of food for a minimum of five months. They get
food by working for other farmers in their neighbourhood, earn daily work in nearby towns,
sell wood, sell their animals, and occasionally get support from food-for-work and aid pro-
grammes. Lack of food for the family directly affects their labour and energy levels to work on
their farms. This was shown in the Konso case where shortage of food has threatened the tra-
ditional labour organisation that is instrumental to the legacy of maintaining the stone
terrace. Thanks to FARM Africa’s food provision, many kilometres of terraces were maintained,
rural roads and ponds constructed, to mention a few. Indirectly, shortage of food forces farm-
ers to carry out land degrading practices irrespective of their knowledge.

210
.tesfaye.thesis 29-01-2003 15:12 Pagina 211

Education
Most of the farmers interviewed were illiterate. For instance, in Wolaita and Wello, those who
have attained up to secondary school were 28% and 22% respectively. These farmers either quit
school over ten years ago or were recent dropouts. Some of them are dependent on their par-
ents while most of them have their own plots. There was no variation between literacy and
illiteracy rates in terms of SWC practices. The farming practice common in the area applies to
each, with few variations that are affected by household socio-economic characteristics. The
same holds true for the Konso case. However, during group discussions in Konso, it was
revealed that educated farmers who have followed farmer-training programmes better under-
stand the functions of nodules in the legumes, for instance.

Risk perceptions
Risk in agriculture emerges from biophysical and socio-economic environments (Ruthenberg,
1985). This study has shown that shortage of rainfall hampers construction and maintenance
of terraces as exemplified by the case study in Konso (Chapter 5). In Wolaita, loss of crops due
to the side effects associated with soil bunds such as wastage of land, mole-rats and weeds
were the main sources of risks. Variation among farmers from different watersheds was
reflected in the increased incidence of mole-rats and hindrances to ox-ploughing (Table 6.12a).
More interestingly, the risk perception prevails among farmers who did not have a soil bund
on their farm, including those from the distant location. Even though the situation was
improved later, Wello farmers also reflected similar side effects from the stone bunds. With
the exception of Konso where there has been stable land tenure, insecurity of land tenure con-
tributes to farmers’ risk perceptions, which alter their land-management practices. This was
seen on hillsides that were shifted to agricultural land in both Wolaita and Wello. In these
cases, it was widely observed that farmers refrain from investing in durable soil and water
conservation structures, apart from annually constructed structures such as diversion ditch-
es. Farmers’ risk aversion behaviours, though mistakenly considered to be a resistance to
change, are rational in view of what they can perceive (Ruthenberg, 1985:20).
The message of the preceding discussion, related to socio-economic variables, is to draw our
attention to the way farmers actually make decisions instead of concluding from the
behaviour of the statistical model or the relationships among variables. The most important
and practical thing is whether and how two or more variables influence farmers’ decisions.
This approach would make such studies more thrustworthy for policy actions.

8.7.5 Characteristics of SWC technology

As already illustrated in the preceding sections, farmers’ decisions on SWC are affected by
several factors. Characteristics of SWC technology also play their role in such decisions, which
merit attention in intervention. The typical characteristics of technology are observability,
complexity, divisibility, compatibility and relative advantages (Rogers; 1983, 1995).

Observability and complexity

Observability and complexity of a practice hinder or enhance farmers’ ability to understand
the practice and its process of implementation, and more importantly its end results. That is,
when farmers cannot understand the essence of the practice, either due to complexity or
lower observability, they cannot visualise its outcomes. In this respect, farmers in Wolaita and
Wello, unlike their counterparts in Konso, did not understand the watershed management

211
.tesfaye.thesis 29-01-2003 15:12 Pagina 212

design. This is partly due to poor farmer participation, resulting from the focus of the SCRP
on technical data collection, which neglected the crucial socio-economic diversities. Thus,
farmers were not able to understand the process and how it would affect their plots and ulti-
mately their livelihood. Farmers’ risk perception as discussed earlier partly stems from the
fact that the practices did not allow farmers to assess their fates adequately.

Divisibility
Divisibility is a characteristic that is considered with respect to factors of production; that is,
the tools and other physical inputs and labour. With respect to SWC, divisibility commonly
affects labour organisation. In Wolaita, the disappearance of labour co-operation for heavy
farm operations has limited the construction of diversion ditches, which is an accepted
practice. The same holds true for watershed-level conservation activities, in addition to other
factors, e.g., land tenure. Apart from other limiting factors, the running cost of Wello’s irri-
gation pump and the maintenance of an irrigation channel that covers 27 ha of land raise the
divisibility issue.

Compatibility
Compatibility of a practice to the existing farming system is essential to its acceptance.
Looking at the technologies or practices that were introduced into Wolaita and Wello, the
earlier or continued resistance by farmers was due to a lack of an ex-ante assessment of the
introduced practices. Soil/stone bunds in Wolaita and Wello were not compatible with the ox-
plough. This led farmers to modify it to fit their farming practices.
In Wello, pigeon pea and vetch were destroyed because of ox-ploughing and livestock
grazing during the dry period. Introduced grass species attracted rats and were therefore
abandoned by farmers. Animal grazing also destroyed forage species planted in the area clo-
sure. In addition, its incompatibility with the existing land-use system and to the shortage of
land and insecure land tenure contributed to its failure. Furthermore, irrigated agriculture in
Wello failed partly because of the lack of social factors (religious institutions and convictions)
and partly because of land tenure and shortage of land.

Relative advantages
Any decision taken by farmers is weighed against their evaluative frames of reference, which
indicates the relative advantage of a given practice. Soil and stone bunds were compared with
the existing soil fertility maintenance practices and faced rejection both in Wolaita and
Wello, though the situation has gradually changed in the latter case because of farmer adap-
tation. Bunds waste land, they require labour to construct and maintain them, they have side
effects such as weed and rat infestations which increase the risk of crop loss. Such is farmers’
economics, which is oftentimes neglected by outsiders. Note that this is closely linked to the
socio-economic variables, as discussed earlier under household characteristics.

8.7.6 Physical factors

Physical factors are well established in the soil-erosion process and for SWC, serving as a foun-
dation for the technical options. In what follows, physical characteristics that played an
important role in shaping farmers’ SWC decisions are outlined below.

212
.tesfaye.thesis 29-01-2003 15:12 Pagina 213

Topography
This study indicates that farmers take SWC practices more seriously when they are farming
on steep land. Farmers in Konso started a unique SWC system because their landscape is dom-
inated by steep slopes. Farmers in Wello began to take erosion seriously with their movement
up the slope that also affected the situation downstream. As indicated earlier, such is not the
case among farmers in Wolaita even though some of them farm on steeper lands than others
(soil depth, organic matters).

Rainfall
Konso is the area of low rainfall (551mm), whereas Wolaita (1,314 mm) and Wello (1,211mm)
are high rainfall regions. In all cases, however, the erratic nature of rainfall affects the farm-
ers’ production systems by reducing yields of both crops and livestock to the extent of
complete failure of crop production, not only for a season, but also for consecutive years, such
as in the case of Konso. In the latter case, however, moisture conservation takes precedence
over erosion, even though erosion increases with slope gradients. Erosive rainfall of course
alerts farmers to take on conservation measures.

Soil factors
Soil is the medium that is the foundation of all of the factors discussed above. Ultimately, it
is what matters to the farmers most. Its depth affects farmers’ SWC both negatively (Wolaita)
and positively (Wello and Konso). Deep soil in Wolaita conceals the level of erosion. Thanks to
the unseen biophysical and chemical processes, farmers continue to complain about soil fer-
tility, though they tend to give more weight to shortage of rainfall for declining yield. In
Wello, while the rainfall perception is the same to that of farmers in Wolaita, shallow soil
depth brings erosion into greater focus than in Wolaita. In Konso, soil and water are taken
very seriously due to the climatic and physical factors.

Summing up: determinants of soil and water conservation

This section summarises determinants of SWC based on the discussions in the preceding sec-
tions. This study found out that SWC can be affected by the following factors (Table 8.2):
Among variables discussed in section 8.7, the determinants of SWC in the case study areas
are land size, livestock ownership, family size, risk perception, land tenure on non-arable
lands, labour organisation, characteristics of technology, indigenous institutions and physical
factors (see also Table 8.2).
This shows that farmers’ soil and water conservation are affected by the interplay of social,
economic, institutional and technical factors. Even though some factors are more important
than others under a given situation, attention should be given to all of them in order to under-
stand what farmers do in soil and water conservation.

213
.tesfaye.thesis 29-01-2003 15:12 Pagina 214

Table 8.2: Summary of determinants of SWC in Konso, Wolaita and Wello, Ethiopia

Relevant variables Practices

Macro-institutional variables
Land tenure 1
Incentives 2
Extension services 1

Micro-institutional variables
Access to land 3
Crop and livestock sharing 3
Indigenous institutions 3
Labour organisation 2

Household characteristics
Knowledge and attitude 1
Family size 3
Land size 3
Livestock ownership 3
Risk perception 3

Characteristics of technology
Observability and complexity 2
Divisibility 2
Compatibility 3
Relative advantage 3

Physical factors
Topography 3
Rainfall 3
Soil factors 3

* This summary is based on a scale of 1-3, where 1=low, 2=medium and 3=high. The score was given by the researcher himself
based on the discussion in the relevant sections of this chapter.

8.8 Research question 5: how can constraints to promote soil and water
conservation be overcome?
In the earlier parts of this chapter, we discussed farmers’ responses to interventions, their
knowledge and attitudes on soil erosion and SWC in relation to their practices. Then we pre-
sented an overview of determinants of farmers’ soil and water conservation practices. Now we
come to suggestions for improvement and future research. The discussion on these issues
starts with two key points that have been reiterated throughout the study. These are the out-
siders’ views on technological options to soil and water conservation; i.e., biological vs physi-
cal conservation and indigenous vs formal knowledge.

214
.tesfaye.thesis 29-01-2003 15:12 Pagina 215

8.8.1 Biological vs physical conservation

Large-scale SWC projects implemented in Africa since colonial times end in failure (Scoones,
et al., 1996). The dominant conservation methods promoted in these projects regarded physi-
cal conservation. Owing to this, scholars nowadays discriminate against physical conservation
measures in favour of biological measures (Young, 1997:55-56; Hudson, 1993; Stocking and
Abel, 1992). The most commonly mentioned reason is the cost related to the initial construc-
tion and periodic maintenance. As shown in the cases of Konso and Wello, the issue was not
that of cost or labour, but farmers’ motivation to keep up with the practices. Once they per-
ceive and understand the benefits of the practice to their livelihoods, they will undertake it
irrespective of the technical cost-benefit analysis.
Even though one measure dominates the others, farmers in Ethiopia practise a combina-
tion of both physical and biological practices. The proportion of biological and physical con-
servation methods was altered over time, due to changing internal and external situations.
Therefore, these practices do not displace each other (Herweg, 1993; Yohannes, 1999: 57; Ludi,
2002:237). However, an intervention can focus on one of the measures depending on the
comparative advantages in the farming system. In Wolaita, farmers are reluctant to practise
physical conservation measures due to soil depth and their soil fertility practices. Therefore,
they tend to focus on biological conservation measures. In Wello, biological conservation is
very difficult if not impossible, under the existing farming system. This issue is what was
neglected by the SCRP when it initiated biological conservation practices at Maybar, Wello,
leading to its dismal failure (7.5.3). At the moment, the emphasis is on physical conservation,
but the research and extension system should do its best to provide farmers with a basket of
choices for conservation that fits their resources and household objectives. Such a basket of
choices can at best be developed through participatory on-farm research in addition to incen-
tives for changing the cropping system, rather than pushing what outsiders see as appropri-
ate.

8.8.2 Indigenous vs formal knowledge

Nowadays, there is widespread appreciation for the integration of indigenous and formal
knowledge. The issue is: at what stage is an external intervention appropriate? While there is
no blueprint as to the right time and place for an intervention, it can be said that an external
practice would be appropriate only when farmers fail to solve their problems. This means that
farmers themselves should be able to perceive their problem, not the outsiders. This dilemma
can be overcome by changing an age-old dichotomy between the technology system (research,
extension and farmers). This can best be done through joint learning arrangements where all
these parties are involved under one umbrella. The soil fertility status of farms in Wolaita was
studied (Eyassu, 1997) which showed agreement with farmers’ observations. Such analysis
should have been an ideal basis for a participatory action-learning process with farmers by
introducing practices that improve soil fertility, taking the original level as a yardstick. Should
the outsiders have a new idea or practice, they should face the challenge of making farmers
understand it first before pushing the practice on them. This is one of the major shortcom-
ings of soil and water conservation in Ethiopia (Belay, 1998; Yohannes, 1999).
The next section of this chapter elaborates on the key issues indicated above by spelling out
the policy options related to land management.

215
.tesfaye.thesis 29-01-2003 15:12 Pagina 216

8.8.3 Policy implications of the study

On farmers’ participation
Many of the shortcomings in the food-for-work projects, including the SCRP, were due to the
low level of local people’s participation in the design, implementation and evaluation of the
projects. A similar project culture has continued in the country where the present Extension
Intervention Programme is no exception. Farmers who are not using fertilisers are pushed
into participating in the project rather than motivated to do so. This problem occurs partly
because of excess quotas enforced on the development agents. When these agents are short of
farmers to willingly participate in their demonstrations, they use the power of the PAs to
involve them in the demonstration. The development agents and their supervisors should
learn to take low participation as a challenge to be met through educational processes rather
than using the power at their disposal to effect what they need. This is however not to state
that this is the case everywhere. There are regions where farmers hardly grow certain crops
without fertilisers because of an ‘addiction’ of the soil to them stemming from the 1970s.

Accountability of projects
Ethiopia and its farmers have seen dozens of projects since the 1960s. The general trend has
been that activities flourish during the lifetime of the project when staff morale is at its peak
and running costs are available. However, a few years after the project closes, the situation is
almost back to square one. The organisational history of the SCRP confirms this gap in the
country. Both the Ministry of Agriculture and the research organisations did not challenge
the project to be integrated with its national counterparts, which later proved to be disad-
vantageous for the project.
It appears that the country has learned little to date to change this situation. The national
counterpart organisations are not strong enough to ensure accountability of projects.
Research and extension organisations hardly act beyond their traditional operations. The
regional and zonal offices are tied to the routine targets. The administration that has all the
power either lacks vision for development, or when it does have vision, it drives genuine
development activities beyond the limits to achieve its political goal. This was seen in the co-
operative era of the socialist government and it continues today by the democratic
government in its ‘package’ approach.

Diversification of production and income

The results of this study show that the resource base in the study areas is deteriorating to pro-
vide an adequate livelihood for rural people who mainly depend on agricultural production.
Notwithstanding relentless efforts by Ethiopian farmers to make ends meet, agricultural
intensification as advocated by Boserup (1965) and her associates is unlikely under the present
Ethiopian conditions, unless a ‘threshold’ of the theory requires much more human tragedy
to trigger intensification. In this regard, it is worthwhile to note small land size, widespread
soil erosion, removal of organic matter, poor use of inorganic fertilisers, high fertiliser price,
vertical expansion of agricultural lands, forest degradation, overgrazing, shortening and
absence of fallow, poor use or absence of irrigation systems, poor access to a wider credit port-
folio and a poor marketing infrastructure, all of which contribute to worsening food insecu-
rity of the farming communities.
Owing to the settlement pattern and the terrain, agriculture in some areas seems unlikely
to improve to the extent that it can provide an adequate livelihood for rural people under any
amount of feasible intensification in the Ethiopian highlands in the foreseeable future. The

216
.tesfaye.thesis 29-01-2003 15:12 Pagina 217

main problems in this regard are small farm size and fragility. The latter is caused by the con-
tinuous cultivation of the same plots without fallowing.
Therefore, there is a dire need for policy-led intervention to diversify the livelihoods of
rural people within and outside agriculture, including valley development schemes. These
involve, in the short run, the provision of credit to initiate off-farm and non-farm employ-
ment, skills training for non-farm employment, development of infrastructure and support to
farm production through diversification between and within crop, livestock and forest
products. The state should create incentives and conditions that encourage decreasing the
population of the rural people in the long run. These include: improving the quality of
education, family planning and a broad-based economic development that can increase
employment in non-agricultural sectors.

Intensification of crop and livestock production

Agricultural development in Ethiopia during the last three decades focused on external
inputs mainly improved seed and fertilisers 71. Apart from isolated efforts of ill-conceived SWC
practices, conservation-based agriculture in the country is more colourful in the national
study papers than in practice 72.
Thanks to the diversity in farming systems and variation in the level of prior intensifica-
tion, some areas still have adequate room for further intensification. Feasible technological
changes under the Ethiopian conditions are those that save land and use more labour. These
include irrigation, high-yielding varieties, fertilisers, pesticides and soil and water conserva-
tion methods.
Since the mid-1990s, the government has been vigorously promoting an extension service
to farmers. The shortcoming of this intervention is its focus on seed and fertiliser packages
through credit schemes paying lip-service to conservation-based agriculture. What is more,
the programme is implemented in the entire country without due consideration of agro-eco-
logical, socio-economic and institutional variations.
Hence, reorienting the agricultural development approach towards intensification
through conservation-based agriculture is highly desirable. More specifically, it is worthwhile
to shift from the soil and water conservation perspective, to a land-husbandry approach that
fits the farmers’ understanding, such that conservation measures are seen as an input. In this
shift, however, the integration of physical and biological conservation measures should be
maintained for a higher success rate (see Section 8.8.1 on this discussion).

Improvements in training of extension agents

Curricula of the higher learning institutions in the country hardly include indigenous
knowledge. Therefore, the extension agents have negative attitudes towards rural life and tra-
ditional farming methods. They are trained to promote what is seen as modern agriculture,
which focuses on the use of external inputs and practices that are ‘packaged’ by the
research/extension system. Due to this, they do not pay attention to rural knowledge, coping
mechanisms of farmers, and to learn with them and adapt the outside technology. Extension
agents should understand that farmers do what they do out of their own will and under-
standing. Moreover, the agents and their supervisors should understand the variability among
farming communities that will lead to a differential response to intervention. Therefore, the

71 Note the Minimum Package approaches that started in 1971, the Peasant Agricultural Development Project that followed
it in 1986 using the Training and Visit System and the New Extension Intervention that replaced the T&V in 1993/94,
which is operational to date.
72 These include the Ethiopian Highland Reclamation Study (1985), the Ethiopian Forestry Action Plan (1994) and the
National Conservation Strategy (1997).

217
.tesfaye.thesis 29-01-2003 15:12 Pagina 218

training institutions should change their approach of training the extension agents so that
they understand farmers’ logic in land management.

Building on local institutions

Because of bias against indigenous knowledge and institutions, the extension system of the
country has not involved local institutions in development initiatives. Instead, it has
discredited them and then negated their role in the community by replacing them with for-
mal institutions. In order to increase the chances of success, this trend should be changed to
revitalise locally developed institutions. These should be involved in the overall community
development programmes.

On incentives
Undertaking soil and water conservation practices entails foregoing some resources for the
sake of future benefits. Owing to their poor socio-economic status, most Ethiopian farmers are
not in a position to do so. Therefore, some form of incentive is required to support them to
undertake relevant SWC practices on their farms and in watersheds until they begin to reap
the benefits from the new practices.
This study argues that the past food-for-work programme was the right instrument but was
wrongly deployed. However, incentives for rural people should not be limited to food.
Alternatives include: cash, tools, seed, seedlings, infrastructure, and other materials and ser-
vices that the community favours. It would be more effective if incentives targeted socio-eco-
nomic variations observed in the community such as land and family size.
The most important instrument is, however, the dialogue and discussions with rural people
on resource management, rather than the specific material or service offered per se, which is
lacking. Therefore, there is an urgent need to integrate incentives and learning processes into
introducing conservation initiatives.

On land tenure
It was observed that land tenure was not a strong determinant to soil and water conservation
in the study areas, particularly on the arable lands. However, all farmers are aware of the
government’s power over land. It would therefore be worthwhile to provide further security
of tenure to remove any doubt in the minds of the public, particularly regarding redistribu-
tion of cropland and on utilisation and management of non-arable land (forested and
grazing). Generally, such a policy would be effective if accompanied by a sound land-use poli-
cy that takes into account the existing farmers’ situations.

8.9 Issues for further research

8.9.1 Action research into a Maybar irrigation system

One of the intriguing encounters in this study was the resistance of the Maybar farmers to irri-
gation in spite of many good reasons to take advantage of the lake and the external facilities.
This study and others (Ludi, 2002) have suggested plausible reasons as to why the farmers were
not motivated to co-operate in an irrigation programme. The ideal test of these findings is to
go back to the community and enter into a dialogue with it and facilitate a re-initiation of irri-
gated agriculture through action research.

218
.tesfaye.thesis 29-01-2003 15:12 Pagina 219

8.9.2 Testing the status of soil fertility

Even though soil fertility is an important aspect of this study, testing the status of soil fertili-
ty under different SWC practices was beyond the scope of this research. In order to understand
farmers better and also to clarify the technical matters with them, a participatory soil fertility
study whereby farmers’ knowledge and perceptions are integrated needs to be carried out to
discuss suitable conservation measures with farmers. For instance, the plots in the homestead
areas might have an excess stock of some nutrients while having very low levels in others. In
such a case, assessing the SWC practices and examining soil fertility under different practices
would make things very clear for all partners.

8.9.3 Replication of the Konso study in Ankober

Excluding the terraced fields of the Axumite Kingdom (100BC –1000AD) that have long disin-
tegrated, Ankober, located in the Amhara Regional State is another important area next to
Konso where an indigenous terrace-based farming system has developed in Ethiopia.
After undertaking the study in Konso, I intended to include Ankober for an indigenous sys-
tem in the northern setting to compare to that of Wello which falls in the same geo-political
and farming systems. Unfortunately, I could not extend my study to Ankober due to logistical
problems.
In his seminal book, ‘People of the Plow … (1800-1990)’, James McCann (1995) states that ‘the
presence of terracing as a dominant element of Ankober agriculture in 1840 indicates a high-
land population density sufficient to generate the needs for new arable land and sufficient
labour to build and maintain the terraces.’ On the other hand, he quotes a traveller who sug-
gested low population pressure with vast area still remaining under pasture. The focus of the
book is on the general agrarian history without specifically going into the historical reasons
that led the 19th century Ankober people into terracing. The reasons behind the existence of
terraces in Ankober is said to be the climatic restrictions, which stimulated the building of
the terraces. The terraces facilitated the production of mid-altitude crops like teff, wheat and
pulses that were restricted by climatic factors such as frost and flooding in the upper plateau.
Therefore, an understanding of the emergence of terrace-based agriculture in the 19th cen-
tury ox-plough Ankober requires further study on the organisation of labour, local institu-
tions, the impact of mixed land tenure that existed during that period, its influence on the
neighbouring areas, its present status and prospects.

8.9.4 Farmers’ motivation in mass mobilisation for SWC

After use of the food-for-work programme for SWC was abolished in the Tigray and Amhara
Regions, the main source of labour has become community mobilisation organised at the
village and sub-village levels. In this practice, every able community member has to provide
20 days of free labour for the community every year.
Farmers in my case study area in Wello have clearly indicated that they were not interest-
ed in mobilising labour for terrace construction and maintenance. However, this view was not
tested in other Peasant Associations. Therefore, it is worthwhile to carry out a large-scale study
in different regions to avoid the mistakes of the former food-for-work programme in which
the government and technicians assumed that farmers would simply accept their ‘orders’. The

219
.tesfaye.thesis 29-01-2003 15:12 Pagina 220

arguments of this study are to first understand farmers’ rationales before pushing a practice
on their farm.

8.9.5 Periodic production and consumption study

As highlighted in Chapter 3, there is a technical gap in measuring land degradation and the
contribution of SWC to production. Due to this, the farmers’ view on soil erosion, soil fertili-
ty and production is often doubted and associated with a dependency syndrome. While refine-
ment in the techniques of land degradation measurement continues, it is worthwhile to carry
out periodic studies on production and consumption, which is a less difficult factor for com-
parison than land degradation. Such studies would be useful for policy-makers, in providing
a better picture of the farmers’ views on land productivity than what the national statistical
survey might provide.

8.9.6 Comparative study of gender factors in SWC

This study dealt with the farmers in general without targeting the impact of gender differ-
ences in soil and water conservation. The few female-headed households who turned up in the
random samples were not sufficient to be treated as a stratum. Therefore, a study with a gen-
der focus would help to understand the specific problems and perspectives of female-headed
households regarding land management, and more specifically, soil and water conservation.

220
.tesfaye.thesis 29-01-2003 15:12 Pagina 221
.tesfaye.thesis 29-01-2003 15:12 Pagina 222
.tesfaye.thesis 29-01-2003 15:12 Pagina 223

References 73

Ajzen, I. (1988). Attitudes, Personality and Behaviour. Open University Press, Milton Keynes.
Ajzen, I. and Fishbein, M. (1980). Understanding Attitudes and Predicting Social Behaviour.
Englewood-Cliffs, Prentice-Hall. New Jersey.
Alemayehu Konde, Data Dea, Ejigu Jonfa, Fannel Folla, Scoones, Ian, Kelsa Kena, Tesfaye
Berhanu and Worku Tessema (2001). Creating Gardens: The Dynamics of Soil Fertility
Management in Wolaita, Southern Ethiopia. In: Scoones, I. (ed). Dynamics and Diversity:
Soil Fertility and Farming Livelihoods in Africa. Earthscan Publishing Ltd., London.
Alemneh Dejene (1990). Environment, Famine and Politics in Ethiopia: A view from the
Village. Lynne Rienner Publishers, Colorado and London.
Amare Gebreegziabhher (1988). Socio-economic and Ecological Analysis of Soil Conservation
in Ethiopia. Cases from Gunno Twin Catchments. A Thesis submitted in Partial Fulfilment
of the Requirement for the Degree of Master of Science (Management of Natural Resources
and Sustainable Agriculture). NORAGRIC: Norwegian Centre for International Agricultural
Development, Agricultural University of Norway.
Amare Getahun (1977). Raising The Productivity of Peasant Agriculture in Ethiopia. AAASA
Journal IV(1): 27-40.
Ambron, H. (1984). History of Events and Internal Development: The Example of the Burji-
Konso Cluster. In: Taddesse Bayene (ed.), Proceedings of the Eighth International
Conference of Ethiopian Studies in Addis Ababa. Institute of Ethiopian Studies. Addis
Ababa.
ANRS (2000). Proclamation No.46/2000: Proclamation issued to determine the administration
and use of the rural land in the Amhara Nation Region. Zikre Hig, ANRS Bahir Dar.
Aster Akalu (1982). The Process of Land Nationalization in Ethiopia: Land Nationalization and
The Peasants. CWK, Gleerup.
Baidu-Forson, J. (1999). Factors Influencing Adoption of Land Enhancing Technology in the
Sahel: Lessons from a case study in Niger. Agricultural Economics 20(3): 231-239.
Barels, Gerard (2002). Behavioural Modifications as a Transaction between Individual,
Behaviour and the Environment. In: Bartels, Gerard and Wil Nelissen (eds), Marketing for
Sustainability: Towards Transactional Policy-making. IOS Press, Amsterdam, The
Netherlands.
Beedell, J. and Tahir Rehman (2000). Using Social-psychology Models to Understand Farmers’
Conservation Behaviour. Journal of Rural Studies 16:117-127.
Bekele Shiferaw (1998). Peasant Agriculture and Sustainable Land Use in Ethiopia: Economic
Analysis of Constraints and Incentives for Soil Conservation. Published PhD Thesis.
Dissertation no. 1998:1. Agricultural University of Norway.
Bekele-Tesema, A. (1997). A Participatory Agroforestry Approach for Soil and Water
Conservation in Ethiopia. PhD. Dissertation, Wageningen Agricultural University,
Wageningen, The Netherlands.
Belay T. (1992). Farmers’ Perception of Erosion Hazards and Attitudes towards Soil
Conservation in Gunno, Wolaita, Southern Ethiopia. Ethiopian Journal of Development
Research 14(2): 31-58.
Belay Tegene (1998). Potential and Limitations of an Indigenous Structural Soil Conservation
Technology of Wello, Ethiopia. EASSRR XIV(1): 1-18.
Berhe Wolde-Aregay (1996). Twenty Years of Soil Conservation in Ethiopia: A Personal

73 Note that Ethiopian authors are referred to with their first names in the text and written without commas in this sec-
tion.

223
.tesfaye.thesis 29-01-2003 15:12 Pagina 224

Overview. Regional Soil Conservation Unit/SIDA. Nairobi, Kenya.

Biot, Y., Blaikie, P.M., Jackson, C. and Palmer-Jones (1995). Rethinking Research on Land
Degradation in Developing Countries. World Bank Discussion Paper No. 289, World Bank,
Washington, D.C.
Bojö, J. and Cassels, D. (1995). Land Degradation and Rehabilitation in Ethiopia: A reassess-
ment. AFTES Working Paper No. 17. World Bank, Washington, D.C.
Boserup, Ester (1965). The Conditions of Agricultural Growth: The Economics of Agrarian
Change under Population Pressure. George Allen & Unwin LTD, London.
Brouwers, J.H.A.M. (1993). Rural People’s Response to Soil Fertility Decline: The Adja Case
(Benin). Published PhD Dissertation, Wageningen Agricultural University.
Bureau of Agriculture (2000). Office Record. Southern Nations, Nationalities and People’s
Region (SNNAP) (unpublished).
Centre for Development and Environment (CDE) (1997). A Memo: Statement of WB Members
Regarding Soil Degradation in Ethiopia. University of Berne.
Cerulli, E. (1956). Peoples of Southeast Ethiopia and Its Borderland. North-Eastern Africa, Part
III. Ethnographic Survey of Africa. International African Institute, London.
Chambers, R. and Jiggins, J. (1987). Agricultural Research for Resource-Poor Farmers. Part I.
Transfer of Technology and Farming Systems Research. Agricultural Administration and
Extension 27:35-52.
Checkland, P. (1985). From Optimising to Learning: A development of systems thinking for the
1990’s. Journal of Operational Research Society 36(9): 757-767.
Checkland, P. (1989). Soft Systems Methodology. Human Systems Management 8, 273-289. IOS
B.V.
Checkland, P. And Scholes, J. (1990). Soft Systems Methodology in Action. John Wiley & Sons,
Chichester, New York, Brisbane, Toronto and Singapore.
Cohen, J.C. and Weintraub, Dov (1975). Land and Peasants in Imperial Ethiopia: The Social
Background to a Revolution. Van Gorcum & Comp. B.V., Assen, The Netherlands.
Collinson, M. and Lightfoot, C. (2000). The Future of Farming Systems Research. In: M.
Collinson (ed.), A History of Farming Systems Research. FAO, and CABI Publishing, Biddles
Ltd, Guildford, UK.
Constable, M. (1985). Ethiopian Highlands Reclamation Study (Summary). Land Use Planning
and Regulatory Department, Ministry of Agriculture and FAO. UTF.ETH.037/ETH. Working
Paper 24. Land Use Planning and Regulatory Department, Ministry of Agriculture, Addis
Ababa.
Conway, D. (2000). Some Aspects of Climate Variability in the North East Ethiopian Highlands-
Wello and Tigray. SINET: Ethiopian Journal of Science. 23(2):139-161.
Crabtree, B.F and Miller, W.L. (1992). Primary Care Research: A Multi-method Typology and
Qualitative Road Map. In: Crabtree, B.F. and Miller, W.L (eds). Doing Qualitative Research:
Research Methods for Primary Care. Volume 3. Sage Publications. Newbury Park, London
and New Delhi.
Cramb, R.A., J.N.M. Garcia, R.V. Gerrits and G.C. Saguiguit (1999). Smallholder Adoption of Soil
Conservation Technologies: Evidence from Upland Projects in The Philippines. Land
Degradation & Development 10: 405-423.
Critchley, W.R.S. (2000). Groundtruthing: New perspectives on soil erosion and conservation in
the tropics. Ph.D Thesis. Vrije University of Amesterdam, The Netherlands.
CSA (1996). The 1994 Population and Housing Census of Ethiopia. Vol. I: Part I. Statistical
report on population size and characteristics. Federal Democratic Republic of Ethiopia.
Office of Population and Housing Census Commission, Addis Ababa.

224
.tesfaye.thesis 29-01-2003 15:12 Pagina 225

CSA (1998). The 1994 Population and Housing Census of Ethiopia: Results at Country Level. Vol.
I. Central Statistical Authority (CSA), Addis Ababa.
Dagnew Eshete (1995). Seasonal and Exceptional Food Shortages, Their Causes and Socio-
Economic Consequences: A Case Study of Wolaita District in Southern Ethiopia. A Paper
Presented on Inaugural and First Annual Conference of the Agricultural Economics Society
of Ethiopia; 8-9 June, 1995, Addis Ababa, Ethiopia.
Data Dea (1998). Soil Fertility Management in its Social Context: A study of local perception
and practices in Wolaita Southern Ethiopia. Managing Africa’s Soils No.1. Farm Africa.
Russell Press, Nottingham.
De Graaf, Jan (1996). The Price of Soil Erosion: An economic evaluation of soil conservation
and watershed development. Tropical Resource Management Papers No. 14. Wageningen
Agricultural University, Department of Irrigation and Soil and Water Conservation,
Wageningen, The Netherlands.
Dejene Aredo (1999). Property Rights: Theory and illustration with response to land tenure in
Ethiopia. In: Taye Assefa (ed.), Food Security Through Sustainable Land Use: Policy on
Institution, Land Tenure, and Extension Issues in Ethiopia. NOVIB Partners Forum. Addis
Ababa.
Dessalegn Rahmato (1984). Agrarian Reform in Ethiopia. Scandinavian Institute of African
Studies, Uppsala, Sweden.
Dessaglegn Rahmato (1990). Co-operatives, State Farms and Smallholder Production. In:
Pausewang, S, Cheru Fantu, S. Brune, S. and Chole Eshetu (eds). Ethiopia: Rural
Development Options. Zed Books Ltd, London and New Jersey.
Dessalegn Rahmato. (1994). Land Tenure and Land Policy in Ethiopia after the Derg:
Proceedings of the Second Workshop of the Land Tenure Project. Representralen AVH.
Trondheim, Norway.
Dessalegn Rahmato (1994). Policy Issues: Land Policy in Ethiopia at the Cross-roads. In:
Dessalegn, Rahmato (ed). Land Tenure and Land Policy After the Derg: Proceedings of the
Second Workshop of the Land Tenure Project. Representralen AVH, Trondheim, Norway,
Dessalgen Rahmato (1998). Littering the Landscape: Environment and Environmental Policy in
Wollo (Northeast Ethiopia). Paper Prepared for the International Symposium on African
Savannas: New Perspectives on Environmental and Social Change held at the University of
Illinois, Urbana-Champaign.
EARO (2000a). Soils and Water Research Programme Strategic Plan. Ethiopia Agricultural
Research Organisation (EARO), Soil and Water Research Directorate. Addis Ababa.
EARO (2000b). Research-Extension-Farmer Linkages Strategy. Ethiopian Agricultural Research
Organisation, Research-Extension-Farmer Linkages Department. Addis Ababa.
EFAP (1994). Synopsis Report. Ethiopian Forestry Action Plan (EFAP) Secretariat, Addis Ababa.
EFAP (1994). The Challenge of Development. Vol. II. Ethiopian Forestry Action Plan Secretariat,
Addis Ababa.
Eggertsson, T. (1990). Economic Behaviour and Institutions. Cambridge University Press,
Cambridge.
Engel, Paul G.H. (1997). The Social Organization of Innovation: A focus on stakeholder inter-
action. KIT, Amsterdam.
Escobar, A. (1995). Encountering Development: The making and unmaking of the Third World.
Princeton University Press, Princeton, New Jersey.
Eswaran, H., R. Lal and P.F. Reich (2001). Land Degradation: An overview. In. Bridges, E. Michael;
Ian D. Hannam; L. Roel, Oldeman; Frits W.T. Penning, de Vries; Sara J., Scherr and Samran,
Sombatpanit (eds) with assistance from Robin, N. Leslie, Tanadol, Compo and Apuntree,

225
.tesfaye.thesis 29-01-2003 15:12 Pagina 226

Prueksapog. Response to Land Degradation. Oxford & IBH Publishing Co. Pvt. Ltd., New
Delhi and Calcutta.
Ethiopia (1992). National Report on Environment and Development: A report prepared for the
United Nations Conference on Environment and Development in Rio de Janeiro, Brazil.
Transitional Government of Ethiopia Addis Ababa, Ethiopia.
Ethiopia (1995). Proclamation of the Constitution of the Federal Democratic Republic of
Ethiopia. Proclamation No.1/1995. Addis Ababa.
Ethiopian Evangelical Church Mekane Yesus Clinic (nd). Children’s Nutritional Status in
Konso (unpublished).
Eyasu, Elias (1997). Soil Fertility Declines and Coping Strategies: the case of Kindo Koisha,
Southern Ethiopia. Ph.D Thesis University of East Anglia, Norwich, UK.
Eyasu Elias (1998). Is Soil Fertility Declining? Perspectives on Environmental Change in
Southern Ethiopia. Managing Africa’s Soils No.2. Farm Africa. IIED, London.
Eyasu Elias (2000). Soil Enrichment and Depletion in Southern Ethiopia. In: Hilhorst, T. and
Muchena, F. (eds), Nutrients on the move: Soil fertility dynamics in African Farming
Systems. International Institute for Environment and Development, London.
Eyob Belete (2002). Consultancy Report on Socio-economic Survey of Eleven Peasant
Associations in Konso, Special Wereda (Draft Report). UMC-Baseline Survey, FARM Africa
Konso Project (unpublished).
FARM Africa (2001). Capacity Building Project in Konso: Emergency Response Programme
(Employment Generation). Terminal Report. Farm Africa, Konso
Farrington, J. (2000). The Development of Diagnostic Methods in FSR. In: M. Collinson (ed.), A
History of Farming Systems Research. FAO, and CABI Publishing, Biddles Ltd, Guildford, UK.
FDRE (1997a). The Conservation Strategy of Ethiopia. The Resources Base, its Utilization and
Planning for Sustainability. Volume I. Environmental Protection Authority in Collaboration
with the Ministry of Economic Development and Co-operation, Addis Ababa.
FDRE (1997b). Ethiopian Agricultural Research Organisation Proclamation. Federal Negarit
Gazeta. No.42. Addis Ababa.
Feder, G. and Feeny, D. (1991). Land Tenure and Property Rights: Theory and Implications for
Development Policy. The World Bank Economic Review 5(1): 135-153.
Ferguson, J. (1990). The Anti-politics Machine: ‘Development’, Depoliticisation and
Bureaucratic Power in Lesotho. Cambridge University Press, Cambridge.
Gebre-wold Ingida-worq (1962). Ethiopia’s Traditional System of Land Tenure and Taxation.
Translated by Mengesha Gessesse from the book ‘Ya Ityopya Maretna Gibir Sim’ Ethiopian
Observer. V (4):302-339.
Giger, M. (1998). Using Incentives and Subsidies for Sustainable Management of Agricultural
Soils: A challenge for projects and policy-makers. In: Blume, H., H. Eger, E. Fleischhauer, A.
Hebel, C. Reij and K.G. Steiner (eds), Towards Sustainable Land Use: Furthering co-operation
between people and institutions. Advances in Geoecology 31, Volume II. ISSS CATENA VER-
LAG GMBH, Reiskirchen, Germany.
Gold, John R. (1980). An introduction to Behavioural Geography. Oxford University Press, New
York, London, Glasgow.
Goldstein, H. (1981). Social Learning and Change: A Cognitive Approach to Human Services.
Tavistock Publications, New York and London.
Gonzalez, R.M. (2000). Platforms and Terraces: Bridging participation and GIS in Joint-learning
for watershed management with the Ifugaos of the Philippines. Published PhD Thesis.
Wageningen University and Research Centre, Wageningen, The Netherlands.
Grillo, R.D. and Stirrat, R.L. (eds) (1997). Discourses of Development: Anthropological perspec-

226
.tesfaye.thesis 29-01-2003 15:12 Pagina 227

tives. BERG, Oxford and New York.

Guba, E.G. and Lincon, Y.S. (1989). Fourth Generation Evaluation. Sage Publications, Newbury,
London and New Delhi.
Gunten, von Anna-Barbara (1993). Soil Erosion Processes in a Twin Catchment Set-up in Gunno
Area, Ethiopia. Soil Conservation Research Project. Report 23. Ministry of Agriculture,
Ethiopia; University of Berne, Switzerland in association with The United Nations
University, Tokyo. Switzerland
Hallpike, C.R. (1972). The Konso of Ethiopia: A Study of the Values of a Cushitic People. Oxford
University Press, London.
Hardin, G.(1968). The Tragedy of the Commons. Science 162:1243-48.
Hayami, Y. and Ruttan, V.W. (1985). Agricultural Development: An international perspective.
Revised and Expanded Edition. Johns Hopkins University Press, Baltimore and London.
Herweg, K. (1993). Problems of Acceptance and Adaptation of Soil Conservation in Ethiopia. In:
Baum, E., Peter Wolff and Micheal A. Zöbisch (eds), Acceptance of Soil and Water
Conservation: Strategies and Technologies. Volume 3. Tropics in Applied Resources
Management in the Tropics. The German Institute for Tropical and Subtropical Agriculture,
Witzenhausen, Federal Republic of Germany.
Hoben, Allan (1996). The Cultural Construction of Environmental Policy: Paradigms & politics
in Ethiopia. In: Leach, M. and Mearns, R. (eds), The Lie of the Land: Challenging received wis-
dom on the African environment. African Issues. The International African Institute,
London.
Hounkonnou, D. (2001). Listen to the Cradle: Building from local dynamics for Africa renais-
sance. Case studies from rural areas of Benin, Burkina Faso and Ghana. PhD Thesis.
Wageningen University and Research Centre, Wageningen, The Netherlands.
Hudson, N.W. (1993). Soil and Water Management for the Nineties: New pressure, new objec-
tives. In: Hudson, Norman and Cheatle, Rodney (eds), Working with Farmers for Better Land
Husbandry. Intermediate Technology Publications in association with World Association of
Soil and Water Conservation, London.
Huntingford, G.W.B. (1955). The Galla (sic) of Ethiopia, The Kingdoms of Kafa and Janjero.
Ethnographic Survey of Africa. International Africa Institute, London.
Hurni, H. (1988). Options for Conservation of Steep Lands in Subsistence Agriculture Systems.
In: Moldenhauer, W.C. and Hudson, W. (eds), Conservation Farming on Steep Lands. Soil and
Water Conservation Society, Ankeny.
Hurni, H. (1993). Land Degradation, Famine and Resource Scenarios in Ethiopia. In: Pimental,
D. (ed.), World Soil Erosion and Conservation. Cambridge University Press, Cambridge.
Hurni, H. (1994). Methodological Evolution of Soil Conservation Research in Ethiopia. IDRC
Currents 9 (June): 17-21. Swedish University of Agricultural Sciences, Uppsala.
Hurtubise, R. (1984). Managing Information Systems: Concepts and Tools. Kumarain Press,
Jabbar, Mohammad A., Mohammad Saleemand Hugo Li-Pun (2000). From Component
Technology to Integrated Resources management: Evolution toward Transdisciplinarity
Research in a Project in Ethiopia. In: The Proceedings of The International
Transdisciplinarity 2000 Conference. Transdisciplinarity Joint Problem-Solving among
Science, Technology and Society, Workshop I: Dialogue Sessions and Idea Market. Swiss
Federal Institute of Technology, Zurich, Switzerland.
Janke, E. and Getachew Asamenew (1983). Ethiopian Highland Reclamation Study. An
Assessment of the Recent Past and Present Livestock Situation in the Ethiopian Highlands.
Land Use Planning and Regulatory Department, Ministry of Agriculture and FAO.
UTF/ETH/037/ETH. Working Paper 7. Addis Ababa.

227
.tesfaye.thesis 29-01-2003 15:12 Pagina 228

Jones, H. (1990). Population Geography. Paul Chapman Publishing Ltd, London. (2nd edition).
Kassaye Goshu (1997). Biological Soil Conservation Techniques for Maybar Area, Ethiopia.
Research Report 36. University of Berne, Switzerland, in association with the Ministry of
Agriculture, Ethiopia.
Kebede Tato and Hurni, H. (1992). Soil Conservation for Survival. In: Kebede, Tato and Hurni,
H. (eds), Soil Conservation for Survival. Soil and Water Conservation Society in Co-operation
with International Soil Conservation Organisation (ISCO) and World Association of Soil and
Water Conservation (WASWC).
Keeley, J. and Scoones, I. (2000). Knowledge, Power and Politics: The environmental policy-mak-
ing process in Ethiopia. Journal of Modern African Studies 38(I): 89-120. Cambridge
University Press, United Kingdom.
Kelsa Kena (1996). Major Soil Groups of Southern Regions of Ethiopia and Their Land Use
Practices. A Paper Presented at the Workshop on the Environmental Awareness Raising,
Held at Awassa and Mizan-Teferi Towns, October 8-11 and 28-31 respectively. Southern
Nations, Nationalities and People’s Regions, Awassa.
Kerr, J.M. and Sanghi, N.K. (1993). Indigenous Soil and Water Conservation in India’s Semi-Arid
Tropics. In: Baum, E., Peter Wolff and Micheal A. Zöbisch (eds). Acceptance of Soil and Water
Conservation: Strategies and Technologies. Volume 3. Tropics in Applied Resources
Management in the Tropics. The German Institute for Tropical and Subtropical Agriculture,
Witzenhausen, Federal Republic of Germany.
King, C. and Jiggins, J. (2002). A Systemic Model and Theory for Facilitating Social Learning. In:
Leeuwis, C. and Pyburn, R. (eds). Wheelbarrows Full of Frogs: Social learning in rural
resources management. Koninklijke Van Gorcum BV, Assen, The Netherlands.
Kolb, D. (1984). Experiential Learning: Experience as The Source of Learning and Development.
Prentice-Hall, Inc., Englewood Cliffs, New Jersey.
Kothari, B. (2002). Theoretical Streams in Marginalized People’s Knowledge(s): Systems,
Asystems, and Subaltern Knowledge(s). Agriculture and Human Values 19: 225-237.
Kruger, Hans-Joachim, Brehanu Fentaw, Yohannes G. Michael and Kefeni Kajela (1996).
Creating an Inventory of Indigenous SWC Measures in Ethiopia. In: C. Reij, et al. (eds),
Sustaining the Soil: Indigenous Soil and Water Conservation in Africa. Earthscan
Publications Ltd., London.
Laing, D. and Ashby, J.A. (1993). Adoption of Improved Land Management Practices by Resource
Poor Farmers. In: Baum, E., Peter Wolff and Micheal A. Zöbisch (eds). Acceptance of Soil and
Water Conservation: Strategies and Technologies. Volume 3. Tropics in Applied Resources
Management in the Tropics. The German Institute for Tropical and Subtropical Agriculture,
Witzenhausen, Federal Republic of Germany.
Lapar, Ma. L.A. and Pandey, S. (1999). Adoption of Soil Conservation: The case of the Philippine
uplands. Agricultural Economics 21(3): 241-256.
Leach, M. and Mearns, R. (eds) (1996). The Lie of the Land: Challenging Received Wisdom in
African Environmental Change. International African Institute, London.
Lee, David R., Paul J. Ferraro and Christopher B. Barrett (2001). Introduction: Changing
Perspectives on Agricultural Intensification, Economic Development and the Environment.
In. Lee, D.R and Barrett, C.B. (eds), Tradeoffs or Synergies? Agricultural Intensification,
Economic Development and Environment. CAB International.
Leeuwis, C. (1993). Of Computers Myths and Modelling: The social construction of diversity,
knowledge, information and communication technology in Dutch horticulture and agri-
cultural extension. Published PhD Thesis. Wageningen Agriculture University, The
Netherlands.

228
.tesfaye.thesis 29-01-2003 15:12 Pagina 229

Leeuwis, C. (2002). Making Explicit the Social Dimensions of Cognition. In: Leeuwis, C. and
Pyburn, R. (eds), Wheelbarrows Full of Frogs: Social learning in rural resources manage-
ment. Koninklijke Van Gorcum BV. Assen, The Netherlands.
Leeuwis, Cees with van den Ban A. (forthcoming). Communication for Innovation in
Agriculture and Rural Resources Management: Building on the tradition of agricultural
extension. Blackwell Science, Oxford.
Lele, Uma and Stone, Steven W. (1989). Population Pressure, the Environment and Agricultural
Intensification. MADIA Discussion Paper 4. The World Bank, Washington, D.C.
Lomborg, B. (2001). Should we worry about erosion? In: The Skeptical Environmentalist: mea-
suring the real state of the world. Cambridge.
Long, Norman (2001). Development Sociology: Actor perspectives. Routledge, London and New
York.
Ludi, Eva (2002). Economic Analysis of Soil Conservation: Case Studies from Highlands of
Amhara Region, Ethiopia. Published PhD Dissertation, Centre for Development and
Environment (CDE), Institute of Geography, University of Bern, Switzerland.
Lynam, J. (2002). A Book Review: A History of Farming Systems Research, edited by M.Collinson.
Agricultural Systems 73(2): 227-232.
Maarleveld, M. and Dangbégnon, C. (1999). Managing Natural Resources: A Social Learning
Perspective. Agriculture and Human Values 16: 267-280.
Marglin, S.A. (1990). Towards the Decolonisation of the Mind. In: Marglin, F.A. and Marglin,
S.A. (eds), Dominating Knowledge: Development, Culture, and Resistance. Clarendon Press,
Oxford.
Markos Ezra (1997). Demographic Responses to Ecological Degradation and Food Insecurity:
Drought Prone Areas in Northern Ethiopia. Published PhD Thesis. University of Groningen.
The Netherlands. PDOD Publication, Prinseneiland, Amsterdam.
Marsden, D. (1994). Indigenous Management and the Management of Indigenous Knowledge.
In: Scoones, I. and Thompson, John (eds), Beyond Farmer First: Rural people’s knowledge,
agricultural research and extension practice. Intermediate Technology Publications,
International Institute for Environment and Development, London.
Maturana, H.R. and Varela, F.J. (1992). The Tree of Knowledge, the Biological Roots of Human
Understanding. Shambala Publications, Boston.
Mazzucato, V. and Niemeijer, D. (2000). Rethinking Soil and Water Conservation in a Changing
Society: A case study from eastern Burkina Faso. Published PhD Thesis, Wageningen
University, The Netherlands.
McCann, James (1987). From Poverty to Famine in Northeast Ethiopia: A Rural History 1900-
1935. The University of Pennsylvania Press, Philadelphia.
McCann, James (1995). People of the Plow: An agricultural history of Ethiopia, 1800-1900.
University of Wisconsin Press, Madison, Wisconsin.
Mengistu Woube. (1986). Problems of Land Reform Implementation in Rural Ethiopia: A case
study of Dejen and Wolmerea districts. University of Uppsala, Uppsala, Sweden.
Merid, W. Aregay (1986). Land Tenure and Agricultural Productivity, 1500-1850. Proceedings
3rd Annual Seminar, Department of History, Addis Ababa University. Addis Ababa. pp 115-
125.
Meseret Lejebo (1990). A Study of Technological, Social, Socio-economic and Environmental
Constraints Affecting the Performance of Ethiopian Agriculture: With Special Reference to
Drought and Famine Crises in Konso. Giessener Schriften sur Wirtschafts-und
Regionalsoziologie, Heft 18, Giessen.
Metasebia Bekele (1997). Notes on Megalithic Sites of Southern Ethiopia, with Highlights on

229
.tesfaye.thesis 29-01-2003 15:12 Pagina 230

Late Paleolithic Occupation Sites. Ethiopia in Broader Perspective: Vol. 1: Papers of the XIII
th International Conference of Ethiopian Studies, 12-17th December. In: Fukui, K.,
Kurimoto, E. and Shigeta, M. Shodao Book Sellers, Kyoto, Japan.
Millar, J. and Curtis, A. (1999). Challenging the Boundaries of Local and Scientific Knowledge
in Australia: Opportunities for Social Learning in Managing Temperate Upland Pastures.
Agriculture and Human Values 16: 389-399.
Ministry of Water Resources (1999). Koka Reservoir Bathymetry Survey Report. Department of
Hydrology, Addis Ababa.
Moser, Christine M. and Barrett, Christopher B. (2002). The System of Rice Intensification in
Practice: Explaining low farmer adoption and high disadoption in Madagascar. Department
of Applied Economics and Management, Cornell University, Ithaca, NY, USA. (First Draft).
Mulat Demeke (1999). The Challenge of Increasing Food Production in Ethiopia. Proceedings
of the Eighth Annual Conference on the Ethiopian Economy; Oct. 30-Nov.1, 1999, Nazareth,
Ethiopia.
Mulugetta Tesfaye (1988). Soil Conservation Experiments on Cultivated Land in the Maybar
Area, Wello Region, Ethiopia. Soil Conservation Research Project. Research Report 16.
University of Berne, Switzerland in association with the United Nations University, Berne.
Napier, T.L. and Sommers, D.G. (1993). Soil Conservation in the Tropics: A prerequisite for
social development. In: Baum, E., Peter Wolff and Micheal A. Zöbisch (eds), Acceptance of
Soil and Water Conservation: Strategies and Technologies. Volume 3. Tropics in Applied
Resources Management in the Tropics. The German Institute for Tropical and Subtropical
Agriculture, Witzenhausen, Federal Republic of Germany.
Negarit Gazette (1975). Public Ownership of Rural Lands: Proclamation No. 31 of 1975. 34 (26).
Addis Ababa, Ethiopia.
Netting, Robert McC. (1993). Smallholders, Householders: Farm Families and the Ecology of
Intensive, Sustainable Agriculture. Stanford University Press, Stanford, California.
Nichola, T. (1985). Agricultural Research and Extension in Ethiopia. The State of Arts. IDR
Research Report, No.22. A.A.U. Addis Ababa.
NMSA (2000). Data Base. National Meteorological Service Agency, Addis Ababa.
O’Keefe, D.J. (1993). Persuasion: Theory and Research. Sage Publications, Newbury Park,
London and New Delhi.
Ogunfiditimi, T.O. (1993). Abandoned Adoption: Why adopters Discontinued use of previously
adopted innovations. Journal of Extension Systems 9 (2): 86-91.
Östberg, W. and Reij, C. (1998). Culture and Local Knowledge: Their roles in soil and water con-
servation. In: Blume, H., H. Eger, E. Fleischhauer, A. Hebel, C. Reij and K.G. Steiner (eds),
Towards Sustainable Land Use: Furthering co-operation between people and institutions.
Advances in Geoecology 31, Volume II. ISSS CATENA VERLAG GMBH, Reiskirchen, Germany.
Otto Shako (1994). The nine clans of the Konso. Sociology Ethnology Bulletin 1(3): 80-92.
Ovuka, M. (2000). More People, More Erosion? Land Use, Soil Erosion and Soil Productivity in
Murang’a District, Kenya. Land Degradation & Development 11:111-124.
Pankhurst, A. (1990). Resettlement: Policy and Practice. In: Pausewang, S, Cheru Fantu, S.
Brune, Chole Eshetu (eds), Ethiopia: Rural Development Options. Zed Books Ltd., London
and New Jersey.
Pankhurst, R. (1966). State and Land in Ethiopian History. Institute of Ethiopian Studies and
The Faculty of Law, Haile Sellassie I University in association with Oxford University Press,
Addis Ababa.
Pender, J.L. and Kerr, J.M. (1998). Determinants of Farmers’ Indigenous Soil and Water
Conservation Investment in Semi-arid India. Agricultural Economics 19(1-2): 113-125.

230
.tesfaye.thesis 29-01-2003 15:12 Pagina 231

Pinstrup-Andersen, P. and Pandya-Lorch, Rajul (eds) (2001). The Unfinished Agenda:

Perspectives on Overcoming Hunger, Poverty, and Environmental Degradation.
International Food Policy Research Institute, Washington, D.C.
Rasmussen, K., Fog Bjarne and Jens E. Madsen (2001). Desertification in Reverse? Observations
from northern Burkina Faso. Global Environmental Change 11: 271-282.
Reij, C. (1998). How to Increase the Adoption of Improved Land Management Practices by
Farmers? In: Blume, H., H. Eger, E. Fleischhauer, A. Hebel, C. Reij and K.G. Steiner (eds),
Towards Sustainable Land Use: Furthering co-operation between people and institutions.
Advances in Geoecology 31, Volume II. ISSS CATENA VERLAG GMBH, Reiskirchen, Germany.
Renaud, F., H.D Bechstedt and Udomchai Na Nakorn (1998). Farming Systems and Soil
Conservation Practices in a Study Area of Northern Thailand. Mountain Research and
Development 18(4): 345-356.
Richards, P. (1985). Indigenous Agricultural Revolution: Ecology and Food Production in West
Africa. Huchinson, London.
Ritler, Alfons (1997). Land Use, Forests and the Landscape of Ethiopia 1699-1865. An Enquiry
into the Historical Geography of Central-Northern Ethiopia. Soil Conservation Research
Programme, Ethiopia. University of Berne, Switzerland, in association with the Ministry of
Agriculture, Berne.
Robert, C., A. Pacey, and Lori Ann Thrupp (eds) (1989). Farmer First: Farmer innovation and
agricultural research. Intermediate Technology Publications, Short Run Press, Great
Britain.
Rogers, E.M. (1983). Diffusion of Innovations. Third Edition. The Free University Press, New
York.
Rogers, E.M. (1995). Diffusion of Innovations. Fourth Edition. The Free University Press, New
York.
Röling, N. (1981). The Rural Poor and Development Intervention: Final report of the working
party on ‘The small farmer and development co-operation.’ IAC, Wageningen, The
Netherlands.
Röling, N. (1988). Extension Science: Information systems in agricultural development.
Cambridge University Press, Cambridge.
Röling, N. (1992). The Emergence of Knowledge Systems Thinking: A changing perception of
relationships among innovation, knowledge process and configuration. Knowledge and
Policy: International Journal of Transfer and Utilisation 5(1): 42-64.
Röling, N. (1994). Platforms for Decision-making about Ecosystem. In: Fresco, L.O,
L.Stroosnijder, J. Bouma and H. van Keulen (eds), The Future of the Land: Mobilizing and
Integrating Knowledge for Land Use Options. John Willey & Sons Ltd, New York.
Röling, N. (1997). The Soft-side of Land. ITC Journal, Special Issue on Geo Information for
Sustainable Land Management. Vol, 1997, No. 3/ 4 PP:248-263.
Röling, N. and J. Jiggins (1998). The Ecological Knowledge System. In: N. Röling and M.A.E
Wagemakers (eds). Facilitating Sustainable Agriculture. Participatory and Adaptive
Management in Time of Environmental Uncertainty. Cambridge University Press,
Cambridge.
Röling, N. and Jiggins, J. (2001). Agents in Adaptive Collaborative Management: The Logic of
Collective Cognition. In: Buck, L.E, Charles C. Geisler, John Schelhas and Eva Wollenberg
(eds), Biological Diversity: Balancing interests through adaptive collaborative management.
CRC Press, Florida, USA.
Ruthenberg, Hans (1980). Farming Systems in the Tropics. Third Edition. Clarendon Press,
Oxford.

231
.tesfaye.thesis 29-01-2003 15:12 Pagina 232

Ruthenberg, Hans (1985). Innovation Policy for Small Farmers in the Tropics: The Economics
of Technical Innovations for Agricultural Development. Edited by Jahnke, H.E. Clarendon
Press, Oxford.
Sain, G.E. and Barreto, H.J. (1996). The Adoption of Soil Conservation Technology in El
Salvador: Lining productivity and conservation. Journal of Soil and Water Conservation 51
(4): 313-321.
Sandford, J. and Sandford, S. (1994). Land Tenure in an Enset Growing Region. In: Dessalegn,
Rahmato (ed.), Land Tenure and Land Policy After the Derg: Proceedings of the Second
Workshop of the Land Tenure Project. Representralen AVH, Trondheim, Norway.
Scherr, S. and Hazell, P. (1994). Sustainable Agricultural Development Strategies in Fragile
Lands. IFRPI-EPTD Discussion Papers No. 1.
Scoones, I. and Thompson, John (1994). Knowledge, power and agriculture: towards a theoret-
ical understanding. In: Scoones, I. And John Thompson (eds). Beyond Farmer First: Rural
people’s Knowledge, agricultural research and extension practice. Intermediate Technology
Publications, International Institute for Environment and Development. London.
Scoones, I., Chris Reij and Camilla Toulmin (1996). Sustaining the Soil: Indigenous soil and
water conservation in Africa. In: Reij, C., Ian Scoones and Camilla Touliman (eds),
Sustaining the Soil: Indigenous Soil and Water Conservation in Africa. International
Institute for Environment and Development. Earthscan Publications Ltd., London.
Scoones, I. and Toulmin, Camilla (1999). Policies for Soil Fertility Management in Africa. IIED,
London.
Scott, J.C. (1998). Seeing Like a State: How certain schemes to improve the human condition
have failed. Yale University Press, New Haven and London.
SCRP (1996a). Soil Conservation Research Project. Data Base Report (1982-1993). Series II.
Gunno Research Unit. University of Berne, Switzerland, in Association with The Ministry of
Agriculture, Ethiopia.
SCRP (1996b). Soil Conservation Research Project. Supplementary Notes on Soil Erosion
Assessment and Recommendations on Soil and Water Conservation. University of Berne,
Switzerland, in Association with the Ministry of Agriculture, Ethiopia.
SCRP (1996c). Soil Conservation Research Project. Data Base Report (1981-1993). Series I. Maybar
Research Unit. University of Berne, Switzerland, in Association with The Ministry of
Agriculture, Ethiopia.
SCRP (2000a). Area of Gunno, Sidamo, Ethiopia: Long-term Monitoring of the Agricultural
Environment 1981-94. Soil Erosion and Conservation Database. Centre for Development
and Environment, University of Berne, Switzerland, in Association with The Ministry of
Agriculture, Ethiopia.
SCRP (2000b). Area of Maybar, Wello, Ethiopia: Long-term Monitoring of the Agricultural
Environment 1981-1994. Soil Erosion and Conservation Database. Centre for Development
and Environment, University of Berne, Switzerland, in Association with the Ministry of
Agriculture, Ethiopia.
Semgalawe, Z.M. (1998). Household Adoption Behaviour and Agricultural Sustainability in the
Northern Mountains of Tanzania: The case of soil conservation in the North Pare and West
Usambara Mountains. Published PhD Thesis, Wageningen University, Wageningen, The
Netherlands.
Setotaw Ferede, Hugo Verkuijl, Douglas G. Tanner and Takele Gebre (2000). Optimising
Fertiliser Use in Ethiopia: Results of a Baseline Survey of Fertiliser Adoption in Arsi Zone.
Sasakawa-Global 2000, Addis Ababa.
Sfeir-Younis, A. and Dragun, A.K. (1993). Land and Soil Management: Technology, Economics

232
.tesfaye.thesis 29-01-2003 15:12 Pagina 233

and Institutions. Westview Press, Oxford.

Sheleme Beyene (forthcoming). Soil Characteristics at Gunno/Areka CND-Research Site,
Ethiopia.
Smaling, Eric, M.A. Stephen, M. Nandwa and Bert H. Janssen (1997). Soil Fertility in Africa is at
Stake. In: American Society of Agronomy and Soil Science Society. Replenishing Soil
Fertility in Africa. SSSA Special Publication No. 51. Madison, USA.
Song, Y. (1998). ‘New’ Seed in ‘Old’ China: Impact of CIMMYT Collaborative Programme on
Maize Breeding in South-western China. Published PhD Thesis, Wageningen Agricultural
University, Wageningen, The Netherlands.
Sonneveld, B.G.J.S. (2002). Land Under Pressure: The impact of water erosion on food produc-
tion in Ethiopia. Published PhD Dissertation, Vrije Universiteit. Shaker Publishing BV,
Maastricht, The Netherlands.
Stahl, M. (1990). Constraints to Environmental Rehabilitation through People’s Participation
in the Northern Ethiopian Highlands. United Nations Research Institute for Social
Development, Geneva.
Stocking, M. and Abel, N. (1992). Labour Costs: a critical element in soil conservation. In:
Kebede, Tato and Hurni, H. (eds), Soil Conservation for Survival, Soil and Water
Conservation Society in Co-operation with International Soil Conservation Organisation
(ISCO) and World Association of Soil and Water Conservation (WASWC).
Strauss, A. And Corbin, J. (1990). Basics of Qualitative Research: Grounded Theory Procedures
and Techniques. Sage Publications. Newbury Park, London and New Delhi.
Stuiver, M., C. Leeuwis and J.D. van der Ploeg (forthcoming). The Force of Experience: Farmers’
Knowledge and Sustainable Innovations in Agriculture. In: J.S.C. Wiskerke & J.D. van der
Ploeg (eds). Seeds of Transitions: essays on novelty production, niches and regimes in agri-
culture. Van Gorcum, Assen.
Tagoe, C. and Dejene Alemayehu (1983). Ethiopian Highlands Reclamation Study: A Tentative
Review of Agriculture in the Highlands. Land Use Planning and Regulatory Department,
Ministry of Agriculture and FAO. UTF/ETH/037/ETH. Working Paper 1. Addis Ababa.
Takele Tilahun, Bashir Hadj and Bashir Barre (1994). From Communal Grazing to Privatised
Enclosure: a case study of changing land tenure in Region 5. In: Dessalegn Rahmato (ed.),
Land Tenure and Land Policy After the Derg: Proceedings of the Second Workshop of the
Land Tenure Project. Representralen AVH, Trondheim, Norway.
Tefera Mekuria (1996). Environment, Resources and Population in SNNPR. A Paper Presented
at the Workshop on the Environmental Awareness Raising, Held at Awassa and Mizan-Teferi
Towns, October 8-11 and 28-31 respectively. Southern Nations, Nationalities and People’s
Regions, Awassa.
Teferi Abate (1992). Resource use in Konso. Sociology Ethnology Bulletin 1(2): 22-25
Terefe Degefa (2001). Death of the Mother Tree: Land Tenure and Environmental Degradation
in the Oromian Highlands, Ethiopia, 1900-1997. Published PhD Thesis. Institute of Social
Studies, The Netherlands. Shaker Publishing, Maastricht.
Tesfaye Beshah (1997). The Role of Extension in Agricultural and Rural Development in
Ethiopia: With a Particular Reference to Extension Approaches and Strategies. African
Regional Conference Proceedings. Addis Ababa.
Tesfaye Beshah (1999). Orienting Agricultural Extension towards Sustainable Land Use and
Food Security. In: Taye Assefa (ed.), Food Security Through Sustainable Land Use: Policy on
Institutional, Land Tenure, and Extension Issues in Ethiopia. Proceedings of the First
National Workshop of NOVIB Partners Forum on Sustainable Land Use. NOVIB Partners
Forum. Addis Ababa.

233
.tesfaye.thesis 29-01-2003 15:12 Pagina 234

Tesfaye Beshah (2002). Lessons from Soil Conservation Intervention in the Highlands of
Wolaita, Ethiopia. Paper Presented at the International Conference on Sustainable
Agriculture and Rural Development in Mountain Regions, June 16-20, Adelboden,
Switzerland.
Tewelde Brehan Gebre Egziabher (1989). Historical and Socio-Economic Basis of the Ethiopian
Ecological Crisis. Asmara University, Asmara.
Tiffen, M., M. Mortimore and F. Gichuki (1994). Population Growth and Environmental
Recovery: Policy Lessons from Kenya. Gatekeepers Series No. 45. IIED, London.
Tilahun Amede, Takele Belachew and Endrias Geta (2001). Reversing the Degradation of Arable
Land in the Ethiopian Highlands. Managing Africa’s Soils No. 23. IIED, London.
Tolcha Thomas (1991). Aspects of Soil Degradation and Conservation Measures in Agucho
Catchment, Western Harerge. Soil Conservation Research Project. Research Report 19.
University of Berne, Switzerland in association with MoAEP, Ethiopia and The United
Nations University, Berne.
Tsehai Berhane-Selassie (1994). Social Survey of the Soil Conservation Areas, Dizi, Anjeni and
Gunno (Ethiopia). Soil Conservation Research Programme. Centre for Development and
Environment, University of Berne, Switzerland, in association with the Ministry of
Agriculture, Ethiopia and The United Nations University. Research Report 24. Berne.
Van den Ban, A. W. and Hawkins, H.S. (1996). Agricultural Extension. Longman, Burnt Mill,
Harlow.
Van Eijk, Toon (1998). Farming Systems Research and Spirituality: An analysis of the founda-
tions of professionalism in developing sustainable farming systems. Published PhD Thesis,
Wageningen Agricultural University, Wageningen, The Netherlands.
WADU (1981). Wolaita Agricultural Development Unit. Phase II Evaluation Report. WADU,
Addis Ababa.
Wana Wagesho (1998). The History of the Wolaita People (written in Amharic). Brehan Printing
House, Addis Ababa.
Warren, D.M and Cashman, Krishtin (1988). Indigenous Knowledge for Sustainable
Agricultural and Rural Development. Gatekeepers Series No SA10. IIED, London.
Warren, D.M. (1991). Indigenous Agricultural Knowledge and Development. Agriculture, Food
and Human Values Society, Gainesville.
Waters-Bayer, Ann; Alemneh Dejene, Diribu Jemal and Mitiku Haile (1998). Report on the Final
External Review of the Soil Conservation Research Project (SCRP). (unpublished).
Watson, E.E. (1998). Ground Truth: Land and Power in Konso, Ethiopia. PhD Dissertation,
University of Cambridge, UK.
WCMC (1991). Bio-diversity Guide to Ethiopia. World Conservation Monitoring Centre,
Cambridge.
Weigal, G. (1986a). The Soils of the Gunno Area, Sidamo Research Unit, Ethiopia. Soil
Conservation Research Project. Report 8. Ministry of Agriculture, Ethiopia, University of
Berne, Switzerland in association with the United Nations University, Tokyo, Berne.
Weigel, G. (1986b). An Agro-ecological Development Plan for the Maybar Area, Wello Research
Unit, Ethiopia. Soil Conservation Research Project. Report 8. Ministry of Agriculture,
Ethiopia; University of Berne, Switzerland in association with The United Nations
University, Tokyo. Berne.
Weigel, G. (1986c). The Soils of the Maybar Area, Wello Research Unit, Ethiopia. Soil
Conservation Research Project. Report 8. Ministry of Agriculture, Ethiopia; University of
Berne, Switzerland in association with The United Nations University, Tokyo. Berne.
Westphal, E. (1974). Pulses in Ethiopia: Their taxonomy and agricultural significance. Centre

234
.tesfaye.thesis 29-01-2003 15:12 Pagina 235

for Agricultural Publishing and Documentation, Wageningen, The Netherlands.

Westphal, E. (1975). Agricultural Systems in Ethiopia. Wageningen University, The Netherlands
and the College of Agriculture, Ethiopia. PUDOC, Wageningen.
Wilson, K. and Morren, G.E.B. (1990). Systems Approaches for Improvement in Agriculture and
Resources Management. Macmillan Publishing Company, New York.
Woodhill, J. and Röling, N. (1998). The Second Wing of the Eagle: the human dimension in
learning our way to more sustainable futures. In Röling, N. and M.A.E. Wagemakers (eds),
Facilitating Sustainable Agriculture: Participatory learning and adaptive management in
times of environmental uncertainty. Cambridge University Press, United Kingdom.
World Bank (1991). World Development Report: The challenge of development. World Bank,
Washington D.C.
World Bank (1992). World Development Report: Development and the environment. World
Bank, Washington D.C.
World Bank (1996). Natural Resources Degradation in Sub-Saharan Africa (SSA) Restoration of
Soil Fertility: A Concept Paper and Action Plan. Africa Region. World Bank, Washington,
D.C.
World Bank Group (2002). World Development Indicators. The World Bank, Washington D.C.
Yeraswork Admassie (2000). Twenty Years to Nowhere: Property Rights, Land Management and
Conservation in Ethiopia. The Red Sea Press, Inc., Lawrenceville and Asmara.
Yigremew Adal (1999). Ethiopian Rural Land Tenure System: Its Impact on Sustainable Land
Use and Food Production. In: Taye, Assefa (ed.), Food Security Through Sustainable Land
Use: Policy on Institution, Land Tenure, and Extension Issues in Ethiopia. NOVIM Partners
Forum, Addis Ababa.
Yin, R.K. (1984). Case Study Research: Design and Methods. Sage Publications. Beverly Hills,
London, and New Delhi.
Yohannes Gebre Micheal (1998). Indigenous Soil and Water Conservation Practices in Ethiopia.
In: Blume, H., H. Eger, E. Fleischhauer, A. Hebel, C. Reij and K.G. Steiner (eds), Towards
Sustainable Land Use: Furthering co-operation between people and institutions. Advances
in Geoecology 31, Volume II. ISSS CATENA VERLAG GMBH, Reiskirchen, Germany.
Yohannes Gebre Micheal (1999). The Use, Maintenance and Development of Soil and Water
Conservation Measures by Small-scale Farming Households in Different Agro-climatic
Zones of Northern and Southern Wello, Ethiopia. Soil Conservation Research Programme.
Centre for Development and Environment, University of Berne, Switzerland, in association
with the Ministry of Agriculture, Ethiopia. Research Report 44.
Young, A. (1997). Agroforestry for Soil Management. Second Edition. CAB International in asso-
ciation with the International Centre for Research in Agroforestry, New York.
Zealbowessen Asfaw (1998). Analysis of Farmers’ Perception and Adoption of Traditional and
Introduced Soil Conservation Measures: A case of Maybar and Atarymestk, South Wello,
Ethiopia. MSc Thesis in Management of Natural Resources and Sustainable Agriculture.
Agricultural University of Norway.
Zemenfes Tsighe (1995). The Political Economy of Land Degradation in Ethiopia. Journal of
Northeast Africa Studies 2(2): 71-98.

235
.tesfaye.thesis 29-01-2003 15:12 Pagina 236
.tesfaye.thesis 29-01-2003 15:12 Pagina 237

Summary
Soil erosion by water is an old problem in Ethiopia. The prevalence of mountainous and undu-
lating landscapes, coupled with the expansion of arable farming on steep areas due to popu-
lation pressure have aggravated the soil erosion problem in the country.
Prompted by one of the great famines in the country in 1973, the international communi-
ty and the Ethiopian government began to carry out massive conservation measures that cov-
ered extensive areas. Since then, the conservation movement has continued. Ethiopia’s major
research partner in soil and water conservation was the Swiss-financed Soil Conservation
Research Project that operated from 1981 to 1998. In spite of this project, farmers’ practices
have not changed markedly nor have they adopted the introduced conservation practices.
This study was initiated to understand farmers’ soil and water conservation behaviour in
order to improve insights into the land-management problems of the country. For this, farm-
ers’ knowledge and attitudes on soil erosion and soil and water conservation were assessed to
find out what influenced their soil and water conservation practices. Where applicable, farm-
ers’ responses to interventions in soil and water conservation were examined. Finally, the
determinants of soil and water conservation behaviours among farmers were identified. Both
qualitative and quantitative research methods were used to gather and analyse data.
The conceptual framework that guided this study was drawn from perspectives on land
degradation and intensification, theories on planned change and development intervention,
farming systems, indigenous knowledge, social learning, and from an attitude-behaviour
model. Efforts were made to review empirical studies on determinants of soil and water con-
servation practices.
This study was carried out in three areas, namely, Konso, Wolaita and Wello. The first site
provided the case of an indigenous soil and water conservation system, whereas the other two
sites were intervention cases in which an intensive soil and water conservation research and
extension programme was carried out for nearly two decades.
The indigenous system in Konso, into which soil and water conservation methods were inte-
grated is exceptional in Ethiopia and has existed for at least four centuries. It was developed
out of societal needs for survival in a rather fragile environment. This community developed
institutions that provide labour, control population numbers and protect the stone terrace-
based farming system. The strong soil and water conservation elements in this system
emerged because of the harsh physical environment and were sustained by the institutions
developed in the society. As evidence of the importance of local institutions for developing
and maintaining soil and water conservation practices, this system has recently undergone
some deterioration due to the abandonment of its population-control system and its local
labour organisations, as well as due to climatic factors.
The two intervention case studies, though they vary from one another, are markedly dif-
ferent from that of the indigenous system. Their physical resources were much more secure
than those of the indigenous system. Therefore, physical conservation practices did not devel-
op in these two sites on their own, but needed to be prompted from outside. One of these sys-
tems is characterised by the use of organic sources of fertiliser, whereas the other system
mainly depends on long fallow periods and the establishment of staggered structures in the
steeper fields.
This study found that farmers in these two areas responded differently to the introduced
interventions. In Wolaita, farmers refrained from maintaining the introduced bunds, and
removed a considerable number that had been installed through a food-for-work programme;
whereas in Wello, farmers first removed the stone bunds and later on replaced them with

237
.tesfaye.thesis 29-01-2003 15:12 Pagina 238

their own modification of the design, introducing innovative bund-management methods.

These differences resulted because of variations in the existing physical factors and in the soil
fertility management practices. In the former case, the soil is deep and yet organic fertiliser
was used together with a supplement of inorganic fertilisers. In the latter case, in which shal-
low soils dominate, making the situation even worse when the conservation structures were
removed, the use of organic fertiliser is limited to garden plots and inorganic fertiliser is not
widely used. ‘Dis-adoption’ in Wolaita took place because of negative side effects of the soil
bunds, such as mole-rat and weed infestations, and hindrance to ox-ploughing. The underly-
ing problem of the side effects is the fact that there is a shortage of land, which was exacer-
bated by the introduction of bunds. At the Wolaita site, the technology failed to outweigh the
negative side effects perceived by the farmers. In addition, the research and extension systems
also failed to be proactive to demonstrate the pros and cons of the alternate practices, other
than to collect physical data.
The case studies show that there is no lack of knowledge about soil erosion and soil and
water conservation among farmers, nor do they have a problematic attitude to these issues.
They are very much aware of the problems and solutions. However, farmers and outsiders dif-
fer in terms of their perceptions of soil erosion. Unlike outsiders, farmers look at the problem
of erosion through its impact on soil fertility rather than through the actual physical process
involved, which takes the soil away. This contrast was reflected in farmers’ prioritisation of
crop production problems in which they ranked soil erosion far lower than other physical
problems.
There was no strong evidence found, of negative effects of state ownership of land in the
case study areas with the exception of the non-arable communal lands (grazing, hillsides).
The past food-for-work incentive, which allowed for the establishment of conservation
structures in the intervention areas, was not found to be a shortcoming in itself. The provision
of food was an appropriate method of payment for people who could not produce their own.
It can be noted that food produced by these people supports them for only 5-7 months.
However, a problem emerged because of a lack of understanding of farmers’ behaviour regard-
ing soil and water conservation that was aggravated by the absence of a learning process that
should complement a technical intervention. In view of this, it is suggested to improve exten-
sion agents’ training to help them understand farmers’ practices.
The study found out that farmers’ behaviour regarding soil and water conservation is a
result of the interplay among social, economic, institutional and technical factors. More
specifically, the determinants of soil and water conservation in this study are: physical factors
(topography, soils and rainfall), local institutions, labour organisation, land size, family size,
livestock ownership, risk perception, food availability, land tenure on non-arable lands and
characteristics of technology (e.g., relative advantages, compatibility, etc.).
The improvement of Ethiopia’s land-management practices requires the promotion of
farmer participation, diversification production and income, intensification of production
systems, improvement in the training of extension agents, integration of learning processes
into conservation incentives and the implementation of a land-tenure system that reflects
diversity in the country and that links management and utilisation of natural resources in
communal holdings.
Finally, the study underlines the need to carry out location-specific research studies on
land management using a holistic approach.

238
.tesfaye.thesis 29-01-2003 15:12 Pagina 239

Samenvatting
Watererosie is sinds lange tijd een serieus probleem in Ethiopië, met name in de bergen en op
heuvelachtig terrein. Het probleem is de laatste decennia groter geworden door de sterke
bevolkingsgroei en de daarmee samenhangende uitbreiding van het areaal aan akkerland.
Na de ernstige hongersnood van 1973 werden grootschalige bodemconserverings-projecten
door de Ethiopische regering en de internationale gemeenschap uitgevoerd. Deze conserver-
ingsactiviteiten zijn sindsdien doorgegaan. Voor Ethiopië was Zwitserland de belangrijkste
buitenlandse partner voor onderzoek naar erosie en bodemconservering. De Zwitsers
financierden het 'Soil Conservation Project' van 1981 tot 1998. Ondanks dit langdurige en
grootschalige project hebben Ethiopische boeren hun bodemmanagement nauwelijks veran-
derd en bleef de adoptie van erosiebeperkende maatregelen laag.
Deze studie werd uitgevoerd om te begrijpen hoe Ethiopische boeren omgaan met erosie en
bodemconservering, zodat een beter inzicht in het bodembeheer verkregen kan worden.
Hiervoor werd onderzocht wat de kennis en de gebruiken van boeren met betrekking tot
erosie en bodemconservering zijn en wat hun keuzes voor erosiebeperkende maatregelen
beïnvloedt. Zowel kwalitatieve als kwantitatieve onderzoeksmethoden werden toegepast. De
basis voor het onderzoek waren theorieën en inzichten over bedrijfssystemen, intensivering
van bodemgebruik, bodemdegradatie, veranderend landgebruik, traditionele kennis en ken-
nisoverdracht.
De studie werd uitgevoerd in drie gebieden, namelijk Konso, Wolaita en Wello. In Konso
werd een traditioneel bodemconserverings-systeem bestudeerd, terwijl in de andere twee
gebieden gedurende de afgelopen 20 jaar intensieve bodemconserverings-programma's wer-
den uitgevoerd.
Het traditionele systeem in Konso is uniek voor Ethiopië en bestaat minstens vier eeuwen.
Het werd uit nood geboren om te kunnen overleven in een fragiele omgeving. De lokale
gemeenschap ontwikkelde instituties die arbeid leverden, de groei van de bevolking binnen
toelaatbare marges hield en de terraslandbouw beschermde. Beperking van erosie was een
belangrijk element in het landbouwsysteem vanwege de moeilijke fysieke omstandigheden.
De bodemconserverings-structuren werden onderhouden door de verschillende instituties in
de gemeenschap. Het belang van deze lokale instituties voor ontwikkeling en onderhoud van
de structuren wordt benadrukt door het recente verval van deze maatregelen. Dit is voor-
namelijk het gevolg van de ongecontroleerde bevolkingstoename en het teloorgaan van de
arbeidsorganisaties. Ook de veranderde klimatologische omstandigheden spelen een rol.
Hoewel de situaties in Wello en Wolaita niet exact gelijk zijn, verschillen ze beide in grote
mate van Konso. De beschikbaarheid en zekerheid van natuurlijke bestaansbronnen is vele
malen grotere dan in Konso. Hierdoor werden lokaal geen bodemconserverings-maatregelen
ontwikkeld, maar werden geïnitieerd door buitenstaanders. Het systeem in Wolaita wordt
gekarakteriseerd door het gebruik van organische mest, terwijl het systeem in Wello is
gebaseerd op lange braakperioden en het gebruik van erosiebeperkende maatregelen op steile
hellingen.
De boeren in Wello en Wolaita reageerden verschillend op geïntroduceerde maatregelen. In
Wolaita wezen de boeren de geadviseerde terrassen af en ze verwijderden een aanzienlijk aan-
tal terrassen die middels een voedsel-voor-arbeid (food-for-work) project waren aangelegd op
hun velden. In Wello daarentegen werden de terrassen eerst verwijderd, maar later werden ze
weer aangebracht in een door de boeren aangepaste vorm. Dit kan worden beschouwd als een
innovatieve vorm van bodemconservering. De oorzaken van deze verschillen in reacties tussen
Wello en Wolaita lagen in andere omgevingsfactoren en het management van de

239
.tesfaye.thesis 29-01-2003 15:12 Pagina 240

bodemvruchtbaarheid. In Wolaita zijn de bodems diep en maken de boeren gebruik van

organische mest, aangevuld met kunstmest. In Wello zijn de bodems minder diep en is der-
halve de noodzaak voor bodemconservering groter. Organische mest wordt alleen gebruikt in
groentetuinen, terwijl kunstmest slechts sporadisch wordt gebruikt. Het niet adopteren van
terrassen in Wolaita is voornamelijk het gevolg van ongewenste neveneffecten, zoals belem-
mering van grondbewerking en de verspreiding van ratten en onkruiden. Deze neveneffecten
zijn belangrijk voor boeren omdat er relatief weinig land beschikbaar is. De aanleg van ter-
rassen verergert dit probleem. Het bleek dat de landbouwvoorlichting had nagelaten aan de
boeren de voor- en nadelen van terrassen te demonstreren.
In de drie studiegebieden beschikten de boeren over voldoende kennis van erosie. Over het
algemeen werd erosie niet als een groot probleem beschouwd in tegenstelling tot andere prob-
lemen die de gewasproductie beïnvloeden (o.a. droogte, ziekten). Boeren zijn voldoende op de
hoogte van de erosieproblemen en de mogelijke maatregelen. Uit deze studie bleek tevens dat
boeren anders over erosie denken dan buitenstaanders, zoals landbouwvoorlichters en onder-
zoekers. Boeren beschouwen erosie als een probleem dat de bodemvruchtbaarheid beïnvloedt.
Ze hebben niet of nauwelijks aandacht voor de processen die tot bodemverlies leiden, terwijl
buitenstaanders hier met name in geïnteresseerd zijn.
De stimulans voedsel-voor-arbeid (food-for-work) uit het recente verleden maakte het
mogelijk dat bodemconserverings-maatregelen werden aangelegd in twee van de drie
studiegebieden. Deze stimulans was op zich niet slecht. De verschaffing van voedsel aan
mensen die zelf niet voldoende konden produceren was een goede methode van betaling.
Maar er ontstond een probleem doordat de projectuitvoerders geen inzicht hadden in de
meningen van boeren met betrekking tot bodemconservering. Dit probleem werd verergerd
door het ontbreken van een leerproces dat zou moeten samengaan met het plannen en
ontwerpen van maatregelen. Het wordt aanbevolen de training van voorlichters en projectu-
itvoerders op dit punt te verbeteren, zodat ze in de toekomst beter in staat zijn de boerenge-
bruiken te begrijpen.
De houding van boeren ten opzichte van erosie en bodemconservering in de drie studiege-
bieden is het resultaat van de interactie tussen sociale, economische, institutionele en tech-
nische factoren. De belangrijkste factoren die bodemconservering beïnvloeden zijn: fysieke
omgeving (bodem, topografie en klimaat), lokale instituties, gezinsgrootte, landbezit, veebez-
it, voedselzekerheid en de beschikbare technische hulpmiddelen. Voor de verbetering van het
bodemmanagement in Ethiopië zijn nodig: participatie van boeren in projecten, diversificatie
van landbouwproductie en -inkomen, intensivering van landbouwsystemen, verbeterde train-
ing van landbouwvoorlichters, integratie van leerprocessen in bodemconserverings-projecten,
invoering van een systeem voor landbezit dat een afspiegeling van de diversiteit in Ethiopië is
en dat management en gebruik van natuurlijke bestaansbronnen combineert.
Tenslotte blijkt uit deze studie het belang van lokatie-specifiek onderzoek naar bodem-
management waarbij gebruik gemaakt wordt van een holistische benadering.

240
.tesfaye.thesis 29-01-2003 15:12 Pagina 241

Appendix 1

Methods of soil loss and runoff data collection 74

Soil loss and runoff assessment within plots

Individual storm or storm-period data on runoff and soil loss from different land uses, soil
types, slope lengths and gradients were assessed in each research station using test plots (TP)
and micro-plots (MP) from an area of 30 m2 (2x15m) and 3 m2 (1x3m) respectively. Corrugated
iron borders were installed. They were inserted 10 cm in the ground and erected 20 cm above
the ground to enable the runoff to be collected in the tank through an inlet tube. Runoff and
sediments are collected in two tanks each with a 250 litre capacity; the first tank (A) takes
most of the soil from the plot, with the second tank (B) taking 1/10th only of the eventual over-
flow from the first tank (A), through a slot divisor. Activities relating to the monitoring of test
and micro-plots were operated by the research assistants. In order to keep a very close approx-
imation to natural conditions in the catchment, instead of simulating the traditional culti-
vation practices, the test-plot borders were removed whenever the farmers were working on
the field (ploughing), and were set up again afterwards. Both test plots and micro-plots were
arranged side by side.
Plots were established near the rainfall meter to link the results of the rainfall and runoff
period. Measurements of erosion losses were made after each rainy day.
The volume of runoff water was measured directly from the sedimentation tanks.
Representative runoff and soil samples were taken for laboratory analysis. For soil losses, 500
grams of soil and 1 litre of suspended sediment were sampled from the sedimentation tank
(A) and the overflow tank (B), if necessary.
After each sampling, the collection system was emptied and cleaned for subsequent storms.
In the station, the 1 litre sample, taken for sediment assessment was filtered and dried. The
500 g sample taken from each plot was also dried in the oven and weighed. Soil loss was cal-
culated on a per hectare basis.

Soil loss and runoff assessment under experimental plots - Soil conservation impact assess-
ment

Different soil conservation measures were applied on different sites with varying soil types
and slope gradients. This was meant for collecting information about the performance of dif-
ferent soil conservation measures, tested on parallel plots of 180 m2 (6x30 m) each. In all sta-
tions except Anjeni and Afdeyu, level bunds, level Fanya juu, graded bunds, graded Fanya juu
and grass strips were installed to monitor the soil/runoff losses versus a control or tradition-
al plot. Methods used for runoff and soil loss data collection from each experimental plot
were similar to those described in section 1.1 above.

74 Source: SCRP (1996c). Data Base Report (1981-1993). Series I Maybar Research Unit. Note that these procedures were uni-
formly applied in other research stations too.

241
.tesfaye.thesis 29-01-2003 15:12 Pagina 242

Appendix 2
Correlation results of Wolaita and Wello

Percentage of bunds removed

Wolaita (n=82 farmers) Wello (n=399 plots)

Age - 0.145 -.112*

0.193 0.026
Livestock (in number) 0.074 0.009
0.510 .861
Labour (active labour force) - 0.010 - 0.068
0.931 0.175
Family size 0.046 0.056
0.685 0.266
Cropland area (ha) - 0.260* - 0.011
0.018 0.827

* Correlation is significant at the 0.05 level (2-tailed)

242
.tesfaye.thesis 29-01-2003 15:12 Pagina 243

Acronyms
ANRS Amhara National Regional State
AP Agro-Pastoral
AUA Alemaya University of Agriculture
CGIAR Consultative Group for International Agricultural Research
CIMMYT Centro Internacional de Mejoramiento de Maiz y Trigo
CSA Central Statistical Authority
CSE Conservation Strategy of Ethiopia
DAP Di-Ammonium Phosphate
EARO Ethiopian Agricultural Research Organisation
EECMY Ethiopian Evangelical Church Mekane Yesus
EFAP Ethiopian Forestry Action Plan
EHRS Ethiopian Highlands Reclamation Study
EPID Extension Projects Implementation Department
EPRDF Ethiopian People’s Revolutionary Democratic Front
FAO Food and Agriculture Organisation
FDRE Federal Democratic Republic of Ethiopia
FFW Food For Work
GDP Gross Domestic Product
HPC High Potential Cereal
HPP High Potential Perennial
IAR Institute of Agricultural Research
ILRI International Livestock Research Institute
LPC Low Potential Cereal
LSD Least Significant Difference
Masl Meters above sea level
MoA Ministry of Agriculture
NGO Non-Governmental Organisation
PA Peasant Association
PADEP Peasant Agricultural Development Project
PADETES Participatory Demonstration and Training Extension System
PC Producers Co-operatives
PLAR Participatory Learning and Action Reserach
PRA Participatory Rural Appraisal
RAAKS Rapid Appraisal of Agricultural Knowledge Systems
RD Research and Development
RELC Research and Extension Liaison Committee
RRC Relief and Rehabilitation Commission
SC Service Co-operatives
SCRP Soil Conservation Research Project
SNNPR Southern Nations, Nationalities and Peoples Region
SWC Soil and Water Conservation
ToT Transfer of Technology
WADU Wolaita Agricultural Development Unit
WCMC World Conservation Monitoring Centre
WFP World Food Programme

243
.tesfaye.thesis 29-01-2003 15:12 Pagina 244

List of Local Terms

Bega: The dry season of Ethiopia covering October to February
Darkua: The fertile zone of farm plots in Wolaita
Derg: The Military Government that ruled Ethiopia from 1974-1991
Dib: A physical structure constructed in the hill-side to protect soil from eroding
Fereyuma: A custom local to Konso, which forbade marriage until about the age of 30 or
more
Kiremt: The main rainy season of Ethiopia covering June-August/September
Shoka: The low fertility zone of farm farm plots in Wolaita
Teff: A small cereal grown and used as human food in Ethiopia
Wereda: An administrative body above the Peasant Association that is equivalent to a
district
Xela: The indigenous administative assembly in Konso
Waaqa: God

244
.tesfaye.thesis 29-01-2003 15:12 Pagina 245

About the Author

Tesfaye Beshah Asfaw was born at Jarso, Hararghie, eastern Ethiopia on January 18 1962. He
completed primary and secondary schools under the aegis of the then Alemaya College of
Agriculture, in Alemaya, Ethiopia. He obtained a BSc degree in Agricultural Economics in July
1986 from the Alemaya University of Agriculture (AUA). Following that, he was employed by
AUA in September 1986. After two years of service as a graduate assistant at Alemaya, he won
a scholarship at the Wageningen Agricultural University, The Netherlands where he obtained
a MSc degree in the Management of Agricultural Knowledge Systems (MAKS) in July 1990.
Returning to AUA, he served the university in different capacities until October 1997, with the
exception of a semester in 1996 when he was granted an in-country study leave. During his
study leave he was associated with the United Nations Economic Commission for Africa
(UNECA) where he participated in the Sustainable Agriculture and Environmental
Rehabilitation Programme (SAERP) in Ethiopia. During his service years at the AUA, he served
as Research Coordinator and head of the Department of Agricultural Extension besides teach-
ing and research responsibilities.
In addition to his duties and responsibilities at the Alemaya University, the author partici-
pates in professional activities in the country through contributions in workshops and semi-
nars in his main area of interest, i.e., agricultural extension and rural development.
Furthermore, he has worked with some local and international NGOs in Ethiopia, in various
consultancy capacities. These opportunities together with the study leave at UNECA have
immensely contributed to the professional development of the author.
In October 1997, he was granted a scholarship to pursue a PhD study at Wageningen
University and Research Centre, with the Communication and Innovation Studies Group,
which is concluding with this volume.

245
.tesfaye.thesis 29-01-2003 15:12 Pagina 246
.tesfaye.thesis 29-01-2003 15:12 Pagina 247
.tesfaye.thesis 29-01-2003 15:12 Pagina 248

All photos and illustrations not acknowledged were made by the author.
Cover and interior design: Luc Dinnissen (studio DS, Nijmegen)