Diss. Phil.-naturwiss. Fak. Bern, 1998.

Universiẗat, Diss., 2002--Hohenheim.

Techn. Universiẗat, Diss., 2004--Berlin.

Zugl.: Dresden, Techn. University, Diss., 2008.

Techn. University, Diss., 2001--Berlin.

Thesis (Ph. D.)--Universität, Hohenheim, 2006.

Universit́e, Diss., 2004--Heidelberg.

Agroforestry (AF) or agroforestry innovation (AFI) production has long been and continues to be a component of the mixed farming system of Ethiopian and smallholder farmers worldwide. Interventions continue introducing new or improved management practices, species, and techniques to raise AFI's livelihood and natural resource management contributions. Despite considerable efforts, the adoption of these AFI continues to be limited, as proved by several adoption studies and development efforts. Formal and informal studies were conducted for decades to understand the problems for the low adoption of various AFI. Nevertheless, these studies generated redundant and marginally growing important information as it has weakly altered the course of development approaches and policy regulations. Learning from previous studies, researchers have been requesting more robust studies that help address existing knowledge gaps on adopting AFI. To respond to these calls, this PhD project examined the factors affecting the adoption of AFI by smallholders and Ethiopian farmers as a case study. The project builds upon previous studies to explore the diverse perspectives that influence the adoption of AFI. Literature assessment of recent studies indicated that several factors belonging to farmers and institutions influence the adoption of AFI. Simultaneously, we discovered that some issues were explored frequently (e.g., socioeconomic factors), whereas others (e.g., psychological factors) were largely ignored. Besides, researchers have followed the static assumption (i.e., adopt or non-adopt) and failed to learn the adoption process beyond a one-time decision. Additionally, the studies focused on discrete factors and activities and failed to comprehend the diverse perspectives and factors and their combined effect on eventual AFI adoption. Ultimately, learning from the larger adoption science and previous studies, we developed a comprehensive framework, 'AFI adoption framework' (chapter 4.1), that supports the meaningful assessment of adoption practices and comprehensively discovers factors influencing AFI adoption. The framework encompassed three compartmentalized and yet interlinked components that influence AFI adoption under smallholder contexts. The framework commended both distinct studies for exhaustive elaboration and simultaneously suggested holistic examination. Besides, it recommended minor and major modifications to the research approaches, such as proper treatment of variables in econometric models, incorporation of variables related to the psychological status, and employment of robust tools such as the real-options approach for profitability analysis. Based on this framework, we designed a project and conducted fieldwork in the Amhara region of Ethiopia, a typical smallholder context. We explored the household contexts (i.e., farm level and psychological), system level features, and innovation characteristics influencing smallholders' AFI adoption decisions. It employed mixed conventional and advanced analytical tools comprising content analysis, econometric models, principal component analysis, and financial discounting methods. Advanced methods comprehend process analysis and adoption dynamism. The results from discrete analysis indicated that socioeconomic factors, psychological constructs, system level features, and innovation attributes influence AFI adoption. Regarding innovation characteristics, the different attributes are foundations for undertaking AFI adoption decisions of smallholder farmers. Beyond adopt-non-adopt concepts, we found farmers continuously undertake follow-up adoptions of varying extents such as reduced, maintained, and increased. Based on our query and comparable to existing frameworks, the newly developed 'AFI adoption framework' is more reasonable to meaningfully investigate factors influencing AFI (and agricultural innovations) adoption under smallholder contexts. However, there is a need for precaution while employing the framework to more clearly discern the adoption process and reflect the integration among the factors and activities involved from the development to the adoption of AFI. This dissertation excluded empirical analysis of profitability and holistic assessment due to the voluminous nature of the dissertation.:PREFACE ii ACKNOWLEDGEMENTS iii SUMMARY iv ZUSAMMENFASSUNG vi LIST OF FIGURES ix LIST OF TABLES xi ACRONYMS xi 1. INTRODUCTION 1 1.1. Agroforestry in Ethiopia 1 1.2. Problem statement 4 1.3. Objectives and research questions 6 1.4. Scope of the study 7 1.5. Dissertation structure 8 2. CONCEPTUAL FRAMEWORK 11 2.1. The adoption concept 11 2.2. Theoretical frameworks on adoption 12 2.3. The critique and research context 16 2.4. The AFI adoption analytical framework 17 2.5. Description of links between objectives and research questions 19 3. RESEARCH METHODOLOGY 21 3.1. Description of the study area 21 3.2. Selection of innovations and farmers 22 3.3. Research methods 23 3.3.1. Data collection methods 23 3.3.2. Sampling technique and sample size 24 3.3.3. Data analysis 25 4. RESULTS 26 4.1. Agroforestry adoption as a systems concept: a review 27 4.2. Can a sequential analysis provide a more robust understanding of farmers’ adoption decisions? an example from an agroforestry adoption study in Ethiopia 58 4.3. Farmers’ intentions towards sustained agroforestry adoption: an application of the theory of planned behavior 88 4.4. Adoption under the influence of innovation attributes: the case of agroforestry innovations from Ethiopia 111 4.5. Influence of system level factors on adoption of agroforestry innovations 141 5. SYNTHESIS and CONCLUSION 170 5.1. Synthesis of key findings 170 5.1.1. State of AFI adoption research in SSA 170 5.1.2. Persistent calls for rigorous research 172 5.1.3. Critical factors affecting AFI adoption 173 5.1.4. Conceptualizing adoption as a complex decision process 175 5.2. Reflections on research method, theoretical framework, and generalization 177 5.2.1. Reflection on research methods and analytical generalization 177 5.2.2. Reflection on the theoretical framework and theoretical contribution 179 5.3. Outlook and suggestions 184 5.4.1. Recommendations for future research 185 5.4.2. Development and policy recommendations 186 5.5. Limitations of the study 186 REFERENCES 187 APPENDICES 192

In this study, farm forestry decision criteria were elicited by adopting a behavioral decision-making study approach from households in ten Peasant Associations of the study district. Major decision objectives, available alternatives, constraints, and the likelihood of the chance events were elicited through a questionnaire survey, participatory observation, detailed discussion, and review of archival information. It was found that farmers generally, plant various tree and shrub species for meeting various household needs and for generating cash incomes. The goal of cash generation is for sustenance of livelihoods through fulfillment of various basic obligations and overcoming unforeseen contingencies. The three most economically important tree/shrub species were eucalypts, coffee, and t?chat. Eucalypt wood covers nearly all the construction needs and constitutes a substantial part of the fuelwood consumptions. Coffee and t?chat are vital sources of cash income and makeup an important part of daily diets. Growing of eucalypts for cash income is mainly constrained by lack of access road, low farm gate prices, high competition with food crops for soil nutrients and moisture, and shortage of land and labor. The household uses and cash values of coffee are generally, undermined by high incidence of berry disease and lack of manure. Financial benefits of t?chat are weakened by high local tax rates. The logistic regression analysis confirms that agro-ecological zone, sex of household head, number of eucalypt trees owned, and age of household head represent important explanatory variables that explain farmers readiness to expand eucalypt woodlots. The model so constructed correctly predicted 84.1 % of the households that established additional eucalypt woodlots mainly for cash generation. The total number of eucalypt trees owned by households is significantly related to the attitude of the household head towards eucalypts, wealth status, and landholding size of the household. Financial viability of eucalypt woodlots was assessed through both methods of conventional economic calculations and Chayanovian calculations. Both methods confirmed the highly lucrative markets of eucalypt poles as compared to agricultural crop production. This is mainly because of lack of access to more profitable production techniques and low productivity of agricultural crops per unit area. Otherwise, farm gate prices of eucalypt poles are far from being attractive and outperforming that of agricultural crops.

In this study, farm forestry decision criteria were elicited by adopting a behavioral decision-making study approach from households in ten Peasant Associations of the study district. Major decision objectives, available alternatives, constraints, and the likelihood of the chance events were elicited through a questionnaire survey, participatory observation, detailed discussion, and review of archival information. It was found that farmers generally, plant various tree and shrub species for meeting various household needs and for generating cash incomes. The goal of cash generation is for sustenance of livelihoods through fulfillment of various basic obligations and overcoming unforeseen contingencies. The three most economically important tree/shrub species were eucalypts, coffee, and t?chat. Eucalypt wood covers nearly all the construction needs and constitutes a substantial part of the fuelwood consumptions. Coffee and t?chat are vital sources of cash income and makeup an important part of daily diets. Growing of eucalypts for cash income is mainly constrained by lack of access road, low farm gate prices, high competition with food crops for soil nutrients and moisture, and shortage of land and labor. The household uses and cash values of coffee are generally, undermined by high incidence of berry disease and lack of manure. Financial benefits of t?chat are weakened by high local tax rates. The logistic regression analysis confirms that agro-ecological zone, sex of household head, number of eucalypt trees owned, and age of household head represent important explanatory variables that explain farmers readiness to expand eucalypt woodlots. The model so constructed correctly predicted 84.1 % of the households that established additional eucalypt woodlots mainly for cash generation. The total number of eucalypt trees owned by households is significantly related to the attitude of the household head towards eucalypts, wealth status, and landholding size of the household. Financial viability of eucalypt woodlots was assessed through both methods of conventional economic calculations and Chayanovian calculations. Both methods confirmed the highly lucrative markets of eucalypt poles as compared to agricultural crop production. This is mainly because of lack of access to more profitable production techniques and low productivity of agricultural crops per unit area. Otherwise, farm gate prices of eucalypt poles are far from being attractive and outperforming that of agricultural crops.