Comparative Analysis of Agroforestry Adoption and Yield Across Regions

Thesis Abstract

Smallholder farmers in timber-poor landscapes increasingly adopt agroforestry to diversify livelihoods, enhance carbon sequestration, and improve ecosystem services; however, adoption rates and realized yields vary across regional contexts, obscuring the generalizable determinants of adoption and production outcomes. This study addresses the problem of heterogeneous agroforestry uptake and performance by conducting a cross-regional analysis that links socio-economic, agronomic, and biophysical factors to adoption status and yield. The aim is to identify the drivers of agroforestry adoption and quantify its impact on crop and tree yields across contrasting regions, thereby informing scalable policy and extension strategies. Specific objectives are to (i) compare adoption prevalence and agroforestry system types across three typologically distinct regions; (ii) assess how farm characteristics (farm size, access to credit, labor, and tenure security), social inputs (extension contact, farmer knowledge, and peer influence), and biophysical conditions (soil fertility, rainfall variability, and slope) influence adoption decisions; (iii) estimate the effect of agroforestry adoption on aggregate yields and on component yields for staple crops and tree outputs; and (iv) test regional differences in the yield effects and identify context-specific mechanisms driving performance. The study adopts a mixed-methods design anchored in the Diffusion of Innovation and Technology Acceptance theories to illuminate how perceptions, incentives, and constraints shape adoption, while integrating a cultivation-performance framework to interpret yield outcomes. The population comprises smallholder farms in three regions with distinct agroecological zones and market access profiles. A stratified random sample of 600 farms (200 per region) is selected, with proportional representation of agroforestry adopters and non-adopters. Data collection combines structured household surveys (n = 600) capturing demographic attributes, land tenure, input use, risk preferences, and market linkage; farm records and yield data for the last three production seasons; and agroecological measurements including soil tests (pH, organic carbon, nutrient status) and climate records (precipitation, temperature). Key instruments include a validated adoption questionnaire, a yield-monitoring logbook, and an expert-assisted soil and landscape assessment protocol. To triangulate quantitative findings, 40 in-depth interviews with farmers, extension agents, and local leaders will be conducted, and 12 focus groups will explore contextual factors and barriers. Quantitative analysis employs multilevel logistic regression to model adoption probability as a function of individual and regional covariates, and a double-hounded fixed-effects model to estimate the causal impact of adoption on yields, controlling for input use and plot characteristics. ANOVA and pairwise comparison tests will examine regional yield gaps between adopters and non-adopters, while propensity score matching will mitigate selection bias. Mediation analysis will explore pathways through soil improvement, microclimate modification, and enhanced soil-water retention as mechanisms linking agroforestry to yield gains. The study will test for cross-regional heterogeneity by including interaction terms between region and key determinants, and by performing region-specific models. Data quality will be ensured through instrument validation, pre-testing, and reliability analysis (Cronbach’s alpha for scales; test-retest reliability for key measures). Expected findings include higher adoption rates in regions with stronger extension networks and secure land tenure, with adopters showing statistically significant improvements in both crop yields and tree-based outputs, particularly where soil fertility and moisture stress are limiting. The study anticipates that yield gains are moderated by farm size and access to credit, with diminishing returns on very small plots. Theoretical contributions will refine diffusion and technology acceptance models in the context of agroforestry by incorporating biophysical feedback and yield-mediated utilities. Practical implications include region-specific policy recommendations for subsidies, credit schemes, and extension curricula, as well as guidelines for selecting agroforestry configurations that optimize both productivity and resilience under climate variability. The study concludes that agroforestry adoption yields net benefits with substantial regional variation, driven by institutional, biophysical, and economic factors; it recommends targeted extension services, secure land tenure reforms, integrated soil and water management practices, and adaptive agroforestry design guided by region-specific constraints to maximize adoption and agronomic performance.