Contextual drivers of climate-smart agroforestry adoption in Bugesera and Rulindo agroecosystems of Rwanda

This study examined 381 farmers from two regions in Rwanda to investigate how contextual factors at the field level interact with climate-smart agroforestry (CSAF) practices. Farmers were categorized as low (LAD), medium (MAD), and high (HAD) adopters based on tree counts. Various contextual factors — notably location, demographics, assets, farm characteristics, and institutional variables — were analyzed using descriptive statistics, Pearson correlation, logit regression, and propensity score matching. Farmers in Bugesera had larger farms and higher tree counts than those in Rulindo, resulting in greater farm income in Bugesera. Positive correlations were found among altitude, slope, erosion class, gender, household size, poverty level, income source, marital status, education, farm area, cropping practices, farm-river distance, changes in CSAF cover, population dynamics, and LAD. CSAF farms outperformed monoculture farms regarding cassava, maize, and bean yields, particularly in Bugesera and Rulindo among larger landholdings. Logit regression analysis showed that combinations of multipurpose trees and crop planting significantly improved farm yields, with household size and farm size being critical factors for CSAF adoption. Propensity score matching confirmed the positive effects of CSAF practices on farm yield and income, contributing to enhanced rural well-being. These findings underscore the crucial role of CSAF in promoting well-being. The results encourage stakeholders to develop strategies for CSAF. While these findings are specific to local contexts, they may hold potential relevance at regional and global levels. This evidence supports the development of government-led policies implemented through extension services to systematize and stabilize CSAF practices across diverse farming systems.