Household livelihood improvements through agroforestry under climate variability in semi-arid Indonesia

Agroforestry is widely promoted as a nature-based solution for enhancing rural resilience under climate change, yet empirical evidence on its relationship with livelihood outcomes across multiple dimensions remains limited, particularly in semi-arid regions. This study examines the association between agroforestry characteristics and household livelihood outcomes through ecological, economic, social, and physical pathways under climate variability in semi-arid eastern Indonesia. Using household survey data from 122 agroforestry households in Sumba Barat Daya, we constructed composite indices of perceived climate variability and multidimensional livelihood outcomes and applied pathway-specific OLS models with HC3-robust standard errors, complemented by a bootstrap-based decomposition of pathway contributions. The results indicate that agroforestry characteristics are positively associated with ecological improvement (coef = 0.1672, p < 0.01), representing the dominant pathway linked to overall livelihood outcomes within the analytical framework and further supported by a significant pathway contribution (4.06, 95% CI 2.60–5.53). Economic outcomes are not directly associated with agroforestry characteristics but are positively associated with perceived climate variability (coef = 0.2151, p < 0.01), suggesting that households experiencing higher perceived variability may engage in adaptive livelihood responses. In contrast, social and physical pathways exhibit limited statistical influence, reflecting contextual constraints related to social homogeneity and uniformly poor infrastructure. These findings highlight the importance of understanding agroforestry as a pathway-based socio-ecological adaptation strategy rather than a uniform livelihood intervention. Policies that build on farmers’ existing adaptive responses while strengthening ecological functions, extension services, and rural connectivity are more likely to support more resilient livelihood outcomes in climate-stressed dryland regions.