Determinants of multidimensional vulnerability in cocoa agroforestry systems across two contrasting ecological landscapes in Central Cameroon

Context: Cocoa agroforestry systems are exposed to interacting climatic, ecological, market, and socioeconomic stressors. However, many studies assess vulnerability of cocoa farmers to climate change, and there is limited knowledge on how multidimensional drivers shape household vulnerability across heterogeneous landscapes.,Objective: This study quantified multidimensional cocoa household vulnerability and identified its key socioeconomic and institutional drivers across two contrasting ecological landscapes.,Methods:Household survey data were collected from 142 cocoa-producing households in the humid forest and forest–savanna transition landscapes of Cameroon. A multidimensional vulnerability index integrating exposure, sensitivity, and adaptive capacity indicators was developed and analysed using linear mixed-effects models.,Results: Vulnerability differed significantly between ecological landscapes (F (1, 99.3) = 10.45, P = 0.0017), with higher vulnerability in the humid forest zone (2.31 ± 0.12) than in the forest–savanna transition zone (1.85 ± 0.07). Exposure factors increased vulnerability, with cocoa price decline emerging as the strongest driver (β = 0.01319, P < 0.0001), followed by declining rainfall, excess rainfall events, and pest incidence. Sensitivity factors, including cocoa yield variability (β = 0.01228, P < 0.0001), dependence on cocoa income (β = 0.01261, P < 0.0001), food insecurity (β = 0.01056, P < 0.0001), and labour constraints (β = 0.01411, P < 0.0001), amplified vulnerability. Conversely, adaptive capacity significantly reduced vulnerability, particularly agricultural training (β = -0.02311, P < 0.0001) and access to credit (β = -0.01336, P < 0.0001).,Significance: Vulnerability in cocoa agroforestry systems emerges from the interaction between external stressors and internal system characteristics.