Weather Extreme Events, Farm Management, and Cocoa Yield in Ecuador

The frequency and intensity of extreme weather events are increasing, yet their implications for perennial tree crops such as cocoa remain understudied. Existing cocoa-climate evidence is based on mean climate variables, discrete shocks, and concentrated in West and Central Africa. This study addresses these gaps with plot-level cocoa evidence for Ecuador, the world's second-largest exporter, linking the 2016 ESPAC survey to parish-level ET-SCI indices. We estimated the associations of three climate extreme indices: consecutive dry days (CDD), very heavy precipitation days (R20mm), and warm days (TX90p), on observed yield. Each index is summarized over 1-, 5-, and 10-year look-back windows to separate recent from cumulative exposure, within cross-sectional Poisson pseudo-maximum-likelihood (PPML) models with province fixed effects, survey weights, parish-clustered standard errors, and farm-level controls including management intensity. CDD showed no significant association, and TX90p estimates were imprecise, whereas R20mm was negatively associated with yield across all windows, with stronger evidence for longer windows. Each additional very heavy precipitation day (in the window average) was associated with 0.5-0.8% lower yield. Adjustment for management intensity attenuated this association in the intermediate window, although the cross-sectional design precludes a causal buffering interpretation. Recurrent very heavy precipitation emerges as a climatic risk for Ecuadorian cocoa.