Spatiotemporal patterns and drivers of bean yield variability in Food-Deficit Countries

• First multi-country, decade-long analysis of bean yield variability in LIFDCs. • Integrated meta-analysis and statistical modelling revealed key yield drivers. • Temporal yield stability contrasts with wide spatial variability among nations. • Climatic effects (rainfall and temperature) dominate bean yield variability. • Principal component analysis explained 96% of bean yield variability. • We provide data-driven guidance for resilience and productivity enhancement. This study examines the spatiotemporal variability of common bean ( Phaseolus vulgaris L.) yields and their drivers across 32 Low-Income Food-Deficit Countries (LIFDCs) from 2013 to 2022, addressing the limited understanding of how yield patterns change across regions and over time. This knowledge is vital for enhancing food and nutritional security in resource-limited settings. The research combined standardized yield data from FAOSTAT with information from published studies and national field trials to conduct a longitudinal meta-analysis. It integrates meta-analysis and modeling across countries to expand understanding of bean yield variability. Statistical methods, including One-Way ANOVA, Kruskal–Wallis tests, and Principal Component Analysis (PCA), were employed to evaluate yield variability and identify key influencing factors. The results showed that temporal variability remained relatively stable, with no significant differences in mean or median yields over the decade (p > 0. 05), while average yields varied within ±1.5 t ha -1 among countries. Spatial differences were more notable, with Mali showing the highest variability (mean = 10.03 t ha -1 , SD = 3.5 t ha -1 ), Eritrea the lowest (mean = 2.15 t ha -1 , SD = 0.5 t ha -1 ), and Tanzania displaying moderate variability (mean = 2.25 t ha -1 , SD = 1.1 t ha -1 ). Outliers, such as Mali’s higher yield values, likely reflect data aggregation from irrigated or experimental systems, requiring cautious interpretation. Rainfall and temperature emerged as the primary climatic factors influencing inter- annual fluctuations. PCA results indicated that broad temporal dynamics (PC 1) explained 96% of yield variability, while localized influences (PC 2) accounted for only 2%. ANOVA confirmed the absence of significant temporal trends, suggesting that long-term stability outweighs short-term variability. The findings suggest that strengthening climate resilience, inputs, and infrastructure enhances bean yield.