MODELING LEGUME CULTIVATION POTENTIAL UNDER CLIMATE CHANGE: A REGIONAL APPROACH FOR POLISSYA AND THE FOREST-STEPPE OF UKRAINE

This study developed spatially detailed suitability models for common bean (Phaseolus vulgaris L.) cultivation in the Polissya and Forest-Steppe regions of Ukraine, taking into account projected climate changes of the 21st century. The current distribution of bean cultivation was determined using global agricultural land mapping databases and aligned with high-resolution climate and soil datasets (approximately one-kilometer spatial resolution). Nineteen bioclimatic variables reflecting annual trends, seasonal variability, and extreme weather conditions were derived from publicly available climate model ensembles. Soil characteristics, including bulk density, particle size distribution, and organic matter content at a depth of 5–15 cm, were extracted from digital soil maps. To reduce multicollinearity, a two-step predictor selection procedure was applied, combining Principal Component Analysis (PCA) to identify uncorrelated environmental gradients with hierarchical clustering of standardized variables, selecting one representative variable per cluster. Using a multivariate linear regression model, the transformed dependent variable was associated with the selected climatic and soil predictors. The model's performance was assessed using the adjusted coefficient of determination and diagnostic plots, which confirmed the absence of significant violations of homoscedasticity and spatial autocorrelation. The empirical results indicate that lower temperatures during frost-prone periods, as well as reduced annual precipitation, negatively affect the suitability of areas for bean cultivation. Conversely, warmer and wetter conditions during the main growing season, along with greater daily temperature ranges, are favorable for this crop. Among soil factors, higher bulk density and increased organic matter content were found to have a significant positive effect on the potential for bean production.