Descriptive, regression, and geostatistical characteristics of vegetation ecological indicators in the floodplain of Khortytsia Island

This supplementary dataset provides descriptive statistics, topographic regression results, and geostatistical characteristics of vegetation ecological indicators in the floodplain of Khortytsia Island (Ukraine), a system that underwent substantial reorganization after the 2023 destruction of the Kakhovka Reservoir. The dataset is based on 876 vegetation relevé plots surveyed in 2024–2025 and includes species richness, Shannon diversity, Raunkiær life forms, pollenochore and diasporochore groups, Didukh and Ellenberg indicator values, naturalness, hemeroby, disturbance indicators, Grime’s CSR strategies, and ilr-derived adversity gradients. For each vegetation variable, the dataset contains summary statistics and the results of topographic regression models based on DEM-derived predictors, including elevation, slope, topographic wetness index, multiscale topographic position index, roughness, and sine/cosine transformations of aspect. Predictor selection was performed separately for each response variable using stepwise AIC-based model selection, and standardized regression coefficients are reported for the retained predictors. The dataset also includes geostatistical outputs derived within a regression–kriging framework. Spatial autocorrelation was assessed from model residuals using Moran’s I, and spatial structure was characterized using empirical and fitted Matérn variograms. The reported parameters include the smoothness coefficient kappa, nugget, sill, range, and spatial dependence level (SDL), together with regression-based trend information and spatial prediction summaries.