Table 1 Results of multiple regression analyses of distance matrices with divergence in colour morph frequencies (colour) as the dependent variable and showing partial correlation coefficients (r _partial) of each predictor variable included in each model (1–8): pairwise F_ST, geographic distance (distance), mean annual aridity index (aridity), topographic relief (topography), proportion cover of vegetation <1 m in height (vegetation), and proportion cover of rock >50 cm in diameter (rock). AIC_C, ΔAIC_C and normalised AIC_C weights were calculated based on the number of variables in the model (k) and the proportion of variance explained by the model (R²). Only models with ΔAIC_C within five units of the best possible model are shown and the relative importance of variables (RIV) was determined from this subset of models

From: Environment, but not genetic divergence, influences geographic variation in colour morph frequencies in a lizard