**3. Results**

Species richness patterns varied across different taxonomic groups. Amphibians and mammals showed similar patterns with higher species richness in Central Europe, while reptiles followed a different pattern showing higher species richness in Southern Europe (Figure 1a–c). Amphibians and reptiles exhibited similar spatial patterns of functional richness in Europe, where they manifest a clear latitudinal pattern with the lowest functional richness values found in northern Europe, that is, regions with low temperatures (Figure 1d,e). Interestingly, arid areas along the coastline of Southeastern Europe were poorer in terms of functional richness for amphibians than for reptiles, with reptiles having the highest values in these regions. Contrastingly, mammal functional richness showed a more uniform pattern across Europe, with moderate to high functional richness values across all of Europe (Figure 1f). Lower values of mammal functional richness were found in the coastline of the Mediterranean region (Figure 1f). Rao's quadratic entropy patterns exhibited similar results, with those of functional richness patterns for all the three examined taxa showing high and significant associations (amphibians: R2 = 0.72, *p* < 0.001; reptiles: R<sup>2</sup> = 0.88, *p* < 0.001; mammals: R<sup>2</sup> = 0.54, *p* < 0.001). Cross-taxon relationships between amphibians and reptiles indicated significant moderate associations for functional richness (R<sup>2</sup> = 0.27, *p* < 0.001), but also for Rao's quadratic entropy (R2 = 0.40, *p* < 0.001). On the other hand, mammal functional diversity was significantly but weakly related with amphibian (functional richness: R<sup>2</sup> = 0.09, *p* < 0.001; Rao's quadratic entropy: R2 = 0.02, *p* < 0.001) and reptile (functional richness: R<sup>2</sup> = 0.13, *p* < 0.001; Rao's quadratic entropy: R<sup>2</sup> = 0.14, *p* < 0.001) functional diversity (Figure 1d–f).

The climatic model performed better than the land–human model in all cases, and the overall model generally outperformed both the climate and land–human models (Table 1). In the case of functional richness, the explanatory power of the climatic model was (in descending order of fit): R<sup>2</sup> = 0.59 for reptiles, R<sup>2</sup> = 0.42 for amphibians, and R2 = 0.25 for mammals. Meanwhile, the corresponding values of the land–human model were (in descending order of fit): R2 = 0.37 for amphibians, R<sup>2</sup> = 0.33 for reptiles, and R<sup>2</sup> = 0.10 for mammals. The overall model including all the predictors was the best-fitting model in all cases, scoring slightly better than the climatic model. The highest explanatory power of the overall model was observed for reptiles, followed by amphibians and mammals, in all diversity metrics (Table 1, Figure 2), indicating that the variables related to landscape and human pressures play a subordinate role to climatic variables.

**Table 1.** Performance (R2) of the generalized additive mixed models explaining the species richness, functional richness and Rao's quadratic entropy patterns of three taxonomic groups (amphibians, reptiles, and mammals) in Europe (grid cell size 50 km × 50 km). Temperature, precipitation, temperature seasonality, and precipitation seasonality were used to quantify the effects of climate in the climatic model, while land-cover diversity, agricultural land area, urban land area, and human population density were used to quantify the effects of land-cover and human pressures in the land–human model. All variables were used in the overall model.


**Figure 1.** The species richness (**a**–**c**), functional richness (**d**–**f**), and Rao's quadratic entropy (**g**–**i**) distribution patterns of amphibians, reptiles, and mammals in Europe (50 km × 50 km grid cell size).

The performance and shape of the relationship between the predictors (overall model) and each functional diversity metric and species richness for each taxon are summarized in Figure 2. Temperature and precipitation seasonality were significantly related to the species richness and functional diversity of all taxa. The species richness (as in the case of both indices of functional diversity) of amphibians and mammals had a unimodal relationship with temperature, while reptile species richness and functional diversity increased with it. Precipitation seasonality exhibited a convex relationship (suggesting the existence of a bimodal relationship, with the curve showing only a part of the variability) with diversity, independently of taxon or metric, although relatively little variation of their values was observed. Temperature seasonality significantly affected the species richness of all taxa, but only mammal and reptile functional richness was positively related to temperature seasonality. Reptile species and functional richness increased significantly with precipitation, and amphibians and mammals showed a unimodal relationship. A unimodal relationship was also found for Rao's quadratic entropy of mammals. Rao's quadratic entropy of amphibians decreased with increasing percentage of urban land area. All richness and diversity measures tended to have negative relationships with the

percentage of urban area. However, only in the case of reptile functional richness and Rao's quadratic entropy of amphibians was this negative relationship significant. Furthermore, the species and functional richness of amphibians and mammals tended to increase with the percentage of agricultural area, while a linear decreasing relationship was observed in the case of reptiles. Species richness for all taxa increased significantly with land-cover diversity. In addition, functional diversity measures tended to have a positive relationship with land-cover diversity, although these relationships were significant for reptiles, as well as for mammal functional richness. Finally, mammal species richness and amphibian Rao's quadratic entropy increased with human population density.

**Figure 2.** Summary plot showing the results of generalized additive mixed models (performance, direction, and significance of the relationship) predicting species richness, functional richness, and Rao's quadratic entropy of amphibian, reptiles, and mammals in Europe as function of climatic variables (temperature, precipitation, temperature seasonality, and precipitation seasonality) and variables related to land use and human pressures (land-cover diversity, agricultural land area, urban land area, and human population density). Continuous lines show significant associations while dotted lines show non-significant associations. Performance (R2) is shown for the overall model in which all variables were used.
