Beta diversity of plants, birds and butterflies is closely associated with climate and habitat structure

Aim: The aim was to investigate the relationship between climate, topography and soil pH, as well as vegetation structure and the beta diversity of plants, butterflies and birds; and to investigate the correlations of (woody) plant beta diversity with animal beta diversity. Location: Switzerland (central Europe). Methods: We used pairwise Sorensen dissimilarity as measure of total beta diversity and partitioned it into its turnover and nestedness components. Variation partitioning was used to assess the independent and cumulative effects of environmental predictors, with vegetation structure being derived from airborne light detection and ranging (LiDAR) data. We also checked for independent effects of plant and woody plant beta diversity on butterfly and bird beta diversity, respectively, and for independent effects of spatial distance on beta diversity. Results: Climate emerged as the strongest statistical predictor of beta diversity across taxonomic groups, with large independent effects on species turnover. Climate effects were most pronounced for plants, followed by butterflies and birds. We also found large independent effects of vegetation structure on total beta diversity and its turnover component across taxonomic groups, particularly for birds. Plant and woody plant beta diversity substantially improved the predictions of butterfly and bird beta diversity, respectively. Spatial distance had hardly any independent effect on beta diversity. Main conclusions: Climate is a stronger filter for plant communities than for butterfly and bird communities, which are more affected by vegetation structure, probably owing to associated resources and niches. Vegetation structure is a crucial predictor of beta diversity, and therefore contiguous and detailed 3-D habitat structure data are highly relevant to further our understanding of niche-based community assembly. Plant and animal beta diversity appear to be nonindependent, suggesting that differences in the response times of interacting taxa should be accounted for in environmental change impact assessments on biodiversity.