Natural habitats matter: Determinants of spatial pattern in the composition of animal assemblages of the Czech Republic
Journal
Acta Oecologica
Journal Volume
59
Pages
7-17
Date Issued
2014
Author(s)
Abstract
Studies that explore species-environment relationships at a broad scale are usually limited by the availability of sufficient habitat description, which is often too coarse to differentiate natural habitat patches. Therefore, it is not well understood how the distribution of natural habitats affects broad-scale patterns in the distribution of animal species. In this study, we evaluate the role of field-mapped natural habitats, land-cover types derived from remote sensing and climate on the composition of assemblages of five distinct animal groups, namely non-volant mammals, birds, reptiles, amphibians and butterflies native to the Czech Republic. First, we used variation partitioning based on redundancy analysis to evaluate the extent to which the environmental variables and their spatial structure might underlie the observed spatial patterns in the composition of animal assemblages. Second, we partitioned variations explained by climate, natural habitats and land-cover to compare their relative importance. Finally, we tested the independent effects of each variable in order to evaluate the significance of their contributions to the environmental model. Our results showed that spatial patterns in the composition of assemblages of almost all the considered animal groups may be ascribed mostly to variations in the environment. Although the shared effects of climatic variables, natural habitats and land-cover types explained the largest proportion of variation in each animal group, the variation explained purely by natural habitats was always higher than the variation explained purely by climate or land-cover. We conclude that most spatial variation in the composition of assemblages of almost all animal groups probably arises from biological processes operating within a spatially structured environment and suggest that natural habitats are important to explain observed patterns because they often perform better than habitat descriptions based on remote sensing. This underlines the value of using appropriate habitat data, for which high-resolution and large-area field-mapping projects are necessary.
Subjects
Animal assemblages; CORINE Land Cover; Habitat mapping; Moran's eigenvector maps; Spatial autocorrelation; Variation partitioning
Type
journal article