Effects of habitat configuration on biodiversity along gradients of forest cover on the Swiss Plateau

The negative effects of habitat loss on biodiversity are undisputed, while the effect of habitat configuration, i.e., the spatial arrangement of habitat area, has been debated for decades. To develop a more comprehensive understanding, it is important to know when and how configuration matters. In this study, we tested whether forest configuration influences the richness of species in groups characterized by varying shade tolerance in different ways and how such effects are related to habitat amount (i.e., the percentage of forest cover) at the landscape scale. Based on 104 survey plots (each measuring 1 ​km²) of vascular plants on the Swiss Plateau, and using two statistical approaches (i.e., multiple regression and path analysis), we modeled the effects of habitat amount and configuration (measured as number of forest patches, total edge length, and proximity index) across all the plots and separately for three habitat amount classes: <10%, 10%–30%, and >30% forest cover. When we modeled all plots together, we found that, after controlling for habitat amount, the forest configuration significantly affected species richness. When we considered the different habitat amount classes separately, most of the significant effects of habitat configuration on species richness occurred only for habitat amounts of <10% forest cover. Additionally, the response to forest configuration differed among species with different shade tolerances. When forest area was <10%, the effects of the number of patches and the total edge length on the species richness of light-demanding forest species were greater than the effect of habitat amount, whereas neither configuration metric affected the richness of shade-tolerant species. In conclusion, our findings highlight the importance of configuration in landscapes with a small amount of habitat. At the same time, they demonstrate that considering the confounding factors (e.g. species traits) is important for understanding the effects of forest configuration on biodiversity and that generalizations remain a challenge for landscape ecology.