A multi-scale analysis on the importance of patch-surroundings for farmland birds

Agricultural intensification has profoundly changed agricultural landscapes with important biodiversity impacts. There is increasing knowledge on the general effects of landscape structure and management practices on plant and animal species but understanding the role of surrounding landscape structure for patch-scale biodiversity is more complex. While it can be reasonably assumed, that adjacent habitats are more important than more distant ones, the importance of landscape structure has often been tested at the landscape scale but rarely at smaller scales such as patch-surroundings. We assessed the influence and interdependences of landscape composition and configuration (LCC) and land use/land cover (LULC) on bird species richness and abundance through a multi-scale analysis with specific focus on the surrounding patches. In two agricultural regions in Switzerland, we collected point data of birds on 36 transects (500 m) and combined them with detailed spatial data on LULC. Bird richness and abundance were correlated to sets of landscape metrics as proxies for LCC computed at the transect-scale as well as for the patch surroundings. We analysed patch LULC as well as the most important patch-surrounding metrics using generalized linear mixed models. The results illustrate that patch LULC is the most important predictor of bird richness and abundance. Woody structures increase bird richness, followed by extensive management on patch scale. On the transect-scale semi-natural structure and heterogeneous LCCs are beneficial for bird richness and abundance. The effect of patch-surrounding structure LCCs is only small and interacts with patch LULC. Birds in grassland benefit from fallows in the surroundings, while in cropland they tend to respond positively to surrounding extensive grassland. Our results highlight that considering sur-roundings can help improve patch-based biodiversity assessments, which will then better predict the conse-quences of farmland management and make the outcome more applicable for practice.