Abstract
To counteract the decline of biodiversity in intensively used agricultural landscapes, a diversity of local conservation interventions, including on-field measures, such as extensification of cereal production, and off-field measures, such as flower strips, is taken. Pollinators, among which wild bees are of particular importance, are a focal group of biodiversity conservation. Although several studies have shown positive local effects of conservation interventions on wild bees, it remains unclear whether these lead to a real increase of wild beess at landscape level. Therefore, we surveyed wild bees in 18 agricultural landscape sections (1 × 1 km) distributed over nine agricultural regions across Germany. In one of the paired landscapes per region, local conservation interventions were implemented, while they were largely absent in the others. We recorded wild bees with a stratified transect sampling on 1000 × 4 m² per landscape where transect sections were distributed proportionally across the edges of all major land-cover types. Further, we conducted GIS-based landscape analyses to calculate the area of conservation interventions and additional landscape metrics, e.g., area of semi natural habitats and edge density. Then, we modelled the effects on landscape-level species richness and abundance of wild bees. Thereby, we also grouped bee species by taxonomy and frequency of occurrence. We found that the area of local conservation interventions was positively related to richness and abundance of solitary bees and rare bees. The abundance of bumble bees and frequent wild bees was positively influenced by crop richness or evenness. In contrast, semi-natural habitats and edge density had no significant effects in the studied landscapes. Our results show that local conservation interventions can promote wild bees at landscape level already few years after their implementation.