Conclusion
In conclusion, we observed contrasting and scale-dependent responses of wild bees and hoverflies to measures of land-use intensification, with no particular effects of local flower diversity of plants. As a consequence, pollination service on dry grasslands should change with the surrounding landscape. In homogeneous landscapes with a high share of arable field cover, insect pollination should occur less frequent, due to lower numbers of wild bee individuals and species. Although hoverflies concentrate particularly in these landscapes at dry grasslands they cannot compensate absence of wild bees, as they are less specialized to particular flower types and appear only in late summer in reasonable amounts (see also Brandt et al. 2017). Moreover, the comparison with historical data indicate that species composition and abundance of both guilds already shrunk in our study area, presumably due to high land-use intensification. In order to achieve a better understanding of how land-use intensification affects pollinators, we advocate to a) acknowledge that landscape effects may differ between landscapes and b) analyze therefore possibly confounding landscape parameters across scales.