The current biodiversity loss in agricultural landscapes calls for a thorough mechanistic understanding of the underlying causes. Our working group has a particular focus on plant communities of grasslands. These habitats belong to the most species-rich habitats in Central Europe, harboring many rare and endangered species that are confined to this vegetation type. We are fascinated by the high diversity within small scales and are especially interested to understand how environmental change affects species coexistence and subsequently the species composition in these communities. Hereby, we aim to disentangle the importance of local and regional factors. We seek a mechanistic understanding for testing general theories and enhancing nature conservation concepts. As study system, we used mainly dry and mesophytic grasslands, but expanded our research to different study systems, like kettle holes, within the past years. Further, we gained first insights of the main drivers of arthropods communities in agricultural landscapes. In our approaches we use trait-based field studies combined with common garden experiments and spatially explicit process-based simulation modelling.