LAUSR

Abstract In agroecosystems, species‐rich habitats, such as linear field margins and flower strips, are beneficial to the overall biodiversity and contribute to pest control. Their effects are thought to be mediated by plant species composition and diversity. However, the management of plant communities with targeted levels of functional diversity has been little investigated. In an open field landscape, we compared the effects of the sown species richness (9, 14, and 29 species) and functional diversity (high vs. low) of eight different seed mixtures, sown in flower strips, on the 4‐year temporal dynamics of their functional diversity. There was a good agreement between the expected and realized species richness and functional diversity at the start of the experiment. All plant assemblages progressively lost species over time, but this decline was lower for assemblages sown with a high initial functional diversity, in which species evenness was maintained at higher levels. Species‐rich assemblages had a higher degree of functional redundancy, and their functional diversity remained higher over time than less rich assemblages. A possible explanation for this is that functional redundancy would have enabled the compensation for the loss of species by functionally equivalent species. The realized functional diversity of the sown species also limited the establishment of spontaneous species, perhaps due to a higher degree of niche occupancy. This study provides useful insight into the creation of functionally diversified plant communities. A high level of initial species and functional diversity is required to guarantee a greater temporal persistence of the communities.