Plant species richness of perennial flower strips on arable land is affected by seed diversity, provenance and seeding density

Land-use intensification has caused drastic declines in the biodiversity of plants and insects in agricultural landscapes. To counteract these losses, the European Union introduced agri-environmental schemes that subsidise the establishment and maintenance of perennial flower strips on arable land. In practice, the large variety of flower strips makes it difficult to monitor their effectiveness, which calls for systematic experiments. This study reports on a field experiment to understand the effects of seed mixture and seeding density on the diversity of sown and unsown plant species. In spring 2020, we set up a field experiment with five seed mixtures differing in diversity, provenance and composition as well as three seed densities. Plant diversity was monitored after one and four years. There were clear effects of the seed mixtures on plant biomass and species richness, while the impact of seeding density was low. The moderately diverse regional mixture resulted in the highest richness; including insect-pollinated species, while the most diverse mixture and high-density seeding showed lowest colonisation by unsown species. Among the commercially available (and thus cheaper) mixtures, the one with lowest diversity performed worst in terms of sown species establishment, species richness and the proportion of unsown plants. Thus, we recommend usage of regional seeds with intermediate diversity at low to moderate seeding density when establishing perennial flower strips and aiming at high plant richness.