Search

Abstract: Grassland restoration on arable land is the second most implemented compensation measure in Germany to counteract impacts of infrastructural projects on nature. Most grassland restoration has been carried out using standardized commercial seed mixtures with large amounts of perennial generalists, cultivars and seeds of non-local origin. To evaluate whether this current practice is appropriate for developing regional types of species-rich mesophile grasslands, we analyzed four widely used treatments in a real-world setting (48 plots): (1) sowing a non-site-specific herb-poor mixture; (2) sowing a non-site-specific herb-enriched mixture; (3) sowing a site-specific herb-enriched mixture; and (4) spontaneous regeneration. After up to nine years, restored sites differed from target grasslands in: (1) number of species; (2) abundance and dominance of target species; and (3) dominance structure. Sown fields were dominated by sown species from the beginning. Because most differences were due to increasing cover of a small number of sown species, we found little development toward regional types of species-rich mesophile grasslands. In contrast, species composition on spontaneously regenerated sites changed to a greater degree and showed gradual development toward target grasslands. The limiting factor for successful restoration on all sites was availability of propagules. On sown sites, dominance patterns – particularly of Festuca rubra cultivars – had a negative effect on immigration and development of target species. For future restoration practice, we strongly recommend avoiding standardized commercial non-local seed mixtures. In particular, highly competitive cultivars should never be used. Even spontaneous regeneration should be preferred over standardized mixtures. However, in species-poor environments enrichment with selected species is necessary to reach target state. [Copyright &y& Elsevier]