Search for a command to run...
Abstract Questions Although the role of landscape context, often interpreted as the available species pool, is generally assumed to be important for the restoration of disturbed sites, not many studies evaluating this role quantitatively in restored dry grasslands have been carried out. There is especially a lack of large‐scale, multi‐site studies. We asked how many unsown target species established spontaneously in restored dry grasslands and how spontaneous establishment depended on their occurrence in the surrounding landscape. Location White Carpathian Mts., eastern Czech Republic (48°50′–49°05′ N, 17°19′–17°55′ E, 250–610 m a.s.l.). Methods We studied spontaneous colonization by unsown target species in dry grasslands on ex‐arable land at 82 sites, sown with a regional seed mixture, commercial clover–grass seed mixtures or left to spontaneous succession. In total, 246 phytosociological relevés in 5‐m × 5‐m plots were recorded. Restoration started 1–31 yr ago. Nearby semi‐natural grasslands were included as reference sites (94 phytosociological relevés). A total of 108 target species were defined and their presence evaluated in the surroundings of the restored grasslands, i.e., in grid cells (3.1 km × 2.8 km) in which the restored sites were situated. Data were processed using multivariate analyses (DCA, CCA) and GLM. Results The number of target species at the restored sites significantly increased with time elapsed since restoration started and the number of target species occurring in the surroundings. The sites, although restored in different ways, converged in their species composition towards ancient grasslands. The maximum relative colonization of a restored grassland by target species from the surroundings was 23.4%, and the restored grasslands contained max. Fifty‐seven percent of the average number of dry grassland target species present in the ancient grasslands. Conclusions Sowing seed mixtures should be seen as not more than a first phase in the restoration of species‐rich grasslands. In a next phase of restoration, we can reasonably rely on gradual spontaneous colonization by target species if they occur in the surroundings, even at sites sown with commercial seed mixtures or restoring spontaneously. However, full restoration probably takes decades or more.