This paper deals with the flora and vegetation of stony walls (wall tops, vertical wall surfaces) in East Bohemia. In total, 207 species of vascular plants and 60 mosses were identified in 114 recorded relevés. Flora of walls is composed of a high number of accidental species. Only two species (Poa compressa, Taraxacum sect. Ruderalia) were frequently recorded on walls. Differences in species’ traits (life strategy, life form, dispersal) and ecological requirements of plants (light, moisture) were analysed between vertical wall surfaces and wall tops. Due to high floristic heterogeneity, many communities can be classified only at the level of higher syntaxa. In total, 10 communities were reported on the studied walls. Communities on wall tops were dominated by Poa compressa, P. palustris subsp. xerotica, P. nemoralis subsp. rigidula, Conyza canadensis and Syringa vulgaris. Four communities dominated by Corydalis lutea, Cymbalaria muralis, Asplenium ruta-muraria and Cystopteris fragilis were identified on vertical wall surfaces. Their structure, species composition, ecology and distribution are briefly discussed.
Urbanization is one of the most extreme forms of land transformation. It is supposed to change the frequencies of species trait states in species assemblages. We hypothesize that the flora of urban and rural areas differs in the frequency of trait states and ask which traits enable a plant to cope with the urban environment. We tested our hypothesis in Germany, which was divided into grid-cells of ca 130 km2. We distinguished urbanized (with more than 33% urban land use; n = 59), agricultural (with more than 50% agricultural land use; n = 1365) and semi-natural (with more than 50% forest and semi-natural land use; n = 312) grid-cells and calculated the proportions of plant species per trait state in each grid-cell. Multiple linear regressions explained the log-transformed ratio of one proportion to another with land use (urban, agricultural, semi-natural) and the environmental parameters (climate, topography, soils and geology). Additionally, linear mixed effect models accounted for the effects of land use and biogeography and differences in sample size of the three grid-cell types. Urbanized and rural areas showed clear differences in the proportion of trait states. Urbanized grid-cells had e.g., higher proportions of wind-pollinated plants, plants with scleromorphic leaves or plants dispersed by animals, and lower proportions of insect-pollinated plants, plants with hygromorphic leaves or plants dispersed by wind than other grid-cells. Our study shows that shifts in land use can change the trait state composition of plant assemblages. Far-reaching urbanization might consequently homogenize our flora with respect to trait state frequency.