Scale dependency of taxonomic and functional diversity in pristine and recovered loess steppic grasslands.

Widespread campaigns on forest restoration and various tree planting actions lower the awareness of the importance of grasslands for carbon sequestration and biodiversity conservation. Even lower attention is given to the conservation of biodiversity and ecosystem functioning in remnants of ancient, so-called pristine grasslands. Pristine grasslands generally harbour high biodiversity, and even small patches can act as important refuges for many plant and animal species in urbanised or agricultural landscapes. Spontaneous succession of grassland is frequently viewed as a cost-effective tool for grassland restoration, but its applicability is strongly dependent on many local to landscape-scale factors, and the recovery is often slow. It is therefore essential to monitor the success of grassland restoration projects that rely on spontaneous succession. We compared the species diversity and functional attributes of pristine and recovered grasslands by studying the taxonomic and functional diversity in thirteen (8 pristine and 5 recovered) loess steppic grasslands using differently sized sampling plots from 0.01 to 100 m. Our results indicate that there are remarkable differences in taxonomic and functional diversity between pristine and recovered grasslands. We also found that during secondary succession there is a likely functional saturation of the species assembly in the first few decades of recovery, and while patterns and structure of recovered grasslands became quite similar to those of pristine grasslands, species richness and diversity still remained much lower. Pristine grasslands support considerable plant diversity, and species composition is slow to recover if destroyed by agricultural land use. This underlines the importance of preserving existing pristine grassland remnants, which might serve as sources of species for future restoration measures.