Conclusions
We detected a humped-back relationship between green biomass and species richness in the studied sand pastures, but the plots characterised by different grazing intensities were not separated along the biomass gradient from each other. This clearly indicated that i) grazing intensity levels assessed by the stocking rate on the pasture level is too robust to assess effects of grazing intensity on the vegetation, and ii) even sheep grazing created a highly patchy vegetation at the scale of the biomass sampling both for species richness and biomass. These results also suggests that in case we would like to reveal the intensity-dependent distribution of biomass samples we need to use multiple scales for sampling, and a fine-scale assessment of grazing intensity. With this paper, we might take a step in revealing that the stocking rate and the grazing frequency can jointly drive the vegetation. In terms of grazing frequency, we formulated assumptions based on the impact of proximity but to find clearer effects, a broader knowledge about each pasture is needed e.g., with studying rotational grazing or elaborating other ideas.