Results
We detected 84 species in the samples consisting of 24 graminoids (8 short-lived and 16 perennial), and 60 forbs (36 short-lived and 24 perennial). Grazing intensity had a significant effect on green biomass + litter, green biomass, litter, the biomass of perennial forbs, perennial graminoids, short-lived forbs as well as on the species richness and disturbance (Table 4). The species richness at the fourth and fifth level of intensity was significantly higher than that of the second level, and we detected the highest richness at the fourth level of grazing intensity (Figure 1A). For disturbance values, the fifth level was significantly higher compared to all other levels (Figure 1B). We detected a humped-back relationship between the amount of green biomass and the species richness (Figure 2).
We also detected significant differences for green biomass + litter, litter (highest scores at the first level of intensity). For litter, we detected higher scores at the first and second level of intensity compared to the other levels (Figure 3A, B, C); moss + lichen biomass did not show a significant difference along the intensity levels (Table 2; Figure 3D). Perennial forb biomass showed the highest scores at the fourth level of intensity (Figure 4A). The perennial graminoids’ biomass was significantly higher at the second level of intensity than at the fourth and fifth levels (Figure 4B). We found no significant difference in the biomass of short-lived forbs between the first and fourth level of intensity while that of at the fifth level was significantly higher compared to the other levels (Figure 4C). The biomass of short-lived graminoids did not show significant differences (Table 4; Figure 4D).
The CCA identified only two significant predictors for the species compositional patterns : soil humus content and disturbance tolerance, but these two predictors were weakly correlated with each other (Figure 5). We found no clear separation of pastures grazed with different levels of intensity based on the biomass composition and sites with increasing levels of grazing intensity were also not separated along the predictors. The samples of the different sites did not show clear separation along either the axis of green biomass quantity or the species richness also in Figure 2.