No abstract

In the context of grazing, the intermediate disturbance hypothesis (IDH) predicts that plant diversity peaks under moderate grazing, resulting in a hump‐shape pattern for the grazing–diversity relationship. Although this has been debated due to contradictory empirical results, the IDH is still widely accepted among rangeland ecologists. The Milchunas–Sala–Lauenroth (MSL) model predicts that in arid areas grazing affects diversity negatively regardless of grazing intensity, whereas in mesic areas grazing effects on plant diversity are the same as predicted by the IDH. Very few studies have attempted to specifically evaluate the applicability of the IDH and MSL to grazing systems at a global scale, accounting for the possible effects of climate. We conducted a meta‐analysis and vote‐counting analysis to evaluate these two hypotheses. The results of both analyses show that the IDH cannot be applied globally and that its application largely depends on aridity. The IDH prediction of a hump‐shape curve is supported in wet areas, while in dry areas there is a slight decrease in species richness with increasing grazing intensity. Overall, the MSL model correctly predicted grazing–diversity relations in both wet and dry areas. Looking at specific ecosystem types, we found that these results hold in grasslands, but not in woodlands. Differences between livestock types, not considered by the MSL model, were found to be important. Mixed sheep and goat grazing in dry areas resulted in a significant decline of species richness with grazing intensity, while grazing by sheep only had little effect on species richness. Cattle grazing and yak grazing in wet areas yielded a clear hump‐shape pattern. Therefore, we conclude that the climate‐specific MSL model better predicts the impact of grazing on diversity than the IDH in rangelands, and that the response patterns of plant richness to grazing are dependent on aridity, grazing intensity and grazer type.

No abstract