Consumers are critical for controlling structure and function in many of the world’s most extensive ecosystems, particularly so in grasslands. However, human activities have greatly altered these top-down forces with consequences for biodiversity and ecosystem services that remain to be fully understood. In part, this uncertainty occurs because effects on consumers on diversity can be highly variable with herbivory increasing diversity in some cases but decreasing diversity in others. Differences in ecosystem productivity, driven by precipitation, may play a key role in determining the effect of herbivores on diversity. However, deviations from this pattern are widely recognized calling into question the generality of the role of productivity in governing herbivore effects on diversity and suggesting that alternative mechanisms may be driving how consumers impact this aspect of community structure. Here we propose that changes in dominance rather than productivity drive the diversity response to herbivory as changes in community dominance can be directly linked to alterations in the abundance of other species. To test this hypothesis, we conducted a meta-level analysis (based on raw species composition data) comparing grazing exclosure experiments in which large herbivores were excluded for more than 3 years from 252 sites around the world.
These sites span a wide range of locations as well as large gradients in mean annual precipitation (MAP; 45 to 1511 mm). Despite strong relationships between precipitation and ANPP in grasslands globally, we found no relationship between grazing induced change in species richness and increasing precipitation (a proxy for productivity). In contrast, we found a strong relationship between grazing induced changes in dominance and changes in species richness. When herbivores decreased dominance, species richness increased, while increased dominance resulted in a decline in species richness. Using path analysis we compared the relative effects of precipitation and changes in dominance on the change in species richness in response to herbivores. Both dominance and the change in dominance were significantly and strongly correlated with the richness response to grazing. That is, as the change in dominance increased, the change in species richness increased. Precipitation, our surrogate for ANPP, had no significant effect. These results demonstrate for the first time across large precipitation gradients and on multiple continents that the mechanism driving grazing effects on richness is dominance.