Herbivore impact on grassland plant diversity depends on habitat productivity and herbivore size

Abstract

Mammalian herbivores can have pronounced effects on plant diversity but are currently declining in many productive ecosystems through direct extirpation, habitat loss and fragmentation, while being simultaneously introduced as livestock in other, often unproductive, ecosystems that lacked such species during recent evolutionary times. The biodiversity consequences of these changes are still poorly understood. We experimentally separated the effects of primary productivity and herbivores of different body size on plant species richness across a 10-fold productivity gradient using a 7-year field experiment at seven grassland sites in North America and Europe. We show that assemblages including large herbivores increased plant diversity at higher productivity but decreased diversity at low productivity, while small herbivores did not have consistent effects along the productivity gradient. The recognition of these large-scale, cross-site patterns in herbivore effects is important for the development of appropriate biodiversity conservation strategies.

Herbivores and nutrients control grassland plant diversity via light limitation

“…Thus, climate, which predicts herbivory intensity, places an ultimate constraint on the effects of herbivores on plant diversity, but local plant diversity is determined primarily through herbivore effects on groundlevel light. These experimental data demonstrate that across a wide range of the world's grasslands, herbivores serve as an important force to maintain plant diversity in grassland ecosystems in which they increase ground-level light availability, consistent with the prediction that light limitation is a critical factor controlling grassland species diversity 9,16 , but counter to the interpretation of nutrient supply or ecosystem productivity as the dominant force constraining herbivore effects on local plant diversity [11][12][13][14][15][16][17] .…”

“…The lack of a consistent effect of herbivore removal on diversity across these globally distributed grassland sites ( Fig. 2a) reflects the broad range of positive and negative effects found in past studies 11,17,29 . However, a critical assumption underlying the hypothesis that grassland diversity is jointly controlled by nutrient supply and consumers is that diversity should be rescued consistently by herbivory.…”

“…For example, in highly productive, eutrophic systems in which plant species extinction is likely to occur owing to a loss of ecological niches 8 , ecological theory predicts that herbivores can act to maintain local-scale plant diversity if they selectively consume the superior resource competitors 9 . Empirical studies in many ecosystem types find highly variable effects of herbivores on plant species diversity [11][12][13] , with the magnitude of herbivore mediation of diversity frequently observed to be greatest in regions of high ecosystem productivity [11][12][13][14][15][16][17] . Thus, ecosystem productivity and its regional climate drivers have been observed to mediate the local-scale effects of herbivores on plant diversity.…”

“…The availability of ground-level light is the commonly postulated mechanism modulating the relationships among plant diversity, herbivory and observed gradients of plant productivity. However, these patterns and predictions have primarily emerged from studies across observed gradients of productivity or reviews and meta-analyses based on an extremely limited number of single-site experiments that manipulate both nutrients and herbivory, often with different methods [11][12][13][14][15]17 . In most of these studies, ground-level light has not been measured.…”

How do sheep affect plant communities and arthropod populations in temperate grasslands?

“…This led us to add seven more references to the first dataset, two more to the second dataset and four more to the third dataset. Given the different reactions of plant communities to grazing intensity depending on habitat productivity (Bakker et al, 2006;de Bello et al, 2006) and salinity levels (Olff and Ritchie, 1998), the few data available on semiarid grasslands (Osem et al, 2002;Li et al, 2008;Campbell et al, 2010) and marshland plant communities (Bouchard et al, 2003;Hofmann and Mason, 2006;Milotić et al, 2010) were excluded from this analysis. Assessment of study quality also led us to exclude surveys dealing with mixed grazing, as well as those where stocking rate could not be calculated from plot size and number of grazing days.…”