Interactive control of global grassland productivity and diversity by consumers and nutrients

Background/Question/Methods: Fossil fuel use and agriculture have increased nutrient deposition even into unmanaged grasslands, and the composition of grassland vertebrate herbivores has been radically altered through both extinctions and introductions. Because of the practical and scientific importance of the grassland biome, a more predictive continental- and global-scale understanding of the forces controlling grassland composition and productivity will inform environmental and agricultural sustainability. Here we use an unprecedented replicated experiment, spanning 35 sites on 5 continents, to examine the independent and interactive effects of vertebrate herbivory and nutrient supply on species diversity, relative abundance, and net primary production (NPP) of grasslands around the world. In this ongoing distributed experiment, all 35 sites implemented a full factorial combination of nutrient addition (Control or All Nutrients) and consumer density (Control or Fenced) for a total of 4 treatment levels, following protocols standardized among sites. Fences (180 cm tall) were designed to exclude aboveground mammalian herbivores (>50 g). We sampled plant species composition, richness, and NPP annually at all sites for two to three years.

Results/Conclusions: After up to three years of manipulation, species richness declined and live biomass increased with fertilization in grasslands around the world, whereas removal of large herbivores had no overall effect on plot-scale richness, evenness, or biomass. Species evenness was not significantly altered by either experimental fertilization or fencing. Reduced herbivory and elevated nutrients interacted to support significantly greater NPP. Species richness declined rapidly in response to fertilization, and after only three years of chronically elevated nutrients, individual plots lost more than 1.5 species on average. These effects on production and diversity were contingent in part on regional climatic variables, soils, and plant community composition. Our empirical results provide new mechanistic insights that improve our understanding - from theory, case studies, and meta-analyses - of the global importance of top-down and bottom-up factors as mediators of producer community diversity and function.