The interactive effects of bottom-up and top-down forces vary across a broad grassland productivity gradient

Background/Question/Methods

Primary production in herbaceous systems is influenced by both grazing and soil nutrient availability. However the interactive effects of these top-down (grazer) and bottom-up (nutrient) forces remain understudied. The magnitude and direction of the effects of nutrient availability, herbivory, and their interactions and the mechanisms driving these effects likely vary across systems and broad resource gradients. Here we examine the effects of multiple nutrient additions (nitrogen, phosphorous, and potassium) in combination with vertebrate herbivore exclusion on primary production at three sites spanning the precipitation gradient of the North American Central Great Plains. These sites are located within shortgrass steppe (xeric), mixed grass prairie, and tallgrass prairie (mesic) ecosystems. We hypothesized that: (1) nutrient availability and herbivory will interact to influence primary production and (2) that these interactive effects will vary across systems along a broad precipitation-productivity gradient. To test these hypotheses, we examined changes in total above-biomass in response to both nutrient additions and vertebrate herbivore removals, in addition to identifying species-specific responses to these treatments. 

Results/Conclusions

The effect of nutrient additions on above-ground production was dependent on water availability across all three sites. Nutrient additions resulted in an increase in above-ground biomass at the mesic end of the precipitation gradient (mixed grass and tallgrass prairie) across all three years of treatment but only resulted in an increase in above-ground biomass at the xeric end of the precipitation gradient (shortgrass steppe) during an above-average precipitation year. Removal of herbivores only resulted in an increase in above-ground biomass in mixed grass prairie. In addition, we observed an additive interactive effect between nutrient additions and herbivore removals across all three treatment years in mixed grass prairie, indicating selective grazing by herbivores in the nutrient addition plots. In tallgrass prairie, we observed a sub-additive interaction between nutrient additions and herbivore removals during a year when the system was unburned (two-year burn cycle) due to high levels of litter built up in plots with both nutrients added and herbivores removed, demonstrating that herbivores likely play a large role in breaking down and distributing litter in this system. Across all three sites the dominant C₄ grasses drove the productivity responses to nutrient additions and herbivore removals. These results have implications for ecosystem function in response to human induced eutrophication and alteration of grazing regimes in natural prairie.