Little information exists on how climate variation may interact with trophic controls on annual net primary production (ANPP), and even less is known about how these interactions will affect community structure. Understanding how communities respond to climate variation and grazing will be crucial for managing grasslands with abundant large herbivores. We conducted a simulated grazing experiment in the Rainfall Manipulation Plots (RaMPs) at Konza Prairie Biological Station, Kansas. In this experiment six plots receive ambient rainfall, and six receive rainfall treatments that decrease event frequency but increase event size without altering total growing season precipitation. This latter treatment represents a more extreme rainfall regime. Within this long-term experiment we examined the effects of grazing by large herbivores (cattle and bison) by clipping all graminoids to 5 cm (leaving all forbs unclipped) each month during the growing season (Jun-Aug). We predicted that (1) a more extreme rainfall regime would decrease grass cover and increase forb cover, (2) grazing would increase grass biomass due to overcompensation in the ambient rainfall treatment, and (3) grazing would decrease grass biomass further due to enhanced stress and increase forb cover in the extreme rainfall treatment.

No treatment effects on species richness occurred in 2008. More extreme rainfall caused an increase in forb cover within both grazed and ungrazed treatments; however, grazing had no effect on forb cover in the first year of the experiment. A more extreme rainfall regime decreased ANPP overall. With no grazing, ANPP was similar in ambient rainfall and extreme rainfall treatments (566.6 + 48.7(SE) g/m² and 549.7 + 48.5 g/m²). Grazing caused a significant decrease in grass ANPP and, moreover, grass ANPP in grazed plots with extreme rainfall was 247.1 + 20.0 g/m² while ANPP in ambient rainfall grazed plots was 387.5 + 16.0 g/m². Thus, when grazed, a more extreme rainfall regime significantly reduced total grass ANPP. We found no differences in forb and total ANPP among treatments. Overall, these results suggest that grass ANPP in grazed grasslands will be more sensitive to changes in rainfall patterns than ungrazed grasslands. As treatments continue, we predict that community responses via increases in forbs will further affect the sensitivity of this grassland to more extreme rainfall regimes.