As climate change alters precipitation regimes globally, many projections suggest the Midwestern United States will experience more intense and longer lasting droughts, increasing plant water stress and potentially altering ecosystem structure and function. To explore the potential consequences of drought stress on a tallgrass prairie community, we established a field experiment at in West Lafayette, IN that examines how both the timing and length of droughts during the growing season impact plant communities.
Summer precipitation was manipulated from April to September 2016 to create droughts that varied in length (7 weeks, 14 weeks, and 21 weeks) and timing over the growing season (beginning (B), middle (M), and end (E), resulting in a total of 8 treatments, B, M, E, BM, ME, BE, BME, and an ambient control (C). During exclusion periods, precipitation was reduced in experimental plots by 100% using rainout shelters. We hypothesized that aboveground primary productivity (ANPP), measured using biomass harvests, would be reduced in all drought treatments and would be cumulative for longer droughts. Using percent cover measurements and biomass sorted into species, we expected to see fewer species in longer droughts, but greater diversity in early droughts as reduced aboveground biomass would reduce competition for light.
We found that this plant community’s aboveground growth was resistant to drought, most notably in the BME treatment. Despite drastic reductions in soil moisture across depths (10cm to 100cm) in BME plots, ANPP and community structure of the vegetation remained similar to that of control plots. In fact, ANPP was not different in any plot after the October harvest. Differences in plant functional type (PFT) composition, as well as the amount of senesced plant material prior to harvest, were detected in some treatments. E treatments had more senesced material in October and higher C4 grass composition in the August harvest. Treatments that included a drought in the middle of the growing season, regardless of their length (M, ME, or BME), had less senesced material in October, but only M and ME had higher forb concentrations.
The timing of drought appears to alter the timing of senescence, where a drought in the middle of the growing season delayed senescence and a drought at the end of the growing season advanced senescence. Our results suggest that the ANPP of this plant community is remarkably resistant to drought, regardless of its timing or length. This resistance could be attributed to shifts in PFT.