Wednesday, August 4, 2010 - 10:50 AM

COS 47-9: Response of semi-arid grasslands to severe multi-year drought

Karie L. Cherwin, Colorado State University

Background/Question/Methods

Global climate change models are predicting an increase in the variability and intensity of extreme weather events, including multi-year droughts, which are already common in semi-arid regions.  Grasslands in general, and the shortgrass-steppe in particular, are among the most vulnerable ecosystems to global climate change, therefore it is critical to determine both the magnitude and the underlying mechanisms of their responses to severe drought and how these mechanisms vary across space and time.  Although annual net primary productivity (ANPP) is heavily influenced by precipitation in grasslands, time series analyses have revealed temporal lags in recovery of ANPP following drought.  These time lags may create a "window of opportunity" for invasive species to colonize shortgrass ecosystems because of the extended post-drought reduction in plant cover and meristem density, and the accumulation of soil nitrogen.  The objective of this project is to better understand the effects of severe, multi-year drought on the invasibility of semi-arid grassland ecosystems across a broad geographic scale.  To simulate drought, rainout shelters were used to impose severe two year droughts at three sites across the shortgrass-steppe ecoregion of the United States.  Recovery of the grassland was assessed in the third growing season.  To determine whether invasibility increased in a post-drought community we added seeds of five locally weedy species to the plots to determine the relative success of seedling establishment. 

Results/Conclusions Results show the rainout shelters effectively created drought at all three sites by reducing daily average soil moisture up to 70% of ambient levels.  Invasibility following severe, multi-year drought increased significantly at one site but showed only marginal increases at the two other sites.   Aboveground net primary productivity, species composition, and plant-available soil nitrogen responded differently between the three sites, indicating that biome-level generalizations will be difficult to make.